id	sid	tid	token	lemma	pos
cana-6247	1	1	communications	communication	NOUN
cana-6247	1	2	on	on	ADP
cana-6247	1	3	applied	apply	VERB
cana-6247	1	4	nonlinear	nonlinear	ADJ
cana-6247	1	5	analysis	analysis	NOUN
cana-6247	1	6	issn	issn	NOUN
cana-6247	1	7	:	:	PUNCT
cana-6247	1	8	1074	1074	NUM
cana-6247	1	9	-	-	PUNCT
cana-6247	1	10	133x	133x	NUM
cana-6247	1	11	vol	vol	NOUN
cana-6247	1	12	.	.	PROPN
cana-6247	2	1	31	31	NUM
cana-6247	3	1	no	no	NOUN
cana-6247	3	2	.	.	PUNCT
cana-6247	4	1	7s	7	NOUN
cana-6247	4	2	(	(	PUNCT
cana-6247	4	3	2024	2024	NUM
cana-6247	4	4	)	)	PUNCT
cana-6247	4	5	comprehensive	comprehensive	ADJ
cana-6247	4	6	survey	survey	NOUN
cana-6247	4	7	of	of	ADP
cana-6247	4	8	noise	noise	NOUN
cana-6247	4	9	strategies	strategy	NOUN
cana-6247	4	10	in	in	ADP
cana-6247	4	11	diffusion	diffusion	NOUN
cana-6247	4	12	model	model	NOUN
cana-6247	4	13	frameworks	framework	NOUN
cana-6247	4	14	sanjay	sanjay	PROPN
cana-6247	4	15	kumar1	kumar1	PROPN
cana-6247	4	16	,	,	PUNCT
cana-6247	4	17	adhisha	adhisha	PROPN
cana-6247	4	18	sahoo2	sahoo2	NOUN
cana-6247	4	19	,	,	PUNCT
cana-6247	4	20	hari	hari	X
cana-6247	4	21	pratap3∗	pratap3∗	PROPN
cana-6247	4	22	1sag	1sag	NUM
cana-6247	4	23	,	,	PUNCT
cana-6247	4	24	drdo	drdo	PROPN
cana-6247	4	25	,	,	PUNCT
cana-6247	4	26	new	new	ADJ
cana-6247	4	27	delhi	delhi	PROPN
cana-6247	4	28	,	,	PUNCT
cana-6247	4	29	delhi-110054	delhi-110054	ADV
cana-6247	4	30	,	,	PUNCT
cana-6247	4	31	india	india	PROPN
cana-6247	4	32	2manipal	2manipal	NUM
cana-6247	4	33	institute	institute	PROPN
cana-6247	4	34	of	of	ADP
cana-6247	4	35	technology	technology	PROPN
cana-6247	4	36	,	,	PUNCT
cana-6247	4	37	manipal	manipal	NOUN
cana-6247	4	38	,	,	PUNCT
cana-6247	4	39	udupi	udupi	NOUN
cana-6247	4	40	,	,	PUNCT
cana-6247	4	41	karnataka-576104	karnataka-576104	NOUN
cana-6247	4	42	,	,	PUNCT
cana-6247	4	43	india	india	PROPN
cana-6247	4	44	3department	3department	PROPN
cana-6247	4	45	of	of	ADP
cana-6247	4	46	mathematics	mathematic	NOUN
cana-6247	4	47	,	,	PUNCT
cana-6247	4	48	pgdav(e	pgdav(e	NOUN
cana-6247	4	49	)	)	PUNCT
cana-6247	4	50	college	college	NOUN
cana-6247	4	51	,	,	PUNCT
cana-6247	4	52	university	university	NOUN
cana-6247	4	53	of	of	ADP
cana-6247	4	54	delhi	delhi	PROPN
cana-6247	4	55	,	,	PUNCT
cana-6247	4	56	delhi-110065	delhi-110065	PRON
cana-6247	4	57	,	,	PUNCT
cana-6247	4	58	india	india	PROPN
cana-6247	4	59	∗corresponding	∗corresponde	VERB
cana-6247	4	60	author	author	NOUN
cana-6247	4	61	:	:	PUNCT
cana-6247	4	62	pratap.pgdavcollege.hari@gmail.com	pratap.pgdavcollege.hari@gmail.com	PRON
cana-6247	4	63	received	receive	VERB
cana-6247	4	64	:	:	PUNCT
cana-6247	4	65	02	02	NUM
cana-6247	4	66	-	-	SYM
cana-6247	4	67	11	11	NUM
cana-6247	4	68	-	-	PUNCT
cana-6247	4	69	2024	2024	NUM
cana-6247	4	70	revised	revise	VERB
cana-6247	4	71	:	:	PUNCT
cana-6247	4	72	07	07	NUM
cana-6247	4	73	-	-	SYM
cana-6247	4	74	11	11	NUM
cana-6247	4	75	-	-	PUNCT
cana-6247	4	76	2024	2024	NUM
cana-6247	4	77	accepted	accept	VERB
cana-6247	4	78	:	:	PUNCT
cana-6247	4	79	11	11	NUM
cana-6247	4	80	-	-	SYM
cana-6247	4	81	11	11	NUM
cana-6247	4	82	-	-	SYM
cana-6247	4	83	2024	2024	NUM
cana-6247	4	84	published	publish	VERB
cana-6247	4	85	:	:	PUNCT
cana-6247	4	86	19	19	NUM
cana-6247	4	87	-	-	SYM
cana-6247	4	88	11	11	NUM
cana-6247	4	89	-	-	PUNCT
cana-6247	4	90	2024	2024	NUM
cana-6247	4	91	abstract	abstract	NOUN
cana-6247	4	92	:	:	PUNCT
cana-6247	4	93	diffusion	diffusion	NOUN
cana-6247	4	94	models	model	NOUN
cana-6247	4	95	are	be	AUX
cana-6247	4	96	swiftly	swiftly	ADV
cana-6247	4	97	evolving	evolve	VERB
cana-6247	4	98	the	the	DET
cana-6247	4	99	state	state	NOUN
cana-6247	4	100	-	-	PUNCT
cana-6247	4	101	of	of	ADP
cana-6247	4	102	-	-	PUNCT
cana-6247	4	103	the	the	DET
cana-6247	4	104	-	-	PUNCT
cana-6247	4	105	art	art	NOUN
cana-6247	4	106	for	for	ADP
cana-6247	4	107	image	image	NOUN
cana-6247	4	108	enhancement	enhancement	NOUN
cana-6247	4	109	applications	application	NOUN
cana-6247	4	110	,	,	PUNCT
cana-6247	4	111	resulting	result	VERB
cana-6247	4	112	in	in	ADP
cana-6247	4	113	a	a	DET
cana-6247	4	114	strong	strong	ADJ
cana-6247	4	115	generalizable	generalizable	ADJ
cana-6247	4	116	framework	framework	NOUN
cana-6247	4	117	for	for	ADP
cana-6247	4	118	restoration	restoration	NOUN
cana-6247	4	119	,	,	PUNCT
cana-6247	4	120	super	super	NOUN
cana-6247	4	121	-	-	NOUN
cana-6247	4	122	resolution	resolution	NOUN
cana-6247	4	123	,	,	PUNCT
cana-6247	4	124	inpainting	inpainting	ADJ
cana-6247	4	125	,	,	PUNCT
cana-6247	4	126	deblurring	deblurring	ADJ
cana-6247	4	127	,	,	PUNCT
cana-6247	4	128	low	low	ADJ
cana-6247	4	129	-	-	PUNCT
cana-6247	4	130	light	light	NOUN
cana-6247	4	131	imaging	imaging	NOUN
cana-6247	4	132	,	,	PUNCT
cana-6247	4	133	etc	etc	X
cana-6247	4	134	.	.	X
cana-6247	5	1	this	this	DET
cana-6247	5	2	survey	survey	NOUN
cana-6247	5	3	provides	provide	VERB
cana-6247	5	4	a	a	DET
cana-6247	5	5	novel	novel	ADJ
cana-6247	5	6	and	and	CCONJ
cana-6247	5	7	inclusive	inclusive	ADJ
cana-6247	5	8	noise	noise	NOUN
cana-6247	5	9	/	/	SYM
cana-6247	5	10	initialization	initialization	NOUN
cana-6247	5	11	-	-	PUNCT
cana-6247	5	12	based	base	VERB
cana-6247	5	13	taxonomy	taxonomy	NOUN
cana-6247	5	14	with	with	ADP
cana-6247	5	15	diffusion	diffusion	NOUN
cana-6247	5	16	models	model	NOUN
cana-6247	5	17	placed	place	VERB
cana-6247	5	18	in	in	ADP
cana-6247	5	19	relationship	relationship	NOUN
cana-6247	5	20	to	to	ADP
cana-6247	5	21	the	the	DET
cana-6247	5	22	type	type	NOUN
cana-6247	5	23	of	of	ADP
cana-6247	5	24	noise	noise	NOUN
cana-6247	5	25	mechanisms	mechanism	NOUN
cana-6247	5	26	exploited	exploit	VERB
cana-6247	5	27	:	:	PUNCT
cana-6247	6	1	additive	additive	ADJ
cana-6247	6	2	gaussian	gaussian	ADJ
cana-6247	6	3	noise	noise	NOUN
cana-6247	6	4	(	(	PUNCT
cana-6247	6	5	agn	agn	NOUN
cana-6247	6	6	)	)	PUNCT
cana-6247	6	7	,	,	PUNCT
cana-6247	6	8	conditional	conditional	ADJ
cana-6247	6	9	noise	noise	NOUN
cana-6247	6	10	injection	injection	NOUN
cana-6247	6	11	(	(	PUNCT
cana-6247	6	12	cni	cni	PROPN
cana-6247	6	13	)	)	PUNCT
cana-6247	6	14	,	,	PUNCT
cana-6247	6	15	learned	learn	VERB
cana-6247	6	16	noise	noise	NOUN
cana-6247	6	17	(	(	PUNCT
cana-6247	6	18	ln	ln	ADJ
cana-6247	6	19	)	)	PUNCT
cana-6247	6	20	,	,	PUNCT
cana-6247	6	21	and	and	CCONJ
cana-6247	6	22	poisson	poisson	PROPN
cana-6247	6	23	/	/	SYM
cana-6247	6	24	signal	signal	NOUN
cana-6247	6	25	dependent	dependent	ADJ
cana-6247	6	26	noise	noise	NOUN
cana-6247	6	27	(	(	PUNCT
cana-6247	6	28	psd	psd	NOUN
cana-6247	6	29	)	)	PUNCT
cana-6247	6	30	.	.	PUNCT
cana-6247	7	1	for	for	ADP
cana-6247	7	2	each	each	DET
cana-6247	7	3	category	category	NOUN
cana-6247	7	4	,	,	PUNCT
cana-6247	7	5	the	the	DET
cana-6247	7	6	main	main	ADJ
cana-6247	7	7	methods	method	NOUN
cana-6247	7	8	and	and	CCONJ
cana-6247	7	9	benefits	benefit	NOUN
cana-6247	7	10	are	be	AUX
cana-6247	7	11	discussed	discuss	VERB
cana-6247	7	12	,	,	PUNCT
cana-6247	7	13	focusing	focus	VERB
cana-6247	7	14	on	on	ADP
cana-6247	7	15	how	how	SCONJ
cana-6247	7	16	dimensional	dimensional	ADJ
cana-6247	7	17	or	or	CCONJ
cana-6247	7	18	hierarchical	hierarchical	ADJ
cana-6247	7	19	control	control	NOUN
cana-6247	7	20	of	of	ADP
cana-6247	7	21	noise	noise	NOUN
cana-6247	7	22	results	result	NOUN
cana-6247	7	23	in	in	ADP
cana-6247	7	24	increased	increase	VERB
cana-6247	7	25	perceptual	perceptual	ADJ
cana-6247	7	26	quality	quality	NOUN
cana-6247	7	27	,	,	PUNCT
cana-6247	7	28	improved	improved	ADJ
cana-6247	7	29	restoration	restoration	NOUN
cana-6247	7	30	accuracy	accuracy	NOUN
cana-6247	7	31	,	,	PUNCT
cana-6247	7	32	and	and	CCONJ
cana-6247	7	33	greater	great	ADJ
cana-6247	7	34	flexibility	flexibility	NOUN
cana-6247	7	35	to	to	PART
cana-6247	7	36	compensate	compensate	VERB
cana-6247	7	37	for	for	ADP
cana-6247	7	38	real	real	ADJ
cana-6247	7	39	-	-	PUNCT
cana-6247	7	40	world	world	NOUN
cana-6247	7	41	degradations	degradation	NOUN
cana-6247	7	42	.	.	PUNCT
cana-6247	8	1	such	such	ADJ
cana-6247	8	2	approaches	approach	NOUN
cana-6247	8	3	include	include	VERB
cana-6247	8	4	recent	recent	ADJ
cana-6247	8	5	advances	advance	NOUN
cana-6247	8	6	that	that	PRON
cana-6247	8	7	combine	combine	VERB
cana-6247	8	8	scheduling	scheduling	NOUN
cana-6247	8	9	,	,	PUNCT
cana-6247	8	10	adaptive	adaptive	ADJ
cana-6247	8	11	conditioning	conditioning	NOUN
cana-6247	8	12	,	,	PUNCT
cana-6247	8	13	and	and	CCONJ
cana-6247	8	14	physical	physical	ADJ
cana-6247	8	15	signal	signal	NOUN
cana-6247	8	16	beliefs	belief	NOUN
cana-6247	8	17	,	,	PUNCT
cana-6247	8	18	such	such	ADJ
cana-6247	8	19	that	that	SCONJ
cana-6247	8	20	the	the	DET
cana-6247	8	21	latter	latter	ADJ
cana-6247	8	22	models	model	NOUN
cana-6247	8	23	generalise	generalise	VERB
cana-6247	8	24	better	well	ADV
cana-6247	8	25	across	across	ADP
cana-6247	8	26	tasks	task	NOUN
cana-6247	8	27	and	and	CCONJ
cana-6247	8	28	domains	domain	NOUN
cana-6247	8	29	.	.	PUNCT
cana-6247	9	1	by	by	ADP
cana-6247	9	2	structurally	structurally	ADV
cana-6247	9	3	regrouping	regroup	VERB
cana-6247	9	4	diffusion	diffusion	NOUN
cana-6247	9	5	models	model	NOUN
cana-6247	9	6	around	around	ADP
cana-6247	9	7	noise	noise	NOUN
cana-6247	9	8	strategy	strategy	NOUN
cana-6247	9	9	,	,	PUNCT
cana-6247	9	10	we	we	PRON
cana-6247	9	11	hope	hope	VERB
cana-6247	9	12	to	to	PART
cana-6247	9	13	provide	provide	VERB
cana-6247	9	14	enough	enough	ADJ
cana-6247	9	15	clarity	clarity	NOUN
cana-6247	9	16	to	to	PART
cana-6247	9	17	rethink	rethink	VERB
cana-6247	9	18	the	the	DET
cana-6247	9	19	future	future	ADJ
cana-6247	9	20	practicality	practicality	NOUN
cana-6247	9	21	of	of	ADP
cana-6247	9	22	developing	develop	VERB
cana-6247	9	23	next	next	ADJ
cana-6247	9	24	-	-	PUNCT
cana-6247	9	25	generation	generation	NOUN
cana-6247	9	26	image	image	NOUN
cana-6247	9	27	enhancement	enhancement	NOUN
cana-6247	9	28	solutions	solution	NOUN
cana-6247	9	29	.	.	PUNCT
cana-6247	10	1	keyword	keyword	NOUN
cana-6247	10	2	:	:	PUNCT
cana-6247	10	3	additive	additive	ADJ
cana-6247	10	4	gaussian	gaussian	NOUN
cana-6247	10	5	noise	noise	NOUN
cana-6247	10	6	,	,	PUNCT
cana-6247	10	7	conditional	conditional	ADJ
cana-6247	10	8	noise	noise	NOUN
cana-6247	10	9	injection	injection	NOUN
cana-6247	10	10	,	,	PUNCT
cana-6247	10	11	gaussian	gaussian	ADJ
cana-6247	10	12	noise	noise	NOUN
cana-6247	10	13	,	,	PUNCT
cana-6247	10	14	diffusion	diffusion	NOUN
cana-6247	10	15	model	model	NOUN
cana-6247	10	16	,	,	PUNCT
cana-6247	10	17	noise	noise	NOUN
cana-6247	10	18	strategies	strategy	NOUN
cana-6247	10	19	1	1	NUM
cana-6247	10	20	introduction	introduction	NOUN
cana-6247	10	21	diffusion	diffusion	NOUN
cana-6247	10	22	models	model	NOUN
cana-6247	10	23	recently	recently	ADV
cana-6247	10	24	became	become	VERB
cana-6247	10	25	one	one	NUM
cana-6247	10	26	of	of	ADP
cana-6247	10	27	the	the	DET
cana-6247	10	28	most	most	ADV
cana-6247	10	29	powerful	powerful	ADJ
cana-6247	10	30	generative	generative	NOUN
cana-6247	10	31	and	and	CCONJ
cana-6247	10	32	restoration	restoration	NOUN
cana-6247	10	33	paradigms	paradigm	NOUN
cana-6247	10	34	in	in	ADP
cana-6247	10	35	computer	computer	NOUN
cana-6247	10	36	vision	vision	NOUN
cana-6247	10	37	.	.	PUNCT
cana-6247	11	1	diffusion	diffusion	NOUN
cana-6247	11	2	models	model	NOUN
cana-6247	11	3	can	can	AUX
cana-6247	11	4	be	be	AUX
cana-6247	11	5	thought	think	VERB
cana-6247	11	6	of	of	ADP
cana-6247	11	7	as	as	ADP
cana-6247	11	8	a	a	DET
cana-6247	11	9	two	two	NUM
cana-6247	11	10	-	-	PUNCT
cana-6247	11	11	stage	stage	NOUN
cana-6247	11	12	markovian	markovian	NOUN
cana-6247	11	13	process	process	NOUN
cana-6247	11	14	:	:	PUNCT
cana-6247	11	15	in	in	ADP
cana-6247	11	16	the	the	DET
cana-6247	11	17	forward	forward	ADJ
cana-6247	11	18	process	process	NOUN
cana-6247	11	19	(	(	PUNCT
cana-6247	11	20	or	or	CCONJ
cana-6247	11	21	the	the	DET
cana-6247	11	22	forward	forward	ADJ
cana-6247	11	23	diffusion	diffusion	NOUN
cana-6247	11	24	process	process	NOUN
cana-6247	11	25	)	)	PUNCT
cana-6247	11	26	,	,	PUNCT
cana-6247	11	27	noise	noise	NOUN
cana-6247	11	28	is	be	AUX
cana-6247	11	29	added	add	VERB
cana-6247	11	30	to	to	ADP
cana-6247	11	31	an	an	DET
cana-6247	11	32	image	image	NOUN
cana-6247	11	33	iteratively	iteratively	ADV
cana-6247	11	34	over	over	ADP
cana-6247	11	35	a	a	DET
cana-6247	11	36	series	series	NOUN
cana-6247	11	37	of	of	ADP
cana-6247	11	38	time	time	NOUN
cana-6247	11	39	steps	step	NOUN
cana-6247	11	40	until	until	SCONJ
cana-6247	11	41	the	the	DET
cana-6247	11	42	time	time	NOUN
cana-6247	11	43	-	-	PUNCT
cana-6247	11	44	dependent	dependent	ADJ
cana-6247	11	45	image	image	NOUN
cana-6247	11	46	is	be	AUX
cana-6247	11	47	indistinguishable	indistinguishable	ADJ
cana-6247	11	48	from	from	ADP
cana-6247	11	49	white	white	ADJ
cana-6247	11	50	noise	noise	NOUN
cana-6247	11	51	;	;	PUNCT
cana-6247	11	52	in	in	ADP
cana-6247	11	53	the	the	DET
cana-6247	11	54	reverse	reverse	NOUN
cana-6247	11	55	set	set	NOUN
cana-6247	11	56	of	of	ADP
cana-6247	11	57	processes	process	NOUN
cana-6247	11	58	(	(	PUNCT
cana-6247	11	59	or	or	CCONJ
cana-6247	11	60	reverse	reverse	VERB
cana-6247	11	61	denoising	denoising	NOUN
cana-6247	11	62	process	process	NOUN
cana-6247	11	63	)	)	PUNCT
cana-6247	11	64	,	,	PUNCT
cana-6247	11	65	a	a	DET
cana-6247	11	66	neural	neural	ADJ
cana-6247	11	67	network	network	NOUN
cana-6247	11	68	(	(	PUNCT
cana-6247	11	69	typically	typically	ADV
cana-6247	11	70	a	a	DET
cana-6247	11	71	u	u	NOUN
cana-6247	11	72	-	-	NOUN
cana-6247	11	73	net	net	NOUN
cana-6247	11	74	)	)	PUNCT
cana-6247	11	75	is	be	AUX
cana-6247	11	76	trained	train	VERB
cana-6247	11	77	to	to	PART
cana-6247	11	78	learn	learn	VERB
cana-6247	11	79	the	the	DET
cana-6247	11	80	reverse	reverse	ADJ
cana-6247	11	81	denoising	denoise	VERB
cana-6247	11	82	evolution	evolution	NOUN
cana-6247	11	83	that	that	PRON
cana-6247	11	84	restores	restore	VERB
cana-6247	11	85	the	the	DET
cana-6247	11	86	image	image	NOUN
cana-6247	11	87	to	to	ADP
cana-6247	11	88	the	the	DET
cana-6247	11	89	moment	moment	NOUN
cana-6247	11	90	before	before	ADP
cana-6247	11	91	all	all	DET
cana-6247	11	92	additive	additive	ADJ
cana-6247	11	93	noise	noise	NOUN
cana-6247	11	94	,	,	PUNCT
cana-6247	11	95	using	use	VERB
cana-6247	11	96	each	each	DET
cana-6247	11	97	progressively	progressively	ADV
cana-6247	11	98	noisier	noisy	ADJ
cana-6247	11	99	image	image	NOUN
cana-6247	11	100	to	to	PART
cana-6247	11	101	reconstruct	reconstruct	VERB
cana-6247	11	102	the	the	DET
cana-6247	11	103	clean	clean	ADJ
cana-6247	11	104	image	image	NOUN
cana-6247	11	105	over	over	ADP
cana-6247	11	106	the	the	DET
cana-6247	11	107	series	series	NOUN
cana-6247	11	108	of	of	ADP
cana-6247	11	109	discrete	discrete	ADJ
cana-6247	11	110	time	time	NOUN
cana-6247	11	111	steps	step	NOUN
cana-6247	11	112	.	.	PUNCT
cana-6247	12	1	this	this	DET
cana-6247	12	2	probabilistic	probabilistic	ADJ
cana-6247	12	3	modeling	modeling	NOUN
cana-6247	12	4	framework	framework	NOUN
cana-6247	12	5	is	be	AUX
cana-6247	12	6	generally	generally	ADV
cana-6247	12	7	parameterized	parameterized	ADJ
cana-6247	12	8	as	as	ADP
cana-6247	12	9	a	a	DET
cana-6247	12	10	known	know	VERB
cana-6247	12	11	(	(	PUNCT
cana-6247	12	12	multivariate	multivariate	NOUN
cana-6247	12	13	)	)	PUNCT
cana-6247	12	14	gaussian	gaussian	ADJ
cana-6247	12	15	distribution	distribution	NOUN
cana-6247	12	16	or	or	CCONJ
cana-6247	12	17	characterized	characterize	VERB
cana-6247	12	18	as	as	ADP
cana-6247	12	19	another	another	DET
cana-6247	12	20	sampling	sampling	NOUN
cana-6247	12	21	process	process	NOUN
cana-6247	12	22	by	by	ADP
cana-6247	12	23	stochastic	stochastic	ADJ
cana-6247	12	24	differential	differential	ADJ
cana-6247	12	25	equations	equation	NOUN
cana-6247	13	1	[	[	X
cana-6247	13	2	25	25	NUM
cana-6247	13	3	]	]	PUNCT
cana-6247	13	4	,	,	PUNCT
cana-6247	13	5	allowing	allow	VERB
cana-6247	13	6	the	the	DET
cana-6247	13	7	diffusion	diffusion	NOUN
cana-6247	13	8	model	model	NOUN
cana-6247	13	9	to	to	PART
cana-6247	13	10	act	act	VERB
cana-6247	13	11	as	as	ADP
cana-6247	13	12	a	a	DET
cana-6247	13	13	probabilistic	probabilistic	ADJ
cana-6247	13	14	data	datum	NOUN
cana-6247	13	15	prior	prior	ADV
cana-6247	13	16	and	and	CCONJ
cana-6247	13	17	allow	allow	VERB
cana-6247	13	18	for	for	ADP
cana-6247	13	19	any	any	DET
cana-6247	13	20	other	other	ADJ
cana-6247	13	21	type	type	NOUN
cana-6247	13	22	of	of	ADP
cana-6247	13	23	image	image	NOUN
cana-6247	13	24	restoration	restoration	NOUN
cana-6247	13	25	process	process	NOUN
cana-6247	13	26	with	with	ADP
cana-6247	13	27	a	a	DET
cana-6247	13	28	very	very	ADV
cana-6247	13	29	few	few	ADJ
cana-6247	13	30	constraints	constraint	NOUN
cana-6247	13	31	across	across	ADP
cana-6247	13	32	a	a	DET
cana-6247	13	33	wide	wide	ADJ
cana-6247	13	34	range	range	NOUN
cana-6247	13	35	of	of	ADP
cana-6247	13	36	image	image	NOUN
cana-6247	13	37	enhancement	enhancement	NOUN
cana-6247	13	38	tasks	task	NOUN
cana-6247	13	39	.	.	PUNCT
cana-6247	14	1	diffusion	diffusion	NOUN
cana-6247	14	2	for	for	ADP
cana-6247	14	3	image	image	NOUN
cana-6247	14	4	restoration	restoration	NOUN
cana-6247	14	5	fundamentally	fundamentally	ADV
cana-6247	14	6	derives	derive	VERB
cana-6247	14	7	from	from	ADP
cana-6247	14	8	classical	classical	ADJ
cana-6247	14	9	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	14	10	978	978	NUM
cana-6247	14	11	communications	communication	NOUN
cana-6247	14	12	on	on	ADP
cana-6247	14	13	applied	apply	VERB
cana-6247	14	14	nonlinear	nonlinear	ADJ
cana-6247	14	15	analysis	analysis	NOUN
cana-6247	14	16	issn	issn	NOUN
cana-6247	14	17	:	:	PUNCT
cana-6247	14	18	1074	1074	NUM
cana-6247	14	19	-	-	PUNCT
cana-6247	14	20	133x	133x	NUM
cana-6247	14	21	vol	vol	NOUN
cana-6247	14	22	.	.	PROPN
cana-6247	15	1	31	31	NUM
cana-6247	16	1	no	no	NOUN
cana-6247	16	2	.	.	PUNCT
cana-6247	17	1	7s	7	NOUN
cana-6247	17	2	(	(	PUNCT
cana-6247	17	3	2024	2024	NUM
cana-6247	17	4	)	)	PUNCT
cana-6247	17	5	pde	pde	NOUN
cana-6247	17	6	-	-	PUNCT
cana-6247	17	7	based	base	VERB
cana-6247	17	8	diffusion	diffusion	NOUN
cana-6247	17	9	methods	method	NOUN
cana-6247	17	10	.	.	PUNCT
cana-6247	18	1	the	the	DET
cana-6247	18	2	development	development	NOUN
cana-6247	18	3	of	of	ADP
cana-6247	18	4	anisotropic	anisotropic	NOUN
cana-6247	18	5	diffusion	diffusion	NOUN
cana-6247	18	6	and	and	CCONJ
cana-6247	18	7	reaction	reaction	NOUN
cana-6247	18	8	–	–	PUNCT
cana-6247	18	9	diffusion	diffusion	NOUN
cana-6247	18	10	models	model	NOUN
cana-6247	18	11	was	be	AUX
cana-6247	18	12	foundational	foundational	ADJ
cana-6247	18	13	in	in	ADP
cana-6247	18	14	demonstrating	demonstrate	VERB
cana-6247	18	15	how	how	SCONJ
cana-6247	18	16	diffusion	diffusion	NOUN
cana-6247	18	17	can	can	AUX
cana-6247	18	18	be	be	AUX
cana-6247	18	19	guided	guide	VERB
cana-6247	18	20	to	to	PART
cana-6247	18	21	remove	remove	VERB
cana-6247	18	22	noise	noise	NOUN
cana-6247	18	23	while	while	SCONJ
cana-6247	18	24	simultaneously	simultaneously	ADV
cana-6247	18	25	preserving	preserve	VERB
cana-6247	18	26	edges	edge	NOUN
cana-6247	18	27	[	[	X
cana-6247	18	28	7	7	X
cana-6247	18	29	]	]	X
cana-6247	18	30	[	[	X
cana-6247	18	31	2	2	NUM
cana-6247	18	32	]	]	PUNCT
cana-6247	18	33	;	;	PUNCT
cana-6247	18	34	because	because	SCONJ
cana-6247	18	35	of	of	ADP
cana-6247	18	36	this	this	PRON
cana-6247	18	37	,	,	PUNCT
cana-6247	18	38	anisotropic	anisotropic	NOUN
cana-6247	18	39	diffusion	diffusion	NOUN
cana-6247	18	40	and	and	CCONJ
cana-6247	18	41	reaction	reaction	NOUN
cana-6247	18	42	–	–	PUNCT
cana-6247	18	43	diffusion	diffusion	NOUN
cana-6247	18	44	models	model	NOUN
cana-6247	18	45	have	have	AUX
cana-6247	18	46	become	become	VERB
cana-6247	18	47	prevalent	prevalent	ADJ
cana-6247	18	48	in	in	ADP
cana-6247	18	49	,	,	PUNCT
cana-6247	18	50	for	for	ADP
cana-6247	18	51	instance	instance	NOUN
cana-6247	18	52	,	,	PUNCT
cana-6247	18	53	medical	medical	ADJ
cana-6247	18	54	imaging	imaging	NOUN
cana-6247	18	55	and	and	CCONJ
cana-6247	18	56	radiography	radiography	NOUN
cana-6247	18	57	.	.	PUNCT
cana-6247	19	1	the	the	DET
cana-6247	19	2	introduction	introduction	NOUN
cana-6247	19	3	of	of	ADP
cana-6247	19	4	fractional	fractional	ADJ
cana-6247	19	5	and	and	CCONJ
cana-6247	19	6	nonlinear	nonlinear	ADJ
cana-6247	19	7	higherorder	higherorder	NOUN
cana-6247	19	8	diffusion	diffusion	NOUN
cana-6247	19	9	processes	process	NOUN
cana-6247	19	10	improved	improve	VERB
cana-6247	19	11	stability	stability	NOUN
cana-6247	19	12	in	in	ADP
cana-6247	19	13	the	the	DET
cana-6247	19	14	diffusion	diffusion	NOUN
cana-6247	19	15	process	process	NOUN
cana-6247	19	16	while	while	SCONJ
cana-6247	19	17	maintaining	maintain	VERB
cana-6247	19	18	a	a	DET
cana-6247	19	19	more	more	ADV
cana-6247	19	20	descriptive	descriptive	ADJ
cana-6247	19	21	framework	framework	NOUN
cana-6247	19	22	to	to	PART
cana-6247	19	23	model	model	VERB
cana-6247	19	24	the	the	DET
cana-6247	19	25	more	more	ADV
cana-6247	19	26	complex	complex	ADJ
cana-6247	19	27	dynamics	dynamic	NOUN
cana-6247	19	28	of	of	ADP
cana-6247	19	29	images	image	NOUN
cana-6247	19	30	[	[	X
cana-6247	19	31	3	3	NUM
cana-6247	19	32	]	]	PUNCT
cana-6247	19	33	.	.	PUNCT
cana-6247	20	1	hybrid	hybrid	ADJ
cana-6247	20	2	processes	process	NOUN
cana-6247	20	3	that	that	PRON
cana-6247	20	4	deployed	deploy	VERB
cana-6247	20	5	isotropic	isotropic	ADJ
cana-6247	20	6	diffusion	diffusion	NOUN
cana-6247	20	7	with	with	ADP
cana-6247	20	8	perona	perona	PROPN
cana-6247	20	9	–	–	PUNCT
cana-6247	20	10	malik	malik	PROPN
cana-6247	20	11	style	style	NOUN
cana-6247	20	12	delivered	deliver	VERB
cana-6247	20	13	superior	superior	ADJ
cana-6247	20	14	denoising	denoising	NOUN
cana-6247	20	15	while	while	SCONJ
cana-6247	20	16	maintaining	maintain	VERB
cana-6247	20	17	structural	structural	ADJ
cana-6247	20	18	integrity	integrity	NOUN
cana-6247	20	19	and	and	CCONJ
cana-6247	20	20	fidelity	fidelity	NOUN
cana-6247	20	21	[	[	X
cana-6247	20	22	40	40	NUM
cana-6247	20	23	]	]	PUNCT
cana-6247	20	24	.	.	PUNCT
cana-6247	21	1	collectively	collectively	ADV
cana-6247	21	2	,	,	PUNCT
cana-6247	21	3	the	the	DET
cana-6247	21	4	classical	classical	ADJ
cana-6247	21	5	models	model	NOUN
cana-6247	21	6	demonstrated	demonstrate	VERB
cana-6247	21	7	the	the	DET
cana-6247	21	8	capacity	capacity	NOUN
cana-6247	21	9	for	for	ADP
cana-6247	21	10	mathematical	mathematical	ADJ
cana-6247	21	11	diffusion	diffusion	NOUN
cana-6247	21	12	to	to	PART
cana-6247	21	13	assist	assist	VERB
cana-6247	21	14	in	in	ADP
cana-6247	21	15	noise	noise	NOUN
cana-6247	21	16	removal	removal	NOUN
cana-6247	21	17	,	,	PUNCT
cana-6247	21	18	contrast	contrast	NOUN
cana-6247	21	19	enhancement	enhancement	NOUN
cana-6247	21	20	,	,	PUNCT
cana-6247	21	21	and	and	CCONJ
cana-6247	21	22	defect	defect	ADJ
cana-6247	21	23	detection	detection	NOUN
cana-6247	21	24	.	.	PUNCT
cana-6247	22	1	the	the	DET
cana-6247	22	2	emergence	emergence	NOUN
cana-6247	22	3	of	of	ADP
cana-6247	22	4	deep	deep	ADJ
cana-6247	22	5	generative	generative	ADJ
cana-6247	22	6	models	model	NOUN
cana-6247	22	7	paved	pave	VERB
cana-6247	22	8	the	the	DET
cana-6247	22	9	way	way	NOUN
cana-6247	22	10	for	for	ADP
cana-6247	22	11	the	the	DET
cana-6247	22	12	creation	creation	NOUN
cana-6247	22	13	of	of	ADP
cana-6247	22	14	denoising	denoise	VERB
cana-6247	22	15	diffusion	diffusion	NOUN
cana-6247	22	16	probabilistic	probabilistic	ADJ
cana-6247	22	17	models	model	NOUN
cana-6247	22	18	(	(	PUNCT
cana-6247	22	19	ddpms	ddpm	NOUN
cana-6247	22	20	)	)	PUNCT
cana-6247	22	21	and	and	CCONJ
cana-6247	22	22	their	their	PRON
cana-6247	22	23	derivatives	derivative	NOUN
cana-6247	22	24	.	.	PUNCT
cana-6247	23	1	ddrm	ddrm	NOUN
cana-6247	23	2	and	and	CCONJ
cana-6247	23	3	sdedit	sdedit	NOUN
cana-6247	23	4	showed	show	VERB
cana-6247	23	5	how	how	SCONJ
cana-6247	23	6	it	it	PRON
cana-6247	23	7	was	be	AUX
cana-6247	23	8	possible	possible	ADJ
cana-6247	23	9	to	to	PART
cana-6247	23	10	apply	apply	VERB
cana-6247	23	11	gaussian	gaussian	ADJ
cana-6247	23	12	noise	noise	NOUN
cana-6247	23	13	schedules	schedule	NOUN
cana-6247	23	14	to	to	PART
cana-6247	23	15	guided	guide	VERB
cana-6247	23	16	denoising	denoising	NOUN
cana-6247	23	17	and	and	CCONJ
cana-6247	23	18	editing	editing	NOUN
cana-6247	23	19	;	;	PUNCT
cana-6247	23	20	score	score	NOUN
cana-6247	23	21	-	-	PUNCT
cana-6247	23	22	based	base	VERB
cana-6247	23	23	generative	generative	ADJ
cana-6247	23	24	models	model	NOUN
cana-6247	23	25	reframed	reframe	VERB
cana-6247	23	26	diffusion	diffusion	NOUN
cana-6247	23	27	in	in	ADP
cana-6247	23	28	terms	term	NOUN
cana-6247	23	29	of	of	ADP
cana-6247	23	30	stochastic	stochastic	ADJ
cana-6247	23	31	differential	differential	ADJ
cana-6247	23	32	equations	equation	NOUN
cana-6247	23	33	[	[	X
cana-6247	23	34	25	25	NUM
cana-6247	23	35	]	]	PUNCT
cana-6247	23	36	.	.	PUNCT
cana-6247	24	1	works	work	VERB
cana-6247	24	2	like	like	ADP
cana-6247	24	3	diffir	diffir	NOUN
cana-6247	24	4	exhibited	exhibit	VERB
cana-6247	24	5	the	the	DET
cana-6247	24	6	utility	utility	NOUN
cana-6247	24	7	of	of	ADP
cana-6247	24	8	these	these	DET
cana-6247	24	9	approaches	approach	NOUN
cana-6247	24	10	in	in	ADP
cana-6247	24	11	image	image	NOUN
cana-6247	24	12	restoration	restoration	NOUN
cana-6247	24	13	,	,	PUNCT
cana-6247	24	14	and	and	CCONJ
cana-6247	24	15	guided	guide	VERB
cana-6247	24	16	diffusion	diffusion	NOUN
cana-6247	24	17	for	for	ADP
cana-6247	24	18	adversarial	adversarial	ADJ
cana-6247	24	19	purification	purification	NOUN
cana-6247	24	20	provided	provide	VERB
cana-6247	24	21	an	an	DET
cana-6247	24	22	appropriate	appropriate	ADJ
cana-6247	24	23	level	level	NOUN
cana-6247	24	24	of	of	ADP
cana-6247	24	25	confidence	confidence	NOUN
cana-6247	24	26	regarding	regard	VERB
cana-6247	24	27	safety	safety	NOUN
cana-6247	24	28	and	and	CCONJ
cana-6247	24	29	robustness	robustness	NOUN
cana-6247	24	30	.	.	PUNCT
cana-6247	25	1	related	relate	VERB
cana-6247	25	2	frameworks	framework	NOUN
cana-6247	25	3	such	such	ADJ
cana-6247	25	4	as	as	ADP
cana-6247	25	5	plug	plug	NOUN
cana-6247	25	6	-	-	PUNCT
cana-6247	25	7	and	and	CCONJ
cana-6247	25	8	-	-	PUNCT
cana-6247	25	9	play	play	NOUN
cana-6247	25	10	diffusion	diffusion	NOUN
cana-6247	25	11	models	model	NOUN
cana-6247	25	12	[	[	X
cana-6247	25	13	56	56	NUM
cana-6247	25	14	]	]	PUNCT
cana-6247	25	15	and	and	CCONJ
cana-6247	25	16	resshift	resshift	PROPN
cana-6247	25	17	further	far	ADV
cana-6247	25	18	improved	improve	VERB
cana-6247	25	19	computational	computational	ADJ
cana-6247	25	20	efficiency	efficiency	NOUN
cana-6247	25	21	by	by	ADP
cana-6247	25	22	using	use	VERB
cana-6247	25	23	priors	prior	NOUN
cana-6247	25	24	along	along	ADP
cana-6247	25	25	with	with	ADP
cana-6247	25	26	early	early	ADJ
cana-6247	25	27	stopping	stopping	NOUN
cana-6247	25	28	strategies	strategy	NOUN
cana-6247	25	29	,	,	PUNCT
cana-6247	25	30	while	while	SCONJ
cana-6247	25	31	residual	residual	ADJ
cana-6247	25	32	noise	noise	NOUN
cana-6247	25	33	priors	prior	NOUN
cana-6247	25	34	[	[	X
cana-6247	25	35	35	35	NUM
cana-6247	25	36	]	]	X
cana-6247	26	1	[	[	X
cana-6247	26	2	50	50	NUM
cana-6247	26	3	]	]	PUNCT
cana-6247	26	4	explicitly	explicitly	ADV
cana-6247	26	5	modelled	model	VERB
cana-6247	26	6	noise	noise	NOUN
cana-6247	26	7	structure	structure	NOUN
cana-6247	26	8	to	to	PART
cana-6247	26	9	enhance	enhance	VERB
cana-6247	26	10	restoration	restoration	NOUN
cana-6247	26	11	.	.	PUNCT
cana-6247	27	1	as	as	SCONJ
cana-6247	27	2	tasks	task	NOUN
cana-6247	27	3	grew	grow	VERB
cana-6247	27	4	more	more	ADV
cana-6247	27	5	varied	varied	ADJ
cana-6247	27	6	,	,	PUNCT
cana-6247	27	7	different	different	ADJ
cana-6247	27	8	applications	application	NOUN
cana-6247	27	9	of	of	ADP
cana-6247	27	10	diffusion	diffusion	NOUN
cana-6247	27	11	were	be	AUX
cana-6247	27	12	developed	develop	VERB
cana-6247	27	13	in	in	ADP
cana-6247	27	14	a	a	DET
cana-6247	27	15	specialized	specialized	ADJ
cana-6247	27	16	manner	manner	NOUN
cana-6247	27	17	.	.	PUNCT
cana-6247	28	1	ddfm	ddfm	PROPN
cana-6247	28	2	applied	apply	VERB
cana-6247	28	3	noise	noise	NOUN
cana-6247	28	4	-	-	PUNCT
cana-6247	28	5	dependent	dependent	ADJ
cana-6247	28	6	blending	blending	NOUN
cana-6247	28	7	for	for	ADP
cana-6247	28	8	multimodal	multimodal	ADJ
cana-6247	28	9	fusion	fusion	NOUN
cana-6247	28	10	(	(	PUNCT
cana-6247	28	11	e.g.	e.g.	ADV
cana-6247	28	12	,	,	PUNCT
cana-6247	28	13	visible	visible	ADJ
cana-6247	28	14	and	and	CCONJ
cana-6247	28	15	infrared	infrared	ADJ
cana-6247	28	16	images	image	NOUN
cana-6247	28	17	)	)	PUNCT
cana-6247	29	1	[	[	X
cana-6247	29	2	54	54	NUM
cana-6247	29	3	]	]	PUNCT
cana-6247	29	4	.	.	PUNCT
cana-6247	30	1	jpeg	jpeg	NOUN
cana-6247	30	2	restoration	restoration	NOUN
cana-6247	30	3	models	model	NOUN
cana-6247	30	4	learned	learn	VERB
cana-6247	30	5	forward	forward	ADP
cana-6247	30	6	degradation	degradation	NOUN
cana-6247	30	7	while	while	SCONJ
cana-6247	30	8	treating	treat	VERB
cana-6247	30	9	compression	compression	NOUN
cana-6247	30	10	artifacts	artifact	NOUN
cana-6247	30	11	as	as	ADP
cana-6247	30	12	statistical	statistical	ADJ
cana-6247	30	13	noise	noise	NOUN
cana-6247	30	14	[	[	X
cana-6247	30	15	15	15	NUM
cana-6247	30	16	]	]	PUNCT
cana-6247	30	17	.	.	PUNCT
cana-6247	31	1	repaint	repaint	NOUN
cana-6247	31	2	generalized	generalized	ADJ
cana-6247	31	3	non	non	ADJ
cana-6247	31	4	-	-	ADJ
cana-6247	31	5	uniform	uniform	ADJ
cana-6247	31	6	corruption	corruption	NOUN
cana-6247	31	7	to	to	ADP
cana-6247	31	8	inpainting	inpainte	VERB
cana-6247	31	9	and	and	CCONJ
cana-6247	31	10	more	more	ADV
cana-6247	31	11	contextually	contextually	ADV
cana-6247	31	12	faithful	faithful	ADJ
cana-6247	31	13	inpainting	inpainting	ADJ
cana-6247	31	14	[	[	X
cana-6247	31	15	22	22	NUM
cana-6247	31	16	]	]	PUNCT
cana-6247	31	17	.	.	PUNCT
cana-6247	32	1	adverse	adverse	ADJ
cana-6247	32	2	weather	weather	NOUN
cana-6247	32	3	restoration	restoration	NOUN
cana-6247	32	4	models	model	NOUN
cana-6247	32	5	used	use	VERB
cana-6247	32	6	patch	patch	NOUN
cana-6247	32	7	-	-	PUNCT
cana-6247	32	8	based	base	VERB
cana-6247	32	9	denoising	denoising	NOUN
cana-6247	32	10	for	for	ADP
cana-6247	32	11	fog	fog	NOUN
cana-6247	32	12	and	and	CCONJ
cana-6247	32	13	haze	haze	NOUN
cana-6247	32	14	removal	removal	NOUN
cana-6247	32	15	[	[	X
cana-6247	32	16	28	28	NUM
cana-6247	32	17	]	]	PUNCT
cana-6247	32	18	.	.	PUNCT
cana-6247	33	1	cold	cold	ADJ
cana-6247	33	2	diffusion	diffusion	NOUN
cana-6247	33	3	generalized	generalize	VERB
cana-6247	33	4	the	the	DET
cana-6247	33	5	forward	forward	ADJ
cana-6247	33	6	process	process	NOUN
cana-6247	33	7	to	to	PART
cana-6247	33	8	replace	replace	VERB
cana-6247	33	9	stochastic	stochastic	ADJ
cana-6247	33	10	noise	noise	NOUN
cana-6247	33	11	with	with	ADP
cana-6247	33	12	deterministic	deterministic	ADJ
cana-6247	33	13	forms	form	NOUN
cana-6247	33	14	of	of	ADP
cana-6247	33	15	degradation	degradation	NOUN
cana-6247	33	16	,	,	PUNCT
cana-6247	33	17	e.g.	e.g.	ADV
cana-6247	33	18	,	,	PUNCT
cana-6247	33	19	blur	blur	NOUN
cana-6247	33	20	,	,	PUNCT
cana-6247	33	21	indicating	indicate	VERB
cana-6247	33	22	the	the	DET
cana-6247	33	23	universality	universality	NOUN
cana-6247	33	24	of	of	ADP
cana-6247	33	25	diffusion	diffusion	NOUN
cana-6247	33	26	mechanisms	mechanism	NOUN
cana-6247	33	27	[	[	X
cana-6247	33	28	4	4	NUM
cana-6247	33	29	]	]	PUNCT
cana-6247	33	30	.	.	PUNCT
cana-6247	34	1	the	the	DET
cana-6247	34	2	field	field	NOUN
cana-6247	34	3	quickly	quickly	ADV
cana-6247	34	4	moved	move	VERB
cana-6247	34	5	toward	toward	ADP
cana-6247	34	6	conditional	conditional	ADJ
cana-6247	34	7	and	and	CCONJ
cana-6247	34	8	context	context	NOUN
cana-6247	34	9	-	-	PUNCT
cana-6247	34	10	aware	aware	ADJ
cana-6247	34	11	diffusion	diffusion	NOUN
cana-6247	34	12	.	.	PUNCT
cana-6247	35	1	inpdiffusion	inpdiffusion	NOUN
cana-6247	35	2	introduced	introduce	VERB
cana-6247	35	3	maskguided	maskguide	VERB
cana-6247	35	4	inpainting	inpainte	VERB
cana-6247	35	5	[	[	X
cana-6247	35	6	39	39	NUM
cana-6247	35	7	]	]	PUNCT
cana-6247	35	8	,	,	PUNCT
cana-6247	35	9	and	and	CCONJ
cana-6247	35	10	gradpaint	gradpaint	NOUN
cana-6247	35	11	employed	employ	VERB
cana-6247	35	12	gradient	gradient	NOUN
cana-6247	35	13	-	-	PUNCT
cana-6247	35	14	based	base	VERB
cana-6247	35	15	conditioning	conditioning	NOUN
cana-6247	35	16	,	,	PUNCT
cana-6247	35	17	giving	give	VERB
cana-6247	35	18	sharper	sharp	ADJ
cana-6247	35	19	contexts	context	NOUN
cana-6247	35	20	for	for	ADP
cana-6247	35	21	inpainted	inpainte	VERB
cana-6247	35	22	images	image	NOUN
cana-6247	35	23	[	[	X
cana-6247	35	24	9	9	NUM
cana-6247	35	25	]	]	PUNCT
cana-6247	35	26	.	.	PUNCT
cana-6247	36	1	noisecollage	noisecollage	NOUN
cana-6247	36	2	was	be	AUX
cana-6247	36	3	a	a	DET
cana-6247	36	4	method	method	NOUN
cana-6247	36	5	that	that	SCONJ
cana-6247	36	6	combined	combine	VERB
cana-6247	36	7	structured	structured	ADJ
cana-6247	36	8	noise	noise	NOUN
cana-6247	36	9	maps	map	NOUN
cana-6247	36	10	and	and	CCONJ
cana-6247	36	11	multi	multi	ADJ
cana-6247	36	12	-	-	ADJ
cana-6247	36	13	scale	scale	ADJ
cana-6247	36	14	noise	noise	NOUN
cana-6247	36	15	for	for	ADP
cana-6247	36	16	layout	layout	NOUN
cana-6247	36	17	-	-	PUNCT
cana-6247	36	18	aware	aware	ADJ
cana-6247	36	19	image	image	NOUN
cana-6247	36	20	generation	generation	NOUN
cana-6247	36	21	[	[	X
cana-6247	36	22	36	36	NUM
cana-6247	36	23	]	]	PUNCT
cana-6247	36	24	.	.	PUNCT
cana-6247	37	1	furthermore	furthermore	ADV
cana-6247	37	2	,	,	PUNCT
cana-6247	37	3	palette	palette	NOUN
cana-6247	37	4	,	,	PUNCT
cana-6247	37	5	pair	pair	NOUN
cana-6247	37	6	diffusion	diffusion	NOUN
cana-6247	37	7	,	,	PUNCT
cana-6247	37	8	latent	latent	NOUN
cana-6247	37	9	paint	paint	NOUN
cana-6247	37	10	,	,	PUNCT
cana-6247	37	11	and	and	CCONJ
cana-6247	37	12	personalized	personalized	ADJ
cana-6247	37	13	face	face	NOUN
cana-6247	37	14	inpainting	inpainte	VERB
cana-6247	37	15	were	be	AUX
cana-6247	37	16	created	create	VERB
cana-6247	37	17	,	,	PUNCT
cana-6247	37	18	extending	extend	VERB
cana-6247	37	19	the	the	DET
cana-6247	37	20	field	field	NOUN
cana-6247	37	21	with	with	ADP
cana-6247	37	22	multimodal	multimodal	NOUN
cana-6247	37	23	priors	prior	NOUN
cana-6247	37	24	and	and	CCONJ
cana-6247	37	25	semantic	semantic	ADJ
cana-6247	37	26	priors	prior	NOUN
cana-6247	37	27	,	,	PUNCT
cana-6247	37	28	giving	give	VERB
cana-6247	37	29	rise	rise	NOUN
cana-6247	37	30	to	to	ADP
cana-6247	37	31	coherent	coherent	ADJ
cana-6247	37	32	object	object	NOUN
cana-6247	37	33	-	-	PUNCT
cana-6247	37	34	aware	aware	ADJ
cana-6247	37	35	editing	editing	NOUN
cana-6247	37	36	and	and	CCONJ
cana-6247	37	37	robust	robust	ADJ
cana-6247	37	38	inpainting	inpainting	NOUN
cana-6247	37	39	/	/	SYM
cana-6247	37	40	outpainting	outpainte	VERB
cana-6247	37	41	[	[	X
cana-6247	37	42	8	8	NUM
cana-6247	37	43	]	]	X
cana-6247	38	1	[	[	X
cana-6247	38	2	6	6	NUM
cana-6247	38	3	]	]	X
cana-6247	38	4	[	[	X
cana-6247	38	5	32	32	NUM
cana-6247	38	6	]	]	PUNCT
cana-6247	39	1	[	[	X
cana-6247	39	2	43	43	NUM
cana-6247	39	3	]	]	PUNCT
cana-6247	39	4	.	.	PUNCT
cana-6247	40	1	super	super	ADJ
cana-6247	40	2	-	-	NOUN
cana-6247	40	3	resolution	resolution	NOUN
cana-6247	40	4	(	(	PUNCT
cana-6247	40	5	sr	sr	PROPN
cana-6247	40	6	)	)	PUNCT
cana-6247	40	7	has	have	AUX
cana-6247	40	8	emerged	emerge	VERB
cana-6247	40	9	as	as	ADP
cana-6247	40	10	another	another	DET
cana-6247	40	11	significant	significant	ADJ
cana-6247	40	12	application	application	NOUN
cana-6247	40	13	area	area	NOUN
cana-6247	40	14	.	.	PUNCT
cana-6247	41	1	cdpmsr	cdpmsr	PROPN
cana-6247	42	1	[	[	X
cana-6247	42	2	27	27	NUM
cana-6247	42	3	]	]	PUNCT
cana-6247	42	4	and	and	CCONJ
cana-6247	42	5	acdmsr	acdmsr	X
cana-6247	43	1	[	[	X
cana-6247	43	2	26	26	NUM
cana-6247	43	3	]	]	PUNCT
cana-6247	43	4	demonstrated	demonstrate	VERB
cana-6247	43	5	conditional	conditional	ADJ
cana-6247	43	6	sr	sr	PROPN
cana-6247	43	7	pipelines	pipeline	NOUN
cana-6247	43	8	,	,	PUNCT
cana-6247	43	9	while	while	SCONJ
cana-6247	43	10	ecdp	ecdp	NOUN
cana-6247	43	11	utilized	utilize	VERB
cana-6247	43	12	the	the	DET
cana-6247	43	13	more	more	ADV
cana-6247	43	14	efficient	efficient	ADJ
cana-6247	43	15	probability	probability	NOUN
cana-6247	43	16	flow	flow	NOUN
cana-6247	43	17	sampling	sample	VERB
cana-6247	43	18	method	method	NOUN
cana-6247	43	19	[	[	X
cana-6247	43	20	47	47	NUM
cana-6247	43	21	]	]	PUNCT
cana-6247	43	22	.	.	PUNCT
cana-6247	44	1	resdiff	resdiff	PROPN
cana-6247	44	2	combined	combine	VERB
cana-6247	44	3	cnn	cnn	PROPN
cana-6247	44	4	residuals	residual	NOUN
cana-6247	44	5	of	of	ADP
cana-6247	44	6	the	the	DET
cana-6247	44	7	model	model	NOUN
cana-6247	44	8	with	with	ADP
cana-6247	44	9	diffusion	diffusion	NOUN
cana-6247	44	10	priors	prior	NOUN
cana-6247	44	11	,	,	PUNCT
cana-6247	44	12	seesr	seesr	X
cana-6247	44	13	integrated	integrate	VERB
cana-6247	44	14	semantic	semantic	ADJ
cana-6247	44	15	priors	prior	NOUN
cana-6247	44	16	for	for	ADP
cana-6247	44	17	real	real	ADJ
cana-6247	44	18	world	world	NOUN
cana-6247	44	19	sr	sr	PROPN
cana-6247	44	20	,	,	PUNCT
cana-6247	44	21	and	and	CCONJ
cana-6247	44	22	superresdiff	superresdiff	PROPN
cana-6247	44	23	gan	gan	PROPN
cana-6247	44	24	integrated	integrate	VERB
cana-6247	44	25	adversarial	adversarial	ADJ
cana-6247	44	26	objectives	objective	NOUN
cana-6247	44	27	with	with	ADP
cana-6247	44	28	the	the	DET
cana-6247	44	29	idea	idea	NOUN
cana-6247	44	30	of	of	ADP
cana-6247	44	31	diffusion	diffusion	NOUN
cana-6247	44	32	noise	noise	NOUN
cana-6247	44	33	adaptation	adaptation	NOUN
cana-6247	44	34	to	to	PART
cana-6247	44	35	attain	attain	VERB
cana-6247	44	36	higher	high	ADJ
cana-6247	44	37	frequency	frequency	NOUN
cana-6247	44	38	detail	detail	NOUN
cana-6247	44	39	preservation	preservation	NOUN
cana-6247	44	40	[	[	X
cana-6247	44	41	17	17	NUM
cana-6247	44	42	]	]	X
cana-6247	45	1	[	[	X
cana-6247	45	2	41	41	NUM
cana-6247	45	3	]	]	PUNCT
cana-6247	45	4	.	.	PUNCT
cana-6247	46	1	multi	multi	ADJ
cana-6247	46	2	-	-	ADJ
cana-6247	46	3	scale	scale	ADJ
cana-6247	46	4	methods	method	NOUN
cana-6247	46	5	like	like	ADP
cana-6247	46	6	partdiff	partdiff	NOUN
cana-6247	46	7	and	and	CCONJ
cana-6247	46	8	multi	multi	ADJ
cana-6247	46	9	-	-	ADJ
cana-6247	46	10	scale	scale	ADJ
cana-6247	46	11	adversarial	adversarial	ADJ
cana-6247	46	12	diffusion	diffusion	NOUN
cana-6247	46	13	networks	network	NOUN
cana-6247	46	14	allowed	allow	VERB
cana-6247	46	15	for	for	ADP
cana-6247	46	16	adaptive	adaptive	ADJ
cana-6247	46	17	noise	noise	NOUN
cana-6247	46	18	distributed	distribute	VERB
cana-6247	46	19	across	across	ADP
cana-6247	46	20	a	a	DET
cana-6247	46	21	variation	variation	NOUN
cana-6247	46	22	of	of	ADP
cana-6247	46	23	image	image	NOUN
cana-6247	46	24	scales	scale	NOUN
cana-6247	46	25	to	to	PART
cana-6247	46	26	ensure	ensure	VERB
cana-6247	46	27	not	not	PART
cana-6247	46	28	only	only	ADV
cana-6247	46	29	the	the	DET
cana-6247	46	30	proper	proper	ADJ
cana-6247	46	31	global	global	ADJ
cana-6247	46	32	consistency	consistency	NOUN
cana-6247	46	33	but	but	CCONJ
cana-6247	46	34	also	also	ADV
cana-6247	46	35	facilitated	facilitate	VERB
cana-6247	46	36	fine	fine	ADJ
cana-6247	46	37	detail	detail	NOUN
cana-6247	46	38	recovery	recovery	NOUN
cana-6247	46	39	[	[	X
cana-6247	46	40	53	53	NUM
cana-6247	46	41	]	]	PUNCT
cana-6247	46	42	.	.	PUNCT
cana-6247	47	1	each	each	PRON
cana-6247	47	2	of	of	ADP
cana-6247	47	3	these	these	DET
cana-6247	47	4	works	work	NOUN
cana-6247	47	5	demonstrated	demonstrate	VERB
cana-6247	47	6	that	that	SCONJ
cana-6247	47	7	learning	learn	VERB
cana-6247	47	8	noise	noise	NOUN
cana-6247	47	9	and	and	CCONJ
cana-6247	47	10	scheduling	scheduling	NOUN
cana-6247	47	11	noise	noise	NOUN
cana-6247	47	12	adaptively	adaptively	ADV
cana-6247	47	13	improves	improve	VERB
cana-6247	47	14	the	the	DET
cana-6247	47	15	perceptual	perceptual	ADJ
cana-6247	47	16	fidelity	fidelity	NOUN
cana-6247	47	17	of	of	ADP
cana-6247	47	18	the	the	DET
cana-6247	47	19	output	output	NOUN
cana-6247	47	20	images	image	NOUN
cana-6247	47	21	in	in	ADP
cana-6247	47	22	sr	sr	PROPN
cana-6247	47	23	tasks	task	NOUN
cana-6247	47	24	.	.	PUNCT
cana-6247	48	1	with	with	ADP
cana-6247	48	2	evidence	evidence	NOUN
cana-6247	48	3	of	of	ADP
cana-6247	48	4	sensor	sensor	NOUN
cana-6247	48	5	-	-	PUNCT
cana-6247	48	6	aware	aware	ADJ
cana-6247	48	7	and	and	CCONJ
cana-6247	48	8	physics	physics	NOUN
cana-6247	48	9	-	-	PUNCT
cana-6247	48	10	grounded	ground	VERB
cana-6247	48	11	environments	environment	NOUN
cana-6247	48	12	,	,	PUNCT
cana-6247	48	13	researchers	researcher	NOUN
cana-6247	48	14	proposed	propose	VERB
cana-6247	48	15	poisson	poisson	NOUN
cana-6247	48	16	and	and	CCONJ
cana-6247	48	17	signal	signal	NOUN
cana-6247	48	18	-	-	PUNCT
cana-6247	48	19	dependent	dependent	ADJ
cana-6247	48	20	diffusion	diffusion	NOUN
cana-6247	48	21	models	model	NOUN
cana-6247	48	22	for	for	ADP
cana-6247	48	23	low	low	ADJ
cana-6247	48	24	-	-	PUNCT
cana-6247	48	25	light	light	NOUN
cana-6247	48	26	imaging	imaging	NOUN
cana-6247	48	27	and	and	CCONJ
cana-6247	48	28	photon	photon	NOUN
cana-6247	48	29	-	-	PUNCT
cana-6247	48	30	limited	limit	VERB
cana-6247	48	31	imaging	imaging	NOUN
cana-6247	48	32	.	.	PUNCT
cana-6247	49	1	poisson	poisson	NOUN
cana-6247	49	2	-	-	PUNCT
cana-6247	49	3	guided	guide	VERB
cana-6247	49	4	decomposition	decomposition	NOUN
cana-6247	49	5	networks	network	NOUN
cana-6247	49	6	and	and	CCONJ
cana-6247	49	7	photon	photon	NOUN
cana-6247	49	8	-	-	PUNCT
cana-6247	49	9	counting	counting	NOUN
cana-6247	49	10	flow	flow	NOUN
cana-6247	49	11	models	model	NOUN
cana-6247	49	12	explicitly	explicitly	ADV
cana-6247	49	13	modelled	model	VERB
cana-6247	49	14	discrete	discrete	ADJ
cana-6247	49	15	photon	photon	NOUN
cana-6247	49	16	statistics	statistic	NOUN
cana-6247	49	17	,	,	PUNCT
cana-6247	49	18	enabling	enable	VERB
cana-6247	49	19	more	more	ADJ
cana-6247	49	20	natural	natural	ADJ
cana-6247	49	21	denoising	denoising	NOUN
cana-6247	49	22	in	in	ADP
cana-6247	49	23	extremely	extremely	ADV
cana-6247	49	24	dark	dark	ADJ
cana-6247	49	25	environments	environment	NOUN
cana-6247	49	26	.	.	PUNCT
cana-6247	50	1	diffretinex	diffretinex	ADV
cana-6247	50	2	[	[	X
cana-6247	50	3	46	46	NUM
cana-6247	50	4	]	]	X
cana-6247	50	5	,	,	PUNCT
cana-6247	50	6	for	for	ADP
cana-6247	50	7	example	example	NOUN
cana-6247	50	8	,	,	PUNCT
cana-6247	50	9	difflle	difflle	NOUN
cana-6247	51	1	[	[	X
cana-6247	51	2	44	44	NUM
cana-6247	51	3	]	]	PUNCT
cana-6247	51	4	,	,	PUNCT
cana-6247	51	5	pyramid	pyramid	NOUN
cana-6247	51	6	diffusion	diffusion	NOUN
cana-6247	51	7	[	[	X
cana-6247	51	8	55	55	NUM
cana-6247	51	9	]	]	PUNCT
cana-6247	51	10	,	,	PUNCT
cana-6247	51	11	rw	rw	NOUN
cana-6247	51	12	-	-	PUNCT
cana-6247	51	13	dm	dm	PROPN
cana-6247	52	1	[	[	X
cana-6247	52	2	18	18	NUM
cana-6247	52	3	]	]	PUNCT
cana-6247	52	4	,	,	PUNCT
cana-6247	52	5	etc	etc	X
cana-6247	52	6	,	,	PUNCT
cana-6247	52	7	provided	provide	VERB
cana-6247	52	8	low	low	ADJ
cana-6247	52	9	-	-	PUNCT
cana-6247	52	10	light	light	NOUN
cana-6247	52	11	image	image	NOUN
cana-6247	52	12	enhancement	enhancement	NOUN
cana-6247	52	13	through	through	ADP
cana-6247	52	14	the	the	DET
cana-6247	52	15	combination	combination	NOUN
cana-6247	52	16	of	of	ADP
cana-6247	52	17	physical	physical	ADJ
cana-6247	52	18	priors	prior	NOUN
cana-6247	52	19	and	and	CCONJ
cana-6247	52	20	diffusion	diffusion	NOUN
cana-6247	52	21	structure	structure	NOUN
cana-6247	52	22	,	,	PUNCT
cana-6247	52	23	while	while	SCONJ
cana-6247	52	24	also	also	ADV
cana-6247	52	25	respecting	respect	VERB
cana-6247	52	26	edges	edge	NOUN
cana-6247	52	27	and	and	CCONJ
cana-6247	52	28	the	the	DET
cana-6247	52	29	original	original	ADJ
cana-6247	52	30	structure	structure	NOUN
cana-6247	52	31	of	of	ADP
cana-6247	52	32	the	the	DET
cana-6247	52	33	image	image	NOUN
cana-6247	52	34	.	.	PUNCT
cana-6247	53	1	noise	noise	NOUN
cana-6247	53	2	synthesis	synthesis	NOUN
cana-6247	53	3	for	for	ADP
cana-6247	53	4	low	low	ADJ
cana-6247	53	5	-	-	PUNCT
cana-6247	53	6	light	light	NOUN
cana-6247	53	7	[	[	X
cana-6247	53	8	20	20	NUM
cana-6247	53	9	]	]	PUNCT
cana-6247	53	10	,	,	PUNCT
cana-6247	53	11	global	global	ADJ
cana-6247	53	12	structure	structure	NOUN
cana-6247	53	13	aware	aware	ADJ
cana-6247	53	14	diffusion	diffusion	NOUN
cana-6247	54	1	[	[	X
cana-6247	54	2	12	12	NUM
cana-6247	54	3	]	]	PUNCT
cana-6247	54	4	,	,	PUNCT
cana-6247	54	5	and	and	CCONJ
cana-6247	54	6	lightendiffusion	lightendiffusion	NOUN
cana-6247	54	7	(	(	PUNCT
cana-6247	54	8	among	among	ADP
cana-6247	54	9	others	other	NOUN
cana-6247	54	10	)	)	PUNCT
cana-6247	55	1	[	[	X
cana-6247	55	2	14	14	NUM
cana-6247	55	3	]	]	PUNCT
cana-6247	55	4	,	,	PUNCT
cana-6247	55	5	all	all	PRON
cana-6247	55	6	showed	show	VERB
cana-6247	55	7	how	how	SCONJ
cana-6247	55	8	signal	signal	NOUN
cana-6247	55	9	-	-	PUNCT
cana-6247	55	10	dependent	dependent	ADJ
cana-6247	55	11	noise	noise	NOUN
cana-6247	55	12	can	can	AUX
cana-6247	55	13	be	be	AUX
cana-6247	55	14	modelled	model	VERB
cana-6247	55	15	in	in	ADP
cana-6247	55	16	the	the	DET
cana-6247	55	17	diffusion	diffusion	NOUN
cana-6247	55	18	method	method	NOUN
cana-6247	55	19	adaptively	adaptively	ADV
cana-6247	55	20	;	;	PUNCT
cana-6247	55	21	preparing	prepare	VERB
cana-6247	55	22	for	for	ADP
cana-6247	55	23	actual	actual	ADJ
cana-6247	55	24	sensor	sensor	NOUN
cana-6247	55	25	signal	signal	NOUN
cana-6247	55	26	distributions	distribution	NOUN
cana-6247	55	27	produced	produce	VERB
cana-6247	55	28	more	more	ADV
cana-6247	55	29	realistic	realistic	ADJ
cana-6247	55	30	reconstructions	reconstruction	NOUN
cana-6247	55	31	than	than	SCONJ
cana-6247	55	32	gaussian	gaussian	NOUN
cana-6247	55	33	-	-	PUNCT
cana-6247	55	34	based	base	VERB
cana-6247	55	35	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	55	36	979	979	NUM
cana-6247	55	37	communications	communication	NOUN
cana-6247	55	38	on	on	ADP
cana-6247	55	39	applied	apply	VERB
cana-6247	55	40	nonlinear	nonlinear	ADJ
cana-6247	55	41	analysis	analysis	NOUN
cana-6247	55	42	issn	issn	NOUN
cana-6247	55	43	:	:	PUNCT
cana-6247	55	44	1074	1074	NUM
cana-6247	55	45	-	-	PUNCT
cana-6247	55	46	133x	133x	NUM
cana-6247	55	47	vol	vol	NOUN
cana-6247	55	48	.	.	PROPN
cana-6247	56	1	31	31	NUM
cana-6247	57	1	no	no	NOUN
cana-6247	57	2	.	.	PUNCT
cana-6247	58	1	7s	7	NOUN
cana-6247	58	2	(	(	PUNCT
cana-6247	58	3	2024	2024	NUM
cana-6247	58	4	)	)	PUNCT
cana-6247	58	5	models	model	NOUN
cana-6247	58	6	.	.	PUNCT
cana-6247	59	1	each	each	PRON
cana-6247	59	2	of	of	ADP
cana-6247	59	3	these	these	DET
cana-6247	59	4	works	work	NOUN
cana-6247	59	5	demonstrated	demonstrate	VERB
cana-6247	59	6	how	how	SCONJ
cana-6247	59	7	the	the	DET
cana-6247	59	8	presence	presence	NOUN
cana-6247	59	9	of	of	ADP
cana-6247	59	10	signal	signal	NOUN
cana-6247	59	11	-	-	PUNCT
cana-6247	59	12	dependent	dependent	ADJ
cana-6247	59	13	noise	noise	NOUN
cana-6247	59	14	found	find	VERB
cana-6247	59	15	in	in	ADP
cana-6247	59	16	many	many	ADJ
cana-6247	59	17	imaging	imaging	NOUN
cana-6247	59	18	applications	application	NOUN
cana-6247	59	19	makes	make	VERB
cana-6247	59	20	diffusion	diffusion	NOUN
cana-6247	59	21	models	model	NOUN
cana-6247	59	22	appropriate	appropriate	ADJ
cana-6247	59	23	for	for	ADP
cana-6247	59	24	cases	case	NOUN
cana-6247	59	25	like	like	ADP
cana-6247	59	26	medical	medical	ADJ
cana-6247	59	27	data	datum	NOUN
cana-6247	59	28	,	,	PUNCT
cana-6247	59	29	ct	ct	PROPN
cana-6247	59	30	,	,	PUNCT
cana-6247	59	31	and	and	CCONJ
cana-6247	59	32	photon	photon	NOUN
cana-6247	59	33	-	-	PUNCT
cana-6247	59	34	limited	limit	VERB
cana-6247	59	35	imaging	imaging	NOUN
cana-6247	59	36	.	.	PUNCT
cana-6247	60	1	together	together	ADV
cana-6247	60	2	,	,	PUNCT
cana-6247	60	3	these	these	DET
cana-6247	60	4	contributions	contribution	NOUN
cana-6247	60	5	establish	establish	VERB
cana-6247	60	6	that	that	SCONJ
cana-6247	60	7	the	the	DET
cana-6247	60	8	noise	noise	NOUN
cana-6247	60	9	injection	injection	NOUN
cana-6247	60	10	method	method	NOUN
cana-6247	60	11	has	have	VERB
cana-6247	60	12	a	a	DET
cana-6247	60	13	significant	significant	ADJ
cana-6247	60	14	impact	impact	NOUN
cana-6247	60	15	on	on	ADP
cana-6247	60	16	the	the	DET
cana-6247	60	17	overall	overall	ADJ
cana-6247	60	18	performance	performance	NOUN
cana-6247	60	19	of	of	ADP
cana-6247	60	20	diffusion	diffusion	NOUN
cana-6247	60	21	models	model	NOUN
cana-6247	60	22	in	in	ADP
cana-6247	60	23	various	various	ADJ
cana-6247	60	24	tasks	task	NOUN
cana-6247	60	25	.	.	PUNCT
cana-6247	61	1	however	however	ADV
cana-6247	61	2	,	,	PUNCT
cana-6247	61	3	existing	exist	VERB
cana-6247	61	4	surveys	survey	NOUN
cana-6247	61	5	tend	tend	VERB
cana-6247	61	6	to	to	PART
cana-6247	61	7	categorize	categorize	VERB
cana-6247	61	8	diffusion	diffusion	NOUN
cana-6247	61	9	models	model	NOUN
cana-6247	61	10	based	base	VERB
cana-6247	61	11	on	on	ADP
cana-6247	61	12	their	their	PRON
cana-6247	61	13	application	application	NOUN
cana-6247	61	14	domain	domain	NOUN
cana-6247	61	15	(	(	PUNCT
cana-6247	61	16	restoration	restoration	NOUN
cana-6247	61	17	,	,	PUNCT
cana-6247	61	18	sr	sr	PROPN
cana-6247	61	19	,	,	PUNCT
cana-6247	61	20	inpainting	inpainte	VERB
cana-6247	61	21	,	,	PUNCT
cana-6247	61	22	fusion	fusion	NOUN
cana-6247	61	23	)	)	PUNCT
cana-6247	61	24	or	or	CCONJ
cana-6247	61	25	their	their	PRON
cana-6247	61	26	architectural	architectural	ADJ
cana-6247	61	27	variant	variant	NOUN
cana-6247	61	28	(	(	PUNCT
cana-6247	61	29	ddpm	ddpm	PROPN
cana-6247	61	30	,	,	PUNCT
cana-6247	61	31	latent	latent	NOUN
cana-6247	61	32	diffusion	diffusion	NOUN
cana-6247	61	33	,	,	PUNCT
cana-6247	61	34	score	score	NOUN
cana-6247	61	35	-	-	PUNCT
cana-6247	61	36	based	base	VERB
cana-6247	61	37	)	)	PUNCT
cana-6247	61	38	.	.	PUNCT
cana-6247	62	1	in	in	ADP
cana-6247	62	2	this	this	DET
cana-6247	62	3	paper	paper	NOUN
cana-6247	62	4	,	,	PUNCT
cana-6247	62	5	we	we	PRON
cana-6247	62	6	present	present	VERB
cana-6247	62	7	a	a	DET
cana-6247	62	8	noise	noise	NOUN
cana-6247	62	9	-	-	PUNCT
cana-6247	62	10	centred	centre	VERB
cana-6247	62	11	taxonomy	taxonomy	NOUN
cana-6247	62	12	that	that	PRON
cana-6247	62	13	categorizes	categorize	VERB
cana-6247	62	14	diffusion	diffusion	NOUN
cana-6247	62	15	models	model	NOUN
cana-6247	62	16	into	into	ADP
cana-6247	62	17	:	:	PUNCT
cana-6247	62	18	•	•	NUM
cana-6247	62	19	additive	additive	ADJ
cana-6247	62	20	gaussian	gaussian	NOUN
cana-6247	62	21	noise	noise	NOUN
cana-6247	62	22	(	(	PUNCT
cana-6247	62	23	agn	agn	NOUN
cana-6247	62	24	)	)	PUNCT
cana-6247	62	25	based	base	VERB
cana-6247	62	26	models	model	NOUN
cana-6247	62	27	,	,	PUNCT
cana-6247	62	28	which	which	PRON
cana-6247	62	29	are	be	AUX
cana-6247	62	30	the	the	DET
cana-6247	62	31	most	most	ADV
cana-6247	62	32	common	common	ADJ
cana-6247	62	33	approach	approach	NOUN
cana-6247	62	34	,	,	PUNCT
cana-6247	62	35	typically	typically	ADV
cana-6247	62	36	used	use	VERB
cana-6247	62	37	for	for	ADP
cana-6247	62	38	restoration	restoration	NOUN
cana-6247	62	39	and	and	CCONJ
cana-6247	62	40	denoising	denoising	NOUN
cana-6247	62	41	;	;	PUNCT
cana-6247	62	42	•	•	NUM
cana-6247	62	43	conditional	conditional	ADJ
cana-6247	62	44	noise	noise	NOUN
cana-6247	62	45	injection	injection	NOUN
cana-6247	62	46	(	(	PUNCT
cana-6247	62	47	cni	cni	PROPN
cana-6247	62	48	)	)	PUNCT
cana-6247	62	49	based	base	VERB
cana-6247	62	50	models	model	NOUN
cana-6247	62	51	,	,	PUNCT
cana-6247	62	52	in	in	ADP
cana-6247	62	53	which	which	PRON
cana-6247	62	54	conditions	condition	NOUN
cana-6247	62	55	from	from	ADP
cana-6247	62	56	external	external	ADJ
cana-6247	62	57	sources	source	NOUN
cana-6247	62	58	such	such	ADJ
cana-6247	62	59	as	as	ADP
cana-6247	62	60	masks	mask	NOUN
cana-6247	62	61	and/or	and/or	CCONJ
cana-6247	62	62	priors	prior	NOUN
cana-6247	62	63	guide	guide	VERB
cana-6247	62	64	the	the	DET
cana-6247	62	65	noise	noise	NOUN
cana-6247	62	66	schedule	schedule	NOUN
cana-6247	62	67	;	;	PUNCT
cana-6247	62	68	•	•	NUM
cana-6247	62	69	learned	learn	VERB
cana-6247	62	70	noise	noise	NOUN
cana-6247	62	71	based	base	VERB
cana-6247	62	72	models	model	NOUN
cana-6247	62	73	,	,	PUNCT
cana-6247	62	74	which	which	PRON
cana-6247	62	75	can	can	AUX
cana-6247	62	76	modify	modify	VERB
cana-6247	62	77	the	the	DET
cana-6247	62	78	noise	noise	NOUN
cana-6247	62	79	process	process	NOUN
cana-6247	62	80	or	or	CCONJ
cana-6247	62	81	learn	learn	VERB
cana-6247	62	82	to	to	PART
cana-6247	62	83	train	train	VERB
cana-6247	62	84	it	it	PRON
cana-6247	62	85	,	,	PUNCT
cana-6247	62	86	to	to	PART
cana-6247	62	87	enable	enable	VERB
cana-6247	62	88	semantic	semantic	ADJ
cana-6247	62	89	or	or	CCONJ
cana-6247	62	90	uncertainty	uncertainty	NOUN
cana-6247	62	91	-	-	PUNCT
cana-6247	62	92	based	base	VERB
cana-6247	62	93	restoration	restoration	NOUN
cana-6247	62	94	;	;	PUNCT
cana-6247	62	95	and	and	CCONJ
cana-6247	62	96	•	•	NUM
cana-6247	62	97	poisson	poisson	NOUN
cana-6247	62	98	/	/	SYM
cana-6247	62	99	signal	signal	NOUN
cana-6247	62	100	-	-	PUNCT
cana-6247	62	101	dependent	dependent	NOUN
cana-6247	62	102	which	which	PRON
cana-6247	62	103	can	can	AUX
cana-6247	62	104	be	be	AUX
cana-6247	62	105	understood	understand	VERB
cana-6247	62	106	as	as	ADP
cana-6247	62	107	being	be	AUX
cana-6247	62	108	physically	physically	ADV
cana-6247	62	109	embedded	embed	VERB
cana-6247	62	110	in	in	ADP
cana-6247	62	111	sensor	sensor	NOUN
cana-6247	62	112	degradation	degradation	NOUN
cana-6247	62	113	processes	process	NOUN
cana-6247	62	114	such	such	ADJ
cana-6247	62	115	as	as	ADP
cana-6247	62	116	photon	photon	NOUN
cana-6247	62	117	counting	counting	NOUN
cana-6247	62	118	.	.	PUNCT
cana-6247	63	1	this	this	DET
cana-6247	63	2	classification	classification	NOUN
cana-6247	63	3	underscores	underscore	VERB
cana-6247	63	4	that	that	SCONJ
cana-6247	63	5	noise	noise	NOUN
cana-6247	63	6	represents	represent	VERB
cana-6247	63	7	the	the	DET
cana-6247	63	8	unifying	unifying	ADJ
cana-6247	63	9	design	design	NOUN
cana-6247	63	10	axis	axis	NOUN
cana-6247	63	11	that	that	PRON
cana-6247	63	12	distinguishes	distinguish	VERB
cana-6247	63	13	diffusion	diffusion	NOUN
cana-6247	63	14	models	model	NOUN
cana-6247	63	15	.	.	PUNCT
cana-6247	64	1	by	by	ADP
cana-6247	64	2	re	re	VERB
cana-6247	64	3	-	-	VERB
cana-6247	64	4	conceptualizing	conceptualize	VERB
cana-6247	64	5	diffusion	diffusion	NOUN
cana-6247	64	6	research	research	NOUN
cana-6247	64	7	around	around	ADP
cana-6247	64	8	noise	noise	NOUN
cana-6247	64	9	mechanisms	mechanism	NOUN
cana-6247	64	10	,	,	PUNCT
cana-6247	64	11	this	this	DET
cana-6247	64	12	taxonomy	taxonomy	NOUN
cana-6247	64	13	elucidates	elucidate	VERB
cana-6247	64	14	theoretical	theoretical	ADJ
cana-6247	64	15	foundations	foundation	NOUN
cana-6247	64	16	,	,	PUNCT
cana-6247	64	17	indicates	indicate	VERB
cana-6247	64	18	why	why	SCONJ
cana-6247	64	19	certain	certain	ADJ
cana-6247	64	20	models	model	NOUN
cana-6247	64	21	perform	perform	VERB
cana-6247	64	22	best	good	ADJ
cana-6247	64	23	for	for	ADP
cana-6247	64	24	specific	specific	ADJ
cana-6247	64	25	types	type	NOUN
cana-6247	64	26	of	of	ADP
cana-6247	64	27	degradations	degradation	NOUN
cana-6247	64	28	,	,	PUNCT
cana-6247	64	29	and	and	CCONJ
cana-6247	64	30	provides	provide	VERB
cana-6247	64	31	a	a	DET
cana-6247	64	32	principled	principled	ADJ
cana-6247	64	33	guide	guide	NOUN
cana-6247	64	34	for	for	ADP
cana-6247	64	35	choosing	choose	VERB
cana-6247	64	36	or	or	CCONJ
cana-6247	64	37	developing	develop	VERB
cana-6247	64	38	new	new	ADJ
cana-6247	64	39	models	model	NOUN
cana-6247	64	40	for	for	ADP
cana-6247	64	41	existing	exist	VERB
cana-6247	64	42	tasks	task	NOUN
cana-6247	64	43	in	in	ADP
cana-6247	64	44	appearance	appearance	NOUN
cana-6247	64	45	enhancement	enhancement	NOUN
cana-6247	64	46	.	.	PUNCT
cana-6247	65	1	figure	figure	NOUN
cana-6247	65	2	1	1	NUM
cana-6247	65	3	:	:	PUNCT
cana-6247	66	1	proposed	propose	VERB
cana-6247	66	2	classification	classification	NOUN
cana-6247	66	3	2	2	NUM
cana-6247	66	4	proposed	propose	VERB
cana-6247	66	5	classification	classification	NOUN
cana-6247	66	6	diffusion	diffusion	NOUN
cana-6247	66	7	models	model	NOUN
cana-6247	66	8	have	have	AUX
cana-6247	66	9	been	be	AUX
cana-6247	66	10	substantially	substantially	ADV
cana-6247	66	11	investigated	investigate	VERB
cana-6247	66	12	over	over	ADP
cana-6247	66	13	the	the	DET
cana-6247	66	14	last	last	ADJ
cana-6247	66	15	decade	decade	NOUN
cana-6247	66	16	to	to	PART
cana-6247	66	17	achieve	achieve	VERB
cana-6247	66	18	optimum	optimum	ADJ
cana-6247	66	19	results	result	NOUN
cana-6247	66	20	in	in	ADP
cana-6247	66	21	the	the	DET
cana-6247	66	22	case	case	NOUN
cana-6247	66	23	of	of	ADP
cana-6247	66	24	image	image	NOUN
cana-6247	66	25	enhancement	enhancement	NOUN
cana-6247	66	26	.	.	PUNCT
cana-6247	67	1	in	in	ADP
cana-6247	67	2	this	this	DET
cana-6247	67	3	review	review	NOUN
cana-6247	67	4	,	,	PUNCT
cana-6247	67	5	we	we	PRON
cana-6247	67	6	have	have	VERB
cana-6247	67	7	classified	classify	VERB
cana-6247	67	8	diffusion	diffusion	NOUN
cana-6247	67	9	models	model	NOUN
cana-6247	67	10	used	use	VERB
cana-6247	67	11	for	for	ADP
cana-6247	67	12	image	image	NOUN
cana-6247	67	13	enhancement	enhancement	NOUN
cana-6247	67	14	into	into	ADP
cana-6247	67	15	additional	additional	ADJ
cana-6247	67	16	gaussian	gaussian	ADJ
cana-6247	67	17	noise	noise	NOUN
cana-6247	67	18	,	,	PUNCT
cana-6247	67	19	learned	learn	VERB
cana-6247	67	20	noise	noise	NOUN
cana-6247	67	21	,	,	PUNCT
cana-6247	67	22	poisson	poisson	NOUN
cana-6247	67	23	/	/	SYM
cana-6247	67	24	signal	signal	NOUN
cana-6247	67	25	dependent	dependent	ADJ
cana-6247	67	26	noise	noise	NOUN
cana-6247	67	27	,	,	PUNCT
cana-6247	67	28	and	and	CCONJ
cana-6247	67	29	conditional	conditional	ADJ
cana-6247	67	30	noise	noise	NOUN
cana-6247	67	31	injection	injection	NOUN
cana-6247	67	32	based	base	VERB
cana-6247	67	33	on	on	ADP
cana-6247	67	34	the	the	DET
cana-6247	67	35	type	type	NOUN
cana-6247	67	36	of	of	ADP
cana-6247	67	37	noise	noise	NOUN
cana-6247	67	38	that	that	PRON
cana-6247	67	39	is	be	AUX
cana-6247	67	40	implemented	implement	VERB
cana-6247	67	41	by	by	ADP
cana-6247	67	42	the	the	DET
cana-6247	67	43	model	model	NOUN
cana-6247	67	44	on	on	ADP
cana-6247	67	45	the	the	DET
cana-6247	67	46	image	image	NOUN
cana-6247	67	47	,	,	PUNCT
cana-6247	67	48	during	during	ADP
cana-6247	67	49	the	the	DET
cana-6247	67	50	forward	forward	ADJ
cana-6247	67	51	diffusion	diffusion	NOUN
cana-6247	67	52	process	process	NOUN
cana-6247	67	53	.	.	PUNCT
cana-6247	68	1	the	the	DET
cana-6247	68	2	details	detail	NOUN
cana-6247	68	3	of	of	ADP
cana-6247	68	4	classification	classification	NOUN
cana-6247	68	5	is	be	AUX
cana-6247	68	6	depicted	depict	VERB
cana-6247	68	7	in	in	ADP
cana-6247	68	8	fig.1	fig.1	ADJ
cana-6247	68	9	1	1	NUM
cana-6247	68	10	additional	additional	ADJ
cana-6247	68	11	gaussian	gaussian	ADJ
cana-6247	68	12	noise	noise	NOUN
cana-6247	68	13	modern	modern	ADJ
cana-6247	68	14	generative	generative	NOUN
cana-6247	68	15	and	and	CCONJ
cana-6247	68	16	image	image	NOUN
cana-6247	68	17	restoration	restoration	NOUN
cana-6247	68	18	frameworks	framework	NOUN
cana-6247	68	19	rely	rely	VERB
cana-6247	68	20	on	on	ADP
cana-6247	68	21	a	a	DET
cana-6247	68	22	carefully	carefully	ADV
cana-6247	68	23	structured	structure	VERB
cana-6247	68	24	stochastic	stochastic	ADJ
cana-6247	68	25	mechanism	mechanism	NOUN
cana-6247	68	26	for	for	ADP
cana-6247	68	27	further	further	ADJ
cana-6247	68	28	gaussian	gaussian	ADJ
cana-6247	68	29	noise	noise	NOUN
cana-6247	68	30	addition	addition	NOUN
cana-6247	68	31	in	in	ADP
cana-6247	68	32	diffusion	diffusion	NOUN
cana-6247	68	33	models	model	NOUN
cana-6247	68	34	.	.	PUNCT
cana-6247	69	1	fundamentally	fundamentally	ADV
cana-6247	69	2	,	,	PUNCT
cana-6247	69	3	the	the	DET
cana-6247	69	4	forward	forward	ADJ
cana-6247	69	5	diffusion	diffusion	NOUN
cana-6247	69	6	process	process	NOUN
cana-6247	69	7	adds	add	VERB
cana-6247	69	8	gaussian	gaussian	ADJ
cana-6247	69	9	noise	noise	NOUN
cana-6247	69	10	iteratively	iteratively	ADV
cana-6247	69	11	at	at	ADP
cana-6247	69	12	each	each	DET
cana-6247	69	13	discrete	discrete	ADJ
cana-6247	69	14	time	time	NOUN
cana-6247	69	15	step	step	NOUN
cana-6247	69	16	,	,	PUNCT
cana-6247	69	17	usually	usually	ADV
cana-6247	69	18	indexed	index	VERB
cana-6247	69	19	as	as	ADP
cana-6247	69	20	t	t	NOUN
cana-6247	69	21	ranging	range	VERB
cana-6247	69	22	from	from	ADP
cana-6247	69	23	1	1	NUM
cana-6247	69	24	to	to	ADP
cana-6247	69	25	t	t	PROPN
cana-6247	69	26	,	,	PUNCT
cana-6247	69	27	in	in	ADP
cana-6247	69	28	order	order	NOUN
cana-6247	69	29	to	to	ADP
cana-6247	69	30	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	69	31	980	980	NUM
cana-6247	69	32	communications	communication	NOUN
cana-6247	69	33	on	on	ADP
cana-6247	69	34	applied	apply	VERB
cana-6247	69	35	nonlinear	nonlinear	ADJ
cana-6247	69	36	analysis	analysis	NOUN
cana-6247	69	37	issn	issn	NOUN
cana-6247	69	38	:	:	PUNCT
cana-6247	69	39	1074	1074	NUM
cana-6247	69	40	-	-	PUNCT
cana-6247	69	41	133x	133x	NUM
cana-6247	69	42	vol	vol	NOUN
cana-6247	69	43	.	.	PROPN
cana-6247	70	1	31	31	NUM
cana-6247	71	1	no	no	NOUN
cana-6247	71	2	.	.	PUNCT
cana-6247	72	1	7s	7	NOUN
cana-6247	72	2	(	(	PUNCT
cana-6247	72	3	2024	2024	NUM
cana-6247	72	4	)	)	PUNCT
cana-6247	72	5	convert	convert	VERB
cana-6247	72	6	a	a	DET
cana-6247	72	7	clean	clean	ADJ
cana-6247	72	8	image	image	NOUN
cana-6247	72	9	into	into	ADP
cana-6247	72	10	a	a	DET
cana-6247	72	11	series	series	NOUN
cana-6247	72	12	of	of	ADP
cana-6247	72	13	increasingly	increasingly	ADV
cana-6247	72	14	noisier	noisy	ADJ
cana-6247	72	15	versions	version	NOUN
cana-6247	72	16	.	.	PUNCT
cana-6247	73	1	this	this	PRON
cana-6247	73	2	is	be	AUX
cana-6247	73	3	controlled	control	VERB
cana-6247	73	4	by	by	ADP
cana-6247	73	5	a	a	DET
cana-6247	73	6	variance	variance	NOUN
cana-6247	73	7	schedule	schedule	NOUN
cana-6247	73	8	,	,	PUNCT
cana-6247	73	9	represented	represent	VERB
cana-6247	73	10	by	by	ADP
cana-6247	73	11	{	{	PUNCT
cana-6247	73	12	βt}tt=1	βt}tt=1	ADV
cana-6247	73	13	,	,	PUNCT
cana-6247	73	14	which	which	PRON
cana-6247	73	15	accurately	accurately	ADV
cana-6247	73	16	regulates	regulate	VERB
cana-6247	73	17	the	the	DET
cana-6247	73	18	amount	amount	NOUN
cana-6247	73	19	of	of	ADP
cana-6247	73	20	noise	noise	NOUN
cana-6247	73	21	added	add	VERB
cana-6247	73	22	at	at	ADP
cana-6247	73	23	each	each	DET
cana-6247	73	24	stage	stage	NOUN
cana-6247	73	25	.	.	PUNCT
cana-6247	74	1	this	this	DET
cana-6247	74	2	process	process	NOUN
cana-6247	74	3	can	can	AUX
cana-6247	74	4	be	be	AUX
cana-6247	74	5	expressed	express	VERB
cana-6247	74	6	mathematically	mathematically	ADV
cana-6247	74	7	as	as	ADP
cana-6247	74	8	xt	xt	NOUN
cana-6247	74	9	=	=	SYM
cana-6247	74	10	√	√	PROPN
cana-6247	74	11	1	1	NUM
cana-6247	74	12	−	−	NOUN
cana-6247	74	13	βt	βt	PUNCT
cana-6247	74	14	xt−1	xt−1	PROPN
cana-6247	75	1	+	+	CCONJ
cana-6247	75	2	√	√	ADJ
cana-6247	75	3	βt	βt	VERB
cana-6247	75	4	ϵt	ϵt	ADP
cana-6247	75	5	where	where	SCONJ
cana-6247	75	6	x0	x0	PROPN
cana-6247	75	7	is	be	AUX
cana-6247	75	8	the	the	DET
cana-6247	75	9	original	original	ADJ
cana-6247	75	10	image	image	NOUN
cana-6247	75	11	,	,	PUNCT
cana-6247	75	12	ϵt	ϵt	NOUN
cana-6247	75	13	is	be	AUX
cana-6247	75	14	a	a	DET
cana-6247	75	15	separately	separately	ADV
cana-6247	75	16	sampled	sample	VERB
cana-6247	75	17	value	value	NOUN
cana-6247	75	18	from	from	ADP
cana-6247	75	19	a	a	DET
cana-6247	75	20	standard	standard	ADJ
cana-6247	75	21	multivariate	multivariate	NOUN
cana-6247	75	22	normal	normal	ADJ
cana-6247	75	23	distribution	distribution	NOUN
cana-6247	75	24	n	n	CCONJ
cana-6247	75	25	(	(	PUNCT
cana-6247	75	26	0	0	NUM
cana-6247	75	27	,	,	PUNCT
cana-6247	75	28	i	i	NOUN
cana-6247	75	29	)	)	PUNCT
cana-6247	75	30	and	and	CCONJ
cana-6247	75	31	βt	βt	VERB
cana-6247	75	32	is	be	AUX
cana-6247	75	33	a	a	DET
cana-6247	75	34	small	small	ADJ
cana-6247	75	35	,	,	PUNCT
cana-6247	75	36	positive	positive	ADJ
cana-6247	75	37	scalar	scalar	NOUN
cana-6247	75	38	that	that	PRON
cana-6247	75	39	increases	increase	VERB
cana-6247	75	40	over	over	ADP
cana-6247	75	41	time	time	NOUN
cana-6247	75	42	.	.	PUNCT
cana-6247	76	1	gaussian	gaussian	ADJ
cana-6247	76	2	noise	noise	NOUN
cana-6247	76	3	is	be	AUX
cana-6247	76	4	used	use	VERB
cana-6247	76	5	owing	owe	VERB
cana-6247	76	6	to	to	ADP
cana-6247	76	7	its	its	PRON
cana-6247	76	8	statistical	statistical	ADJ
cana-6247	76	9	features	feature	NOUN
cana-6247	76	10	,	,	PUNCT
cana-6247	76	11	which	which	PRON
cana-6247	76	12	include	include	VERB
cana-6247	76	13	the	the	DET
cana-6247	76	14	characteristics	characteristic	NOUN
cana-6247	76	15	of	of	ADP
cana-6247	76	16	being	be	AUX
cana-6247	76	17	memoryless	memoryless	NOUN
cana-6247	76	18	,	,	PUNCT
cana-6247	76	19	additive	additive	NOUN
cana-6247	76	20	,	,	PUNCT
cana-6247	76	21	and	and	CCONJ
cana-6247	76	22	,	,	PUNCT
cana-6247	76	23	after	after	ADP
cana-6247	76	24	sufficient	sufficient	ADJ
cana-6247	76	25	iterations	iteration	NOUN
cana-6247	76	26	,	,	PUNCT
cana-6247	76	27	generating	generate	VERB
cana-6247	76	28	a	a	DET
cana-6247	76	29	final	final	ADJ
cana-6247	76	30	state	state	NOUN
cana-6247	76	31	xt	xt	NOUN
cana-6247	76	32	,	,	PUNCT
cana-6247	76	33	that	that	PRON
cana-6247	76	34	is	be	AUX
cana-6247	76	35	nearly	nearly	ADV
cana-6247	76	36	identical	identical	ADJ
cana-6247	76	37	to	to	ADP
cana-6247	76	38	pure	pure	ADJ
cana-6247	76	39	white	white	ADJ
cana-6247	76	40	noise	noise	NOUN
cana-6247	76	41	,	,	PUNCT
cana-6247	76	42	i.e.	i.e.	X
cana-6247	76	43	xt	xt	ADP
cana-6247	76	44	n	n	PROPN
cana-6247	76	45	(	(	PUNCT
cana-6247	76	46	0	0	NUM
cana-6247	76	47	,	,	PUNCT
cana-6247	76	48	i	i	PROPN
cana-6247	76	49	)	)	PUNCT
cana-6247	76	50	.	.	PUNCT
cana-6247	77	1	this	this	DET
cana-6247	77	2	property	property	NOUN
cana-6247	77	3	is	be	AUX
cana-6247	77	4	vital	vital	ADJ
cana-6247	77	5	because	because	SCONJ
cana-6247	77	6	it	it	PRON
cana-6247	77	7	makes	make	VERB
cana-6247	77	8	the	the	DET
cana-6247	77	9	reverse	reverse	ADJ
cana-6247	77	10	process	process	NOUN
cana-6247	77	11	mathematically	mathematically	ADV
cana-6247	77	12	tractable	tractable	ADJ
cana-6247	77	13	,	,	PUNCT
cana-6247	77	14	as	as	ADV
cana-6247	77	15	well	well	ADV
cana-6247	77	16	as	as	ADP
cana-6247	77	17	making	make	VERB
cana-6247	77	18	it	it	PRON
cana-6247	77	19	feasible	feasible	ADJ
cana-6247	77	20	to	to	PART
cana-6247	77	21	learn	learn	VERB
cana-6247	77	22	the	the	DET
cana-6247	77	23	denoising	denoising	NOUN
cana-6247	77	24	trajectory	trajectory	NOUN
cana-6247	77	25	using	use	VERB
cana-6247	77	26	powerful	powerful	ADJ
cana-6247	77	27	neural	neural	ADJ
cana-6247	77	28	approximators	approximator	NOUN
cana-6247	77	29	[	[	X
cana-6247	77	30	11	11	NUM
cana-6247	77	31	]	]	PUNCT
cana-6247	77	32	.	.	PUNCT
cana-6247	78	1	in	in	ADP
cana-6247	78	2	the	the	DET
cana-6247	78	3	opposite	opposite	ADJ
cana-6247	78	4	procedure	procedure	NOUN
cana-6247	78	5	,	,	PUNCT
cana-6247	78	6	a	a	DET
cana-6247	78	7	neural	neural	ADJ
cana-6247	78	8	network	network	NOUN
cana-6247	78	9	—	—	PUNCT
cana-6247	78	10	typically	typically	ADV
cana-6247	78	11	a	a	DET
cana-6247	78	12	u	u	NOUN
cana-6247	78	13	-	-	NOUN
cana-6247	78	14	net	net	ADJ
cana-6247	78	15	or	or	CCONJ
cana-6247	78	16	similar	similar	ADJ
cana-6247	78	17	architecture	architecture	NOUN
cana-6247	78	18	—	—	PUNCT
cana-6247	78	19	is	be	AUX
cana-6247	78	20	trained	train	VERB
cana-6247	78	21	to	to	PART
cana-6247	78	22	determine	determine	VERB
cana-6247	78	23	the	the	DET
cana-6247	78	24	clean	clean	ADJ
cana-6247	78	25	image	image	NOUN
cana-6247	78	26	x0	x0	PROPN
cana-6247	78	27	straight	straight	ADV
cana-6247	78	28	from	from	ADP
cana-6247	78	29	the	the	DET
cana-6247	78	30	noisy	noisy	ADJ
cana-6247	78	31	observation	observation	NOUN
cana-6247	78	32	x	x	NOUN
cana-6247	78	33	,	,	PUNCT
cana-6247	78	34	or	or	CCONJ
cana-6247	78	35	the	the	DET
cana-6247	78	36	noise	noise	NOUN
cana-6247	78	37	component	component	NOUN
cana-6247	78	38	ϵt	ϵt	AUX
cana-6247	78	39	added	add	VERB
cana-6247	78	40	at	at	ADP
cana-6247	78	41	each	each	DET
cana-6247	78	42	step	step	NOUN
cana-6247	78	43	.	.	PUNCT
cana-6247	79	1	high	high	ADJ
cana-6247	79	2	-	-	PUNCT
cana-6247	79	3	quality	quality	NOUN
cana-6247	79	4	reconstructions	reconstruction	NOUN
cana-6247	79	5	and	and	CCONJ
cana-6247	79	6	generative	generative	ADJ
cana-6247	79	7	samples	sample	NOUN
cana-6247	79	8	are	be	AUX
cana-6247	79	9	empirically	empirically	ADV
cana-6247	79	10	generated	generate	VERB
cana-6247	79	11	when	when	SCONJ
cana-6247	79	12	the	the	DET
cana-6247	79	13	training	training	NOUN
cana-6247	79	14	objective	objective	NOUN
cana-6247	79	15	intends	intend	VERB
cana-6247	79	16	to	to	PART
cana-6247	79	17	minimize	minimize	VERB
cana-6247	79	18	the	the	DET
cana-6247	79	19	mean	mean	ADJ
cana-6247	79	20	squared	square	VERB
cana-6247	79	21	error	error	NOUN
cana-6247	79	22	between	between	ADP
cana-6247	79	23	the	the	DET
cana-6247	79	24	predicted	predict	VERB
cana-6247	79	25	and	and	CCONJ
cana-6247	79	26	actual	actual	ADJ
cana-6247	79	27	noise	noise	NOUN
cana-6247	79	28	.	.	PUNCT
cana-6247	80	1	the	the	DET
cana-6247	80	2	broad	broad	ADJ
cana-6247	80	3	application	application	NOUN
cana-6247	80	4	of	of	ADP
cana-6247	80	5	this	this	DET
cana-6247	80	6	forward	forward	ADV
cana-6247	80	7	-	-	PUNCT
cana-6247	80	8	noising	noise	VERB
cana-6247	80	9	and	and	CCONJ
cana-6247	80	10	reverse	reverse	NOUN
cana-6247	80	11	-	-	PUNCT
cana-6247	80	12	denoising	denoise	VERB
cana-6247	80	13	markov	markov	NOUN
cana-6247	80	14	chain	chain	NOUN
cana-6247	80	15	structure	structure	NOUN
cana-6247	80	16	in	in	ADP
cana-6247	80	17	restoration	restoration	NOUN
cana-6247	80	18	of	of	ADP
cana-6247	80	19	pictures	picture	NOUN
cana-6247	80	20	,	,	PUNCT
cana-6247	80	21	fusion	fusion	NOUN
cana-6247	80	22	,	,	PUNCT
cana-6247	80	23	artifact	artifact	ADJ
cana-6247	80	24	correction	correction	NOUN
cana-6247	80	25	,	,	PUNCT
cana-6247	80	26	and	and	CCONJ
cana-6247	80	27	enhancement	enhancement	NOUN
cana-6247	80	28	tasks	task	NOUN
cana-6247	80	29	displays	display	VERB
cana-6247	80	30	both	both	PRON
cana-6247	80	31	its	its	PRON
cana-6247	80	32	theoretical	theoretical	ADJ
cana-6247	80	33	sophistication	sophistication	NOUN
cana-6247	80	34	and	and	CCONJ
cana-6247	80	35	empirical	empirical	ADJ
cana-6247	80	36	reliability	reliability	NOUN
cana-6247	80	37	.	.	PUNCT
cana-6247	81	1	[	[	X
cana-6247	81	2	42	42	NUM
cana-6247	81	3	]	]	PUNCT
cana-6247	81	4	for	for	ADP
cana-6247	81	5	instance	instance	NOUN
cana-6247	81	6	,	,	PUNCT
cana-6247	81	7	by	by	ADP
cana-6247	81	8	using	use	VERB
cana-6247	81	9	the	the	DET
cana-6247	81	10	noise	noise	NOUN
cana-6247	81	11	process	process	NOUN
cana-6247	81	12	to	to	PART
cana-6247	81	13	blend	blend	VERB
cana-6247	81	14	and	and	CCONJ
cana-6247	81	15	then	then	ADV
cana-6247	81	16	denoise	denoise	VERB
cana-6247	81	17	the	the	DET
cana-6247	81	18	fused	fuse	VERB
cana-6247	81	19	representation	representation	NOUN
cana-6247	81	20	,	,	PUNCT
cana-6247	81	21	the	the	DET
cana-6247	81	22	diffusion	diffusion	NOUN
cana-6247	81	23	process	process	NOUN
cana-6247	81	24	in	in	ADP
cana-6247	81	25	image	image	NOUN
cana-6247	81	26	fusion	fusion	NOUN
cana-6247	81	27	models	model	NOUN
cana-6247	81	28	such	such	ADJ
cana-6247	81	29	as	as	ADP
cana-6247	81	30	ddfm	ddfm	NOUN
cana-6247	81	31	enables	enable	VERB
cana-6247	81	32	the	the	DET
cana-6247	81	33	smooth	smooth	ADJ
cana-6247	81	34	integration	integration	NOUN
cana-6247	81	35	of	of	ADP
cana-6247	81	36	multi	multi	ADJ
cana-6247	81	37	-	-	ADJ
cana-6247	81	38	modal	modal	ADJ
cana-6247	81	39	data	datum	NOUN
cana-6247	81	40	(	(	PUNCT
cana-6247	81	41	such	such	ADJ
cana-6247	81	42	as	as	ADP
cana-6247	81	43	visible	visible	ADJ
cana-6247	81	44	and	and	CCONJ
cana-6247	81	45	infrared	infrared	ADJ
cana-6247	81	46	images	image	NOUN
cana-6247	81	47	)	)	PUNCT
cana-6247	82	1	[	[	X
cana-6247	82	2	54	54	NUM
cana-6247	82	3	]	]	PUNCT
cana-6247	82	4	.	.	PUNCT
cana-6247	83	1	similarly	similarly	ADV
cana-6247	83	2	,	,	PUNCT
cana-6247	83	3	the	the	DET
cana-6247	83	4	model	model	NOUN
cana-6247	83	5	learns	learn	VERB
cana-6247	83	6	to	to	PART
cana-6247	83	7	reverse	reverse	VERB
cana-6247	83	8	complex	complex	ADJ
cana-6247	83	9	degradations	degradation	NOUN
cana-6247	83	10	by	by	ADP
cana-6247	83	11	simulating	simulate	VERB
cana-6247	83	12	them	they	PRON
cana-6247	83	13	as	as	ADP
cana-6247	83	14	a	a	DET
cana-6247	83	15	series	series	NOUN
cana-6247	83	16	of	of	ADP
cana-6247	83	17	gaussian	gaussian	ADJ
cana-6247	83	18	noise	noise	NOUN
cana-6247	83	19	additions	addition	NOUN
cana-6247	83	20	in	in	ADP
cana-6247	83	21	restoration	restoration	NOUN
cana-6247	83	22	tasks	task	NOUN
cana-6247	83	23	like	like	ADP
cana-6247	83	24	contrast	contrast	NOUN
cana-6247	83	25	enhancement	enhancement	NOUN
cana-6247	83	26	or	or	CCONJ
cana-6247	83	27	jpeg	jpeg	NOUN
cana-6247	83	28	artifact	artifact	ADJ
cana-6247	83	29	correction	correction	NOUN
cana-6247	83	30	[	[	X
cana-6247	83	31	15	15	NUM
cana-6247	83	32	]	]	PUNCT
cana-6247	83	33	.	.	PUNCT
cana-6247	84	1	this	this	DET
cana-6247	84	2	process	process	NOUN
cana-6247	84	3	is	be	AUX
cana-6247	84	4	further	far	ADV
cana-6247	84	5	refined	refine	VERB
cana-6247	84	6	by	by	ADP
cana-6247	84	7	sophisticated	sophisticated	ADJ
cana-6247	84	8	diffusion	diffusion	NOUN
cana-6247	84	9	model	model	NOUN
cana-6247	84	10	variations	variation	NOUN
cana-6247	84	11	.	.	PUNCT
cana-6247	85	1	while	while	SCONJ
cana-6247	85	2	some	some	PRON
cana-6247	85	3	use	use	VERB
cana-6247	85	4	mixtures	mixture	NOUN
cana-6247	85	5	of	of	ADP
cana-6247	85	6	gaussians	gaussian	NOUN
cana-6247	85	7	or	or	CCONJ
cana-6247	85	8	even	even	ADV
cana-6247	85	9	hybridize	hybridize	VERB
cana-6247	85	10	with	with	ADP
cana-6247	85	11	reaction	reaction	NOUN
cana-6247	85	12	-	-	PUNCT
cana-6247	85	13	diffusion	diffusion	NOUN
cana-6247	85	14	or	or	CCONJ
cana-6247	85	15	anisotropic	anisotropic	NOUN
cana-6247	85	16	diffusion	diffusion	NOUN
cana-6247	85	17	models	model	NOUN
cana-6247	85	18	to	to	PART
cana-6247	85	19	capture	capture	VERB
cana-6247	85	20	more	more	ADV
cana-6247	85	21	nuanced	nuanced	ADJ
cana-6247	85	22	image	image	NOUN
cana-6247	85	23	statistics	statistic	NOUN
cana-6247	86	1	[	[	X
cana-6247	86	2	2	2	X
cana-6247	86	3	]	]	X
cana-6247	86	4	[	[	X
cana-6247	86	5	7	7	NUM
cana-6247	86	6	]	]	PUNCT
cana-6247	86	7	,	,	PUNCT
cana-6247	86	8	others	other	NOUN
cana-6247	86	9	introduce	introduce	VERB
cana-6247	86	10	non	non	ADJ
cana-6247	86	11	-	-	ADJ
cana-6247	86	12	isotropic	isotropic	ADJ
cana-6247	86	13	or	or	CCONJ
cana-6247	86	14	correlated	correlate	VERB
cana-6247	86	15	noise	noise	NOUN
cana-6247	86	16	(	(	PUNCT
cana-6247	86	17	such	such	ADJ
cana-6247	86	18	as	as	ADP
cana-6247	86	19	blue	blue	ADJ
cana-6247	86	20	noise	noise	NOUN
cana-6247	86	21	)	)	PUNCT
cana-6247	86	22	to	to	PART
cana-6247	86	23	better	well	ADV
cana-6247	86	24	match	match	VERB
cana-6247	86	25	the	the	DET
cana-6247	86	26	frequency	frequency	NOUN
cana-6247	86	27	characteristics	characteristic	NOUN
cana-6247	86	28	of	of	ADP
cana-6247	86	29	natural	natural	ADJ
cana-6247	86	30	images	image	NOUN
cana-6247	86	31	.	.	PUNCT
cana-6247	87	1	to	to	PART
cana-6247	87	2	significantly	significantly	ADV
cana-6247	87	3	speed	speed	VERB
cana-6247	87	4	up	up	ADP
cana-6247	87	5	inference	inference	NOUN
cana-6247	87	6	without	without	ADP
cana-6247	87	7	affecting	affect	VERB
cana-6247	87	8	restoration	restoration	NOUN
cana-6247	87	9	quality	quality	NOUN
cana-6247	87	10	,	,	PUNCT
cana-6247	87	11	early	early	ADJ
cana-6247	87	12	stopping	stopping	NOUN
cana-6247	87	13	strategies	strategy	NOUN
cana-6247	87	14	have	have	AUX
cana-6247	87	15	also	also	ADV
cana-6247	87	16	been	be	AUX
cana-6247	87	17	proposed	propose	VERB
cana-6247	87	18	[	[	X
cana-6247	87	19	23	23	NUM
cana-6247	87	20	]	]	PUNCT
cana-6247	87	21	,	,	PUNCT
cana-6247	87	22	in	in	ADP
cana-6247	87	23	which	which	PRON
cana-6247	87	24	the	the	DET
cana-6247	87	25	forward	forward	ADJ
cana-6247	87	26	process	process	NOUN
cana-6247	87	27	is	be	AUX
cana-6247	87	28	truncated	truncate	VERB
cana-6247	87	29	before	before	ADP
cana-6247	87	30	reaching	reach	VERB
cana-6247	87	31	full	full	ADJ
cana-6247	87	32	noise	noise	NOUN
cana-6247	87	33	and	and	CCONJ
cana-6247	87	34	the	the	DET
cana-6247	87	35	model	model	NOUN
cana-6247	87	36	is	be	AUX
cana-6247	87	37	trained	train	VERB
cana-6247	87	38	to	to	PART
cana-6247	87	39	denoise	denoise	VERB
cana-6247	87	40	from	from	ADP
cana-6247	87	41	these	these	DET
cana-6247	87	42	intermediary	intermediary	ADJ
cana-6247	87	43	states	state	NOUN
cana-6247	87	44	.	.	PUNCT
cana-6247	88	1	another	another	DET
cana-6247	88	2	extension	extension	NOUN
cana-6247	88	3	is	be	AUX
cana-6247	88	4	score	score	NOUN
cana-6247	88	5	-	-	PUNCT
cana-6247	88	6	based	base	VERB
cana-6247	88	7	diffusion	diffusion	NOUN
cana-6247	88	8	models	model	NOUN
cana-6247	88	9	[	[	X
cana-6247	88	10	25	25	NUM
cana-6247	88	11	]	]	PUNCT
cana-6247	88	12	,	,	PUNCT
cana-6247	88	13	in	in	ADP
cana-6247	88	14	which	which	PRON
cana-6247	88	15	the	the	DET
cana-6247	88	16	neural	neural	ADJ
cana-6247	88	17	network	network	NOUN
cana-6247	88	18	uses	use	VERB
cana-6247	88	19	stochastic	stochastic	ADJ
cana-6247	88	20	differential	differential	ADJ
cana-6247	88	21	equations	equation	NOUN
cana-6247	88	22	to	to	PART
cana-6247	88	23	guide	guide	VERB
cana-6247	88	24	the	the	DET
cana-6247	88	25	denoising	denoising	NOUN
cana-6247	88	26	process	process	NOUN
cana-6247	88	27	after	after	ADP
cana-6247	88	28	learning	learn	VERB
cana-6247	88	29	the	the	DET
cana-6247	88	30	gradient	gradient	NOUN
cana-6247	88	31	(	(	PUNCT
cana-6247	88	32	score	score	NOUN
cana-6247	88	33	)	)	PUNCT
cana-6247	88	34	of	of	ADP
cana-6247	88	35	the	the	DET
cana-6247	88	36	data	datum	NOUN
cana-6247	88	37	distribution	distribution	NOUN
cana-6247	88	38	.	.	PUNCT
cana-6247	89	1	1.1	1.1	NUM
cana-6247	89	2	image	image	NOUN
cana-6247	89	3	restoration	restoration	NOUN
cana-6247	89	4	/	/	SYM
cana-6247	89	5	fusion	fusion	NOUN
cana-6247	89	6	image	image	NOUN
cana-6247	89	7	restoration	restoration	NOUN
cana-6247	89	8	is	be	AUX
cana-6247	89	9	a	a	DET
cana-6247	89	10	key	key	ADJ
cana-6247	89	11	reference	reference	NOUN
cana-6247	89	12	topic	topic	NOUN
cana-6247	89	13	in	in	ADP
cana-6247	89	14	both	both	PRON
cana-6247	89	15	computer	computer	NOUN
cana-6247	89	16	vision	vision	NOUN
cana-6247	89	17	and	and	CCONJ
cana-6247	89	18	image	image	NOUN
cana-6247	89	19	processing	processing	NOUN
cana-6247	89	20	.	.	PUNCT
cana-6247	90	1	image	image	NOUN
cana-6247	90	2	restoration	restoration	NOUN
cana-6247	90	3	aims	aim	VERB
cana-6247	90	4	to	to	PART
cana-6247	90	5	obtain	obtain	VERB
cana-6247	90	6	a	a	DET
cana-6247	90	7	high	high	ADJ
cana-6247	90	8	-	-	PUNCT
cana-6247	90	9	quality	quality	NOUN
cana-6247	90	10	image	image	NOUN
cana-6247	90	11	from	from	ADP
cana-6247	90	12	a	a	DET
cana-6247	90	13	low	low	ADJ
cana-6247	90	14	-	-	PUNCT
cana-6247	90	15	quality	quality	NOUN
cana-6247	90	16	observation	observation	NOUN
cana-6247	90	17	or	or	CCONJ
cana-6247	90	18	reconstruction	reconstruction	NOUN
cana-6247	90	19	.	.	PUNCT
cana-6247	91	1	degradations	degradation	NOUN
cana-6247	91	2	can	can	AUX
cana-6247	91	3	result	result	VERB
cana-6247	91	4	from	from	ADP
cana-6247	91	5	noise	noise	NOUN
cana-6247	91	6	,	,	PUNCT
cana-6247	91	7	blurring	blurring	NOUN
cana-6247	91	8	,	,	PUNCT
cana-6247	91	9	compression	compression	NOUN
cana-6247	91	10	artifacts	artifact	NOUN
cana-6247	91	11	,	,	PUNCT
cana-6247	91	12	missing	miss	VERB
cana-6247	91	13	data	datum	NOUN
cana-6247	91	14	,	,	PUNCT
cana-6247	91	15	or	or	CCONJ
cana-6247	91	16	other	other	ADJ
cana-6247	91	17	forms	form	NOUN
cana-6247	91	18	of	of	ADP
cana-6247	91	19	distortion	distortion	NOUN
cana-6247	91	20	while	while	SCONJ
cana-6247	91	21	an	an	DET
cana-6247	91	22	image	image	NOUN
cana-6247	91	23	is	be	AUX
cana-6247	91	24	obtained	obtain	VERB
cana-6247	91	25	,	,	PUNCT
cana-6247	91	26	communicated	communicate	VERB
cana-6247	91	27	,	,	PUNCT
cana-6247	91	28	and	and	CCONJ
cana-6247	91	29	stored	store	VERB
cana-6247	91	30	.	.	PUNCT
cana-6247	92	1	the	the	DET
cana-6247	92	2	goal	goal	NOUN
cana-6247	92	3	in	in	ADP
cana-6247	92	4	restoration	restoration	NOUN
cana-6247	92	5	is	be	AUX
cana-6247	92	6	to	to	PART
cana-6247	92	7	create	create	VERB
cana-6247	92	8	an	an	DET
cana-6247	92	9	image	image	NOUN
cana-6247	92	10	that	that	PRON
cana-6247	92	11	appears	appear	VERB
cana-6247	92	12	and/or	and/or	CCONJ
cana-6247	92	13	measures	measure	NOUN
cana-6247	92	14	as	as	ADV
cana-6247	92	15	close	close	ADJ
cana-6247	92	16	to	to	ADP
cana-6247	92	17	that	that	PRON
cana-6247	92	18	of	of	ADP
cana-6247	92	19	the	the	DET
cana-6247	92	20	original	original	ADJ
cana-6247	92	21	or	or	CCONJ
cana-6247	92	22	unaltered	unaltered	ADJ
cana-6247	92	23	scene	scene	NOUN
cana-6247	92	24	as	as	ADP
cana-6247	92	25	possible	possible	ADJ
cana-6247	92	26	.	.	PUNCT
cana-6247	93	1	typical	typical	ADJ
cana-6247	93	2	applications	application	NOUN
cana-6247	93	3	of	of	ADP
cana-6247	93	4	image	image	NOUN
cana-6247	93	5	restoration	restoration	NOUN
cana-6247	93	6	include	include	VERB
cana-6247	93	7	noise	noise	NOUN
cana-6247	93	8	removal	removal	NOUN
cana-6247	93	9	,	,	PUNCT
cana-6247	93	10	blur	blur	NOUN
cana-6247	93	11	correction	correction	NOUN
cana-6247	93	12	,	,	PUNCT
cana-6247	93	13	haze	haze	NOUN
cana-6247	93	14	removal	removal	NOUN
cana-6247	93	15	,	,	PUNCT
cana-6247	93	16	artifact	artifact	ADJ
cana-6247	93	17	removal	removal	NOUN
cana-6247	93	18	,	,	PUNCT
cana-6247	93	19	super	super	NOUN
cana-6247	93	20	-	-	NOUN
cana-6247	93	21	resolution	resolution	NOUN
cana-6247	93	22	/	/	SYM
cana-6247	93	23	postrecovery	postrecovery	NOUN
cana-6247	93	24	super	super	NOUN
cana-6247	93	25	-	-	NOUN
cana-6247	93	26	resolution	resolution	NOUN
cana-6247	93	27	,	,	PUNCT
cana-6247	93	28	inpainting	inpainting	NOUN
cana-6247	93	29	,	,	PUNCT
cana-6247	93	30	and	and	CCONJ
cana-6247	93	31	other	other	ADJ
cana-6247	93	32	similar	similar	ADJ
cana-6247	93	33	enhancements	enhancement	NOUN
cana-6247	93	34	.	.	PUNCT
cana-6247	94	1	additive	additive	ADJ
cana-6247	94	2	gaussian	gaussian	ADJ
cana-6247	94	3	noise	noise	NOUN
cana-6247	94	4	(	(	PUNCT
cana-6247	94	5	agn	agn	NOUN
cana-6247	94	6	)	)	PUNCT
cana-6247	94	7	is	be	AUX
cana-6247	94	8	considered	consider	VERB
cana-6247	94	9	to	to	PART
cana-6247	94	10	be	be	AUX
cana-6247	94	11	the	the	DET
cana-6247	94	12	most	most	ADV
cana-6247	94	13	suitable	suitable	ADJ
cana-6247	94	14	noise	noise	NOUN
cana-6247	94	15	type	type	NOUN
cana-6247	94	16	for	for	ADP
cana-6247	94	17	image	image	NOUN
cana-6247	94	18	restoration	restoration	NOUN
cana-6247	94	19	using	use	VERB
cana-6247	94	20	diffusion	diffusion	NOUN
cana-6247	94	21	models	model	NOUN
cana-6247	94	22	,	,	PUNCT
cana-6247	94	23	due	due	ADP
cana-6247	94	24	to	to	ADP
cana-6247	94	25	several	several	ADJ
cana-6247	94	26	mathematical	mathematical	ADJ
cana-6247	94	27	,	,	PUNCT
cana-6247	94	28	statistical	statistical	ADJ
cana-6247	94	29	and	and	CCONJ
cana-6247	94	30	practical	practical	ADJ
cana-6247	94	31	reasons	reason	NOUN
cana-6247	94	32	.	.	PUNCT
cana-6247	95	1	it	it	PRON
cana-6247	95	2	is	be	AUX
cana-6247	95	3	isotropic	isotropic	ADJ
cana-6247	95	4	in	in	ADP
cana-6247	95	5	nature	nature	NOUN
cana-6247	95	6	,	,	PUNCT
cana-6247	95	7	which	which	PRON
cana-6247	95	8	means	mean	VERB
cana-6247	95	9	that	that	SCONJ
cana-6247	95	10	the	the	DET
cana-6247	95	11	noise	noise	NOUN
cana-6247	95	12	uniformly	uniformly	ADV
cana-6247	95	13	affects	affect	VERB
cana-6247	95	14	the	the	DET
cana-6247	95	15	data	data	NOUN
cana-6247	95	16	space	space	NOUN
cana-6247	95	17	,	,	PUNCT
cana-6247	95	18	making	make	VERB
cana-6247	95	19	sure	sure	ADJ
cana-6247	95	20	that	that	SCONJ
cana-6247	95	21	the	the	DET
cana-6247	95	22	noise	noise	NOUN
cana-6247	95	23	addition	addition	NOUN
cana-6247	95	24	process	process	NOUN
cana-6247	95	25	does	do	AUX
cana-6247	95	26	not	not	PART
cana-6247	95	27	introduce	introduce	VERB
cana-6247	95	28	any	any	DET
cana-6247	95	29	bias	bias	NOUN
cana-6247	95	30	or	or	CCONJ
cana-6247	95	31	artifacts	artifact	NOUN
cana-6247	95	32	in	in	ADP
cana-6247	95	33	the	the	DET
cana-6247	95	34	original	original	ADJ
cana-6247	95	35	image	image	NOUN
cana-6247	95	36	.	.	PUNCT
cana-6247	96	1	this	this	DET
cana-6247	96	2	property	property	NOUN
cana-6247	96	3	is	be	AUX
cana-6247	96	4	crucial	crucial	ADJ
cana-6247	96	5	for	for	ADP
cana-6247	96	6	maintaining	maintain	VERB
cana-6247	96	7	the	the	DET
cana-6247	96	8	structural	structural	ADJ
cana-6247	96	9	integrity	integrity	NOUN
cana-6247	96	10	of	of	ADP
cana-6247	96	11	the	the	DET
cana-6247	96	12	images	image	NOUN
cana-6247	96	13	as	as	SCONJ
cana-6247	96	14	they	they	PRON
cana-6247	96	15	are	be	AUX
cana-6247	96	16	corrupted	corrupt	VERB
cana-6247	96	17	stepwise	stepwise	ADV
cana-6247	96	18	through	through	ADP
cana-6247	96	19	a	a	DET
cana-6247	96	20	markov	markov	NOUN
cana-6247	96	21	chain	chain	NOUN
cana-6247	96	22	.	.	PUNCT
cana-6247	97	1	analytically	analytically	ADV
cana-6247	97	2	,	,	PUNCT
cana-6247	97	3	gaussian	gaussian	ADJ
cana-6247	97	4	noise	noise	NOUN
cana-6247	97	5	is	be	AUX
cana-6247	97	6	tractable	tractable	ADJ
cana-6247	97	7	.	.	PUNCT
cana-6247	98	1	stability	stability	NOUN
cana-6247	98	2	under	under	ADP
cana-6247	98	3	addition	addition	NOUN
cana-6247	98	4	and	and	CCONJ
cana-6247	98	5	the	the	DET
cana-6247	98	6	central	central	ADJ
cana-6247	98	7	limit	limit	NOUN
cana-6247	98	8	theorem	theorem	VERB
cana-6247	98	9	make	make	VERB
cana-6247	98	10	agn	agn	PROPN
cana-6247	98	11	a	a	DET
cana-6247	98	12	go	go	VERB
cana-6247	98	13	-	-	PUNCT
cana-6247	98	14	to	to	NOUN
cana-6247	98	15	choice	choice	NOUN
cana-6247	98	16	for	for	ADP
cana-6247	98	17	modeling	model	VERB
cana-6247	98	18	the	the	DET
cana-6247	98	19	set	set	NOUN
cana-6247	98	20	of	of	ADP
cana-6247	98	21	many	many	ADJ
cana-6247	98	22	small	small	ADJ
cana-6247	98	23	,	,	PUNCT
cana-6247	98	24	independent	independent	ADJ
cana-6247	98	25	noise	noise	NOUN
cana-6247	98	26	sources	source	NOUN
cana-6247	98	27	[	[	X
cana-6247	98	28	4	4	NUM
cana-6247	98	29	]	]	PUNCT
cana-6247	98	30	.	.	PUNCT
cana-6247	99	1	this	this	PRON
cana-6247	99	2	allows	allow	VERB
cana-6247	99	3	the	the	DET
cana-6247	99	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	99	5	981	981	NUM
cana-6247	99	6	communications	communication	NOUN
cana-6247	99	7	on	on	ADP
cana-6247	99	8	applied	apply	VERB
cana-6247	99	9	nonlinear	nonlinear	ADJ
cana-6247	99	10	analysis	analysis	NOUN
cana-6247	99	11	issn	issn	NOUN
cana-6247	99	12	:	:	PUNCT
cana-6247	99	13	1074	1074	NUM
cana-6247	99	14	-	-	PUNCT
cana-6247	99	15	133x	133x	NUM
cana-6247	99	16	vol	vol	NOUN
cana-6247	99	17	.	.	PROPN
cana-6247	99	18	31	31	NUM
cana-6247	100	1	no	no	NOUN
cana-6247	100	2	.	.	PUNCT
cana-6247	101	1	7s	7	NOUN
cana-6247	101	2	(	(	PUNCT
cana-6247	101	3	2024	2024	NUM
cana-6247	101	4	)	)	PUNCT
cana-6247	101	5	figure	figure	NOUN
cana-6247	101	6	2	2	NUM
cana-6247	101	7	:	:	PUNCT
cana-6247	101	8	forward	forward	ADJ
cana-6247	101	9	diffusion	diffusion	NOUN
cana-6247	101	10	process	process	NOUN
cana-6247	101	11	in	in	ADP
cana-6247	101	12	additive	additive	ADJ
cana-6247	101	13	gaussian	gaussian	NOUN
cana-6247	101	14	noise	noise	NOUN
cana-6247	101	15	type	type	NOUN
cana-6247	101	16	diffusion	diffusion	NOUN
cana-6247	101	17	model	model	NOUN
cana-6247	101	18	forward	forward	PROPN
cana-6247	101	19	diffusion	diffusion	NOUN
cana-6247	101	20	process	process	NOUN
cana-6247	101	21	to	to	PART
cana-6247	101	22	be	be	AUX
cana-6247	101	23	carried	carry	VERB
cana-6247	101	24	out	out	ADP
cana-6247	101	25	in	in	ADP
cana-6247	101	26	the	the	DET
cana-6247	101	27	form	form	NOUN
cana-6247	101	28	of	of	ADP
cana-6247	101	29	a	a	DET
cana-6247	101	30	simple	simple	ADJ
cana-6247	101	31	markov	markov	NOUN
cana-6247	101	32	chain	chain	NOUN
cana-6247	101	33	that	that	PRON
cana-6247	101	34	finally	finally	ADV
cana-6247	101	35	results	result	VERB
cana-6247	101	36	in	in	ADP
cana-6247	101	37	a	a	DET
cana-6247	101	38	standard	standard	ADJ
cana-6247	101	39	normal	normal	ADJ
cana-6247	101	40	distribution	distribution	NOUN
cana-6247	101	41	.	.	PUNCT
cana-6247	102	1	the	the	DET
cana-6247	102	2	reverse	reverse	ADJ
cana-6247	102	3	diffusion	diffusion	NOUN
cana-6247	102	4	process	process	NOUN
cana-6247	102	5	,	,	PUNCT
cana-6247	102	6	which	which	PRON
cana-6247	102	7	is	be	AUX
cana-6247	102	8	basically	basically	ADV
cana-6247	102	9	denoising	denoise	VERB
cana-6247	102	10	the	the	DET
cana-6247	102	11	noisy	noisy	ADJ
cana-6247	102	12	image	image	NOUN
cana-6247	102	13	,	,	PUNCT
cana-6247	102	14	step	step	NOUN
cana-6247	102	15	by	by	ADP
cana-6247	102	16	step	step	NOUN
cana-6247	102	17	,	,	PUNCT
cana-6247	102	18	becomes	become	VERB
cana-6247	102	19	efficiently	efficiently	ADV
cana-6247	102	20	parametrized	parametrized	ADJ
cana-6247	102	21	and	and	CCONJ
cana-6247	102	22	optimized	optimize	VERB
cana-6247	102	23	.	.	PUNCT
cana-6247	103	1	a	a	DET
cana-6247	103	2	wide	wide	ADJ
cana-6247	103	3	variety	variety	NOUN
cana-6247	103	4	of	of	ADP
cana-6247	103	5	diffusion	diffusion	NOUN
cana-6247	103	6	models	model	NOUN
cana-6247	103	7	,	,	PUNCT
cana-6247	103	8	using	use	VERB
cana-6247	103	9	agn	agn	NOUN
cana-6247	103	10	specifically	specifically	ADV
cana-6247	103	11	,	,	PUNCT
cana-6247	103	12	have	have	AUX
cana-6247	103	13	been	be	AUX
cana-6247	103	14	developed	develop	VERB
cana-6247	103	15	for	for	ADP
cana-6247	103	16	image	image	NOUN
cana-6247	103	17	restoration	restoration	NOUN
cana-6247	103	18	.	.	PUNCT
cana-6247	104	1	some	some	PRON
cana-6247	104	2	of	of	ADP
cana-6247	104	3	the	the	DET
cana-6247	104	4	most	most	ADV
cana-6247	104	5	notable	notable	ADJ
cana-6247	104	6	types	type	NOUN
cana-6247	104	7	include	include	VERB
cana-6247	104	8	:	:	PUNCT
cana-6247	104	9	•	•	ADP
cana-6247	104	10	denoising	denoising	NOUN
cana-6247	104	11	diffusion	diffusion	NOUN
cana-6247	104	12	probabilistic	probabilistic	ADJ
cana-6247	104	13	model	model	NOUN
cana-6247	104	14	(	(	PUNCT
cana-6247	104	15	ddpm	ddpm	PROPN
cana-6247	104	16	):	):	PUNCT
cana-6247	104	17	these	these	PRON
cana-6247	104	18	form	form	VERB
cana-6247	104	19	the	the	DET
cana-6247	104	20	foundational	foundational	ADJ
cana-6247	104	21	framework	framework	NOUN
cana-6247	104	22	of	of	ADP
cana-6247	104	23	diffusion	diffusion	NOUN
cana-6247	104	24	models	model	NOUN
cana-6247	104	25	used	use	VERB
cana-6247	104	26	for	for	ADP
cana-6247	104	27	image	image	NOUN
cana-6247	104	28	enhancement	enhancement	NOUN
cana-6247	104	29	.	.	PUNCT
cana-6247	105	1	for	for	ADP
cana-6247	105	2	image	image	NOUN
cana-6247	105	3	restorations	restoration	NOUN
cana-6247	105	4	,	,	PUNCT
cana-6247	105	5	ddpms	ddpm	NOUN
cana-6247	105	6	are	be	AUX
cana-6247	105	7	well	well	ADV
cana-6247	105	8	-	-	PUNCT
cana-6247	105	9	suited	suited	ADJ
cana-6247	105	10	as	as	SCONJ
cana-6247	105	11	they	they	PRON
cana-6247	105	12	can	can	AUX
cana-6247	105	13	robustly	robustly	ADV
cana-6247	105	14	model	model	VERB
cana-6247	105	15	the	the	DET
cana-6247	105	16	complex	complex	ADJ
cana-6247	105	17	distribution	distribution	NOUN
cana-6247	105	18	of	of	ADP
cana-6247	105	19	natural	natural	ADJ
cana-6247	105	20	images	image	NOUN
cana-6247	105	21	and	and	CCONJ
cana-6247	105	22	systematically	systematically	ADV
cana-6247	105	23	remove	remove	VERB
cana-6247	105	24	the	the	DET
cana-6247	105	25	noise	noise	NOUN
cana-6247	105	26	added	add	VERB
cana-6247	105	27	in	in	ADP
cana-6247	105	28	the	the	DET
cana-6247	105	29	forward	forward	ADJ
cana-6247	105	30	diffusion	diffusion	NOUN
cana-6247	105	31	process	process	NOUN
cana-6247	105	32	.	.	PUNCT
cana-6247	106	1	[	[	X
cana-6247	106	2	25	25	NUM
cana-6247	106	3	]	]	SYM
cana-6247	106	4	•	•	NUM
cana-6247	106	5	score	score	NOUN
cana-6247	106	6	-	-	PUNCT
cana-6247	106	7	based	base	VERB
cana-6247	106	8	generative	generative	ADJ
cana-6247	106	9	models	model	NOUN
cana-6247	106	10	(	(	PUNCT
cana-6247	106	11	sbgm	sbgm	ADV
cana-6247	106	12	):	):	PUNCT
cana-6247	106	13	instead	instead	ADV
cana-6247	106	14	of	of	ADP
cana-6247	106	15	directly	directly	ADV
cana-6247	106	16	modeling	model	VERB
cana-6247	106	17	the	the	DET
cana-6247	106	18	denoising	denoising	NOUN
cana-6247	106	19	process	process	NOUN
cana-6247	106	20	,	,	PUNCT
cana-6247	106	21	these	these	DET
cana-6247	106	22	models	model	NOUN
cana-6247	106	23	learn	learn	VERB
cana-6247	106	24	the	the	DET
cana-6247	106	25	gradient	gradient	NOUN
cana-6247	106	26	/	/	SYM
cana-6247	106	27	score	score	NOUN
cana-6247	106	28	of	of	ADP
cana-6247	106	29	the	the	DET
cana-6247	106	30	data	datum	NOUN
cana-6247	106	31	distribution	distribution	NOUN
cana-6247	106	32	.	.	PUNCT
cana-6247	107	1	they	they	PRON
cana-6247	107	2	guide	guide	VERB
cana-6247	107	3	the	the	DET
cana-6247	107	4	reverse	reverse	ADJ
cana-6247	107	5	diffusion	diffusion	NOUN
cana-6247	107	6	process	process	NOUN
cana-6247	107	7	into	into	ADP
cana-6247	107	8	generating	generate	VERB
cana-6247	107	9	a	a	DET
cana-6247	107	10	high	high	ADJ
cana-6247	107	11	-	-	PUNCT
cana-6247	107	12	quality	quality	NOUN
cana-6247	107	13	restoration	restoration	NOUN
cana-6247	107	14	by	by	ADP
cana-6247	107	15	estimating	estimate	VERB
cana-6247	107	16	how	how	SCONJ
cana-6247	107	17	the	the	DET
cana-6247	107	18	probability	probability	NOUN
cana-6247	107	19	density	density	NOUN
cana-6247	107	20	of	of	ADP
cana-6247	107	21	clear	clear	ADJ
cana-6247	107	22	images	image	NOUN
cana-6247	107	23	changes	change	NOUN
cana-6247	107	24	in	in	ADP
cana-6247	107	25	the	the	DET
cana-6247	107	26	presence	presence	NOUN
cana-6247	107	27	of	of	ADP
cana-6247	107	28	noise	noise	NOUN
cana-6247	107	29	.	.	PUNCT
cana-6247	108	1	this	this	DET
cana-6247	108	2	approach	approach	NOUN
cana-6247	108	3	is	be	AUX
cana-6247	108	4	extremely	extremely	ADV
cana-6247	108	5	powerful	powerful	ADJ
cana-6247	108	6	as	as	SCONJ
cana-6247	108	7	it	it	PRON
cana-6247	108	8	leverages	leverage	VERB
cana-6247	108	9	the	the	DET
cana-6247	108	10	underlying	underlie	VERB
cana-6247	108	11	geometry	geometry	NOUN
cana-6247	108	12	of	of	ADP
cana-6247	108	13	the	the	DET
cana-6247	108	14	data	datum	NOUN
cana-6247	108	15	manifold	manifold	ADJ
cana-6247	108	16	.	.	PUNCT
cana-6247	109	1	[	[	X
cana-6247	109	2	42	42	NUM
cana-6247	109	3	]	]	X
cana-6247	109	4	[	[	X
cana-6247	109	5	38	38	NUM
cana-6247	109	6	]	]	SYM
cana-6247	109	7	•	•	NOUN
cana-6247	109	8	plug	plug	VERB
cana-6247	109	9	and	and	CCONJ
cana-6247	109	10	play	play	VERB
cana-6247	109	11	model	model	NOUN
cana-6247	109	12	:	:	PUNCT
cana-6247	109	13	these	these	DET
cana-6247	109	14	models	model	NOUN
cana-6247	109	15	achieve	achieve	VERB
cana-6247	109	16	a	a	DET
cana-6247	109	17	natural	natural	ADJ
cana-6247	109	18	balance	balance	NOUN
cana-6247	109	19	of	of	ADP
cana-6247	109	20	flexibility	flexibility	NOUN
cana-6247	109	21	and	and	CCONJ
cana-6247	109	22	performance	performance	NOUN
cana-6247	109	23	by	by	ADP
cana-6247	109	24	using	use	VERB
cana-6247	109	25	traditional	traditional	ADJ
cana-6247	109	26	image	image	NOUN
cana-6247	109	27	restoration	restoration	NOUN
cana-6247	109	28	models	model	NOUN
cana-6247	109	29	or	or	CCONJ
cana-6247	109	30	priors	prior	NOUN
cana-6247	109	31	in	in	ADP
cana-6247	109	32	conjunction	conjunction	NOUN
cana-6247	109	33	with	with	ADP
cana-6247	109	34	the	the	DET
cana-6247	109	35	diffusion	diffusion	NOUN
cana-6247	109	36	sampling	sampling	NOUN
cana-6247	109	37	process	process	NOUN
cana-6247	109	38	,	,	PUNCT
cana-6247	109	39	to	to	PART
cana-6247	109	40	provide	provide	VERB
cana-6247	109	41	a	a	DET
cana-6247	109	42	more	more	ADV
cana-6247	109	43	modular	modular	ADJ
cana-6247	109	44	way	way	NOUN
cana-6247	109	45	to	to	PART
cana-6247	109	46	adapt	adapt	VERB
cana-6247	109	47	to	to	ADP
cana-6247	109	48	different	different	ADJ
cana-6247	109	49	types	type	NOUN
cana-6247	109	50	of	of	ADP
cana-6247	109	51	image	image	NOUN
cana-6247	109	52	corruption	corruption	NOUN
cana-6247	109	53	without	without	ADP
cana-6247	109	54	having	have	VERB
cana-6247	109	55	to	to	PART
cana-6247	109	56	retrain	retrain	VERB
cana-6247	109	57	the	the	DET
cana-6247	109	58	entire	entire	ADJ
cana-6247	109	59	system	system	NOUN
cana-6247	109	60	from	from	ADP
cana-6247	109	61	scratch	scratch	NOUN
cana-6247	109	62	.	.	PUNCT
cana-6247	110	1	this	this	PRON
cana-6247	110	2	is	be	AUX
cana-6247	110	3	especially	especially	ADV
cana-6247	110	4	versatile	versatile	ADJ
cana-6247	110	5	because	because	SCONJ
cana-6247	110	6	this	this	PRON
cana-6247	110	7	provides	provide	VERB
cana-6247	110	8	the	the	DET
cana-6247	110	9	flexibility	flexibility	NOUN
cana-6247	110	10	to	to	PART
cana-6247	110	11	adapt	adapt	VERB
cana-6247	110	12	in	in	ADP
cana-6247	110	13	situations	situation	NOUN
cana-6247	110	14	where	where	SCONJ
cana-6247	110	15	the	the	DET
cana-6247	110	16	form	form	NOUN
cana-6247	110	17	of	of	ADP
cana-6247	110	18	the	the	DET
cana-6247	110	19	image	image	NOUN
cana-6247	110	20	corruption	corruption	NOUN
cana-6247	110	21	is	be	AUX
cana-6247	110	22	not	not	PART
cana-6247	110	23	well	well	ADV
cana-6247	110	24	-	-	PUNCT
cana-6247	110	25	defined	define	VERB
cana-6247	110	26	.	.	PUNCT
cana-6247	111	1	therefore	therefore	ADV
cana-6247	111	2	,	,	PUNCT
cana-6247	111	3	the	the	DET
cana-6247	111	4	diffusion	diffusion	NOUN
cana-6247	111	5	models	model	NOUN
cana-6247	111	6	deliver	deliver	VERB
cana-6247	111	7	restoration	restoration	NOUN
cana-6247	111	8	and	and	CCONJ
cana-6247	111	9	manipulation	manipulation	NOUN
cana-6247	111	10	solutions	solution	NOUN
cana-6247	111	11	with	with	ADP
cana-6247	111	12	improved	improved	ADJ
cana-6247	111	13	performance	performance	NOUN
cana-6247	111	14	when	when	SCONJ
cana-6247	111	15	combined	combine	VERB
cana-6247	111	16	with	with	ADP
cana-6247	111	17	taskspecific	taskspecific	ADJ
cana-6247	111	18	tools	tool	NOUN
cana-6247	111	19	.	.	PUNCT
cana-6247	112	1	[	[	X
cana-6247	112	2	56	56	NUM
cana-6247	112	3	]	]	X
cana-6247	113	1	[	[	X
cana-6247	113	2	23	23	NUM
cana-6247	113	3	]	]	PUNCT
cana-6247	114	1	[	[	X
cana-6247	114	2	51	51	NUM
cana-6247	114	3	]	]	PUNCT
cana-6247	114	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	114	5	982	982	NUM
cana-6247	114	6	communications	communication	NOUN
cana-6247	114	7	on	on	ADP
cana-6247	114	8	applied	apply	VERB
cana-6247	114	9	nonlinear	nonlinear	ADJ
cana-6247	114	10	analysis	analysis	NOUN
cana-6247	114	11	issn	issn	NOUN
cana-6247	114	12	:	:	PUNCT
cana-6247	114	13	1074	1074	NUM
cana-6247	114	14	-	-	PUNCT
cana-6247	114	15	133x	133x	NUM
cana-6247	114	16	vol	vol	NOUN
cana-6247	114	17	.	.	PROPN
cana-6247	115	1	31	31	NUM
cana-6247	116	1	no	no	NOUN
cana-6247	116	2	.	.	PUNCT
cana-6247	117	1	7s	7	NOUN
cana-6247	117	2	(	(	PUNCT
cana-6247	117	3	2024	2024	NUM
cana-6247	117	4	)	)	PUNCT
cana-6247	117	5	•	•	NUM
cana-6247	117	6	accelerated	accelerated	ADJ
cana-6247	117	7	and	and	CCONJ
cana-6247	117	8	efficient	efficient	ADJ
cana-6247	117	9	diffusion	diffusion	NOUN
cana-6247	117	10	models	model	NOUN
cana-6247	117	11	:	:	PUNCT
cana-6247	117	12	these	these	DET
cana-6247	117	13	models	model	NOUN
cana-6247	117	14	combat	combat	VERB
cana-6247	117	15	the	the	DET
cana-6247	117	16	inherent	inherent	ADJ
cana-6247	117	17	computation	computation	NOUN
cana-6247	117	18	demands	demand	NOUN
cana-6247	117	19	of	of	ADP
cana-6247	117	20	diffusion	diffusion	NOUN
cana-6247	117	21	methods	method	NOUN
cana-6247	117	22	,	,	PUNCT
cana-6247	117	23	which	which	PRON
cana-6247	117	24	can	can	AUX
cana-6247	117	25	take	take	VERB
cana-6247	117	26	hundreds	hundred	NOUN
cana-6247	117	27	,	,	PUNCT
cana-6247	117	28	if	if	SCONJ
cana-6247	117	29	not	not	PART
cana-6247	117	30	thousands	thousand	NOUN
cana-6247	117	31	,	,	PUNCT
cana-6247	117	32	of	of	ADP
cana-6247	117	33	iterations	iteration	NOUN
cana-6247	117	34	to	to	PART
cana-6247	117	35	compute	compute	VERB
cana-6247	117	36	.	.	PUNCT
cana-6247	118	1	there	there	PRON
cana-6247	118	2	have	have	AUX
cana-6247	118	3	been	be	AUX
cana-6247	118	4	methods	method	NOUN
cana-6247	118	5	of	of	ADP
cana-6247	118	6	optimization	optimization	NOUN
cana-6247	118	7	,	,	PUNCT
cana-6247	118	8	such	such	ADJ
cana-6247	118	9	as	as	ADP
cana-6247	118	10	residual	residual	ADJ
cana-6247	118	11	shifting	shifting	NOUN
cana-6247	118	12	,	,	PUNCT
cana-6247	118	13	that	that	PRON
cana-6247	118	14	have	have	AUX
cana-6247	118	15	reduced	reduce	VERB
cana-6247	118	16	the	the	DET
cana-6247	118	17	inference	inference	NOUN
cana-6247	118	18	time	time	NOUN
cana-6247	118	19	,	,	PUNCT
cana-6247	118	20	but	but	CCONJ
cana-6247	118	21	not	not	PART
cana-6247	118	22	at	at	ADP
cana-6247	118	23	the	the	DET
cana-6247	118	24	expense	expense	NOUN
cana-6247	118	25	of	of	ADP
cana-6247	118	26	restoration	restoration	NOUN
cana-6247	118	27	capabilities	capability	NOUN
cana-6247	118	28	.	.	PUNCT
cana-6247	119	1	these	these	PRON
cana-6247	119	2	are	be	AUX
cana-6247	119	3	important	important	ADJ
cana-6247	119	4	methods	method	NOUN
cana-6247	119	5	in	in	ADP
cana-6247	119	6	situations	situation	NOUN
cana-6247	119	7	where	where	SCONJ
cana-6247	119	8	rapid	rapid	ADJ
cana-6247	119	9	and	and	CCONJ
cana-6247	119	10	efficient	efficient	ADJ
cana-6247	119	11	results	result	NOUN
cana-6247	119	12	are	be	AUX
cana-6247	119	13	required	require	VERB
cana-6247	119	14	,	,	PUNCT
cana-6247	119	15	without	without	ADP
cana-6247	119	16	affecting	affect	VERB
cana-6247	119	17	the	the	DET
cana-6247	119	18	fidelity	fidelity	NOUN
cana-6247	119	19	of	of	ADP
cana-6247	119	20	restored	restore	VERB
cana-6247	119	21	images	image	NOUN
cana-6247	119	22	.	.	PUNCT
cana-6247	120	1	[	[	X
cana-6247	120	2	50	50	NUM
cana-6247	120	3	]	]	PUNCT
cana-6247	120	4	[	[	X
cana-6247	120	5	24	24	NUM
cana-6247	120	6	]	]	SYM
cana-6247	120	7	•	•	PRON
cana-6247	120	8	hybrid	hybrid	NOUN
cana-6247	120	9	/	/	SYM
cana-6247	120	10	specialized	specialized	ADJ
cana-6247	120	11	models	model	NOUN
cana-6247	120	12	:	:	PUNCT
cana-6247	120	13	these	these	DET
cana-6247	120	14	models	model	NOUN
cana-6247	120	15	take	take	VERB
cana-6247	120	16	advantage	advantage	NOUN
cana-6247	120	17	of	of	ADP
cana-6247	120	18	diffusion	diffusion	NOUN
cana-6247	120	19	processes	process	NOUN
cana-6247	120	20	in	in	ADP
cana-6247	120	21	conjunction	conjunction	NOUN
cana-6247	120	22	with	with	ADP
cana-6247	120	23	a	a	DET
cana-6247	120	24	prior	prior	ADJ
cana-6247	120	25	or	or	CCONJ
cana-6247	120	26	mathematical	mathematical	ADJ
cana-6247	120	27	framework	framework	NOUN
cana-6247	120	28	.	.	PUNCT
cana-6247	121	1	for	for	ADP
cana-6247	121	2	example	example	NOUN
cana-6247	121	3	,	,	PUNCT
cana-6247	121	4	models	model	NOUN
cana-6247	121	5	that	that	PRON
cana-6247	121	6	incorporate	incorporate	VERB
cana-6247	121	7	residual	residual	ADJ
cana-6247	121	8	noise	noise	NOUN
cana-6247	121	9	priors	prior	NOUN
cana-6247	121	10	,	,	PUNCT
cana-6247	121	11	gaussian	gaussian	ADJ
cana-6247	121	12	mixture	mixture	NOUN
cana-6247	121	13	models	model	NOUN
cana-6247	121	14	,	,	PUNCT
cana-6247	121	15	or	or	CCONJ
cana-6247	121	16	reaction	reaction	NOUN
cana-6247	121	17	-	-	PUNCT
cana-6247	121	18	diffusion	diffusion	NOUN
cana-6247	121	19	equations	equation	NOUN
cana-6247	121	20	for	for	ADP
cana-6247	121	21	specific	specific	ADJ
cana-6247	121	22	restoration	restoration	NOUN
cana-6247	121	23	tasks	task	NOUN
cana-6247	121	24	(	(	PUNCT
cana-6247	121	25	e.g.	e.g.	ADV
cana-6247	121	26	,	,	PUNCT
cana-6247	121	27	multimodality	multimodality	NOUN
cana-6247	121	28	fusion	fusion	NOUN
cana-6247	121	29	,	,	PUNCT
cana-6247	121	30	adversarial	adversarial	ADJ
cana-6247	121	31	purification	purification	NOUN
cana-6247	121	32	,	,	PUNCT
cana-6247	121	33	or	or	CCONJ
cana-6247	121	34	artifact	artifact	ADJ
cana-6247	121	35	correction	correction	NOUN
cana-6247	121	36	)	)	PUNCT
cana-6247	121	37	.	.	PUNCT
cana-6247	122	1	the	the	DET
cana-6247	122	2	goal	goal	NOUN
cana-6247	122	3	of	of	ADP
cana-6247	122	4	these	these	DET
cana-6247	122	5	hybrid	hybrid	ADJ
cana-6247	122	6	methods	method	NOUN
cana-6247	122	7	is	be	AUX
cana-6247	122	8	to	to	PART
cana-6247	122	9	improve	improve	VERB
cana-6247	122	10	restoration	restoration	NOUN
cana-6247	122	11	effectiveness	effectiveness	NOUN
cana-6247	122	12	in	in	ADP
cana-6247	122	13	special	special	ADJ
cana-6247	122	14	cases	case	NOUN
cana-6247	122	15	that	that	PRON
cana-6247	122	16	leverage	leverage	VERB
cana-6247	122	17	both	both	DET
cana-6247	122	18	domain	domain	NOUN
cana-6247	122	19	knowledge	knowledge	NOUN
cana-6247	122	20	and	and	CCONJ
cana-6247	122	21	associated	associate	VERB
cana-6247	122	22	techniques	technique	NOUN
cana-6247	122	23	to	to	PART
cana-6247	122	24	tackle	tackle	VERB
cana-6247	122	25	complex	complex	ADJ
cana-6247	122	26	or	or	CCONJ
cana-6247	122	27	non	non	ADJ
cana-6247	122	28	-	-	ADJ
cana-6247	122	29	standard	standard	ADJ
cana-6247	122	30	degradations	degradation	NOUN
cana-6247	122	31	.	.	PUNCT
cana-6247	123	1	[	[	X
cana-6247	123	2	54	54	NUM
cana-6247	123	3	]	]	X
cana-6247	123	4	[	[	X
cana-6247	123	5	15	15	NUM
cana-6247	123	6	]	]	SYM
cana-6247	123	7	•	•	NOUN
cana-6247	123	8	zero	zero	NUM
cana-6247	123	9	-	-	PUNCT
cana-6247	123	10	shot	shoot	VERB
cana-6247	123	11	/	/	SYM
cana-6247	123	12	blind	blind	ADJ
cana-6247	123	13	models	model	NOUN
cana-6247	123	14	:	:	PUNCT
cana-6247	123	15	these	these	DET
cana-6247	123	16	methods	method	NOUN
cana-6247	123	17	,	,	PUNCT
cana-6247	123	18	based	base	VERB
cana-6247	123	19	on	on	ADP
cana-6247	123	20	stochastic	stochastic	ADJ
cana-6247	123	21	differential	differential	ADJ
cana-6247	123	22	equations	equation	NOUN
cana-6247	123	23	or	or	CCONJ
cana-6247	123	24	dual	dual	ADJ
cana-6247	123	25	priors	prior	NOUN
cana-6247	123	26	,	,	PUNCT
cana-6247	123	27	are	be	AUX
cana-6247	123	28	trained	train	VERB
cana-6247	123	29	to	to	PART
cana-6247	123	30	produce	produce	VERB
cana-6247	123	31	unseen	unseen	ADJ
cana-6247	123	32	degradations	degradation	NOUN
cana-6247	123	33	and	and	CCONJ
cana-6247	123	34	generalizations	generalization	NOUN
cana-6247	123	35	for	for	ADP
cana-6247	123	36	new	new	ADJ
cana-6247	123	37	tasks	task	NOUN
cana-6247	123	38	without	without	ADP
cana-6247	123	39	the	the	DET
cana-6247	123	40	need	need	NOUN
cana-6247	123	41	for	for	ADP
cana-6247	123	42	taskspecific	taskspecific	ADJ
cana-6247	123	43	training	training	NOUN
cana-6247	123	44	.	.	PUNCT
cana-6247	124	1	zero	zero	NUM
cana-6247	124	2	-	-	PUNCT
cana-6247	124	3	shot	shot	NOUN
cana-6247	124	4	restoration	restoration	NOUN
cana-6247	124	5	models	model	NOUN
cana-6247	124	6	leverage	leverage	VERB
cana-6247	124	7	the	the	DET
cana-6247	124	8	generalization	generalization	NOUN
cana-6247	124	9	ability	ability	NOUN
cana-6247	124	10	of	of	ADP
cana-6247	124	11	learned	learn	VERB
cana-6247	124	12	diffusion	diffusion	NOUN
cana-6247	124	13	priors	prior	NOUN
cana-6247	124	14	to	to	PART
cana-6247	124	15	apply	apply	VERB
cana-6247	124	16	robust	robust	ADJ
cana-6247	124	17	restoration	restoration	NOUN
cana-6247	124	18	even	even	ADV
cana-6247	124	19	when	when	SCONJ
cana-6247	124	20	little	little	ADJ
cana-6247	124	21	is	be	AUX
cana-6247	124	22	known	know	VERB
cana-6247	124	23	about	about	ADP
cana-6247	124	24	the	the	DET
cana-6247	124	25	image	image	NOUN
cana-6247	124	26	degradation	degradation	NOUN
cana-6247	124	27	.	.	PUNCT
cana-6247	125	1	this	this	PRON
cana-6247	125	2	promotes	promote	VERB
cana-6247	125	3	versatility	versatility	NOUN
cana-6247	125	4	and	and	CCONJ
cana-6247	125	5	flexibility	flexibility	NOUN
cana-6247	125	6	for	for	ADP
cana-6247	125	7	real	real	ADJ
cana-6247	125	8	-	-	PUNCT
cana-6247	125	9	world	world	NOUN
cana-6247	125	10	and	and	CCONJ
cana-6247	125	11	open	open	ADJ
cana-6247	125	12	-	-	PUNCT
cana-6247	125	13	source	source	NOUN
cana-6247	125	14	applications	application	NOUN
cana-6247	125	15	.	.	PUNCT
cana-6247	126	1	[	[	X
cana-6247	126	2	25	25	NUM
cana-6247	126	3	]	]	PUNCT
cana-6247	126	4	in	in	ADP
cana-6247	126	5	sum	sum	NOUN
cana-6247	126	6	.	.	PUNCT
cana-6247	127	1	additive	additive	ADJ
cana-6247	127	2	gaussian	gaussian	ADJ
cana-6247	127	3	noise	noise	NOUN
cana-6247	127	4	(	(	PUNCT
cana-6247	127	5	agn	agn	NOUN
cana-6247	127	6	)	)	PUNCT
cana-6247	127	7	is	be	AUX
cana-6247	127	8	the	the	DET
cana-6247	127	9	primary	primary	ADJ
cana-6247	127	10	underlying	underlying	ADJ
cana-6247	127	11	mechanism	mechanism	NOUN
cana-6247	127	12	in	in	ADP
cana-6247	127	13	diffusion	diffusion	NOUN
cana-6247	127	14	approaches	approach	NOUN
cana-6247	127	15	to	to	ADP
cana-6247	127	16	image	image	NOUN
cana-6247	127	17	restoration	restoration	NOUN
cana-6247	127	18	and	and	CCONJ
cana-6247	127	19	enhancement	enhancement	NOUN
cana-6247	127	20	.	.	PUNCT
cana-6247	128	1	it	it	PRON
cana-6247	128	2	maintains	maintain	VERB
cana-6247	128	3	statistical	statistical	ADJ
cana-6247	128	4	tractability	tractability	NOUN
cana-6247	128	5	while	while	SCONJ
cana-6247	128	6	uniformly	uniformly	ADV
cana-6247	128	7	and	and	CCONJ
cana-6247	128	8	unbiasedly	unbiasedly	ADV
cana-6247	128	9	corrupting	corrupt	VERB
cana-6247	128	10	images	image	NOUN
cana-6247	128	11	,	,	PUNCT
cana-6247	128	12	making	make	VERB
cana-6247	128	13	denoising	denoise	VERB
cana-6247	128	14	a	a	DET
cana-6247	128	15	mathematically	mathematically	ADV
cana-6247	128	16	convenient	convenient	ADJ
cana-6247	128	17	process	process	NOUN
cana-6247	128	18	with	with	ADP
cana-6247	128	19	neural	neural	ADJ
cana-6247	128	20	networks	network	NOUN
cana-6247	128	21	,	,	PUNCT
cana-6247	128	22	such	such	ADJ
cana-6247	128	23	as	as	ADP
cana-6247	128	24	unet	unet	NOUN
cana-6247	128	25	[	[	X
cana-6247	128	26	42	42	NUM
cana-6247	128	27	]	]	PUNCT
cana-6247	128	28	.	.	PUNCT
cana-6247	129	1	classic	classic	ADJ
cana-6247	129	2	ddpm	ddpm	ADJ
cana-6247	129	3	models	model	NOUN
cana-6247	129	4	effectively	effectively	ADV
cana-6247	129	5	express	express	VERB
cana-6247	129	6	complex	complex	ADJ
cana-6247	129	7	noise	noise	NOUN
cana-6247	129	8	properties	property	NOUN
cana-6247	129	9	throughout	throughout	ADP
cana-6247	129	10	their	their	PRON
cana-6247	129	11	restoration	restoration	NOUN
cana-6247	129	12	.	.	PUNCT
cana-6247	130	1	in	in	ADP
cana-6247	130	2	contrast	contrast	NOUN
cana-6247	130	3	,	,	PUNCT
cana-6247	130	4	score	score	NOUN
cana-6247	130	5	-	-	PUNCT
cana-6247	130	6	based	base	VERB
cana-6247	130	7	models	model	NOUN
cana-6247	130	8	help	help	VERB
cana-6247	130	9	improve	improve	VERB
cana-6247	130	10	restoration	restoration	NOUN
cana-6247	130	11	quality	quality	NOUN
cana-6247	130	12	,	,	PUNCT
cana-6247	130	13	indirectly	indirectly	ADV
cana-6247	130	14	’	'	PUNCT
cana-6247	130	15	gradually	gradually	ADV
cana-6247	130	16	’	'	PUNCT
cana-6247	130	17	learning	learn	VERB
cana-6247	130	18	the	the	DET
cana-6247	130	19	probability	probability	NOUN
cana-6247	130	20	gradients	gradient	NOUN
cana-6247	130	21	that	that	PRON
cana-6247	130	22	exist	exist	VERB
cana-6247	130	23	in	in	ADP
cana-6247	130	24	the	the	DET
cana-6247	130	25	data	data	NOUN
cana-6247	130	26	manifold	manifold	ADJ
cana-6247	131	1	[	[	X
cana-6247	131	2	11	11	NUM
cana-6247	131	3	]	]	X
cana-6247	131	4	[	[	X
cana-6247	131	5	25	25	NUM
cana-6247	131	6	]	]	X
cana-6247	131	7	[	[	X
cana-6247	131	8	38	38	NUM
cana-6247	131	9	]	]	PUNCT
cana-6247	131	10	.	.	PUNCT
cana-6247	132	1	although	although	SCONJ
cana-6247	132	2	plug	plug	NOUN
cana-6247	132	3	-	-	PUNCT
cana-6247	132	4	and	and	CCONJ
cana-6247	132	5	-	-	PUNCT
cana-6247	132	6	play	play	NOUN
cana-6247	132	7	and	and	CCONJ
cana-6247	132	8	hybrid	hybrid	NOUN
cana-6247	132	9	models	model	NOUN
cana-6247	132	10	permit	permit	VERB
cana-6247	132	11	greater	great	ADJ
cana-6247	132	12	flexibility	flexibility	NOUN
cana-6247	132	13	by	by	ADP
cana-6247	132	14	combining	combine	VERB
cana-6247	132	15	external	external	ADJ
cana-6247	132	16	priors	prior	NOUN
cana-6247	132	17	or	or	CCONJ
cana-6247	132	18	residual	residual	ADJ
cana-6247	132	19	shifting	shifting	NOUN
cana-6247	132	20	,	,	PUNCT
cana-6247	132	21	these	these	PRON
cana-6247	132	22	also	also	ADV
cana-6247	132	23	promote	promote	VERB
cana-6247	132	24	intensive	intensive	ADJ
cana-6247	132	25	domain	domain	NOUN
cana-6247	132	26	specificity	specificity	NOUN
cana-6247	132	27	and	and	CCONJ
cana-6247	132	28	customisation	customisation	NOUN
cana-6247	132	29	—	—	PUNCT
cana-6247	132	30	able	able	ADJ
cana-6247	132	31	to	to	PART
cana-6247	132	32	flexibly	flexibly	ADV
cana-6247	132	33	handle	handle	VERB
cana-6247	132	34	tasks	task	NOUN
cana-6247	132	35	such	such	ADJ
cana-6247	132	36	as	as	ADP
cana-6247	132	37	fusion	fusion	NOUN
cana-6247	132	38	,	,	PUNCT
cana-6247	132	39	artifact	artifact	NOUN
cana-6247	132	40	or	or	CCONJ
cana-6247	132	41	damage	damage	NOUN
cana-6247	132	42	correction	correction	NOUN
cana-6247	132	43	,	,	PUNCT
cana-6247	132	44	and	and	CCONJ
cana-6247	132	45	adversarial	adversarial	ADJ
cana-6247	132	46	purification	purification	NOUN
cana-6247	132	47	[	[	X
cana-6247	132	48	56	56	NUM
cana-6247	132	49	]	]	X
cana-6247	133	1	[	[	X
cana-6247	133	2	54	54	NUM
cana-6247	133	3	]	]	X
cana-6247	134	1	[	[	X
cana-6247	134	2	15	15	NUM
cana-6247	134	3	]	]	X
cana-6247	135	1	[	[	X
cana-6247	135	2	24	24	NUM
cana-6247	135	3	]	]	X
cana-6247	136	1	[	[	X
cana-6247	136	2	51	51	NUM
cana-6247	136	3	]	]	X
cana-6247	137	1	[	[	X
cana-6247	137	2	50	50	NUM
cana-6247	137	3	]	]	PUNCT
cana-6247	137	4	.	.	PUNCT
cana-6247	137	5	accelerated	accelerate	VERB
cana-6247	137	6	models	model	NOUN
cana-6247	137	7	(	(	PUNCT
cana-6247	137	8	including	include	VERB
cana-6247	137	9	those	those	PRON
cana-6247	137	10	that	that	PRON
cana-6247	137	11	leverage	leverage	VERB
cana-6247	137	12	early	early	ADJ
cana-6247	137	13	stopping	stopping	NOUN
cana-6247	137	14	)	)	PUNCT
cana-6247	137	15	satisfy	satisfy	VERB
cana-6247	137	16	an	an	DET
cana-6247	137	17	increasing	increase	VERB
cana-6247	137	18	and	and	CCONJ
cana-6247	137	19	widespread	widespread	ADJ
cana-6247	137	20	demand	demand	NOUN
cana-6247	137	21	for	for	ADP
cana-6247	137	22	expedient	expedient	ADJ
cana-6247	137	23	results	result	NOUN
cana-6247	137	24	without	without	ADP
cana-6247	137	25	sacrificing	sacrifice	VERB
cana-6247	137	26	faithfulness	faithfulness	NOUN
cana-6247	137	27	or	or	CCONJ
cana-6247	137	28	realism	realism	NOUN
cana-6247	137	29	,	,	PUNCT
cana-6247	137	30	while	while	SCONJ
cana-6247	137	31	zero	zero	NUM
cana-6247	137	32	-	-	PUNCT
cana-6247	137	33	shot	shoot	VERB
cana-6247	137	34	/	/	SYM
cana-6247	137	35	blind	blind	ADJ
cana-6247	137	36	models	model	NOUN
cana-6247	137	37	using	use	VERB
cana-6247	137	38	stochastic	stochastic	ADJ
cana-6247	137	39	differential	differential	ADJ
cana-6247	137	40	equations	equation	NOUN
cana-6247	137	41	,	,	PUNCT
cana-6247	137	42	when	when	SCONJ
cana-6247	137	43	trained	train	VERB
cana-6247	137	44	,	,	PUNCT
cana-6247	137	45	create	create	VERB
cana-6247	137	46	a	a	DET
cana-6247	137	47	strong	strong	ADJ
cana-6247	137	48	confidence	confidence	NOUN
cana-6247	137	49	to	to	PART
cana-6247	137	50	generalize	generalize	VERB
cana-6247	137	51	restoration	restoration	NOUN
cana-6247	137	52	to	to	ADP
cana-6247	137	53	unseen	unseen	ADJ
cana-6247	137	54	image	image	NOUN
cana-6247	137	55	degradations	degradation	NOUN
cana-6247	137	56	.	.	PUNCT
cana-6247	138	1	while	while	SCONJ
cana-6247	138	2	agn	agn	PROPN
cana-6247	138	3	consistently	consistently	ADV
cana-6247	138	4	outperforms	outperform	VERB
cana-6247	138	5	other	other	ADJ
cana-6247	138	6	methods	method	NOUN
cana-6247	138	7	like	like	ADP
cana-6247	138	8	impulse	impulse	ADJ
cana-6247	138	9	as	as	ADP
cana-6247	138	10	a	a	DET
cana-6247	138	11	standard	standard	NOUN
cana-6247	138	12	for	for	ADP
cana-6247	138	13	restoration	restoration	NOUN
cana-6247	138	14	across	across	ADP
cana-6247	138	15	a	a	DET
cana-6247	138	16	diversity	diversity	NOUN
cana-6247	138	17	of	of	ADP
cana-6247	138	18	imaging	imaging	NOUN
cana-6247	138	19	tasks	task	NOUN
cana-6247	138	20	,	,	PUNCT
cana-6247	138	21	its	its	PRON
cana-6247	138	22	implications	implication	NOUN
cana-6247	138	23	are	be	AUX
cana-6247	138	24	varied	varied	ADJ
cana-6247	138	25	.	.	PUNCT
cana-6247	139	1	agn	agn	PROPN
cana-6247	139	2	possesses	possess	VERB
cana-6247	139	3	stability	stability	NOUN
cana-6247	139	4	,	,	PUNCT
cana-6247	139	5	flexibility	flexibility	NOUN
cana-6247	139	6	,	,	PUNCT
cana-6247	139	7	and	and	CCONJ
cana-6247	139	8	superior	superior	ADJ
cana-6247	139	9	restoration	restoration	NOUN
cana-6247	139	10	performance	performance	NOUN
cana-6247	139	11	across	across	ADP
cana-6247	139	12	diverse	diverse	ADJ
cana-6247	139	13	imaging	imaging	NOUN
cana-6247	139	14	tasks	task	NOUN
cana-6247	139	15	,	,	PUNCT
cana-6247	139	16	such	such	ADJ
cana-6247	139	17	as	as	ADP
cana-6247	139	18	denoising	denoise	VERB
cana-6247	139	19	of	of	ADP
cana-6247	139	20	raw	raw	ADJ
cana-6247	139	21	data	datum	NOUN
cana-6247	139	22	or	or	CCONJ
cana-6247	139	23	fused	fuse	VERB
cana-6247	139	24	data	datum	NOUN
cana-6247	139	25	,	,	PUNCT
cana-6247	139	26	bistatic	bistatic	ADJ
cana-6247	139	27	,	,	PUNCT
cana-6247	139	28	artificially	artificially	ADV
cana-6247	139	29	included	include	VERB
cana-6247	139	30	noise	noise	NOUN
cana-6247	139	31	removal	removal	NOUN
cana-6247	139	32	from	from	ADP
cana-6247	139	33	images	image	NOUN
cana-6247	139	34	,	,	PUNCT
cana-6247	139	35	generative	generative	ADJ
cana-6247	139	36	noise	noise	NOUN
cana-6247	139	37	removal	removal	NOUN
cana-6247	139	38	,	,	PUNCT
cana-6247	139	39	and	and	CCONJ
cana-6247	139	40	deleting	delete	VERB
cana-6247	139	41	or	or	CCONJ
cana-6247	139	42	transforming	transform	VERB
cana-6247	139	43	the	the	DET
cana-6247	139	44	image	image	NOUN
cana-6247	139	45	scene	scene	NOUN
cana-6247	139	46	itself	itself	PRON
cana-6247	140	1	[	[	X
cana-6247	140	2	25	25	NUM
cana-6247	140	3	]	]	PUNCT
cana-6247	140	4	.	.	PUNCT
cana-6247	141	1	2	2	NUM
cana-6247	141	2	conditional	conditional	ADJ
cana-6247	141	3	noise	noise	NOUN
cana-6247	141	4	injection	injection	NOUN
cana-6247	141	5	(	(	PUNCT
cana-6247	141	6	cni	cni	PROPN
cana-6247	141	7	)	)	PUNCT
cana-6247	141	8	the	the	DET
cana-6247	141	9	core	core	ADJ
cana-6247	141	10	idea	idea	NOUN
cana-6247	141	11	of	of	ADP
cana-6247	141	12	cni	cni	PROPN
cana-6247	141	13	is	be	AUX
cana-6247	141	14	a	a	DET
cana-6247	141	15	complex	complex	ADJ
cana-6247	141	16	one	one	NUM
cana-6247	141	17	in	in	ADP
cana-6247	141	18	diffusion	diffusion	NOUN
cana-6247	141	19	models	model	NOUN
cana-6247	141	20	,	,	PUNCT
cana-6247	141	21	wherein	wherein	SCONJ
cana-6247	141	22	the	the	DET
cana-6247	141	23	noise	noise	NOUN
cana-6247	141	24	that	that	PRON
cana-6247	141	25	is	be	AUX
cana-6247	141	26	usually	usually	ADV
cana-6247	141	27	injected	inject	VERB
cana-6247	141	28	during	during	ADP
cana-6247	141	29	the	the	DET
cana-6247	141	30	forward	forward	ADJ
cana-6247	141	31	diffusion	diffusion	NOUN
cana-6247	141	32	process	process	NOUN
cana-6247	141	33	comes	come	VERB
cana-6247	141	34	under	under	ADP
cana-6247	141	35	the	the	DET
cana-6247	141	36	influence	influence	NOUN
cana-6247	141	37	of	of	ADP
cana-6247	141	38	extrinsic	extrinsic	ADJ
cana-6247	141	39	conditions	condition	NOUN
cana-6247	141	40	,	,	PUNCT
cana-6247	141	41	such	such	ADJ
cana-6247	141	42	as	as	ADP
cana-6247	141	43	degraded	degraded	ADJ
cana-6247	141	44	images	image	NOUN
cana-6247	141	45	,	,	PUNCT
cana-6247	141	46	a	a	DET
cana-6247	141	47	semantic	semantic	ADJ
cana-6247	141	48	mask	mask	NOUN
cana-6247	141	49	,	,	PUNCT
cana-6247	141	50	or	or	CCONJ
cana-6247	141	51	other	other	ADJ
cana-6247	141	52	ancillary	ancillary	ADJ
cana-6247	141	53	information	information	NOUN
cana-6247	141	54	.	.	PUNCT
cana-6247	142	1	this	this	DET
cana-6247	142	2	approach	approach	NOUN
cana-6247	142	3	is	be	AUX
cana-6247	142	4	the	the	DET
cana-6247	142	5	generalized	generalized	ADJ
cana-6247	142	6	form	form	NOUN
cana-6247	142	7	of	of	ADP
cana-6247	142	8	the	the	DET
cana-6247	142	9	usual	usual	ADJ
cana-6247	142	10	diffusion	diffusion	NOUN
cana-6247	142	11	process	process	NOUN
cana-6247	142	12	;	;	PUNCT
cana-6247	142	13	instead	instead	ADV
cana-6247	142	14	of	of	ADP
cana-6247	142	15	a	a	DET
cana-6247	142	16	crude	crude	ADJ
cana-6247	142	17	and	and	CCONJ
cana-6247	142	18	pure	pure	ADJ
cana-6247	142	19	random	random	ADJ
cana-6247	142	20	corruption	corruption	NOUN
cana-6247	142	21	,	,	PUNCT
cana-6247	142	22	it	it	PRON
cana-6247	142	23	is	be	AUX
cana-6247	142	24	now	now	ADV
cana-6247	142	25	an	an	DET
cana-6247	142	26	organised	organised	ADJ
cana-6247	142	27	,	,	PUNCT
cana-6247	142	28	information	information	NOUN
cana-6247	142	29	-	-	PUNCT
cana-6247	142	30	preserving	preserve	VERB
cana-6247	142	31	process	process	NOUN
cana-6247	142	32	linked	link	VERB
cana-6247	142	33	to	to	ADP
cana-6247	142	34	the	the	DET
cana-6247	142	35	task	task	NOUN
cana-6247	142	36	[	[	X
cana-6247	142	37	48	48	NUM
cana-6247	142	38	]	]	PUNCT
cana-6247	142	39	.	.	PUNCT
cana-6247	143	1	in	in	ADP
cana-6247	143	2	traditional	traditional	ADJ
cana-6247	143	3	diffusion	diffusion	NOUN
cana-6247	143	4	models	model	NOUN
cana-6247	143	5	,	,	PUNCT
cana-6247	143	6	the	the	DET
cana-6247	143	7	forward	forward	ADJ
cana-6247	143	8	process	process	NOUN
cana-6247	143	9	corrupts	corrupt	VERB
cana-6247	143	10	the	the	DET
cana-6247	143	11	input	input	NOUN
cana-6247	143	12	image	image	NOUN
cana-6247	143	13	by	by	ADP
cana-6247	143	14	adding	add	VERB
cana-6247	143	15	isotropic	isotropic	ADJ
cana-6247	143	16	gaussian	gaussian	NOUN
cana-6247	143	17	noise	noise	NOUN
cana-6247	143	18	at	at	ADP
cana-6247	143	19	each	each	DET
cana-6247	143	20	timestep	timestep	NOUN
cana-6247	143	21	until	until	SCONJ
cana-6247	143	22	it	it	PRON
cana-6247	143	23	is	be	AUX
cana-6247	143	24	completely	completely	ADV
cana-6247	143	25	noise	noise	ADJ
cana-6247	143	26	,	,	PUNCT
cana-6247	143	27	completely	completely	ADV
cana-6247	143	28	uninformed	uninformed	ADJ
cana-6247	143	29	and	and	CCONJ
cana-6247	143	30	regards	regard	VERB
cana-6247	143	31	all	all	DET
cana-6247	143	32	pixels	pixel	NOUN
cana-6247	143	33	equally	equally	ADV
cana-6247	143	34	.	.	PUNCT
cana-6247	144	1	this	this	PRON
cana-6247	144	2	is	be	AUX
cana-6247	144	3	a	a	DET
cana-6247	144	4	practical	practical	ADJ
cana-6247	144	5	approach	approach	NOUN
cana-6247	144	6	when	when	SCONJ
cana-6247	144	7	the	the	DET
cana-6247	144	8	goal	goal	NOUN
cana-6247	144	9	is	be	AUX
cana-6247	144	10	unconditional	unconditional	ADJ
cana-6247	144	11	generation	generation	NOUN
cana-6247	144	12	,	,	PUNCT
cana-6247	144	13	but	but	CCONJ
cana-6247	144	14	suboptimal	suboptimal	ADJ
cana-6247	144	15	for	for	ADP
cana-6247	144	16	tasks	task	NOUN
cana-6247	144	17	that	that	PRON
cana-6247	144	18	require	require	VERB
cana-6247	144	19	guidance	guidance	NOUN
cana-6247	144	20	,	,	PUNCT
cana-6247	144	21	such	such	ADJ
cana-6247	144	22	as	as	ADP
cana-6247	144	23	picture	picture	NOUN
cana-6247	144	24	inpainting	inpainting	NOUN
cana-6247	144	25	.	.	PUNCT
cana-6247	145	1	thus	thus	ADV
cana-6247	145	2	,	,	PUNCT
cana-6247	145	3	conditional	conditional	ADJ
cana-6247	145	4	noise	noise	NOUN
cana-6247	145	5	injection	injection	NOUN
cana-6247	145	6	completely	completely	ADV
cana-6247	145	7	alters	alter	VERB
cana-6247	145	8	this	this	DET
cana-6247	145	9	paradigm	paradigm	NOUN
cana-6247	145	10	by	by	ADP
cana-6247	145	11	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	145	12	983	983	NUM
cana-6247	145	13	communications	communication	NOUN
cana-6247	145	14	on	on	ADP
cana-6247	145	15	applied	apply	VERB
cana-6247	145	16	nonlinear	nonlinear	ADJ
cana-6247	145	17	analysis	analysis	NOUN
cana-6247	145	18	issn	issn	NOUN
cana-6247	145	19	:	:	PUNCT
cana-6247	145	20	1074	1074	NUM
cana-6247	145	21	-	-	PUNCT
cana-6247	145	22	133x	133x	NUM
cana-6247	145	23	vol	vol	NOUN
cana-6247	145	24	.	.	PROPN
cana-6247	146	1	31	31	NUM
cana-6247	147	1	no	no	NOUN
cana-6247	147	2	.	.	PUNCT
cana-6247	148	1	7s	7	NOUN
cana-6247	148	2	(	(	PUNCT
cana-6247	148	3	2024	2024	NUM
cana-6247	148	4	)	)	PUNCT
cana-6247	148	5	permitting	permit	VERB
cana-6247	148	6	noise	noise	NOUN
cana-6247	148	7	addition	addition	NOUN
cana-6247	148	8	to	to	PART
cana-6247	148	9	depend	depend	VERB
cana-6247	148	10	on	on	ADP
cana-6247	148	11	a	a	DET
cana-6247	148	12	given	give	VERB
cana-6247	148	13	condition	condition	NOUN
cana-6247	148	14	.	.	PUNCT
cana-6247	149	1	theoretically	theoretically	ADV
cana-6247	149	2	,	,	PUNCT
cana-6247	149	3	this	this	PRON
cana-6247	149	4	would	would	AUX
cana-6247	149	5	indicate	indicate	VERB
cana-6247	149	6	that	that	SCONJ
cana-6247	149	7	the	the	DET
cana-6247	149	8	forward	forward	ADJ
cana-6247	149	9	process	process	NOUN
cana-6247	149	10	is	be	AUX
cana-6247	149	11	not	not	PART
cana-6247	149	12	only	only	ADV
cana-6247	149	13	destroying	destroy	VERB
cana-6247	149	14	information	information	NOUN
cana-6247	149	15	but	but	CCONJ
cana-6247	149	16	also	also	ADV
cana-6247	149	17	encoding	encode	VERB
cana-6247	149	18	into	into	ADP
cana-6247	149	19	the	the	DET
cana-6247	149	20	noisy	noisy	ADJ
cana-6247	149	21	representation	representation	NOUN
cana-6247	149	22	those	those	DET
cana-6247	149	23	cues	cue	NOUN
cana-6247	149	24	that	that	PRON
cana-6247	149	25	are	be	AUX
cana-6247	149	26	relevant	relevant	ADJ
cana-6247	149	27	for	for	ADP
cana-6247	149	28	the	the	DET
cana-6247	149	29	task	task	NOUN
cana-6247	149	30	at	at	ADP
cana-6247	149	31	hand	hand	NOUN
cana-6247	149	32	.	.	PUNCT
cana-6247	150	1	[	[	X
cana-6247	150	2	39].for	39].for	NUM
cana-6247	150	3	example	example	NOUN
cana-6247	150	4	,	,	PUNCT
cana-6247	150	5	a	a	DET
cana-6247	150	6	violent	violent	ADJ
cana-6247	150	7	corruption	corruption	NOUN
cana-6247	150	8	in	in	ADP
cana-6247	150	9	the	the	DET
cana-6247	150	10	masked	mask	VERB
cana-6247	150	11	areas	area	NOUN
cana-6247	150	12	(	(	PUNCT
cana-6247	150	13	missing	miss	VERB
cana-6247	150	14	content	content	NOUN
cana-6247	150	15	)	)	PUNCT
cana-6247	150	16	could	could	AUX
cana-6247	150	17	be	be	AUX
cana-6247	150	18	combined	combine	VERB
cana-6247	150	19	with	with	ADP
cana-6247	150	20	a	a	DET
cana-6247	150	21	gentle	gentle	ADJ
cana-6247	150	22	noise	noise	NOUN
cana-6247	150	23	injection	injection	NOUN
cana-6247	150	24	in	in	ADP
cana-6247	150	25	the	the	DET
cana-6247	150	26	known	know	VERB
cana-6247	150	27	ones	one	NOUN
cana-6247	150	28	—	—	PUNCT
cana-6247	150	29	this	this	PRON
cana-6247	150	30	would	would	AUX
cana-6247	150	31	lead	lead	VERB
cana-6247	150	32	the	the	DET
cana-6247	150	33	reverse	reverse	ADJ
cana-6247	150	34	processgenerative	processgenerative	NOUN
cana-6247	150	35	power	power	NOUN
cana-6247	150	36	to	to	PART
cana-6247	150	37	be	be	AUX
cana-6247	150	38	focused	focus	VERB
cana-6247	150	39	wherever	wherever	SCONJ
cana-6247	150	40	it	it	PRON
cana-6247	150	41	is	be	AUX
cana-6247	150	42	most	most	ADV
cana-6247	150	43	needed	need	VERB
cana-6247	150	44	.	.	PUNCT
cana-6247	151	1	for	for	ADP
cana-6247	151	2	layout	layout	NOUN
cana-6247	151	3	-	-	PUNCT
cana-6247	151	4	aware	aware	ADJ
cana-6247	151	5	synthesis	synthesis	NOUN
cana-6247	151	6	,	,	PUNCT
cana-6247	151	7	the	the	DET
cana-6247	151	8	noise	noise	NOUN
cana-6247	151	9	map	map	NOUN
cana-6247	151	10	could	could	AUX
cana-6247	151	11	then	then	ADV
cana-6247	151	12	be	be	AUX
cana-6247	151	13	spatially	spatially	ADV
cana-6247	151	14	structured	structure	VERB
cana-6247	151	15	to	to	PART
cana-6247	151	16	conform	conform	VERB
cana-6247	151	17	to	to	ADP
cana-6247	151	18	a	a	DET
cana-6247	151	19	desired	desire	VERB
cana-6247	151	20	layout	layout	NOUN
cana-6247	151	21	so	so	SCONJ
cana-6247	151	22	that	that	SCONJ
cana-6247	151	23	spatial	spatial	ADJ
cana-6247	151	24	information	information	NOUN
cana-6247	151	25	is	be	AUX
cana-6247	151	26	directly	directly	ADV
cana-6247	151	27	embedded	embed	VERB
cana-6247	151	28	into	into	ADP
cana-6247	151	29	the	the	DET
cana-6247	151	30	corrupted	corrupted	ADJ
cana-6247	151	31	input	input	NOUN
cana-6247	151	32	.	.	PUNCT
cana-6247	152	1	[	[	X
cana-6247	152	2	36	36	NUM
cana-6247	152	3	]	]	PUNCT
cana-6247	152	4	this	this	DET
cana-6247	152	5	noise	noise	NOUN
cana-6247	152	6	addition	addition	NOUN
cana-6247	152	7	method	method	NOUN
cana-6247	152	8	allows	allow	VERB
cana-6247	152	9	the	the	DET
cana-6247	152	10	reverse	reverse	NOUN
cana-6247	152	11	(	(	PUNCT
cana-6247	152	12	denoising	denoising	NOUN
cana-6247	152	13	)	)	PUNCT
cana-6247	152	14	strategy	strategy	NOUN
cana-6247	152	15	to	to	PART
cana-6247	152	16	capitalize	capitalize	VERB
cana-6247	152	17	on	on	ADP
cana-6247	152	18	both	both	CCONJ
cana-6247	152	19	the	the	DET
cana-6247	152	20	noisy	noisy	ADJ
cana-6247	152	21	input	input	NOUN
cana-6247	152	22	itself	itself	PRON
cana-6247	152	23	,	,	PUNCT
cana-6247	152	24	as	as	ADV
cana-6247	152	25	well	well	ADV
cana-6247	152	26	as	as	ADP
cana-6247	152	27	the	the	DET
cana-6247	152	28	structure	structure	NOUN
cana-6247	152	29	within	within	ADP
cana-6247	152	30	the	the	DET
cana-6247	152	31	noise	noise	NOUN
cana-6247	152	32	,	,	PUNCT
cana-6247	152	33	to	to	PART
cana-6247	152	34	produce	produce	VERB
cana-6247	152	35	outputs	output	NOUN
cana-6247	152	36	that	that	PRON
cana-6247	152	37	are	be	AUX
cana-6247	152	38	better	well	ADV
cana-6247	152	39	aligned	align	VERB
cana-6247	152	40	with	with	ADP
cana-6247	152	41	the	the	DET
cana-6247	152	42	condition	condition	NOUN
cana-6247	152	43	and	and	CCONJ
cana-6247	152	44	more	more	ADV
cana-6247	152	45	suited	suited	ADJ
cana-6247	152	46	to	to	ADP
cana-6247	152	47	the	the	DET
cana-6247	152	48	restoration	restoration	NOUN
cana-6247	152	49	or	or	CCONJ
cana-6247	152	50	generation	generation	NOUN
cana-6247	152	51	task	task	NOUN
cana-6247	152	52	.	.	PUNCT
cana-6247	153	1	the	the	DET
cana-6247	153	2	standard	standard	ADJ
cana-6247	153	3	forward	forward	ADJ
cana-6247	153	4	process	process	NOUN
cana-6247	153	5	in	in	ADP
cana-6247	153	6	a	a	DET
cana-6247	153	7	diffusion	diffusion	NOUN
cana-6247	153	8	model	model	NOUN
cana-6247	153	9	is	be	AUX
cana-6247	153	10	a	a	DET
cana-6247	153	11	markov	markov	NOUN
cana-6247	153	12	chain	chain	NOUN
cana-6247	153	13	defined	define	VERB
cana-6247	153	14	as	as	ADP
cana-6247	153	15	:	:	PUNCT
cana-6247	153	16	q(xt	q(xt	NOUN
cana-6247	153	17	|	|	ADV
cana-6247	153	18	xt−1	xt−1	PROPN
cana-6247	153	19	)	)	PUNCT
cana-6247	154	1	=	=	SYM
cana-6247	154	2	n	n	PROPN
cana-6247	154	3	(	(	PUNCT
cana-6247	154	4	xt	xt	X
cana-6247	154	5	;	;	PUNCT
cana-6247	154	6	√	√	NUM
cana-6247	154	7	1−	1−	NUM
cana-6247	154	8	βt	βt	PROPN
cana-6247	154	9	xt−1	xt−1	PROPN
cana-6247	154	10	,	,	PUNCT
cana-6247	154	11	βti	βti	NOUN
cana-6247	154	12	)	)	PUNCT
cana-6247	154	13	,	,	PUNCT
cana-6247	154	14	where	where	SCONJ
cana-6247	154	15	x0	x0	PROPN
cana-6247	154	16	is	be	AUX
cana-6247	154	17	the	the	DET
cana-6247	154	18	original	original	ADJ
cana-6247	154	19	image	image	NOUN
cana-6247	154	20	,	,	PUNCT
cana-6247	154	21	xt	xt	PROPN
cana-6247	154	22	is	be	AUX
cana-6247	154	23	the	the	DET
cana-6247	154	24	image	image	NOUN
cana-6247	154	25	at	at	ADP
cana-6247	154	26	time	time	NOUN
cana-6247	154	27	step	step	PROPN
cana-6247	154	28	t	t	PROPN
cana-6247	154	29	,	,	PUNCT
cana-6247	154	30	βt	βt	PROPN
cana-6247	154	31	is	be	AUX
cana-6247	154	32	the	the	DET
cana-6247	154	33	variance	variance	NOUN
cana-6247	154	34	schedule	schedule	NOUN
cana-6247	154	35	,	,	PUNCT
cana-6247	154	36	and	and	CCONJ
cana-6247	154	37	n	n	PRON
cana-6247	154	38	denotes	denote	VERB
cana-6247	154	39	a	a	DET
cana-6247	154	40	gaussian	gaussian	ADJ
cana-6247	154	41	distribution	distribution	NOUN
cana-6247	154	42	.	.	PUNCT
cana-6247	155	1	this	this	DET
cana-6247	155	2	process	process	NOUN
cana-6247	155	3	can	can	AUX
cana-6247	155	4	be	be	AUX
cana-6247	155	5	expressed	express	VERB
cana-6247	155	6	in	in	ADP
cana-6247	155	7	closed	closed	ADJ
cana-6247	155	8	form	form	NOUN
cana-6247	155	9	as	as	ADP
cana-6247	155	10	:	:	PUNCT
cana-6247	155	11	xt	xt	PROPN
cana-6247	155	12	=	=	SYM
cana-6247	155	13	√	√	PROPN
cana-6247	155	14	ᾱt	ᾱt	VERB
cana-6247	155	15	x0	x0	PROPN
cana-6247	156	1	+	+	CCONJ
cana-6247	156	2	√	√	PROPN
cana-6247	156	3	1−	1−	NUM
cana-6247	157	1	ᾱt	ᾱt	VERB
cana-6247	157	2	z	z	PROPN
cana-6247	157	3	,	,	PUNCT
cana-6247	157	4	where	where	SCONJ
cana-6247	157	5	ᾱt	ᾱt	VERB
cana-6247	157	6	=	=	SYM
cana-6247	157	7	∏t	∏t	NOUN
cana-6247	157	8	s=1(1−	s=1(1−	PROPN
cana-6247	157	9	βs	βs	NUM
cana-6247	157	10	)	)	PUNCT
cana-6247	157	11	and	and	CCONJ
cana-6247	157	12	z	z	NOUN
cana-6247	157	13	∼	∼	NOUN
cana-6247	157	14	n	n	CCONJ
cana-6247	157	15	(	(	PUNCT
cana-6247	157	16	0	0	NUM
cana-6247	157	17	,	,	PUNCT
cana-6247	157	18	i	i	NOUN
cana-6247	157	19	)	)	PUNCT
cana-6247	157	20	.	.	PUNCT
cana-6247	158	1	in	in	ADP
cana-6247	158	2	conditional	conditional	ADJ
cana-6247	158	3	noise	noise	NOUN
cana-6247	158	4	injection	injection	NOUN
cana-6247	158	5	,	,	PUNCT
cana-6247	158	6	the	the	DET
cana-6247	158	7	noise	noise	NOUN
cana-6247	158	8	addition	addition	NOUN
cana-6247	158	9	is	be	AUX
cana-6247	158	10	parameterized	parameterize	VERB
cana-6247	158	11	by	by	ADP
cana-6247	158	12	a	a	DET
cana-6247	158	13	condition	condition	NOUN
cana-6247	158	14	c	c	NOUN
cana-6247	158	15	:	:	PUNCT
cana-6247	158	16	q(xt	q(xt	NOUN
cana-6247	158	17	|	|	ADV
cana-6247	158	18	xt−1	xt−1	PROPN
cana-6247	158	19	,	,	PUNCT
cana-6247	158	20	c	c	NOUN
cana-6247	158	21	)	)	PUNCT
cana-6247	158	22	=	=	SYM
cana-6247	158	23	n	n	CCONJ
cana-6247	158	24	(	(	PUNCT
cana-6247	158	25	xt	xt	X
cana-6247	158	26	;	;	PUNCT
cana-6247	158	27	√	√	PROPN
cana-6247	158	28	1−	1−	NUM
cana-6247	158	29	βt(c)xt−1	βt(c)xt−1	NOUN
cana-6247	158	30	,	,	PUNCT
cana-6247	158	31	βt(c)i	βt(c)i	PRON
cana-6247	158	32	)	)	PUNCT
cana-6247	158	33	(	(	PUNCT
cana-6247	158	34	1	1	X
cana-6247	158	35	)	)	PUNCT
cana-6247	158	36	or	or	CCONJ
cana-6247	158	37	more	more	ADV
cana-6247	158	38	generally	generally	ADV
cana-6247	158	39	:	:	PUNCT
cana-6247	158	40	xt	xt	PROPN
cana-6247	158	41	=	=	SYM
cana-6247	159	1	√	√	ADP
cana-6247	159	2	ᾱt(c)x0	ᾱt(c)x0	PROPN
cana-6247	159	3	+	+	CCONJ
cana-6247	159	4	√	√	PROPN
cana-6247	159	5	1−	1−	NUM
cana-6247	159	6	ᾱt(c	ᾱt(c	VERB
cana-6247	159	7	)	)	PUNCT
cana-6247	159	8	z(c	z(c	PROPN
cana-6247	159	9	)	)	PUNCT
cana-6247	159	10	(	(	PUNCT
cana-6247	159	11	2	2	X
cana-6247	159	12	)	)	PUNCT
cana-6247	159	13	here	here	ADV
cana-6247	159	14	,	,	PUNCT
cana-6247	159	15	the	the	DET
cana-6247	159	16	condition	condition	NOUN
cana-6247	159	17	c	c	NOUN
cana-6247	159	18	can	can	AUX
cana-6247	159	19	affect	affect	VERB
cana-6247	159	20	:	:	PUNCT
cana-6247	159	21	•	•	ADP
cana-6247	159	22	the	the	DET
cana-6247	159	23	variance	variance	NOUN
cana-6247	159	24	schedule	schedule	NOUN
cana-6247	159	25	βt(c	βt(c	PUNCT
cana-6247	159	26	)	)	PUNCT
cana-6247	159	27	,	,	PUNCT
cana-6247	159	28	making	make	VERB
cana-6247	159	29	the	the	DET
cana-6247	159	30	noise	noise	NOUN
cana-6247	159	31	addition	addition	NOUN
cana-6247	159	32	adaptive	adaptive	ADJ
cana-6247	159	33	to	to	ADP
cana-6247	159	34	the	the	DET
cana-6247	159	35	condition	condition	NOUN
cana-6247	159	36	;	;	PUNCT
cana-6247	159	37	•	•	NUM
cana-6247	159	38	the	the	DET
cana-6247	159	39	noise	noise	NOUN
cana-6247	159	40	sample	sample	NOUN
cana-6247	159	41	z(c	z(c	PROPN
cana-6247	159	42	)	)	PUNCT
cana-6247	159	43	,	,	PUNCT
cana-6247	159	44	allowing	allow	VERB
cana-6247	159	45	for	for	ADP
cana-6247	159	46	spatially	spatially	ADV
cana-6247	159	47	or	or	CCONJ
cana-6247	159	48	semantically	semantically	ADV
cana-6247	159	49	structured	structure	VERB
cana-6247	159	50	noise	noise	NOUN
cana-6247	159	51	;	;	PUNCT
cana-6247	159	52	•	•	NOUN
cana-6247	159	53	or	or	CCONJ
cana-6247	159	54	the	the	DET
cana-6247	159	55	noise	noise	NOUN
cana-6247	159	56	distribution	distribution	NOUN
cana-6247	159	57	itself	itself	PRON
cana-6247	159	58	,	,	PUNCT
cana-6247	159	59	which	which	PRON
cana-6247	159	60	may	may	AUX
cana-6247	159	61	be	be	AUX
cana-6247	159	62	replaced	replace	VERB
cana-6247	159	63	by	by	ADP
cana-6247	159	64	a	a	DET
cana-6247	159	65	non	non	ADJ
cana-6247	159	66	-	-	ADJ
cana-6247	159	67	gaussian	gaussian	ADJ
cana-6247	159	68	form	form	NOUN
cana-6247	159	69	parameterized	parameterize	VERB
cana-6247	159	70	by	by	ADP
cana-6247	159	71	c.	c.	PROPN
cana-6247	160	1	[	[	X
cana-6247	160	2	48	48	NUM
cana-6247	160	3	]	]	PUNCT
cana-6247	160	4	implementation	implementation	NOUN
cana-6247	160	5	variants	variant	NOUN
cana-6247	160	6	of	of	ADP
cana-6247	160	7	conditioned	condition	VERB
cana-6247	160	8	augmentation	augmentation	NOUN
cana-6247	160	9	fall	fall	VERB
cana-6247	160	10	into	into	ADP
cana-6247	160	11	three	three	NUM
cana-6247	160	12	primary	primary	ADJ
cana-6247	160	13	categories	category	NOUN
cana-6247	160	14	:	:	PUNCT
cana-6247	160	15	•	•	NUM
cana-6247	160	16	adaptive	adaptive	ADJ
cana-6247	160	17	variance	variance	NOUN
cana-6247	160	18	,	,	PUNCT
cana-6247	160	19	wherein	wherein	SCONJ
cana-6247	160	20	βt	βt	PROPN
cana-6247	160	21	is	be	AUX
cana-6247	160	22	set	set	VERB
cana-6247	160	23	as	as	ADP
cana-6247	160	24	a	a	DET
cana-6247	160	25	function	function	NOUN
cana-6247	160	26	of	of	ADP
cana-6247	160	27	the	the	DET
cana-6247	160	28	conditioning	conditioning	NOUN
cana-6247	160	29	c	c	X
cana-6247	160	30	—	—	PUNCT
cana-6247	160	31	for	for	ADP
cana-6247	160	32	example	example	NOUN
cana-6247	160	33	,	,	PUNCT
cana-6247	160	34	applying	apply	VERB
cana-6247	160	35	higher	high	ADJ
cana-6247	160	36	noise	noise	NOUN
cana-6247	160	37	to	to	ADP
cana-6247	160	38	masked	mask	VERB
cana-6247	160	39	regions	region	NOUN
cana-6247	160	40	in	in	ADP
cana-6247	160	41	inpainting	inpainte	VERB
cana-6247	160	42	;	;	PUNCT
cana-6247	160	43	•	•	NUM
cana-6247	160	44	conditioned	condition	VERB
cana-6247	160	45	noise	noise	NOUN
cana-6247	160	46	sampling	sampling	NOUN
cana-6247	160	47	,	,	PUNCT
cana-6247	160	48	where	where	SCONJ
cana-6247	160	49	one	one	PRON
cana-6247	160	50	generates	generate	VERB
cana-6247	160	51	a	a	DET
cana-6247	160	52	noise	noise	NOUN
cana-6247	160	53	sample	sample	NOUN
cana-6247	160	54	that	that	PRON
cana-6247	160	55	reflects	reflect	VERB
cana-6247	160	56	spatial	spatial	ADJ
cana-6247	160	57	or	or	CCONJ
cana-6247	160	58	semantic	semantic	ADJ
cana-6247	160	59	patterns	pattern	NOUN
cana-6247	160	60	from	from	ADP
cana-6247	160	61	the	the	DET
cana-6247	160	62	condition	condition	NOUN
cana-6247	160	63	,	,	PUNCT
cana-6247	160	64	such	such	ADJ
cana-6247	160	65	as	as	ADP
cana-6247	160	66	cropping	crop	VERB
cana-6247	160	67	and	and	CCONJ
cana-6247	160	68	merging	merge	VERB
cana-6247	160	69	noise	noise	NOUN
cana-6247	160	70	maps	map	NOUN
cana-6247	160	71	according	accord	VERB
cana-6247	160	72	to	to	ADP
cana-6247	160	73	a	a	DET
cana-6247	160	74	layout	layout	NOUN
cana-6247	160	75	;	;	PUNCT
cana-6247	160	76	•	•	NUM
cana-6247	160	77	distribution	distribution	NOUN
cana-6247	160	78	substitution	substitution	NOUN
cana-6247	160	79	,	,	PUNCT
cana-6247	160	80	in	in	ADP
cana-6247	160	81	which	which	PRON
cana-6247	160	82	the	the	DET
cana-6247	160	83	noise	noise	NOUN
cana-6247	160	84	distribution	distribution	NOUN
cana-6247	160	85	is	be	AUX
cana-6247	160	86	replaced	replace	VERB
cana-6247	160	87	with	with	ADP
cana-6247	160	88	a	a	DET
cana-6247	160	89	mixture	mixture	NOUN
cana-6247	160	90	of	of	ADP
cana-6247	160	91	gaussians	gaussian	NOUN
cana-6247	160	92	or	or	CCONJ
cana-6247	160	93	another	another	DET
cana-6247	160	94	form	form	NOUN
cana-6247	160	95	,	,	PUNCT
cana-6247	160	96	with	with	ADP
cana-6247	160	97	parameters	parameter	NOUN
cana-6247	160	98	determined	determine	VERB
cana-6247	160	99	by	by	ADP
cana-6247	160	100	c.	c.	PROPN
cana-6247	160	101	the	the	DET
cana-6247	160	102	mathematical	mathematical	ADJ
cana-6247	160	103	consequence	consequence	NOUN
cana-6247	160	104	of	of	ADP
cana-6247	160	105	these	these	DET
cana-6247	160	106	approaches	approach	NOUN
cana-6247	160	107	is	be	AUX
cana-6247	160	108	that	that	SCONJ
cana-6247	160	109	the	the	DET
cana-6247	160	110	forward	forward	ADJ
cana-6247	160	111	process	process	NOUN
cana-6247	160	112	remains	remain	VERB
cana-6247	160	113	a	a	DET
cana-6247	160	114	markov	markov	NOUN
cana-6247	160	115	chain	chain	NOUN
cana-6247	160	116	,	,	PUNCT
cana-6247	160	117	while	while	SCONJ
cana-6247	160	118	the	the	DET
cana-6247	160	119	transition	transition	NOUN
cana-6247	160	120	kernels	kernel	NOUN
cana-6247	160	121	become	become	VERB
cana-6247	160	122	parameterized	parameterized	ADJ
cana-6247	160	123	by	by	ADP
cana-6247	160	124	the	the	DET
cana-6247	160	125	conditioning	conditioning	NOUN
cana-6247	160	126	variable	variable	NOUN
cana-6247	160	127	.	.	PUNCT
cana-6247	161	1	if	if	SCONJ
cana-6247	161	2	the	the	DET
cana-6247	161	3	noise	noise	NOUN
cana-6247	161	4	remains	remain	VERB
cana-6247	161	5	gaussian	gaussian	ADJ
cana-6247	161	6	,	,	PUNCT
cana-6247	161	7	the	the	DET
cana-6247	161	8	marginal	marginal	ADJ
cana-6247	161	9	distribution	distribution	NOUN
cana-6247	161	10	at	at	ADP
cana-6247	161	11	each	each	DET
cana-6247	161	12	step	step	NOUN
cana-6247	161	13	becomes	become	VERB
cana-6247	161	14	:	:	PUNCT
cana-6247	161	15	q(xt	q(xt	NOUN
cana-6247	161	16	|	|	ADV
cana-6247	161	17	x0	x0	PROPN
cana-6247	161	18	,	,	PUNCT
cana-6247	161	19	c	c	X
cana-6247	161	20	)	)	PUNCT
cana-6247	161	21	=	=	SYM
cana-6247	162	1	n	n	CCONJ
cana-6247	162	2	(	(	PUNCT
cana-6247	162	3	xt	xt	X
cana-6247	162	4	;	;	PUNCT
cana-6247	162	5	√	√	ADP
cana-6247	162	6	ᾱt(c)x0	ᾱt(c)x0	PROPN
cana-6247	162	7	,	,	PUNCT
cana-6247	162	8	(	(	PUNCT
cana-6247	162	9	1−	1−	NUM
cana-6247	162	10	ᾱt(c))i	ᾱt(c))i	PROPN
cana-6247	162	11	)	)	PUNCT
cana-6247	162	12	,	,	PUNCT
cana-6247	162	13	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	162	14	984	984	NUM
cana-6247	162	15	communications	communication	NOUN
cana-6247	162	16	on	on	ADP
cana-6247	162	17	applied	apply	VERB
cana-6247	162	18	nonlinear	nonlinear	ADJ
cana-6247	162	19	analysis	analysis	NOUN
cana-6247	162	20	issn	issn	NOUN
cana-6247	162	21	:	:	PUNCT
cana-6247	162	22	1074	1074	NUM
cana-6247	162	23	-	-	PUNCT
cana-6247	162	24	133x	133x	NUM
cana-6247	162	25	vol	vol	NOUN
cana-6247	162	26	.	.	PROPN
cana-6247	163	1	31	31	NUM
cana-6247	164	1	no	no	NOUN
cana-6247	164	2	.	.	PUNCT
cana-6247	165	1	7s	7	NOUN
cana-6247	165	2	(	(	PUNCT
cana-6247	165	3	2024	2024	NUM
cana-6247	165	4	)	)	PUNCT
cana-6247	165	5	or	or	CCONJ
cana-6247	165	6	otherwise	otherwise	ADV
cana-6247	165	7	,	,	PUNCT
cana-6247	165	8	a	a	DET
cana-6247	165	9	more	more	ADV
cana-6247	165	10	general	general	ADJ
cana-6247	165	11	distribution	distribution	NOUN
cana-6247	165	12	depending	depend	VERB
cana-6247	165	13	on	on	ADP
cana-6247	165	14	the	the	DET
cana-6247	165	15	noise	noise	NOUN
cana-6247	165	16	formulation	formulation	NOUN
cana-6247	165	17	.	.	PUNCT
cana-6247	166	1	the	the	DET
cana-6247	166	2	reverse	reverse	ADJ
cana-6247	166	3	process	process	NOUN
cana-6247	166	4	must	must	AUX
cana-6247	166	5	then	then	ADV
cana-6247	166	6	learn	learn	VERB
cana-6247	166	7	to	to	PART
cana-6247	166	8	denoise	denoise	VERB
cana-6247	166	9	with	with	ADP
cana-6247	166	10	respect	respect	NOUN
cana-6247	166	11	to	to	ADP
cana-6247	166	12	the	the	DET
cana-6247	166	13	condition	condition	NOUN
cana-6247	166	14	,	,	PUNCT
cana-6247	166	15	which	which	PRON
cana-6247	166	16	is	be	AUX
cana-6247	166	17	often	often	ADV
cana-6247	166	18	implemented	implement	VERB
cana-6247	166	19	by	by	ADP
cana-6247	166	20	either	either	CCONJ
cana-6247	166	21	concatenating	concatenate	VERB
cana-6247	166	22	the	the	DET
cana-6247	166	23	condition	condition	NOUN
cana-6247	166	24	c	c	NOUN
cana-6247	166	25	with	with	ADP
cana-6247	166	26	xt	xt	PROPN
cana-6247	166	27	as	as	ADP
cana-6247	166	28	input	input	NOUN
cana-6247	166	29	to	to	ADP
cana-6247	166	30	the	the	DET
cana-6247	166	31	denoising	denoising	NOUN
cana-6247	166	32	network	network	NOUN
cana-6247	166	33	,	,	PUNCT
cana-6247	166	34	or	or	CCONJ
cana-6247	166	35	by	by	ADP
cana-6247	166	36	incorporating	incorporate	VERB
cana-6247	166	37	attention	attention	NOUN
cana-6247	166	38	mechanisms	mechanism	NOUN
cana-6247	166	39	to	to	PART
cana-6247	166	40	modulate	modulate	VERB
cana-6247	166	41	the	the	DET
cana-6247	166	42	denoising	denoising	NOUN
cana-6247	166	43	process	process	NOUN
cana-6247	166	44	based	base	VERB
cana-6247	166	45	on	on	ADP
cana-6247	166	46	c.	c.	PROPN
cana-6247	166	47	a	a	DET
cana-6247	166	48	conceptual	conceptual	ADJ
cana-6247	166	49	example	example	NOUN
cana-6247	166	50	can	can	AUX
cana-6247	166	51	be	be	AUX
cana-6247	166	52	used	use	VERB
cana-6247	166	53	to	to	PART
cana-6247	166	54	illustrate	illustrate	VERB
cana-6247	166	55	this	this	DET
cana-6247	166	56	approach	approach	NOUN
cana-6247	166	57	:	:	PUNCT
cana-6247	166	58	suppose	suppose	VERB
cana-6247	166	59	the	the	DET
cana-6247	166	60	task	task	NOUN
cana-6247	166	61	is	be	AUX
cana-6247	166	62	image	image	NOUN
cana-6247	166	63	inpainting	inpainte	VERB
cana-6247	166	64	,	,	PUNCT
cana-6247	166	65	with	with	ADP
cana-6247	166	66	the	the	DET
cana-6247	166	67	mask	mask	NOUN
cana-6247	166	68	being	be	AUX
cana-6247	166	69	the	the	DET
cana-6247	166	70	condition	condition	NOUN
cana-6247	166	71	.	.	PUNCT
cana-6247	167	1	in	in	ADP
cana-6247	167	2	standard	standard	ADJ
cana-6247	167	3	diffusion	diffusion	NOUN
cana-6247	167	4	processes	process	NOUN
cana-6247	167	5	,	,	PUNCT
cana-6247	167	6	noise	noise	NOUN
cana-6247	167	7	is	be	AUX
cana-6247	167	8	generated	generate	VERB
cana-6247	167	9	uniformly	uniformly	ADV
cana-6247	167	10	everywhere	everywhere	ADV
cana-6247	167	11	,	,	PUNCT
cana-6247	167	12	sometimes	sometimes	ADV
cana-6247	167	13	with	with	ADP
cana-6247	167	14	the	the	DET
cana-6247	167	15	consequence	consequence	NOUN
cana-6247	167	16	of	of	ADP
cana-6247	167	17	destroying	destroy	VERB
cana-6247	167	18	useful	useful	ADJ
cana-6247	167	19	information	information	NOUN
cana-6247	167	20	even	even	ADV
cana-6247	167	21	in	in	ADP
cana-6247	167	22	known	know	VERB
cana-6247	167	23	areas	area	NOUN
cana-6247	167	24	.	.	PUNCT
cana-6247	168	1	with	with	ADP
cana-6247	168	2	conditional	conditional	ADJ
cana-6247	168	3	noise	noise	NOUN
cana-6247	168	4	injection	injection	NOUN
cana-6247	168	5	,	,	PUNCT
cana-6247	168	6	the	the	DET
cana-6247	168	7	model	model	NOUN
cana-6247	168	8	injects	inject	VERB
cana-6247	168	9	more	more	ADJ
cana-6247	168	10	noise	noise	NOUN
cana-6247	168	11	in	in	ADP
cana-6247	168	12	masked	mask	VERB
cana-6247	168	13	(	(	PUNCT
cana-6247	168	14	unknown	unknown	ADJ
cana-6247	168	15	)	)	PUNCT
cana-6247	168	16	regions	region	NOUN
cana-6247	168	17	and	and	CCONJ
cana-6247	168	18	less	less	ADJ
cana-6247	168	19	in	in	ADP
cana-6247	168	20	unmasked	unmasked	ADJ
cana-6247	168	21	(	(	PUNCT
cana-6247	168	22	known	known	ADJ
cana-6247	168	23	)	)	PUNCT
cana-6247	168	24	regions	region	NOUN
cana-6247	168	25	.	.	PUNCT
cana-6247	169	1	during	during	ADP
cana-6247	169	2	denoising	denoising	NOUN
cana-6247	169	3	,	,	PUNCT
cana-6247	169	4	the	the	DET
cana-6247	169	5	model	model	NOUN
cana-6247	169	6	learns	learn	VERB
cana-6247	169	7	to	to	PART
cana-6247	169	8	dedicate	dedicate	VERB
cana-6247	169	9	its	its	PRON
cana-6247	169	10	generative	generative	ADJ
cana-6247	169	11	capacity	capacity	NOUN
cana-6247	169	12	to	to	ADP
cana-6247	169	13	reconstructing	reconstruct	VERB
cana-6247	169	14	missing	miss	VERB
cana-6247	169	15	regions	region	NOUN
cana-6247	169	16	while	while	SCONJ
cana-6247	169	17	keeping	keep	VERB
cana-6247	169	18	intact	intact	ADJ
cana-6247	169	19	the	the	DET
cana-6247	169	20	areas	area	NOUN
cana-6247	169	21	treated	treat	VERB
cana-6247	169	22	as	as	ADP
cana-6247	169	23	known	know	VERB
cana-6247	169	24	originally	originally	ADV
cana-6247	169	25	.	.	PUNCT
cana-6247	170	1	hence	hence	ADV
cana-6247	170	2	,	,	PUNCT
cana-6247	170	3	the	the	DET
cana-6247	170	4	noise	noise	NOUN
cana-6247	170	5	map	map	NOUN
cana-6247	170	6	encodes	encodes	PROPN
cana-6247	170	7	,	,	PUNCT
cana-6247	170	8	from	from	ADP
cana-6247	170	9	the	the	DET
cana-6247	170	10	onset	onset	NOUN
cana-6247	170	11	,	,	PUNCT
cana-6247	170	12	where	where	SCONJ
cana-6247	170	13	restoration	restoration	NOUN
cana-6247	170	14	is	be	AUX
cana-6247	170	15	required	require	VERB
cana-6247	170	16	and	and	CCONJ
cana-6247	170	17	guides	guide	VERB
cana-6247	170	18	the	the	DET
cana-6247	170	19	reverse	reverse	ADJ
cana-6247	170	20	process	process	NOUN
cana-6247	170	21	.	.	PUNCT
cana-6247	171	1	figure	figure	VERB
cana-6247	171	2	3	3	NUM
cana-6247	171	3	:	:	PUNCT
cana-6247	171	4	forward	forward	ADJ
cana-6247	171	5	diffusion	diffusion	NOUN
cana-6247	171	6	process	process	NOUN
cana-6247	171	7	in	in	ADP
cana-6247	171	8	conditional	conditional	ADJ
cana-6247	171	9	noise	noise	NOUN
cana-6247	171	10	injection	injection	NOUN
cana-6247	171	11	type	type	NOUN
cana-6247	171	12	diffusion	diffusion	NOUN
cana-6247	171	13	model	model	NOUN
cana-6247	171	14	2.1	2.1	NUM
cana-6247	171	15	image	image	NOUN
cana-6247	171	16	inpainting	inpainte	VERB
cana-6247	171	17	/	/	SYM
cana-6247	171	18	outpainting	outpainte	VERB
cana-6247	171	19	cni	cni	NOUN
cana-6247	171	20	is	be	AUX
cana-6247	171	21	the	the	DET
cana-6247	171	22	most	most	ADV
cana-6247	171	23	favorable	favorable	ADJ
cana-6247	171	24	noise	noise	NOUN
cana-6247	171	25	addition	addition	NOUN
cana-6247	171	26	mode	mode	NOUN
cana-6247	171	27	(	(	PUNCT
cana-6247	171	28	when	when	SCONJ
cana-6247	171	29	image	image	NOUN
cana-6247	171	30	corruption	corruption	NOUN
cana-6247	171	31	is	be	AUX
cana-6247	171	32	performed	perform	VERB
cana-6247	171	33	)	)	PUNCT
cana-6247	171	34	,	,	PUNCT
cana-6247	171	35	for	for	ADP
cana-6247	171	36	image	image	NOUN
cana-6247	171	37	inpainting	inpainte	VERB
cana-6247	171	38	/	/	SYM
cana-6247	171	39	outpainting	outpainte	VERB
cana-6247	171	40	,	,	PUNCT
cana-6247	171	41	as	as	SCONJ
cana-6247	171	42	it	it	PRON
cana-6247	171	43	closely	closely	ADV
cana-6247	171	44	links	link	VERB
cana-6247	171	45	the	the	DET
cana-6247	171	46	forward	forward	ADJ
cana-6247	171	47	corruption	corruption	NOUN
cana-6247	171	48	process	process	NOUN
cana-6247	171	49	to	to	ADP
cana-6247	171	50	the	the	DET
cana-6247	171	51	image	image	NOUN
cana-6247	171	52	’s	’s	PART
cana-6247	171	53	structure	structure	NOUN
cana-6247	171	54	,	,	PUNCT
cana-6247	171	55	mask	mask	NOUN
cana-6247	171	56	,	,	PUNCT
cana-6247	171	57	or	or	CCONJ
cana-6247	171	58	context	context	NOUN
cana-6247	171	59	.	.	PUNCT
cana-6247	172	1	such	such	ADJ
cana-6247	172	2	combined	combined	ADJ
cana-6247	172	3	advantage	advantage	NOUN
cana-6247	172	4	brings	bring	VERB
cana-6247	172	5	great	great	ADJ
cana-6247	172	6	enhancement	enhancement	NOUN
cana-6247	172	7	in	in	ADP
cana-6247	172	8	visual	visual	ADJ
cana-6247	172	9	consistency	consistency	NOUN
cana-6247	172	10	,	,	PUNCT
cana-6247	172	11	semantic	semantic	ADJ
cana-6247	172	12	consistency	consistency	NOUN
cana-6247	172	13	and	and	CCONJ
cana-6247	172	14	task	task	NOUN
cana-6247	172	15	-	-	PUNCT
cana-6247	172	16	agnostic	agnostic	ADJ
cana-6247	172	17	property	property	NOUN
cana-6247	172	18	.	.	PUNCT
cana-6247	173	1	[	[	X
cana-6247	173	2	39	39	NUM
cana-6247	173	3	]	]	PUNCT
cana-6247	174	1	[	[	X
cana-6247	174	2	36	36	NUM
cana-6247	174	3	]	]	X
cana-6247	175	1	[	[	X
cana-6247	175	2	48	48	NUM
cana-6247	175	3	]	]	X
cana-6247	175	4	[	[	X
cana-6247	175	5	8	8	X
cana-6247	175	6	]	]	PUNCT
cana-6247	175	7	the	the	DET
cana-6247	175	8	principal	principal	ADJ
cana-6247	175	9	theoretical	theoretical	ADJ
cana-6247	175	10	benefit	benefit	NOUN
cana-6247	175	11	of	of	ADP
cana-6247	175	12	cni	cni	PROPN
cana-6247	175	13	is	be	AUX
cana-6247	175	14	the	the	DET
cana-6247	175	15	ability	ability	NOUN
cana-6247	175	16	to	to	PART
cana-6247	175	17	schedule	schedule	VERB
cana-6247	175	18	the	the	DET
cana-6247	175	19	noise	noise	NOUN
cana-6247	175	20	region	region	NOUN
cana-6247	175	21	-	-	PUNCT
cana-6247	175	22	wise	wise	ADJ
cana-6247	175	23	or	or	CCONJ
cana-6247	175	24	in	in	ADP
cana-6247	175	25	context	context	NOUN
cana-6247	175	26	:	:	PUNCT
cana-6247	175	27	•	•	NOUN
cana-6247	175	28	in	in	ADP
cana-6247	175	29	inpainting	inpainte	VERB
cana-6247	175	30	,	,	PUNCT
cana-6247	175	31	cni	cni	PROPN
cana-6247	175	32	adds	add	VERB
cana-6247	175	33	noise	noise	NOUN
cana-6247	175	34	to	to	ADP
cana-6247	175	35	the	the	DET
cana-6247	175	36	masked	masked	ADJ
cana-6247	175	37	or	or	CCONJ
cana-6247	175	38	missing	missing	ADJ
cana-6247	175	39	pixels	pixel	NOUN
cana-6247	175	40	while	while	SCONJ
cana-6247	175	41	preserving	preserve	VERB
cana-6247	175	42	detail	detail	NOUN
cana-6247	175	43	and	and	CCONJ
cana-6247	175	44	limiting	limit	VERB
cana-6247	175	45	distortion	distortion	NOUN
cana-6247	175	46	in	in	ADP
cana-6247	175	47	non	non	ADJ
cana-6247	175	48	-	-	ADJ
cana-6247	175	49	spanned	spanned	ADJ
cana-6247	175	50	regions	region	NOUN
cana-6247	175	51	.	.	PUNCT
cana-6247	176	1	this	this	DET
cana-6247	176	2	strategy	strategy	NOUN
cana-6247	176	3	avoids	avoid	VERB
cana-6247	176	4	the	the	DET
cana-6247	176	5	model	model	NOUN
cana-6247	176	6	wasting	waste	VERB
cana-6247	176	7	its	its	PRON
cana-6247	176	8	capacity	capacity	NOUN
cana-6247	176	9	and	and	CCONJ
cana-6247	176	10	resources	resource	NOUN
cana-6247	176	11	trying	try	VERB
cana-6247	176	12	to	to	PART
cana-6247	176	13	reconstruct	reconstruct	VERB
cana-6247	176	14	regions	region	NOUN
cana-6247	176	15	that	that	PRON
cana-6247	176	16	do	do	AUX
cana-6247	176	17	not	not	PART
cana-6247	176	18	need	need	VERB
cana-6247	176	19	to	to	PART
cana-6247	176	20	change	change	VERB
cana-6247	176	21	,	,	PUNCT
cana-6247	176	22	and	and	CCONJ
cana-6247	176	23	instead	instead	ADV
cana-6247	176	24	focuses	focus	VERB
cana-6247	176	25	the	the	DET
cana-6247	176	26	generative	generative	ADJ
cana-6247	176	27	effort	effort	NOUN
cana-6247	176	28	and	and	CCONJ
cana-6247	176	29	diversity	diversity	NOUN
cana-6247	176	30	into	into	ADP
cana-6247	176	31	the	the	DET
cana-6247	176	32	inpainted	inpainte	VERB
cana-6247	176	33	areas	area	NOUN
cana-6247	176	34	.	.	PUNCT
cana-6247	177	1	[	[	X
cana-6247	177	2	39	39	NUM
cana-6247	177	3	]	]	PUNCT
cana-6247	177	4	[	[	X
cana-6247	177	5	9	9	NUM
cana-6247	177	6	]	]	SYM
cana-6247	177	7	•	•	NOUN
cana-6247	177	8	in	in	ADP
cana-6247	177	9	multimodal	multimodal	NOUN
cana-6247	177	10	and	and	CCONJ
cana-6247	177	11	layout	layout	NOUN
cana-6247	177	12	-	-	PUNCT
cana-6247	177	13	aware	aware	ADJ
cana-6247	177	14	tasks	task	NOUN
cana-6247	177	15	,	,	PUNCT
cana-6247	177	16	noise	noise	NOUN
cana-6247	177	17	maps	map	NOUN
cana-6247	177	18	are	be	AUX
cana-6247	177	19	divided	divide	VERB
cana-6247	177	20	,	,	PUNCT
cana-6247	177	21	cropped	cropped	ADJ
cana-6247	177	22	or	or	CCONJ
cana-6247	177	23	combined	combine	VERB
cana-6247	177	24	based	base	VERB
cana-6247	177	25	on	on	ADP
cana-6247	177	26	object	object	NOUN
cana-6247	177	27	masks	mask	NOUN
cana-6247	177	28	or	or	CCONJ
cana-6247	177	29	layout	layout	NOUN
cana-6247	177	30	commands	command	NOUN
cana-6247	177	31	,	,	PUNCT
cana-6247	177	32	resulting	result	VERB
cana-6247	177	33	in	in	ADP
cana-6247	177	34	the	the	DET
cana-6247	177	35	model	model	NOUN
cana-6247	177	36	being	be	AUX
cana-6247	177	37	provided	provide	VERB
cana-6247	177	38	with	with	ADP
cana-6247	177	39	explicit	explicit	ADJ
cana-6247	177	40	per	per	ADP
cana-6247	177	41	-	-	PUNCT
cana-6247	177	42	object	object	NOUN
cana-6247	177	43	or	or	CCONJ
cana-6247	177	44	per	per	ADP
cana-6247	177	45	-	-	PUNCT
cana-6247	177	46	region	region	NOUN
cana-6247	177	47	stochasticity	stochasticity	NOUN
cana-6247	177	48	.	.	PUNCT
cana-6247	178	1	this	this	PRON
cana-6247	178	2	effectively	effectively	ADV
cana-6247	178	3	disentangles	disentangle	VERB
cana-6247	178	4	object	object	NOUN
cana-6247	178	5	attributes	attribute	NOUN
cana-6247	178	6	,	,	PUNCT
cana-6247	178	7	eliminates	eliminate	VERB
cana-6247	178	8	blending	blend	VERB
cana-6247	178	9	artifacts	artifact	NOUN
cana-6247	178	10	,	,	PUNCT
cana-6247	178	11	and	and	CCONJ
cana-6247	178	12	enables	enable	VERB
cana-6247	178	13	spatially	spatially	ADV
cana-6247	178	14	selective	selective	ADJ
cana-6247	178	15	generation	generation	NOUN
cana-6247	178	16	.	.	PUNCT
cana-6247	179	1	[	[	X
cana-6247	179	2	36	36	NUM
cana-6247	179	3	]	]	SYM
cana-6247	179	4	•	•	NOUN
cana-6247	179	5	joint	joint	ADJ
cana-6247	179	6	conditioning	conditioning	NOUN
cana-6247	179	7	with	with	ADP
cana-6247	179	8	degradation	degradation	NOUN
cana-6247	179	9	maps	map	NOUN
cana-6247	179	10	,	,	PUNCT
cana-6247	179	11	reference	reference	NOUN
cana-6247	179	12	guidance	guidance	NOUN
cana-6247	179	13	and	and	CCONJ
cana-6247	179	14	user	user	NOUN
cana-6247	179	15	input	input	NOUN
cana-6247	179	16	facilitates	facilitate	VERB
cana-6247	179	17	the	the	DET
cana-6247	179	18	setting	setting	NOUN
cana-6247	179	19	of	of	ADP
cana-6247	179	20	the	the	DET
cana-6247	179	21	magnitude	magnitude	NOUN
cana-6247	179	22	and	and	CCONJ
cana-6247	179	23	the	the	DET
cana-6247	179	24	structure	structure	NOUN
cana-6247	179	25	of	of	ADP
cana-6247	179	26	the	the	DET
cana-6247	179	27	noise	noise	NOUN
cana-6247	179	28	,	,	PUNCT
cana-6247	179	29	making	make	VERB
cana-6247	179	30	the	the	DET
cana-6247	179	31	model	model	NOUN
cana-6247	179	32	more	more	ADV
cana-6247	179	33	optimal	optimal	ADJ
cana-6247	179	34	for	for	ADP
cana-6247	179	35	occlusion	occlusion	NOUN
cana-6247	179	36	handling	handling	NOUN
cana-6247	179	37	in	in	ADP
cana-6247	179	38	the	the	DET
cana-6247	179	39	real	real	ADJ
cana-6247	179	40	-	-	PUNCT
cana-6247	179	41	world	world	NOUN
cana-6247	179	42	mixed	mixed	ADJ
cana-6247	179	43	degradation	degradation	NOUN
cana-6247	179	44	or	or	CCONJ
cana-6247	179	45	multi	multi	ADJ
cana-6247	179	46	-	-	ADJ
cana-6247	179	47	modal	modal	ADJ
cana-6247	179	48	cases	case	NOUN
cana-6247	179	49	.	.	PUNCT
cana-6247	180	1	[	[	X
cana-6247	180	2	48	48	NUM
cana-6247	180	3	]	]	X
cana-6247	180	4	[	[	X
cana-6247	180	5	31	31	NUM
cana-6247	180	6	]	]	PUNCT
cana-6247	180	7	mathematically	mathematically	ADV
cana-6247	180	8	,	,	PUNCT
cana-6247	180	9	the	the	DET
cana-6247	180	10	diffusion	diffusion	NOUN
cana-6247	180	11	process	process	NOUN
cana-6247	180	12	is	be	AUX
cana-6247	180	13	represented	represent	VERB
cana-6247	180	14	as	as	ADP
cana-6247	180	15	:	:	PUNCT
cana-6247	180	16	xt	xt	PROPN
cana-6247	180	17	=	=	SYM
cana-6247	180	18	√	√	PROPN
cana-6247	180	19	ᾱt	ᾱt	VERB
cana-6247	180	20	x0	x0	PROPN
cana-6247	181	1	+	+	CCONJ
cana-6247	181	2	√	√	PROPN
cana-6247	181	3	1−	1−	NUM
cana-6247	182	1	ᾱt	ᾱt	VERB
cana-6247	182	2	z	z	PROPN
cana-6247	182	3	,	,	PUNCT
cana-6247	182	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	182	5	985	985	NUM
cana-6247	182	6	communications	communication	NOUN
cana-6247	182	7	on	on	ADP
cana-6247	182	8	applied	apply	VERB
cana-6247	182	9	nonlinear	nonlinear	ADJ
cana-6247	182	10	analysis	analysis	NOUN
cana-6247	182	11	issn	issn	NOUN
cana-6247	182	12	:	:	PUNCT
cana-6247	182	13	1074	1074	NUM
cana-6247	182	14	-	-	PUNCT
cana-6247	182	15	133x	133x	NUM
cana-6247	182	16	vol	vol	NOUN
cana-6247	182	17	.	.	PROPN
cana-6247	183	1	31	31	NUM
cana-6247	184	1	no	no	NOUN
cana-6247	184	2	.	.	PUNCT
cana-6247	185	1	7s	7	NOUN
cana-6247	185	2	(	(	PUNCT
cana-6247	185	3	2024	2024	NUM
cana-6247	185	4	)	)	PUNCT
cana-6247	185	5	where	where	SCONJ
cana-6247	185	6	classical	classical	ADJ
cana-6247	185	7	models	model	NOUN
cana-6247	185	8	assume	assume	VERB
cana-6247	185	9	z	z	NOUN
cana-6247	185	10	∼	∼	NOUN
cana-6247	185	11	n	n	CCONJ
cana-6247	185	12	(	(	PUNCT
cana-6247	185	13	0	0	NUM
cana-6247	185	14	,	,	PUNCT
cana-6247	185	15	i	i	NOUN
cana-6247	185	16	)	)	PUNCT
cana-6247	185	17	as	as	ADP
cana-6247	185	18	standard	standard	ADJ
cana-6247	185	19	gaussian	gaussian	ADJ
cana-6247	185	20	noise	noise	NOUN
cana-6247	185	21	.	.	PUNCT
cana-6247	186	1	in	in	ADP
cana-6247	186	2	the	the	DET
cana-6247	186	3	case	case	NOUN
cana-6247	186	4	of	of	ADP
cana-6247	186	5	conditional	conditional	ADJ
cana-6247	186	6	noise	noise	NOUN
cana-6247	186	7	injection	injection	NOUN
cana-6247	186	8	(	(	PUNCT
cana-6247	186	9	cni	cni	PROPN
cana-6247	186	10	)	)	PUNCT
cana-6247	186	11	,	,	PUNCT
cana-6247	186	12	both	both	CCONJ
cana-6247	186	13	the	the	DET
cana-6247	186	14	noise	noise	NOUN
cana-6247	186	15	sample	sample	NOUN
cana-6247	186	16	z	z	NOUN
cana-6247	186	17	and	and	CCONJ
cana-6247	186	18	the	the	DET
cana-6247	186	19	variance	variance	NOUN
cana-6247	186	20	schedule	schedule	NOUN
cana-6247	186	21	βt	βt	AUX
cana-6247	186	22	become	become	VERB
cana-6247	186	23	functions	function	NOUN
cana-6247	186	24	of	of	ADP
cana-6247	186	25	a	a	DET
cana-6247	186	26	conditional	conditional	ADJ
cana-6247	186	27	signal	signal	NOUN
cana-6247	186	28	c	c	PROPN
cana-6247	186	29	,	,	PUNCT
cana-6247	186	30	which	which	PRON
cana-6247	186	31	may	may	AUX
cana-6247	186	32	represent	represent	VERB
cana-6247	186	33	a	a	DET
cana-6247	186	34	mask	mask	NOUN
cana-6247	186	35	,	,	PUNCT
cana-6247	186	36	the	the	DET
cana-6247	186	37	observed	observed	ADJ
cana-6247	186	38	context	context	NOUN
cana-6247	186	39	,	,	PUNCT
cana-6247	186	40	or	or	CCONJ
cana-6247	186	41	semantic	semantic	ADJ
cana-6247	186	42	maps	map	NOUN
cana-6247	186	43	.	.	PUNCT
cana-6247	187	1	this	this	PRON
cana-6247	187	2	allows	allow	VERB
cana-6247	187	3	for	for	ADP
cana-6247	187	4	spatially	spatially	ADV
cana-6247	187	5	-	-	PUNCT
cana-6247	187	6	local	local	ADJ
cana-6247	187	7	or	or	CCONJ
cana-6247	187	8	structure	structure	NOUN
cana-6247	187	9	-	-	PUNCT
cana-6247	187	10	aware	aware	ADJ
cana-6247	187	11	corruption	corruption	NOUN
cana-6247	187	12	that	that	PRON
cana-6247	187	13	effectively	effectively	ADV
cana-6247	187	14	primes	prime	VERB
cana-6247	187	15	the	the	DET
cana-6247	187	16	reverse	reverse	ADJ
cana-6247	187	17	process	process	NOUN
cana-6247	187	18	to	to	PART
cana-6247	187	19	generate	generate	VERB
cana-6247	187	20	content	content	NOUN
cana-6247	187	21	that	that	PRON
cana-6247	187	22	is	be	AUX
cana-6247	187	23	contextually	contextually	ADV
cana-6247	187	24	plausible	plausible	ADJ
cana-6247	187	25	and	and	CCONJ
cana-6247	187	26	consistent	consistent	ADJ
cana-6247	187	27	across	across	ADP
cana-6247	187	28	boundaries	boundary	NOUN
cana-6247	187	29	.	.	PUNCT
cana-6247	188	1	[	[	X
cana-6247	188	2	36	36	NUM
cana-6247	188	3	]	]	X
cana-6247	188	4	[	[	X
cana-6247	188	5	39]colour	39]colour	NUM
cana-6247	188	6	in	in	ADP
cana-6247	188	7	practice	practice	NOUN
cana-6247	188	8	,	,	PUNCT
cana-6247	188	9	conditional	conditional	ADJ
cana-6247	188	10	noise	noise	NOUN
cana-6247	188	11	injection	injection	NOUN
cana-6247	188	12	(	(	PUNCT
cana-6247	188	13	cni	cni	PROPN
cana-6247	188	14	)	)	PUNCT
cana-6247	188	15	yields	yield	VERB
cana-6247	188	16	substantial	substantial	ADJ
cana-6247	188	17	benefits	benefit	NOUN
cana-6247	188	18	:	:	PUNCT
cana-6247	188	19	•	•	NUM
cana-6247	188	20	more	more	ADV
cana-6247	188	21	accurate	accurate	ADJ
cana-6247	188	22	semantic	semantic	ADJ
cana-6247	188	23	blending	blending	NOUN
cana-6247	188	24	with	with	ADP
cana-6247	188	25	fewer	few	ADJ
cana-6247	188	26	artifacts	artifact	NOUN
cana-6247	188	27	:	:	PUNCT
cana-6247	188	28	cni	cni	ADJ
cana-6247	188	29	couples	couple	NOUN
cana-6247	188	30	noise	noise	VERB
cana-6247	188	31	patterns	pattern	NOUN
cana-6247	188	32	to	to	ADP
cana-6247	188	33	the	the	DET
cana-6247	188	34	editable	editable	ADJ
cana-6247	188	35	regions	region	NOUN
cana-6247	188	36	,	,	PUNCT
cana-6247	188	37	helping	help	VERB
cana-6247	188	38	the	the	DET
cana-6247	188	39	model	model	NOUN
cana-6247	188	40	generate	generate	VERB
cana-6247	188	41	color	color	NOUN
cana-6247	188	42	,	,	PUNCT
cana-6247	188	43	gradients	gradient	NOUN
cana-6247	188	44	,	,	PUNCT
cana-6247	188	45	and	and	CCONJ
cana-6247	188	46	textures	texture	NOUN
cana-6247	188	47	that	that	PRON
cana-6247	188	48	are	be	AUX
cana-6247	188	49	coherent	coherent	ADJ
cana-6247	188	50	with	with	ADP
cana-6247	188	51	the	the	DET
cana-6247	188	52	surrounding	surround	VERB
cana-6247	188	53	context	context	NOUN
cana-6247	188	54	.	.	PUNCT
cana-6247	189	1	[	[	X
cana-6247	189	2	9	9	NUM
cana-6247	189	3	]	]	SYM
cana-6247	189	4	•	•	NOUN
cana-6247	189	5	adaptable	adaptable	ADJ
cana-6247	189	6	and	and	CCONJ
cana-6247	189	7	flexible	flexible	ADJ
cana-6247	189	8	editing	editing	NOUN
cana-6247	189	9	:	:	PUNCT
cana-6247	189	10	our	our	PRON
cana-6247	189	11	multimodal	multimodal	ADJ
cana-6247	189	12	editors	editor	NOUN
cana-6247	189	13	can	can	AUX
cana-6247	189	14	demonstrate	demonstrate	VERB
cana-6247	189	15	precise	precise	ADJ
cana-6247	189	16	object	object	NOUN
cana-6247	189	17	-	-	PUNCT
cana-6247	189	18	level	level	NOUN
cana-6247	189	19	control	control	NOUN
cana-6247	189	20	by	by	ADP
cana-6247	189	21	biasing	bias	VERB
cana-6247	189	22	noise	noise	NOUN
cana-6247	189	23	between	between	ADP
cana-6247	189	24	shape	shape	NOUN
cana-6247	189	25	and	and	CCONJ
cana-6247	189	26	appearance	appearance	NOUN
cana-6247	189	27	.	.	PUNCT
cana-6247	190	1	this	this	PRON
cana-6247	190	2	enables	enable	VERB
cana-6247	190	3	robust	robust	ADJ
cana-6247	190	4	strategies	strategy	NOUN
cana-6247	190	5	for	for	ADP
cana-6247	190	6	outpainting	outpainte	VERB
cana-6247	190	7	,	,	PUNCT
cana-6247	190	8	recoloring	recoloring	NOUN
cana-6247	190	9	,	,	PUNCT
cana-6247	190	10	and	and	CCONJ
cana-6247	190	11	inpainting	inpainte	VERB
cana-6247	190	12	—	—	PUNCT
cana-6247	190	13	all	all	PRON
cana-6247	190	14	within	within	ADP
cana-6247	190	15	a	a	DET
cana-6247	190	16	single	single	ADJ
cana-6247	190	17	,	,	PUNCT
cana-6247	190	18	generalizable	generalizable	ADJ
cana-6247	190	19	model	model	NOUN
cana-6247	190	20	.	.	PUNCT
cana-6247	191	1	[	[	X
cana-6247	191	2	8	8	NUM
cana-6247	191	3	]	]	X
cana-6247	191	4	[	[	X
cana-6247	191	5	32	32	NUM
cana-6247	191	6	]	]	SYM
cana-6247	191	7	•	•	ADV
cana-6247	191	8	better	well	ADJ
cana-6247	191	9	boundary	boundary	ADJ
cana-6247	191	10	integrity	integrity	NOUN
cana-6247	191	11	:	:	PUNCT
cana-6247	191	12	edge	edge	NOUN
cana-6247	191	13	and	and	CCONJ
cana-6247	191	14	context	context	NOUN
cana-6247	191	15	-	-	PUNCT
cana-6247	191	16	sensitive	sensitive	ADJ
cana-6247	191	17	noise	noise	NOUN
cana-6247	191	18	assignment	assignment	NOUN
cana-6247	191	19	helps	help	VERB
cana-6247	191	20	diffusion	diffusion	NOUN
cana-6247	191	21	models	model	NOUN
cana-6247	191	22	not	not	PART
cana-6247	191	23	to	to	PART
cana-6247	191	24	"	"	PUNCT
cana-6247	191	25	bleed	bleed	VERB
cana-6247	191	26	"	"	PUNCT
cana-6247	191	27	information	information	NOUN
cana-6247	191	28	across	across	ADP
cana-6247	191	29	mask	mask	NOUN
cana-6247	191	30	/	/	SYM
cana-6247	191	31	boundaries	boundary	NOUN
cana-6247	191	32	,	,	PUNCT
cana-6247	191	33	leading	lead	VERB
cana-6247	191	34	to	to	ADP
cana-6247	191	35	more	more	ADV
cana-6247	191	36	realistic	realistic	ADJ
cana-6247	191	37	inpainted	inpainte	VERB
cana-6247	191	38	and	and	CCONJ
cana-6247	191	39	outpainted	outpainte	VERB
cana-6247	191	40	regions	region	NOUN
cana-6247	191	41	.	.	PUNCT
cana-6247	192	1	[	[	X
cana-6247	192	2	39	39	NUM
cana-6247	192	3	]	]	PUNCT
cana-6247	192	4	[	[	X
cana-6247	192	5	9	9	NUM
cana-6247	192	6	]	]	SYM
cana-6247	192	7	2.1.1	2.1.1	NUM
cana-6247	192	8	evidence	evidence	NOUN
cana-6247	192	9	from	from	ADP
cana-6247	192	10	recent	recent	ADJ
cana-6247	192	11	research	research	NOUN
cana-6247	192	12	several	several	ADJ
cana-6247	192	13	recent	recent	ADJ
cana-6247	192	14	works	work	NOUN
cana-6247	192	15	have	have	AUX
cana-6247	192	16	explored	explore	VERB
cana-6247	192	17	the	the	DET
cana-6247	192	18	use	use	NOUN
cana-6247	192	19	of	of	ADP
cana-6247	192	20	conditional	conditional	ADJ
cana-6247	192	21	noise	noise	NOUN
cana-6247	192	22	injection	injection	NOUN
cana-6247	192	23	(	(	PUNCT
cana-6247	192	24	cni	cni	ADJ
cana-6247	192	25	):	):	PUNCT
cana-6247	192	26	•	•	NUM
cana-6247	192	27	inpdiffusion	inpdiffusion	NOUN
cana-6247	192	28	proposed	propose	VERB
cana-6247	192	29	an	an	DET
cana-6247	192	30	adaptive	adaptive	ADJ
cana-6247	192	31	conditional	conditional	ADJ
cana-6247	192	32	network	network	NOUN
cana-6247	192	33	to	to	PART
cana-6247	192	34	extract	extract	VERB
cana-6247	192	35	semantic	semantic	ADJ
cana-6247	192	36	edge	edge	NOUN
cana-6247	192	37	features	feature	NOUN
cana-6247	192	38	,	,	PUNCT
cana-6247	192	39	and	and	CCONJ
cana-6247	192	40	utilizes	utilize	VERB
cana-6247	192	41	conditional	conditional	ADJ
cana-6247	192	42	noise	noise	NOUN
cana-6247	192	43	to	to	PART
cana-6247	192	44	selectively	selectively	ADV
cana-6247	192	45	guide	guide	VERB
cana-6247	192	46	the	the	DET
cana-6247	192	47	denoising	denoising	NOUN
cana-6247	192	48	inferred	infer	VERB
cana-6247	192	49	by	by	ADP
cana-6247	192	50	the	the	DET
cana-6247	192	51	masked	mask	VERB
cana-6247	192	52	regions	region	NOUN
cana-6247	192	53	of	of	ADP
cana-6247	192	54	the	the	DET
cana-6247	192	55	inputs	input	NOUN
cana-6247	192	56	,	,	PUNCT
cana-6247	192	57	thus	thus	ADV
cana-6247	192	58	achieving	achieve	VERB
cana-6247	192	59	better	well	ADJ
cana-6247	192	60	mask	mask	NOUN
cana-6247	192	61	refinement	refinement	NOUN
cana-6247	192	62	and	and	CCONJ
cana-6247	192	63	boundary	boundary	ADJ
cana-6247	192	64	fidelity	fidelity	NOUN
cana-6247	192	65	.	.	PUNCT
cana-6247	193	1	[	[	X
cana-6247	193	2	39	39	NUM
cana-6247	193	3	]	]	SYM
cana-6247	193	4	•	•	NUM
cana-6247	193	5	noisecollage	noisecollage	NOUN
cana-6247	193	6	extends	extend	VERB
cana-6247	193	7	crop	crop	NOUN
cana-6247	193	8	-	-	PUNCT
cana-6247	193	9	and	and	CCONJ
cana-6247	193	10	-	-	PUNCT
cana-6247	193	11	merge	merge	VERB
cana-6247	193	12	and	and	CCONJ
cana-6247	193	13	demonstrates	demonstrate	VERB
cana-6247	193	14	that	that	PRON
cana-6247	193	15	noise	noise	NOUN
cana-6247	193	16	patches	patch	NOUN
cana-6247	193	17	for	for	ADP
cana-6247	193	18	object	object	NOUN
cana-6247	193	19	and	and	CCONJ
cana-6247	193	20	background	background	NOUN
cana-6247	193	21	can	can	AUX
cana-6247	193	22	be	be	AUX
cana-6247	193	23	composed	compose	VERB
cana-6247	193	24	separately	separately	ADV
cana-6247	193	25	while	while	SCONJ
cana-6247	193	26	maintaining	maintain	VERB
cana-6247	193	27	parts	part	NOUN
cana-6247	193	28	of	of	ADP
cana-6247	193	29	stochasticity	stochasticity	NOUN
cana-6247	193	30	optimal	optimal	ADJ
cana-6247	193	31	for	for	ADP
cana-6247	193	32	a	a	DET
cana-6247	193	33	flexible	flexible	ADJ
cana-6247	193	34	layout	layout	NOUN
cana-6247	193	35	-	-	PUNCT
cana-6247	193	36	aware	aware	ADJ
cana-6247	193	37	image	image	NOUN
cana-6247	193	38	synthesis	synthesis	NOUN
cana-6247	193	39	.	.	PUNCT
cana-6247	194	1	[	[	X
cana-6247	194	2	36	36	NUM
cana-6247	194	3	]	]	SYM
cana-6247	194	4	•	•	NUM
cana-6247	194	5	gradpaint	gradpaint	NOUN
cana-6247	194	6	has	have	AUX
cana-6247	194	7	shown	show	VERB
cana-6247	194	8	that	that	SCONJ
cana-6247	194	9	the	the	DET
cana-6247	194	10	conventional	conventional	ADJ
cana-6247	194	11	uniform	uniform	NOUN
cana-6247	194	12	noise	noise	NOUN
cana-6247	194	13	degrades	degrade	VERB
cana-6247	194	14	the	the	DET
cana-6247	194	15	harmony	harmony	NOUN
cana-6247	194	16	between	between	ADP
cana-6247	194	17	the	the	DET
cana-6247	194	18	inpainted	inpainte	VERB
cana-6247	194	19	masks	mask	NOUN
cana-6247	194	20	and	and	CCONJ
cana-6247	194	21	their	their	PRON
cana-6247	194	22	context	context	NOUN
cana-6247	194	23	while	while	SCONJ
cana-6247	194	24	cni	cni	PROPN
cana-6247	194	25	constrains	constrain	VERB
cana-6247	194	26	the	the	DET
cana-6247	194	27	global	global	ADJ
cana-6247	194	28	and	and	CCONJ
cana-6247	194	29	local	local	ADJ
cana-6247	194	30	consistencies	consistency	NOUN
cana-6247	194	31	.	.	PUNCT
cana-6247	195	1	[	[	X
cana-6247	195	2	9	9	NUM
cana-6247	195	3	]	]	SYM
cana-6247	195	4	•	•	NOUN
cana-6247	195	5	the	the	DET
cana-6247	195	6	joint	joint	ADJ
cana-6247	195	7	conditional	conditional	ADJ
cana-6247	195	8	diffusion	diffusion	NOUN
cana-6247	195	9	model	model	NOUN
cana-6247	195	10	outperformed	outperform	VERB
cana-6247	195	11	single	single	ADJ
cana-6247	195	12	-	-	PUNCT
cana-6247	195	13	degradation	degradation	NOUN
cana-6247	195	14	or	or	CCONJ
cana-6247	195	15	unconditional	unconditional	ADJ
cana-6247	195	16	techniques	technique	NOUN
cana-6247	195	17	for	for	ADP
cana-6247	195	18	mixed	mixed	ADJ
cana-6247	195	19	degradations	degradation	NOUN
cana-6247	195	20	by	by	ADP
cana-6247	195	21	conditioning	condition	VERB
cana-6247	195	22	both	both	CCONJ
cana-6247	195	23	the	the	DET
cana-6247	195	24	forward	forward	ADJ
cana-6247	195	25	and	and	CCONJ
cana-6247	195	26	reverse	reverse	ADJ
cana-6247	195	27	processes	process	NOUN
cana-6247	195	28	on	on	ADP
cana-6247	195	29	degraded	degraded	ADJ
cana-6247	195	30	images	image	NOUN
cana-6247	195	31	and	and	CCONJ
cana-6247	195	32	masks	mask	NOUN
cana-6247	195	33	.	.	PUNCT
cana-6247	196	1	this	this	PRON
cana-6247	196	2	allows	allow	VERB
cana-6247	196	3	the	the	DET
cana-6247	196	4	model	model	NOUN
cana-6247	196	5	to	to	PART
cana-6247	196	6	adaptively	adaptively	ADV
cana-6247	196	7	adjust	adjust	VERB
cana-6247	196	8	noise	noise	NOUN
cana-6247	196	9	levels	level	NOUN
cana-6247	196	10	.	.	PUNCT
cana-6247	197	1	[	[	X
cana-6247	197	2	48	48	NUM
cana-6247	197	3	]	]	SYM
cana-6247	197	4	•	•	NOUN
cana-6247	197	5	in	in	ADP
cana-6247	197	6	order	order	NOUN
cana-6247	197	7	to	to	PART
cana-6247	197	8	ensure	ensure	VERB
cana-6247	197	9	that	that	SCONJ
cana-6247	197	10	the	the	DET
cana-6247	197	11	injected	inject	VERB
cana-6247	197	12	noise	noise	NOUN
cana-6247	197	13	respects	respect	VERB
cana-6247	197	14	high	high	ADJ
cana-6247	197	15	-	-	PUNCT
cana-6247	197	16	level	level	NOUN
cana-6247	197	17	structure	structure	NOUN
cana-6247	197	18	and	and	CCONJ
cana-6247	197	19	appearance	appearance	NOUN
cana-6247	197	20	,	,	PUNCT
cana-6247	197	21	pair	pair	NOUN
cana-6247	197	22	diffusion	diffusion	NOUN
cana-6247	197	23	and	and	CCONJ
cana-6247	197	24	palette	palette	NOUN
cana-6247	197	25	both	both	PRON
cana-6247	197	26	made	make	VERB
cana-6247	197	27	use	use	NOUN
cana-6247	197	28	of	of	ADP
cana-6247	197	29	multimodal	multimodal	NOUN
cana-6247	197	30	conditions	condition	NOUN
cana-6247	197	31	(	(	PUNCT
cana-6247	197	32	such	such	ADJ
cana-6247	197	33	as	as	ADP
cana-6247	197	34	text	text	NOUN
cana-6247	197	35	,	,	PUNCT
cana-6247	197	36	panoptic	panoptic	ADJ
cana-6247	197	37	maps	map	NOUN
cana-6247	197	38	,	,	PUNCT
cana-6247	197	39	and	and	CCONJ
cana-6247	197	40	reference	reference	NOUN
cana-6247	197	41	images	image	NOUN
cana-6247	197	42	)	)	PUNCT
cana-6247	197	43	.	.	PUNCT
cana-6247	198	1	this	this	PRON
cana-6247	198	2	greatly	greatly	ADV
cana-6247	198	3	increased	increase	VERB
cana-6247	198	4	controllability	controllability	NOUN
cana-6247	198	5	and	and	CCONJ
cana-6247	198	6	realism	realism	NOUN
cana-6247	198	7	for	for	ADP
cana-6247	198	8	editing	editing	NOUN
cana-6247	198	9	and	and	CCONJ
cana-6247	198	10	outpainting	outpainte	VERB
cana-6247	198	11	jobs	job	NOUN
cana-6247	198	12	.	.	PUNCT
cana-6247	199	1	[	[	X
cana-6247	199	2	8	8	NUM
cana-6247	199	3	]	]	X
cana-6247	199	4	[	[	X
cana-6247	199	5	32	32	NUM
cana-6247	199	6	]	]	PUNCT
cana-6247	199	7	•	•	NUM
cana-6247	199	8	according	accord	VERB
cana-6247	199	9	to	to	ADP
cana-6247	199	10	latentpaint	latentpaint	NOUN
cana-6247	199	11	,	,	PUNCT
cana-6247	199	12	mask	mask	NOUN
cana-6247	199	13	-	-	PUNCT
cana-6247	199	14	independent	independent	ADJ
cana-6247	199	15	inpainting	inpainting	NOUN
cana-6247	199	16	is	be	AUX
cana-6247	199	17	made	make	VERB
cana-6247	199	18	possible	possible	ADJ
cana-6247	199	19	by	by	ADP
cana-6247	199	20	combining	combine	VERB
cana-6247	199	21	unconditional	unconditional	ADJ
cana-6247	199	22	latent	latent	NOUN
cana-6247	199	23	diffusion	diffusion	NOUN
cana-6247	199	24	with	with	ADP
cana-6247	199	25	mask	mask	NOUN
cana-6247	199	26	-	-	PUNCT
cana-6247	199	27	based	base	VERB
cana-6247	199	28	and	and	CCONJ
cana-6247	199	29	semantic	semantic	ADJ
cana-6247	199	30	cni	cni	NOUN
cana-6247	199	31	,	,	PUNCT
cana-6247	199	32	producing	produce	VERB
cana-6247	199	33	outcomes	outcome	NOUN
cana-6247	199	34	that	that	PRON
cana-6247	199	35	closely	closely	ADV
cana-6247	199	36	resemble	resemble	VERB
cana-6247	199	37	global	global	ADJ
cana-6247	199	38	statistics	statistic	NOUN
cana-6247	199	39	and	and	CCONJ
cana-6247	199	40	textures	texture	NOUN
cana-6247	199	41	.	.	PUNCT
cana-6247	200	1	[	[	X
cana-6247	200	2	6	6	NUM
cana-6247	200	3	]	]	PUNCT
cana-6247	200	4	in	in	ADP
cana-6247	200	5	sum	sum	NOUN
cana-6247	200	6	.	.	PUNCT
cana-6247	201	1	cni	cni	PROPN
cana-6247	201	2	greatly	greatly	ADV
cana-6247	201	3	improves	improve	VERB
cana-6247	201	4	diffusion	diffusion	NOUN
cana-6247	201	5	models	model	NOUN
cana-6247	201	6	for	for	ADP
cana-6247	201	7	image	image	NOUN
cana-6247	201	8	inpainting	inpainte	VERB
cana-6247	201	9	and	and	CCONJ
cana-6247	201	10	outpainting	outpainte	VERB
cana-6247	201	11	by	by	ADP
cana-6247	201	12	adding	add	VERB
cana-6247	201	13	a	a	DET
cana-6247	201	14	condition	condition	NOUN
cana-6247	201	15	to	to	PART
cana-6247	201	16	noise	noise	VERB
cana-6247	201	17	injection	injection	NOUN
cana-6247	201	18	based	base	VERB
cana-6247	201	19	on	on	ADP
cana-6247	201	20	outside	outside	ADJ
cana-6247	201	21	information	information	NOUN
cana-6247	201	22	,	,	PUNCT
cana-6247	201	23	such	such	ADJ
cana-6247	201	24	as	as	ADP
cana-6247	201	25	masks	mask	NOUN
cana-6247	201	26	,	,	PUNCT
cana-6247	201	27	semantic	semantic	ADJ
cana-6247	201	28	maps	map	NOUN
cana-6247	201	29	,	,	PUNCT
cana-6247	201	30	or	or	CCONJ
cana-6247	201	31	context	context	NOUN
cana-6247	201	32	,	,	PUNCT
cana-6247	201	33	allowing	allow	VERB
cana-6247	201	34	different	different	ADJ
cana-6247	201	35	amounts	amount	NOUN
cana-6247	201	36	of	of	ADP
cana-6247	201	37	noise	noise	NOUN
cana-6247	201	38	to	to	PART
cana-6247	201	39	be	be	AUX
cana-6247	201	40	injected	inject	VERB
cana-6247	201	41	across	across	ADP
cana-6247	201	42	different	different	ADJ
cana-6247	201	43	image	image	NOUN
cana-6247	201	44	regions	region	NOUN
cana-6247	201	45	[	[	X
cana-6247	201	46	48	48	NUM
cana-6247	201	47	]	]	PUNCT
cana-6247	201	48	.	.	PUNCT
cana-6247	202	1	as	as	ADP
cana-6247	202	2	an	an	DET
cana-6247	202	3	example	example	NOUN
cana-6247	202	4	,	,	PUNCT
cana-6247	202	5	inpdiffusion	inpdiffusion	NOUN
cana-6247	202	6	allows	allow	VERB
cana-6247	202	7	for	for	ADP
cana-6247	202	8	refinement	refinement	NOUN
cana-6247	202	9	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	202	10	986	986	NUM
cana-6247	202	11	communications	communication	NOUN
cana-6247	202	12	on	on	ADP
cana-6247	202	13	applied	apply	VERB
cana-6247	202	14	nonlinear	nonlinear	ADJ
cana-6247	202	15	analysis	analysis	NOUN
cana-6247	202	16	issn	issn	NOUN
cana-6247	202	17	:	:	PUNCT
cana-6247	202	18	1074	1074	NUM
cana-6247	202	19	-	-	PUNCT
cana-6247	202	20	133x	133x	NUM
cana-6247	202	21	vol	vol	NOUN
cana-6247	202	22	.	.	PROPN
cana-6247	203	1	31	31	NUM
cana-6247	204	1	no	no	NOUN
cana-6247	204	2	.	.	PUNCT
cana-6247	205	1	7s	7	NOUN
cana-6247	205	2	(	(	PUNCT
cana-6247	205	3	2024	2024	NUM
cana-6247	205	4	)	)	PUNCT
cana-6247	205	5	of	of	ADP
cana-6247	205	6	mask	mask	NOUN
cana-6247	205	7	boundaries	boundary	NOUN
cana-6247	205	8	by	by	ADP
cana-6247	205	9	only	only	ADV
cana-6247	205	10	injecting	inject	VERB
cana-6247	205	11	noise	noise	NOUN
cana-6247	205	12	for	for	ADP
cana-6247	205	13	conditions	condition	NOUN
cana-6247	205	14	on	on	ADP
cana-6247	205	15	masked	mask	VERB
cana-6247	205	16	images	image	NOUN
cana-6247	205	17	while	while	SCONJ
cana-6247	205	18	keeping	keep	VERB
cana-6247	205	19	detail	detail	NOUN
cana-6247	205	20	in	in	ADP
cana-6247	205	21	important	important	ADJ
cana-6247	205	22	edge	edge	NOUN
cana-6247	205	23	areas	area	NOUN
cana-6247	205	24	[	[	X
cana-6247	205	25	39	39	NUM
cana-6247	205	26	]	]	PUNCT
cana-6247	205	27	,	,	PUNCT
cana-6247	205	28	and	and	CCONJ
cana-6247	205	29	gradpaint	gradpaint	NOUN
cana-6247	205	30	improves	improve	VERB
cana-6247	205	31	on	on	ADP
cana-6247	205	32	the	the	DET
cana-6247	205	33	limitations	limitation	NOUN
cana-6247	205	34	of	of	ADP
cana-6247	205	35	inpainting	inpainte	VERB
cana-6247	205	36	by	by	ADP
cana-6247	205	37	non	non	ADJ
cana-6247	205	38	-	-	ADJ
cana-6247	205	39	uniform	uniform	ADJ
cana-6247	205	40	noise	noise	NOUN
cana-6247	205	41	injection	injection	NOUN
cana-6247	205	42	better	well	ADJ
cana-6247	205	43	consistency	consistency	NOUN
cana-6247	205	44	across	across	ADP
cana-6247	205	45	inpainted	inpainte	VERB
cana-6247	205	46	and	and	CCONJ
cana-6247	205	47	known	know	VERB
cana-6247	205	48	regions	region	NOUN
cana-6247	206	1	[	[	X
cana-6247	206	2	9	9	NUM
cana-6247	206	3	]	]	PUNCT
cana-6247	206	4	.	.	PUNCT
cana-6247	207	1	noisecollage	noisecollage	NOUN
cana-6247	207	2	also	also	ADV
cana-6247	207	3	builds	build	VERB
cana-6247	207	4	upon	upon	SCONJ
cana-6247	207	5	many	many	ADJ
cana-6247	207	6	of	of	ADP
cana-6247	207	7	these	these	DET
cana-6247	207	8	ideas	idea	NOUN
cana-6247	207	9	,	,	PUNCT
cana-6247	207	10	but	but	CCONJ
cana-6247	207	11	allows	allow	VERB
cana-6247	207	12	the	the	DET
cana-6247	207	13	noise	noise	NOUN
cana-6247	207	14	to	to	PART
cana-6247	207	15	be	be	AUX
cana-6247	207	16	composed	compose	VERB
cana-6247	207	17	modularly	modularly	ADV
cana-6247	207	18	as	as	ADP
cana-6247	207	19	patches	patch	NOUN
cana-6247	207	20	for	for	ADP
cana-6247	207	21	objects	object	NOUN
cana-6247	207	22	and	and	CCONJ
cana-6247	207	23	the	the	DET
cana-6247	207	24	background	background	NOUN
cana-6247	207	25	,	,	PUNCT
cana-6247	207	26	allowing	allow	VERB
cana-6247	207	27	for	for	ADP
cana-6247	207	28	layoutaware	layoutaware	ADJ
cana-6247	207	29	synthesis	synthesis	NOUN
cana-6247	207	30	that	that	PRON
cana-6247	207	31	includes	include	VERB
cana-6247	207	32	some	some	DET
cana-6247	207	33	degree	degree	NOUN
cana-6247	207	34	of	of	ADP
cana-6247	207	35	spatial	spatial	ADJ
cana-6247	207	36	stochasticity	stochasticity	NOUN
cana-6247	207	37	[	[	X
cana-6247	207	38	36	36	NUM
cana-6247	207	39	]	]	PUNCT
cana-6247	207	40	.	.	PUNCT
cana-6247	208	1	pair	pair	NOUN
cana-6247	208	2	diffusion	diffusion	NOUN
cana-6247	208	3	also	also	ADV
cana-6247	208	4	builds	build	VERB
cana-6247	208	5	upon	upon	SCONJ
cana-6247	208	6	these	these	DET
cana-6247	208	7	concepts	concept	NOUN
cana-6247	208	8	by	by	ADP
cana-6247	208	9	introducing	introduce	VERB
cana-6247	208	10	multimodal	multimodal	NOUN
cana-6247	208	11	conditions	condition	NOUN
cana-6247	208	12	(	(	PUNCT
cana-6247	208	13	e.g.	e.g.	ADV
cana-6247	208	14	,	,	PUNCT
cana-6247	208	15	panoptic	panoptic	ADJ
cana-6247	208	16	maps	map	NOUN
cana-6247	208	17	,	,	PUNCT
cana-6247	208	18	text	text	NOUN
cana-6247	208	19	)	)	PUNCT
cana-6247	208	20	,	,	PUNCT
cana-6247	208	21	potentially	potentially	ADV
cana-6247	208	22	improving	improve	VERB
cana-6247	208	23	control	control	NOUN
cana-6247	208	24	and	and	CCONJ
cana-6247	208	25	realism	realism	NOUN
cana-6247	208	26	in	in	ADP
cana-6247	208	27	the	the	DET
cana-6247	208	28	generated	generate	VERB
cana-6247	208	29	content	content	NOUN
cana-6247	208	30	[	[	X
cana-6247	208	31	8	8	NUM
cana-6247	208	32	]	]	PUNCT
cana-6247	208	33	.	.	PUNCT
cana-6247	209	1	finally	finally	ADV
cana-6247	209	2	,	,	PUNCT
cana-6247	209	3	models	model	NOUN
cana-6247	209	4	like	like	ADP
cana-6247	209	5	the	the	DET
cana-6247	209	6	joint	joint	ADJ
cana-6247	209	7	conditional	conditional	ADJ
cana-6247	209	8	diffusion	diffusion	NOUN
cana-6247	209	9	model	model	NOUN
cana-6247	209	10	demonstrate	demonstrate	VERB
cana-6247	209	11	strong	strong	ADJ
cana-6247	209	12	adaptability	adaptability	NOUN
cana-6247	209	13	to	to	ADP
cana-6247	209	14	cases	case	NOUN
cana-6247	209	15	of	of	ADP
cana-6247	209	16	mixed	mixed	ADJ
cana-6247	209	17	degradation	degradation	NOUN
cana-6247	209	18	,	,	PUNCT
cana-6247	209	19	conditioning	condition	VERB
cana-6247	209	20	both	both	CCONJ
cana-6247	209	21	forward	forward	ADV
cana-6247	209	22	and	and	CCONJ
cana-6247	209	23	reverse	reverse	ADJ
cana-6247	209	24	diffusion	diffusion	NOUN
cana-6247	209	25	on	on	ADP
cana-6247	209	26	the	the	DET
cana-6247	209	27	degraded	degraded	ADJ
cana-6247	209	28	inputs	input	NOUN
cana-6247	209	29	[	[	X
cana-6247	209	30	48	48	NUM
cana-6247	209	31	]	]	PUNCT
cana-6247	209	32	,	,	PUNCT
cana-6247	209	33	and	and	CCONJ
cana-6247	209	34	models	model	NOUN
cana-6247	209	35	like	like	ADP
cana-6247	209	36	latentpaint	latentpaint	NOUN
cana-6247	209	37	even	even	ADV
cana-6247	209	38	successfully	successfully	ADV
cana-6247	209	39	integrated	integrate	VERB
cana-6247	209	40	unconditional	unconditional	ADJ
cana-6247	209	41	latent	latent	NOUN
cana-6247	209	42	diffusion	diffusion	NOUN
cana-6247	209	43	with	with	ADP
cana-6247	209	44	semantic	semantic	ADJ
cana-6247	209	45	cni	cni	NOUN
cana-6247	209	46	to	to	PART
cana-6247	209	47	enable	enable	VERB
cana-6247	209	48	mask	mask	NOUN
cana-6247	209	49	-	-	PUNCT
cana-6247	209	50	independent	independent	ADJ
cana-6247	209	51	inpainting	inpainting	NOUN
cana-6247	209	52	,	,	PUNCT
cana-6247	209	53	successfully	successfully	ADV
cana-6247	209	54	preserving	preserve	VERB
cana-6247	209	55	both	both	CCONJ
cana-6247	209	56	global	global	ADJ
cana-6247	209	57	textures	texture	NOUN
cana-6247	209	58	and	and	CCONJ
cana-6247	209	59	local	local	ADJ
cana-6247	209	60	structures	structure	NOUN
cana-6247	209	61	[	[	X
cana-6247	209	62	6	6	NUM
cana-6247	209	63	]	]	PUNCT
cana-6247	209	64	.	.	PUNCT
cana-6247	210	1	in	in	ADP
cana-6247	210	2	all	all	PRON
cana-6247	210	3	of	of	ADP
cana-6247	210	4	these	these	DET
cana-6247	210	5	examples	example	NOUN
cana-6247	210	6	of	of	ADP
cana-6247	210	7	diffusion	diffusion	NOUN
cana-6247	210	8	models	model	NOUN
cana-6247	210	9	conditioned	condition	VERB
cana-6247	210	10	on	on	ADP
cana-6247	210	11	how	how	SCONJ
cana-6247	210	12	to	to	PART
cana-6247	210	13	add	add	VERB
cana-6247	210	14	noise	noise	NOUN
cana-6247	210	15	,	,	PUNCT
cana-6247	210	16	the	the	DET
cana-6247	210	17	noise	noise	NOUN
cana-6247	210	18	schedules	schedule	NOUN
cana-6247	210	19	,	,	PUNCT
cana-6247	210	20	which	which	PRON
cana-6247	210	21	allow	allow	VERB
cana-6247	210	22	for	for	ADP
cana-6247	210	23	or	or	CCONJ
cana-6247	210	24	tailor	tailor	NOUN
cana-6247	210	25	in	in	ADP
cana-6247	210	26	a	a	DET
cana-6247	210	27	region	region	NOUN
cana-6247	210	28	-	-	PUNCT
cana-6247	210	29	specific	specific	ADJ
cana-6247	210	30	manner	manner	NOUN
cana-6247	210	31	,	,	PUNCT
cana-6247	210	32	show	show	VERB
cana-6247	210	33	considerable	considerable	ADJ
cana-6247	210	34	improvements	improvement	NOUN
cana-6247	210	35	in	in	ADP
cana-6247	210	36	the	the	DET
cana-6247	210	37	integrity	integrity	NOUN
cana-6247	210	38	of	of	ADP
cana-6247	210	39	the	the	DET
cana-6247	210	40	boundary	boundary	ADJ
cana-6247	210	41	,	,	PUNCT
cana-6247	210	42	semantic	semantic	ADJ
cana-6247	210	43	coherence	coherence	NOUN
cana-6247	210	44	,	,	PUNCT
cana-6247	210	45	and	and	CCONJ
cana-6247	210	46	flexibility	flexibility	NOUN
cana-6247	210	47	of	of	ADP
cana-6247	210	48	the	the	DET
cana-6247	210	49	image	image	NOUN
cana-6247	210	50	editing	edit	VERB
cana-6247	210	51	process	process	NOUN
cana-6247	210	52	,	,	PUNCT
cana-6247	210	53	securing	secure	VERB
cana-6247	210	54	cni	cni	PROPN
cana-6247	210	55	as	as	ADP
cana-6247	210	56	the	the	DET
cana-6247	210	57	preferred	preferred	ADJ
cana-6247	210	58	noise	noise	NOUN
cana-6247	210	59	process	process	NOUN
cana-6247	210	60	for	for	ADP
cana-6247	210	61	inpainting	inpainte	VERB
cana-6247	210	62	and	and	CCONJ
cana-6247	210	63	outpainting	outpainte	VERB
cana-6247	210	64	.	.	PUNCT
cana-6247	211	1	2.2	2.2	NUM
cana-6247	211	2	deblurring	deblurre	VERB
cana-6247	211	3	the	the	DET
cana-6247	211	4	process	process	NOUN
cana-6247	211	5	of	of	ADP
cana-6247	211	6	restoring	restore	VERB
cana-6247	211	7	a	a	DET
cana-6247	211	8	crisp	crisp	ADJ
cana-6247	211	9	,	,	PUNCT
cana-6247	211	10	clear	clear	ADJ
cana-6247	211	11	image	image	NOUN
cana-6247	211	12	from	from	ADP
cana-6247	211	13	a	a	DET
cana-6247	211	14	blurry	blurry	ADJ
cana-6247	211	15	input	input	NOUN
cana-6247	211	16	is	be	AUX
cana-6247	211	17	known	know	VERB
cana-6247	211	18	as	as	ADP
cana-6247	211	19	image	image	NOUN
cana-6247	211	20	deblurring	deblurring	NOUN
cana-6247	211	21	.	.	PUNCT
cana-6247	212	1	motion	motion	NOUN
cana-6247	212	2	during	during	ADP
cana-6247	212	3	image	image	NOUN
cana-6247	212	4	capture	capture	NOUN
cana-6247	212	5	,	,	PUNCT
cana-6247	212	6	camera	camera	NOUN
cana-6247	212	7	shake	shake	NOUN
cana-6247	212	8	,	,	PUNCT
cana-6247	212	9	defocusing	defocuse	VERB
cana-6247	212	10	,	,	PUNCT
cana-6247	212	11	optical	optical	ADJ
cana-6247	212	12	flaws	flaw	NOUN
cana-6247	212	13	,	,	PUNCT
cana-6247	212	14	and	and	CCONJ
cana-6247	212	15	environmental	environmental	ADJ
cana-6247	212	16	factors	factor	NOUN
cana-6247	212	17	are	be	AUX
cana-6247	212	18	some	some	PRON
cana-6247	212	19	of	of	ADP
cana-6247	212	20	the	the	DET
cana-6247	212	21	causes	cause	NOUN
cana-6247	212	22	of	of	ADP
cana-6247	212	23	blurring	blur	VERB
cana-6247	212	24	.	.	PUNCT
cana-6247	213	1	to	to	PART
cana-6247	213	2	restore	restore	VERB
cana-6247	213	3	the	the	DET
cana-6247	213	4	original	original	ADJ
cana-6247	213	5	image	image	NOUN
cana-6247	213	6	f	f	PROPN
cana-6247	213	7	while	while	SCONJ
cana-6247	213	8	reducing	reduce	VERB
cana-6247	213	9	noise	noise	NOUN
cana-6247	213	10	n	n	CCONJ
cana-6247	213	11	,	,	PUNCT
cana-6247	213	12	the	the	DET
cana-6247	213	13	mathematical	mathematical	ADJ
cana-6247	213	14	model	model	NOUN
cana-6247	213	15	commonly	commonly	ADV
cana-6247	213	16	represented	represent	VERB
cana-6247	213	17	as	as	ADP
cana-6247	213	18	g	g	NOUN
cana-6247	213	19	=	=	NOUN
cana-6247	213	20	hf+n	hf+n	PROPN
cana-6247	213	21	deconvolutes	deconvolute	VERB
cana-6247	213	22	the	the	DET
cana-6247	213	23	blurred	blurred	ADJ
cana-6247	213	24	image	image	NOUN
cana-6247	213	25	g	g	NOUN
cana-6247	213	26	using	use	VERB
cana-6247	213	27	a	a	DET
cana-6247	213	28	point	point	NOUN
cana-6247	213	29	spread	spread	NOUN
cana-6247	213	30	function	function	NOUN
cana-6247	213	31	(	(	PUNCT
cana-6247	213	32	psf	psf	NOUN
cana-6247	213	33	)	)	PUNCT
cana-6247	213	34	h.	h.	NOUN
cana-6247	213	35	effectively	effectively	ADV
cana-6247	213	36	reversing	reverse	VERB
cana-6247	213	37	the	the	DET
cana-6247	213	38	blur	blur	NOUN
cana-6247	213	39	’s	’s	PART
cana-6247	213	40	effect	effect	NOUN
cana-6247	213	41	(	(	PUNCT
cana-6247	213	42	often	often	ADV
cana-6247	213	43	represented	represent	VERB
cana-6247	213	44	as	as	ADP
cana-6247	213	45	a	a	DET
cana-6247	213	46	convolution	convolution	NOUN
cana-6247	213	47	)	)	PUNCT
cana-6247	213	48	while	while	SCONJ
cana-6247	213	49	preventing	prevent	VERB
cana-6247	213	50	noise	noise	NOUN
cana-6247	213	51	amplification	amplification	NOUN
cana-6247	213	52	and	and	CCONJ
cana-6247	213	53	artifacts	artifact	NOUN
cana-6247	213	54	is	be	AUX
cana-6247	213	55	the	the	DET
cana-6247	213	56	primary	primary	ADJ
cana-6247	213	57	issue	issue	NOUN
cana-6247	213	58	in	in	ADP
cana-6247	213	59	deblurring	deblurre	VERB
cana-6247	213	60	,	,	PUNCT
cana-6247	213	61	particularly	particularly	ADV
cana-6247	213	62	when	when	SCONJ
cana-6247	213	63	the	the	DET
cana-6247	213	64	blur	blur	NOUN
cana-6247	213	65	’s	’s	PART
cana-6247	213	66	properties	property	NOUN
cana-6247	213	67	are	be	AUX
cana-6247	213	68	unknown	unknown	ADJ
cana-6247	213	69	or	or	CCONJ
cana-6247	213	70	differ	differ	VERB
cana-6247	213	71	between	between	ADP
cana-6247	213	72	images	image	NOUN
cana-6247	213	73	.	.	PUNCT
cana-6247	214	1	recent	recent	ADJ
cana-6247	214	2	research	research	NOUN
cana-6247	214	3	in	in	ADP
cana-6247	214	4	image	image	NOUN
cana-6247	214	5	restoration	restoration	NOUN
cana-6247	214	6	has	have	AUX
cana-6247	214	7	shown	show	VERB
cana-6247	214	8	that	that	SCONJ
cana-6247	214	9	making	make	VERB
cana-6247	214	10	the	the	DET
cana-6247	214	11	deblurring	deblurre	VERB
cana-6247	214	12	process	process	NOUN
cana-6247	214	13	data	datum	NOUN
cana-6247	214	14	-	-	PUNCT
cana-6247	214	15	dependent	dependent	ADJ
cana-6247	214	16	can	can	AUX
cana-6247	214	17	greatly	greatly	ADV
cana-6247	214	18	improve	improve	VERB
cana-6247	214	19	the	the	DET
cana-6247	214	20	results	result	NOUN
cana-6247	214	21	.	.	PUNCT
cana-6247	215	1	in	in	ADP
cana-6247	215	2	the	the	DET
cana-6247	215	3	case	case	NOUN
cana-6247	215	4	of	of	ADP
cana-6247	215	5	diffusion	diffusion	NOUN
cana-6247	215	6	models	model	NOUN
cana-6247	215	7	,	,	PUNCT
cana-6247	215	8	this	this	DET
cana-6247	215	9	data	data	NOUN
cana-6247	215	10	dependency	dependency	NOUN
cana-6247	215	11	means	mean	VERB
cana-6247	215	12	the	the	DET
cana-6247	215	13	focus	focus	NOUN
cana-6247	215	14	of	of	ADP
cana-6247	215	15	the	the	DET
cana-6247	215	16	restoration	restoration	NOUN
cana-6247	215	17	process	process	NOUN
cana-6247	215	18	can	can	AUX
cana-6247	215	19	vary	vary	VERB
cana-6247	215	20	based	base	VERB
cana-6247	215	21	on	on	ADP
cana-6247	215	22	input	input	NOUN
cana-6247	215	23	edge	edge	NOUN
cana-6247	215	24	information	information	NOUN
cana-6247	215	25	,	,	PUNCT
cana-6247	215	26	the	the	DET
cana-6247	215	27	blur	blur	PROPN
cana-6247	215	28	statistics	statistic	NOUN
cana-6247	215	29	,	,	PUNCT
cana-6247	215	30	and	and	CCONJ
cana-6247	215	31	the	the	DET
cana-6247	215	32	granularity	granularity	NOUN
cana-6247	215	33	of	of	ADP
cana-6247	215	34	the	the	DET
cana-6247	215	35	semantic	semantic	ADJ
cana-6247	215	36	context	context	NOUN
cana-6247	215	37	,	,	PUNCT
cana-6247	215	38	rather	rather	ADV
cana-6247	215	39	than	than	ADP
cana-6247	215	40	deblurring	deblurre	VERB
cana-6247	215	41	every	every	DET
cana-6247	215	42	part	part	NOUN
cana-6247	215	43	of	of	ADP
cana-6247	215	44	the	the	DET
cana-6247	215	45	image	image	NOUN
cana-6247	215	46	the	the	DET
cana-6247	215	47	same	same	ADJ
cana-6247	215	48	.	.	PUNCT
cana-6247	216	1	this	this	DET
cana-6247	216	2	intelligent	intelligent	ADJ
cana-6247	216	3	approach	approach	NOUN
cana-6247	216	4	makes	make	VERB
cana-6247	216	5	for	for	ADP
cana-6247	216	6	better	well	ADJ
cana-6247	216	7	reconstructions	reconstruction	NOUN
cana-6247	216	8	,	,	PUNCT
cana-6247	216	9	particularly	particularly	ADV
cana-6247	216	10	for	for	ADP
cana-6247	216	11	hard	hard	ADJ
cana-6247	216	12	and	and	CCONJ
cana-6247	216	13	uneven	uneven	ADJ
cana-6247	216	14	blurs	blur	VERB
cana-6247	216	15	.	.	PUNCT
cana-6247	217	1	conditional	conditional	ADJ
cana-6247	217	2	noise	noise	NOUN
cana-6247	217	3	injection	injection	NOUN
cana-6247	217	4	(	(	PUNCT
cana-6247	217	5	cni	cni	PROPN
cana-6247	217	6	)	)	PUNCT
cana-6247	217	7	uses	use	VERB
cana-6247	217	8	input	input	NOUN
cana-6247	217	9	-	-	PUNCT
cana-6247	217	10	specific	specific	ADJ
cana-6247	217	11	clues	clue	NOUN
cana-6247	217	12	,	,	PUNCT
cana-6247	217	13	like	like	ADP
cana-6247	217	14	estimated	estimate	VERB
cana-6247	217	15	blur	blur	NOUN
cana-6247	217	16	kernels	kernel	NOUN
cana-6247	217	17	or	or	CCONJ
cana-6247	217	18	degradation	degradation	NOUN
cana-6247	217	19	maps	map	NOUN
cana-6247	217	20	,	,	PUNCT
cana-6247	217	21	by	by	ADP
cana-6247	217	22	directing	direct	VERB
cana-6247	217	23	noise	noise	NOUN
cana-6247	217	24	from	from	ADP
cana-6247	217	25	the	the	DET
cana-6247	217	26	generative	generative	ADJ
cana-6247	217	27	capacity	capacity	NOUN
cana-6247	217	28	of	of	ADP
cana-6247	217	29	a	a	DET
cana-6247	217	30	diffusion	diffusion	NOUN
cana-6247	217	31	model	model	NOUN
cana-6247	217	32	to	to	ADP
cana-6247	217	33	image	image	NOUN
cana-6247	217	34	regions	region	NOUN
cana-6247	217	35	impacted	impact	VERB
cana-6247	217	36	most	most	ADV
cana-6247	217	37	by	by	ADP
cana-6247	217	38	blur	blur	NOUN
cana-6247	217	39	or	or	CCONJ
cana-6247	217	40	uncertainty	uncertainty	NOUN
cana-6247	217	41	.	.	PUNCT
cana-6247	218	1	in	in	ADP
cana-6247	218	2	turn	turn	NOUN
cana-6247	218	3	,	,	PUNCT
cana-6247	218	4	allowing	allow	VERB
cana-6247	218	5	restoration	restoration	NOUN
cana-6247	218	6	to	to	PART
cana-6247	218	7	occur	occur	VERB
cana-6247	218	8	only	only	ADV
cana-6247	218	9	in	in	ADP
cana-6247	218	10	the	the	DET
cana-6247	218	11	corrupted	corrupted	ADJ
cana-6247	218	12	areas	area	NOUN
cana-6247	218	13	of	of	ADP
cana-6247	218	14	the	the	DET
cana-6247	218	15	image	image	NOUN
cana-6247	218	16	,	,	PUNCT
cana-6247	218	17	it	it	PRON
cana-6247	218	18	allows	allow	VERB
cana-6247	218	19	the	the	DET
cana-6247	218	20	model	model	NOUN
cana-6247	218	21	to	to	PART
cana-6247	218	22	effectively	effectively	ADV
cana-6247	218	23	reconstruct	reconstruct	VERB
cana-6247	218	24	sharp	sharp	ADJ
cana-6247	218	25	content	content	NOUN
cana-6247	218	26	wherever	wherever	SCONJ
cana-6247	218	27	the	the	DET
cana-6247	218	28	image	image	NOUN
cana-6247	218	29	was	be	AUX
cana-6247	218	30	previously	previously	ADV
cana-6247	218	31	degraded	degrade	VERB
cana-6247	218	32	and	and	CCONJ
cana-6247	218	33	maintain	maintain	VERB
cana-6247	218	34	fidelity	fidelity	NOUN
cana-6247	218	35	in	in	ADP
cana-6247	218	36	additional	additional	ADJ
cana-6247	218	37	regions	region	NOUN
cana-6247	218	38	of	of	ADP
cana-6247	218	39	the	the	DET
cana-6247	218	40	image	image	NOUN
cana-6247	218	41	where	where	SCONJ
cana-6247	218	42	fidelity	fidelity	PROPN
cana-6247	218	43	was	be	AUX
cana-6247	218	44	already	already	ADV
cana-6247	218	45	high	high	ADJ
cana-6247	218	46	.	.	PUNCT
cana-6247	219	1	the	the	DET
cana-6247	219	2	adaptive	adaptive	ADJ
cana-6247	219	3	accessibility	accessibility	NOUN
cana-6247	219	4	of	of	ADP
cana-6247	219	5	cni	cni	PROPN
cana-6247	219	6	virtues	virtue	NOUN
cana-6247	219	7	is	be	AUX
cana-6247	219	8	important	important	ADJ
cana-6247	219	9	for	for	ADP
cana-6247	219	10	the	the	DET
cana-6247	219	11	protection	protection	NOUN
cana-6247	219	12	of	of	ADP
cana-6247	219	13	necessary	necessary	ADJ
cana-6247	219	14	structural	structural	ADJ
cana-6247	219	15	information	information	NOUN
cana-6247	219	16	and	and	CCONJ
cana-6247	219	17	texture	texture	NOUN
cana-6247	219	18	.	.	PUNCT
cana-6247	220	1	by	by	ADP
cana-6247	220	2	controlling	control	VERB
cana-6247	220	3	where	where	SCONJ
cana-6247	220	4	to	to	PART
cana-6247	220	5	inject	inject	VERB
cana-6247	220	6	noise	noise	NOUN
cana-6247	220	7	and	and	CCONJ
cana-6247	220	8	limiting	limit	VERB
cana-6247	220	9	the	the	DET
cana-6247	220	10	amount	amount	NOUN
cana-6247	220	11	added	add	VERB
cana-6247	220	12	,	,	PUNCT
cana-6247	220	13	cni	cni	PROPN
cana-6247	220	14	can	can	AUX
cana-6247	220	15	minimise	minimise	VERB
cana-6247	220	16	the	the	DET
cana-6247	220	17	unnecessary	unnecessary	ADJ
cana-6247	220	18	corruption	corruption	NOUN
cana-6247	220	19	of	of	ADP
cana-6247	220	20	non	non	ADJ
cana-6247	220	21	-	-	ADJ
cana-6247	220	22	blurry	blurry	ADJ
cana-6247	220	23	,	,	PUNCT
cana-6247	220	24	sharp	sharp	ADJ
cana-6247	220	25	or	or	CCONJ
cana-6247	220	26	clear	clear	ADJ
cana-6247	220	27	areas	area	NOUN
cana-6247	220	28	of	of	ADP
cana-6247	220	29	the	the	DET
cana-6247	220	30	image	image	NOUN
cana-6247	220	31	.	.	PUNCT
cana-6247	221	1	this	this	PRON
cana-6247	221	2	allows	allow	VERB
cana-6247	221	3	the	the	DET
cana-6247	221	4	model	model	NOUN
cana-6247	221	5	to	to	PART
cana-6247	221	6	exert	exert	VERB
cana-6247	221	7	less	less	ADJ
cana-6247	221	8	effort	effort	NOUN
cana-6247	221	9	in	in	ADP
cana-6247	221	10	trying	try	VERB
cana-6247	221	11	to	to	PART
cana-6247	221	12	’	'	PUNCT
cana-6247	221	13	fix	fix	VERB
cana-6247	221	14	’	'	PUNCT
cana-6247	221	15	more	more	ADV
cana-6247	221	16	well	well	ADV
cana-6247	221	17	-	-	PUNCT
cana-6247	221	18	preserved	preserve	VERB
cana-6247	221	19	content	content	NOUN
cana-6247	221	20	,	,	PUNCT
cana-6247	221	21	and	and	CCONJ
cana-6247	221	22	allows	allow	VERB
cana-6247	221	23	more	more	ADJ
cana-6247	221	24	of	of	ADP
cana-6247	221	25	its	its	PRON
cana-6247	221	26	’	'	PUNCT
cana-6247	221	27	effort	effort	NOUN
cana-6247	221	28	’	'	PUNCT
cana-6247	221	29	to	to	PART
cana-6247	221	30	be	be	AUX
cana-6247	221	31	directed	direct	VERB
cana-6247	221	32	towards	towards	ADP
cana-6247	221	33	fixing	fix	VERB
cana-6247	221	34	the	the	DET
cana-6247	221	35	blur	blur	NOUN
cana-6247	221	36	and	and	CCONJ
cana-6247	221	37	reconstructing	reconstruct	VERB
cana-6247	221	38	detail	detail	NOUN
cana-6247	221	39	,	,	PUNCT
cana-6247	221	40	only	only	ADV
cana-6247	221	41	in	in	ADP
cana-6247	221	42	the	the	DET
cana-6247	221	43	places	place	NOUN
cana-6247	221	44	where	where	SCONJ
cana-6247	221	45	it	it	PRON
cana-6247	221	46	was	be	AUX
cana-6247	221	47	actually	actually	ADV
cana-6247	221	48	needed	need	VERB
cana-6247	221	49	.	.	PUNCT
cana-6247	222	1	in	in	ADP
cana-6247	222	2	general	general	ADJ
cana-6247	222	3	,	,	PUNCT
cana-6247	222	4	there	there	PRON
cana-6247	222	5	is	be	VERB
cana-6247	222	6	a	a	DET
cana-6247	222	7	trade	trade	NOUN
cana-6247	222	8	-	-	PUNCT
cana-6247	222	9	off	off	NOUN
cana-6247	222	10	,	,	PUNCT
cana-6247	222	11	or	or	CCONJ
cana-6247	222	12	compromise	compromise	NOUN
cana-6247	222	13	made	make	VERB
cana-6247	222	14	by	by	ADP
cana-6247	222	15	the	the	DET
cana-6247	222	16	model	model	NOUN
cana-6247	222	17	,	,	PUNCT
cana-6247	222	18	and	and	CCONJ
cana-6247	222	19	the	the	DET
cana-6247	222	20	resulting	result	VERB
cana-6247	222	21	reconstructions	reconstruction	NOUN
cana-6247	222	22	are	be	AUX
cana-6247	222	23	sharper	sharp	ADJ
cana-6247	222	24	and	and	CCONJ
cana-6247	222	25	artifact	artifact	NOUN
cana-6247	222	26	-	-	PUNCT
cana-6247	222	27	free	free	ADJ
cana-6247	222	28	with	with	ADP
cana-6247	222	29	structures	structure	NOUN
cana-6247	222	30	and	and	CCONJ
cana-6247	222	31	natural	natural	ADJ
cana-6247	222	32	textures	texture	NOUN
cana-6247	222	33	protected	protect	VERB
cana-6247	222	34	to	to	ADP
cana-6247	222	35	the	the	DET
cana-6247	222	36	maximum	maximum	ADJ
cana-6247	222	37	extent	extent	NOUN
cana-6247	222	38	possible	possible	ADJ
cana-6247	222	39	and	and	CCONJ
cana-6247	222	40	remained	remain	VERB
cana-6247	222	41	intact	intact	ADJ
cana-6247	222	42	through	through	ADP
cana-6247	222	43	the	the	DET
cana-6247	222	44	overall	overall	ADJ
cana-6247	222	45	restoration	restoration	NOUN
cana-6247	222	46	process	process	NOUN
cana-6247	222	47	.	.	PUNCT
cana-6247	223	1	another	another	DET
cana-6247	223	2	advantage	advantage	NOUN
cana-6247	223	3	of	of	ADP
cana-6247	223	4	this	this	DET
cana-6247	223	5	noise	noise	NOUN
cana-6247	223	6	adding	add	VERB
cana-6247	223	7	method	method	NOUN
cana-6247	223	8	is	be	AUX
cana-6247	223	9	the	the	DET
cana-6247	223	10	ability	ability	NOUN
cana-6247	223	11	to	to	PART
cana-6247	223	12	generalize	generalize	VERB
cana-6247	223	13	to	to	ADP
cana-6247	223	14	a	a	DET
cana-6247	223	15	larger	large	ADJ
cana-6247	223	16	range	range	NOUN
cana-6247	223	17	of	of	ADP
cana-6247	223	18	blur	blur	NOUN
cana-6247	223	19	patterns	pattern	NOUN
cana-6247	223	20	.	.	PUNCT
cana-6247	224	1	the	the	DET
cana-6247	224	2	conditional	conditional	ADJ
cana-6247	224	3	noise	noise	NOUN
cana-6247	224	4	process	process	NOUN
cana-6247	224	5	quickly	quickly	ADV
cana-6247	224	6	adapts	adapt	VERB
cana-6247	224	7	to	to	ADP
cana-6247	224	8	unknown	unknown	ADJ
cana-6247	224	9	and	and	CCONJ
cana-6247	224	10	diverse	diverse	ADJ
cana-6247	224	11	blur	blur	NOUN
cana-6247	224	12	conditions	condition	NOUN
cana-6247	224	13	for	for	ADP
cana-6247	224	14	each	each	DET
cana-6247	224	15	image	image	NOUN
cana-6247	224	16	,	,	PUNCT
cana-6247	224	17	including	include	VERB
cana-6247	224	18	uniform	uniform	NOUN
cana-6247	224	19	,	,	PUNCT
cana-6247	224	20	spatially	spatially	ADV
cana-6247	224	21	varying	varying	ADJ
cana-6247	224	22	,	,	PUNCT
cana-6247	224	23	or	or	CCONJ
cana-6247	224	24	unusual	unusual	ADJ
cana-6247	224	25	artifact	artifact	ADJ
cana-6247	224	26	causes	cause	NOUN
cana-6247	224	27	.	.	PUNCT
cana-6247	225	1	as	as	ADP
cana-6247	225	2	a	a	DET
cana-6247	225	3	result	result	NOUN
cana-6247	225	4	,	,	PUNCT
cana-6247	225	5	models	model	NOUN
cana-6247	225	6	with	with	ADP
cana-6247	225	7	cni	cni	PROPN
cana-6247	225	8	recover	recover	NOUN
cana-6247	225	9	better	well	ADV
cana-6247	225	10	for	for	ADP
cana-6247	225	11	real	real	ADJ
cana-6247	225	12	-	-	PUNCT
cana-6247	225	13	world	world	NOUN
cana-6247	225	14	,	,	PUNCT
cana-6247	225	15	non	non	ADJ
cana-6247	225	16	-	-	ADJ
cana-6247	225	17	uniform	uniform	ADJ
cana-6247	225	18	,	,	PUNCT
cana-6247	225	19	and	and	CCONJ
cana-6247	225	20	spatially	spatially	ADV
cana-6247	225	21	varying	vary	VERB
cana-6247	225	22	blur	blur	NOUN
cana-6247	225	23	situations	situation	NOUN
cana-6247	225	24	for	for	ADP
cana-6247	225	25	which	which	PRON
cana-6247	225	26	more	more	ADV
cana-6247	225	27	inflexible	inflexible	ADJ
cana-6247	225	28	or	or	CCONJ
cana-6247	225	29	unconditional	unconditional	ADJ
cana-6247	225	30	methods	method	NOUN
cana-6247	225	31	can	can	AUX
cana-6247	225	32	fail	fail	VERB
cana-6247	225	33	.	.	PUNCT
cana-6247	226	1	cni	cni	PROPN
cana-6247	226	2	is	be	AUX
cana-6247	226	3	also	also	ADV
cana-6247	226	4	important	important	ADJ
cana-6247	226	5	for	for	ADP
cana-6247	226	6	strong	strong	ADJ
cana-6247	226	7	performance	performance	NOUN
cana-6247	226	8	in	in	ADP
cana-6247	226	9	blind	blind	ADJ
cana-6247	226	10	deblurring	deblurring	NOUN
cana-6247	226	11	situations	situation	NOUN
cana-6247	226	12	where	where	SCONJ
cana-6247	226	13	the	the	DET
cana-6247	226	14	blur	blur	PROPN
cana-6247	226	15	kernel	kernel	PROPN
cana-6247	226	16	is	be	AUX
cana-6247	226	17	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	226	18	987	987	NUM
cana-6247	226	19	communications	communication	NOUN
cana-6247	226	20	on	on	ADP
cana-6247	226	21	applied	apply	VERB
cana-6247	226	22	nonlinear	nonlinear	ADJ
cana-6247	226	23	analysis	analysis	NOUN
cana-6247	226	24	issn	issn	NOUN
cana-6247	226	25	:	:	PUNCT
cana-6247	226	26	1074	1074	NUM
cana-6247	226	27	-	-	PUNCT
cana-6247	226	28	133x	133x	NUM
cana-6247	226	29	vol	vol	NOUN
cana-6247	226	30	.	.	PROPN
cana-6247	226	31	31	31	NUM
cana-6247	227	1	no	no	NOUN
cana-6247	227	2	.	.	PUNCT
cana-6247	228	1	7s	7	NOUN
cana-6247	228	2	(	(	PUNCT
cana-6247	228	3	2024	2024	NUM
cana-6247	228	4	)	)	PUNCT
cana-6247	228	5	not	not	PART
cana-6247	228	6	known	know	VERB
cana-6247	228	7	a	a	DET
cana-6247	228	8	priori	priori	NOUN
cana-6247	228	9	or	or	CCONJ
cana-6247	228	10	the	the	DET
cana-6247	228	11	degradation	degradation	NOUN
cana-6247	228	12	function	function	NOUN
cana-6247	228	13	has	have	VERB
cana-6247	228	14	some	some	DET
cana-6247	228	15	other	other	ADJ
cana-6247	228	16	form	form	NOUN
cana-6247	228	17	of	of	ADP
cana-6247	228	18	uncertainty	uncertainty	NOUN
cana-6247	228	19	.	.	PUNCT
cana-6247	229	1	by	by	ADP
cana-6247	229	2	adding	add	VERB
cana-6247	229	3	noise	noise	NOUN
cana-6247	229	4	degradation	degradation	NOUN
cana-6247	229	5	-	-	PUNCT
cana-6247	229	6	aware	aware	ADJ
cana-6247	229	7	—	—	PUNCT
cana-6247	229	8	using	use	VERB
cana-6247	229	9	information	information	NOUN
cana-6247	229	10	that	that	PRON
cana-6247	229	11	is	be	AUX
cana-6247	229	12	derived	derive	VERB
cana-6247	229	13	from	from	ADP
cana-6247	229	14	the	the	DET
cana-6247	229	15	input	input	NOUN
cana-6247	229	16	itself	itself	PRON
cana-6247	229	17	—	—	PUNCT
cana-6247	229	18	cni	cni	PROPN
cana-6247	229	19	helps	help	VERB
cana-6247	229	20	orient	orient	VERB
cana-6247	229	21	the	the	DET
cana-6247	229	22	restoration	restoration	NOUN
cana-6247	229	23	towards	towards	ADP
cana-6247	229	24	plausible	plausible	ADJ
cana-6247	229	25	sharp	sharp	ADJ
cana-6247	229	26	solutions	solution	NOUN
cana-6247	229	27	even	even	ADV
cana-6247	229	28	when	when	SCONJ
cana-6247	229	29	the	the	DET
cana-6247	229	30	blur	blur	NOUN
cana-6247	229	31	itself	itself	PRON
cana-6247	229	32	is	be	AUX
cana-6247	229	33	manifestly	manifestly	ADV
cana-6247	229	34	different	different	ADJ
cana-6247	229	35	or	or	CCONJ
cana-6247	229	36	completely	completely	ADV
cana-6247	229	37	uninformed	uninformed	ADJ
cana-6247	229	38	.	.	PUNCT
cana-6247	230	1	overall	overall	ADV
cana-6247	230	2	,	,	PUNCT
cana-6247	230	3	this	this	DET
cana-6247	230	4	accommodation	accommodation	NOUN
cana-6247	230	5	enables	enable	VERB
cana-6247	230	6	diffusion	diffusion	NOUN
cana-6247	230	7	-	-	PUNCT
cana-6247	230	8	based	base	VERB
cana-6247	230	9	models	model	NOUN
cana-6247	230	10	to	to	PART
cana-6247	230	11	perform	perform	VERB
cana-6247	230	12	exceedingly	exceedingly	ADV
cana-6247	230	13	well	well	INTJ
cana-6247	230	14	even	even	ADV
cana-6247	230	15	when	when	SCONJ
cana-6247	230	16	dealing	deal	VERB
cana-6247	230	17	with	with	ADP
cana-6247	230	18	complicated	complicated	ADJ
cana-6247	230	19	or	or	CCONJ
cana-6247	230	20	compound	compound	NOUN
cana-6247	230	21	degradations	degradation	NOUN
cana-6247	230	22	while	while	SCONJ
cana-6247	230	23	imposing	impose	VERB
cana-6247	230	24	more	more	ADV
cana-6247	230	25	reliable	reliable	ADJ
cana-6247	230	26	and	and	CCONJ
cana-6247	230	27	improved	improve	VERB
cana-6247	230	28	quality	quality	NOUN
cana-6247	230	29	in	in	ADP
cana-6247	230	30	arduous	arduous	ADJ
cana-6247	230	31	situations	situation	NOUN
cana-6247	230	32	.	.	PUNCT
cana-6247	231	1	finally	finally	ADV
cana-6247	231	2	,	,	PUNCT
cana-6247	231	3	implementing	implement	VERB
cana-6247	231	4	conditional	conditional	ADJ
cana-6247	231	5	noise	noise	NOUN
cana-6247	231	6	injection	injection	NOUN
cana-6247	231	7	in	in	ADP
cana-6247	231	8	hierarchical	hierarchical	ADJ
cana-6247	231	9	or	or	CCONJ
cana-6247	231	10	plug	plug	VERB
cana-6247	231	11	-	-	PUNCT
cana-6247	231	12	and	and	CCONJ
cana-6247	231	13	-	-	PUNCT
cana-6247	231	14	play	play	NOUN
cana-6247	231	15	designs	design	NOUN
cana-6247	231	16	optimizes	optimize	VERB
cana-6247	231	17	the	the	DET
cana-6247	231	18	deblurring	deblurring	NOUN
cana-6247	231	19	procedures	procedure	NOUN
cana-6247	231	20	and	and	CCONJ
cana-6247	231	21	also	also	ADV
cana-6247	231	22	improves	improve	VERB
cana-6247	231	23	the	the	DET
cana-6247	231	24	efficiency	efficiency	NOUN
cana-6247	231	25	and	and	CCONJ
cana-6247	231	26	adaptivity	adaptivity	NOUN
cana-6247	231	27	.	.	PUNCT
cana-6247	232	1	by	by	ADP
cana-6247	232	2	allowing	allow	VERB
cana-6247	232	3	models	model	NOUN
cana-6247	232	4	to	to	PART
cana-6247	232	5	seamlessly	seamlessly	ADV
cana-6247	232	6	combine	combine	VERB
cana-6247	232	7	and	and	CCONJ
cana-6247	232	8	include	include	VERB
cana-6247	232	9	various	various	ADJ
cana-6247	232	10	types	type	NOUN
cana-6247	232	11	of	of	ADP
cana-6247	232	12	external	external	ADJ
cana-6247	232	13	priors	prior	NOUN
cana-6247	232	14	,	,	PUNCT
cana-6247	232	15	adaptively	adaptively	ADV
cana-6247	232	16	fuse	fuse	NOUN
cana-6247	232	17	information	information	NOUN
cana-6247	232	18	in	in	ADP
cana-6247	232	19	a	a	DET
cana-6247	232	20	range	range	NOUN
cana-6247	232	21	of	of	ADP
cana-6247	232	22	scales	scale	NOUN
cana-6247	232	23	,	,	PUNCT
cana-6247	232	24	and	and	CCONJ
cana-6247	232	25	fasten	fasten	VERB
cana-6247	232	26	the	the	DET
cana-6247	232	27	search	search	NOUN
cana-6247	232	28	to	to	PART
cana-6247	232	29	achieve	achieve	VERB
cana-6247	232	30	convergence	convergence	NOUN
cana-6247	232	31	,	,	PUNCT
cana-6247	232	32	cni	cni	NOUN
cana-6247	232	33	-	-	PUNCT
cana-6247	232	34	based	base	VERB
cana-6247	232	35	frameworks	framework	NOUN
cana-6247	232	36	decrease	decrease	VERB
cana-6247	232	37	computational	computational	ADJ
cana-6247	232	38	cost	cost	NOUN
cana-6247	232	39	and	and	CCONJ
cana-6247	232	40	make	make	VERB
cana-6247	232	41	operationalizing	operationalize	VERB
cana-6247	232	42	deblurring	deblurring	NOUN
cana-6247	232	43	methods	method	NOUN
cana-6247	232	44	into	into	ADP
cana-6247	232	45	downstream	downstream	ADJ
cana-6247	232	46	applications	application	NOUN
cana-6247	232	47	more	more	ADV
cana-6247	232	48	feasible	feasible	ADJ
cana-6247	232	49	.	.	PUNCT
cana-6247	233	1	this	this	DET
cana-6247	233	2	flexibility	flexibility	NOUN
cana-6247	233	3	ensures	ensure	VERB
cana-6247	233	4	that	that	SCONJ
cana-6247	233	5	state	state	NOUN
cana-6247	233	6	-	-	PUNCT
cana-6247	233	7	of	of	ADP
cana-6247	233	8	-	-	PUNCT
cana-6247	233	9	the	the	DET
cana-6247	233	10	-	-	PUNCT
cana-6247	233	11	art	art	NOUN
cana-6247	233	12	restoration	restoration	NOUN
cana-6247	233	13	performance	performance	NOUN
cana-6247	233	14	can	can	AUX
cana-6247	233	15	be	be	AUX
cana-6247	233	16	reasonably	reasonably	ADV
cana-6247	233	17	achieved	achieve	VERB
cana-6247	233	18	across	across	ADP
cana-6247	233	19	a	a	DET
cana-6247	233	20	diverse	diverse	ADJ
cana-6247	233	21	set	set	NOUN
cana-6247	233	22	of	of	ADP
cana-6247	233	23	images	image	NOUN
cana-6247	233	24	and	and	CCONJ
cana-6247	233	25	degradation	degradation	NOUN
cana-6247	233	26	forms	form	NOUN
cana-6247	233	27	.	.	PUNCT
cana-6247	234	1	2.2.1	2.2.1	NUM
cana-6247	234	2	evidence	evidence	NOUN
cana-6247	234	3	from	from	ADP
cana-6247	234	4	recent	recent	ADJ
cana-6247	234	5	research	research	NOUN
cana-6247	234	6	•	•	ADP
cana-6247	234	7	cold	cold	ADJ
cana-6247	234	8	diffusion	diffusion	NOUN
cana-6247	234	9	developed	develop	VERB
cana-6247	234	10	a	a	DET
cana-6247	234	11	formalism	formalism	NOUN
cana-6247	234	12	which	which	PRON
cana-6247	234	13	reverses	reverse	VERB
cana-6247	234	14	arbitrary	arbitrary	ADJ
cana-6247	234	15	image	image	NOUN
cana-6247	234	16	transforms	transform	VERB
cana-6247	234	17	,	,	PUNCT
cana-6247	234	18	not	not	PART
cana-6247	234	19	only	only	ADV
cana-6247	234	20	noise	noise	NOUN
cana-6247	234	21	,	,	PUNCT
cana-6247	234	22	sometimes	sometimes	ADV
cana-6247	234	23	substituting	substitute	VERB
cana-6247	234	24	noise	noise	NOUN
cana-6247	234	25	with	with	ADP
cana-6247	234	26	deterministic	deterministic	ADJ
cana-6247	234	27	degradations	degradation	NOUN
cana-6247	234	28	(	(	PUNCT
cana-6247	234	29	e.g.	e.g.	ADV
cana-6247	234	30	,	,	PUNCT
cana-6247	234	31	blur	blur	NOUN
cana-6247	234	32	)	)	PUNCT
cana-6247	234	33	.	.	PUNCT
cana-6247	235	1	this	this	DET
cana-6247	235	2	work	work	NOUN
cana-6247	235	3	emphasizes	emphasize	VERB
cana-6247	235	4	that	that	SCONJ
cana-6247	235	5	the	the	DET
cana-6247	235	6	forward	forward	ADJ
cana-6247	235	7	process	process	NOUN
cana-6247	235	8	of	of	ADP
cana-6247	235	9	diffusion	diffusion	NOUN
cana-6247	235	10	can	can	AUX
cana-6247	235	11	be	be	AUX
cana-6247	235	12	generalized	generalize	VERB
cana-6247	235	13	,	,	PUNCT
cana-6247	235	14	and	and	CCONJ
cana-6247	235	15	conditioning	condition	VERB
cana-6247	235	16	on	on	ADP
cana-6247	235	17	,	,	PUNCT
cana-6247	235	18	in	in	ADP
cana-6247	235	19	this	this	DET
cana-6247	235	20	case	case	NOUN
cana-6247	235	21	,	,	PUNCT
cana-6247	235	22	image	image	NOUN
cana-6247	235	23	-	-	PUNCT
cana-6247	235	24	specific	specific	ADJ
cana-6247	235	25	,	,	PUNCT
cana-6247	235	26	degradations	degradation	NOUN
cana-6247	235	27	can	can	AUX
cana-6247	235	28	lead	lead	VERB
cana-6247	235	29	to	to	ADP
cana-6247	235	30	high	high	ADJ
cana-6247	235	31	fidelity	fidelity	NOUN
cana-6247	235	32	restoration	restoration	NOUN
cana-6247	235	33	,	,	PUNCT
cana-6247	235	34	even	even	ADV
cana-6247	235	35	when	when	SCONJ
cana-6247	235	36	the	the	DET
cana-6247	235	37	complications	complication	NOUN
cana-6247	235	38	of	of	ADP
cana-6247	235	39	noise	noise	NOUN
cana-6247	235	40	are	be	AUX
cana-6247	235	41	omitted	omit	VERB
cana-6247	235	42	.	.	PUNCT
cana-6247	236	1	•	•	NUM
cana-6247	236	2	diffblur	diffblur	PROPN
cana-6247	236	3	utilizes	utilize	VERB
cana-6247	236	4	a	a	DET
cana-6247	236	5	latent	latent	ADJ
cana-6247	236	6	kernel	kernel	NOUN
cana-6247	236	7	prediction	prediction	NOUN
cana-6247	236	8	network	network	NOUN
cana-6247	236	9	(	(	PUNCT
cana-6247	236	10	lkpn	lkpn	ADV
cana-6247	236	11	)	)	PUNCT
cana-6247	236	12	to	to	PART
cana-6247	236	13	adaptively	adaptively	ADV
cana-6247	236	14	model	model	VERB
cana-6247	236	15	pixelwise	pixelwise	ADJ
cana-6247	236	16	blur	blur	NOUN
cana-6247	236	17	kernels	kernel	NOUN
cana-6247	236	18	for	for	ADP
cana-6247	236	19	each	each	DET
cana-6247	236	20	input	input	NOUN
cana-6247	236	21	,	,	PUNCT
cana-6247	236	22	and	and	CCONJ
cana-6247	236	23	uses	use	VERB
cana-6247	236	24	this	this	DET
cana-6247	236	25	information	information	NOUN
cana-6247	236	26	in	in	ADP
cana-6247	236	27	the	the	DET
cana-6247	236	28	diffusion	diffusion	NOUN
cana-6247	236	29	process	process	NOUN
cana-6247	236	30	.	.	PUNCT
cana-6247	237	1	the	the	DET
cana-6247	237	2	noise	noise	NOUN
cana-6247	237	3	injection	injection	NOUN
cana-6247	237	4	then	then	ADV
cana-6247	237	5	adapts	adapt	VERB
cana-6247	237	6	accordingly	accordingly	ADV
cana-6247	237	7	,	,	PUNCT
cana-6247	237	8	preserving	preserve	VERB
cana-6247	237	9	sharp	sharp	ADJ
cana-6247	237	10	structures	structure	NOUN
cana-6247	237	11	to	to	PART
cana-6247	237	12	target	target	VERB
cana-6247	237	13	severely	severely	ADV
cana-6247	237	14	blurry	blurry	ADJ
cana-6247	237	15	areas	area	NOUN
cana-6247	237	16	for	for	ADP
cana-6247	237	17	restoration	restoration	NOUN
cana-6247	237	18	,	,	PUNCT
cana-6247	237	19	allowing	allow	VERB
cana-6247	237	20	for	for	ADP
cana-6247	237	21	better	well	ADJ
cana-6247	237	22	detail	detail	NOUN
cana-6247	237	23	and	and	CCONJ
cana-6247	237	24	fidelity	fidelity	NOUN
cana-6247	237	25	across	across	ADP
cana-6247	237	26	both	both	CCONJ
cana-6247	237	27	synthetic	synthetic	ADJ
cana-6247	237	28	and	and	CCONJ
cana-6247	237	29	real	real	ADJ
cana-6247	237	30	-	-	PUNCT
cana-6247	237	31	world	world	NOUN
cana-6247	237	32	benchmarks	benchmark	NOUN
cana-6247	237	33	.	.	PUNCT
cana-6247	238	1	[	[	X
cana-6247	238	2	16	16	NUM
cana-6247	238	3	]	]	SYM
cana-6247	238	4	•	•	NOUN
cana-6247	238	5	difpir	difpir	PROPN
cana-6247	238	6	uses	use	VERB
cana-6247	238	7	a	a	DET
cana-6247	238	8	plug	plug	VERB
cana-6247	238	9	-	-	PUNCT
cana-6247	238	10	and	and	CCONJ
cana-6247	238	11	-	-	PUNCT
cana-6247	238	12	play	play	VERB
cana-6247	238	13	denoising	denoising	NOUN
cana-6247	238	14	process	process	NOUN
cana-6247	238	15	in	in	ADP
cana-6247	238	16	the	the	DET
cana-6247	238	17	diffusion	diffusion	NOUN
cana-6247	238	18	framework	framework	NOUN
cana-6247	238	19	where	where	SCONJ
cana-6247	238	20	the	the	DET
cana-6247	238	21	model	model	NOUN
cana-6247	238	22	can	can	AUX
cana-6247	238	23	use	use	VERB
cana-6247	238	24	prior	prior	ADJ
cana-6247	238	25	knowledge	knowledge	NOUN
cana-6247	238	26	(	(	PUNCT
cana-6247	238	27	e.g.	e.g.	ADV
cana-6247	238	28	,	,	PUNCT
cana-6247	238	29	degradation	degradation	NOUN
cana-6247	238	30	maps	map	NOUN
cana-6247	238	31	/	/	SYM
cana-6247	238	32	blur	blur	NOUN
cana-6247	238	33	)	)	PUNCT
cana-6247	238	34	flexibly	flexibly	ADV
cana-6247	238	35	during	during	ADP
cana-6247	238	36	both	both	CCONJ
cana-6247	238	37	forward	forward	ADV
cana-6247	238	38	noise	noise	NOUN
cana-6247	238	39	scheduling	scheduling	NOUN
cana-6247	238	40	and	and	CCONJ
cana-6247	238	41	reverse	reverse	ADJ
cana-6247	238	42	inference	inference	NOUN
cana-6247	238	43	.	.	PUNCT
cana-6247	239	1	this	this	DET
cana-6247	239	2	results	result	VERB
cana-6247	239	3	in	in	ADP
cana-6247	239	4	a	a	DET
cana-6247	239	5	lossless	lossless	ADJ
cana-6247	239	6	way	way	NOUN
cana-6247	239	7	to	to	PART
cana-6247	239	8	enhance	enhance	VERB
cana-6247	239	9	robustness	robustness	NOUN
cana-6247	239	10	and	and	CCONJ
cana-6247	239	11	perceptual	perceptual	ADJ
cana-6247	239	12	quality	quality	NOUN
cana-6247	239	13	of	of	ADP
cana-6247	239	14	the	the	DET
cana-6247	239	15	source	source	NOUN
cana-6247	239	16	,	,	PUNCT
cana-6247	239	17	while	while	SCONJ
cana-6247	239	18	keeping	keep	VERB
cana-6247	239	19	inference	inference	NOUN
cana-6247	239	20	efficient	efficient	ADJ
cana-6247	239	21	.	.	PUNCT
cana-6247	240	1	[	[	X
cana-6247	240	2	57	57	NUM
cana-6247	240	3	]	]	SYM
cana-6247	240	4	•	•	NUM
cana-6247	240	5	difface	difface	NOUN
cana-6247	240	6	addresses	address	NOUN
cana-6247	240	7	blind	blind	ADJ
cana-6247	240	8	face	face	NOUN
cana-6247	240	9	restoration	restoration	NOUN
cana-6247	240	10	by	by	ADP
cana-6247	240	11	minimizing	minimize	VERB
cana-6247	240	12	restoration	restoration	NOUN
cana-6247	240	13	backbone	backbone	NOUN
cana-6247	240	14	errors	error	NOUN
cana-6247	240	15	through	through	ADP
cana-6247	240	16	a	a	DET
cana-6247	240	17	welldesigned	welldesigne	VERB
cana-6247	240	18	posterior	posterior	ADJ
cana-6247	240	19	transition	transition	NOUN
cana-6247	240	20	(	(	PUNCT
cana-6247	240	21	conditioned	condition	VERB
cana-6247	240	22	on	on	ADP
cana-6247	240	23	the	the	DET
cana-6247	240	24	degraded	degraded	ADJ
cana-6247	240	25	input	input	NOUN
cana-6247	240	26	)	)	PUNCT
cana-6247	240	27	,	,	PUNCT
cana-6247	240	28	resulting	result	VERB
cana-6247	240	29	in	in	ADP
cana-6247	240	30	a	a	DET
cana-6247	240	31	reliable	reliable	ADJ
cana-6247	240	32	restoration	restoration	NOUN
cana-6247	240	33	process	process	NOUN
cana-6247	240	34	even	even	ADV
cana-6247	240	35	for	for	ADP
cana-6247	240	36	unseen	unseen	ADJ
cana-6247	240	37	or	or	CCONJ
cana-6247	240	38	severe	severe	ADJ
cana-6247	240	39	degradations	degradation	NOUN
cana-6247	240	40	that	that	PRON
cana-6247	240	41	highlight	highlight	VERB
cana-6247	240	42	the	the	DET
cana-6247	240	43	power	power	NOUN
cana-6247	240	44	of	of	ADP
cana-6247	240	45	condition	condition	NOUN
cana-6247	240	46	-	-	PUNCT
cana-6247	240	47	aware	aware	ADJ
cana-6247	240	48	diffusion	diffusion	NOUN
cana-6247	240	49	.	.	PUNCT
cana-6247	241	1	[	[	X
cana-6247	241	2	49	49	NUM
cana-6247	241	3	]	]	SYM
cana-6247	241	4	•	•	NUM
cana-6247	241	5	diffbir	diffbir	NOUN
cana-6247	241	6	adopts	adopt	VERB
cana-6247	241	7	a	a	DET
cana-6247	241	8	two	two	NUM
cana-6247	241	9	-	-	PUNCT
cana-6247	241	10	phased	phase	VERB
cana-6247	241	11	approach	approach	NOUN
cana-6247	241	12	:	:	PUNCT
cana-6247	241	13	first	first	ADV
cana-6247	241	14	,	,	PUNCT
cana-6247	241	15	the	the	DET
cana-6247	241	16	image	image	NOUN
cana-6247	241	17	-	-	PUNCT
cana-6247	241	18	specific	specific	ADJ
cana-6247	241	19	degradation	degradation	NOUN
cana-6247	241	20	must	must	AUX
cana-6247	241	21	be	be	AUX
cana-6247	241	22	removed	remove	VERB
cana-6247	241	23	(	(	PUNCT
cana-6247	241	24	based	base	VERB
cana-6247	241	25	on	on	ADP
cana-6247	241	26	the	the	DET
cana-6247	241	27	input	input	NOUN
cana-6247	241	28	)	)	PUNCT
cana-6247	241	29	;	;	PUNCT
cana-6247	241	30	second	second	X
cana-6247	241	31	,	,	PUNCT
cana-6247	241	32	the	the	DET
cana-6247	241	33	missing	miss	VERB
cana-6247	241	34	information	information	NOUN
cana-6247	241	35	is	be	AUX
cana-6247	241	36	generated	generate	VERB
cana-6247	241	37	through	through	ADP
cana-6247	241	38	a	a	DET
cana-6247	241	39	latent	latent	NOUN
cana-6247	241	40	diffusion	diffusion	NOUN
cana-6247	241	41	model	model	NOUN
cana-6247	241	42	and	and	CCONJ
cana-6247	241	43	reconstructed	reconstruct	VERB
cana-6247	241	44	over	over	ADP
cana-6247	241	45	a	a	DET
cana-6247	241	46	diffusion	diffusion	NOUN
cana-6247	241	47	chain	chain	NOUN
cana-6247	241	48	based	base	VERB
cana-6247	241	49	degradation	degradation	NOUN
cana-6247	241	50	-	-	PUNCT
cana-6247	241	51	path	path	NOUN
cana-6247	241	52	controlled	control	VERB
cana-6247	241	53	via	via	ADP
cana-6247	241	54	noise	noise	NOUN
cana-6247	241	55	and	and	CCONJ
cana-6247	241	56	restoration	restoration	NOUN
cana-6247	241	57	modulation	modulation	NOUN
cana-6247	241	58	functions	function	NOUN
cana-6247	241	59	to	to	PART
cana-6247	241	60	allow	allow	VERB
cana-6247	241	61	flexible	flexible	ADJ
cana-6247	241	62	region	region	NOUN
cana-6247	241	63	/	/	SYM
cana-6247	241	64	adaptive	adaptive	ADJ
cana-6247	241	65	denoising	denoising	NOUN
cana-6247	241	66	and	and	CCONJ
cana-6247	241	67	high	high	ADJ
cana-6247	241	68	quality	quality	NOUN
cana-6247	241	69	synthesis	synthesis	NOUN
cana-6247	241	70	for	for	ADP
cana-6247	241	71	blind	blind	ADJ
cana-6247	241	72	restoration	restoration	NOUN
cana-6247	241	73	tasks	task	NOUN
cana-6247	241	74	.	.	PUNCT
cana-6247	242	1	[	[	X
cana-6247	242	2	19	19	NUM
cana-6247	242	3	]	]	SYM
cana-6247	242	4	•	•	NUM
cana-6247	242	5	hierarchical	hierarchical	ADJ
cana-6247	242	6	integration	integration	NOUN
cana-6247	242	7	diffusion	diffusion	NOUN
cana-6247	242	8	model	model	NOUN
cana-6247	242	9	(	(	PUNCT
cana-6247	242	10	hi	hi	NOUN
cana-6247	242	11	-	-	NOUN
cana-6247	242	12	diff	diff	NOUN
cana-6247	242	13	)	)	PUNCT
cana-6247	242	14	executes	execute	VERB
cana-6247	242	15	diffusion	diffusion	NOUN
cana-6247	242	16	in	in	ADP
cana-6247	242	17	a	a	DET
cana-6247	242	18	highly	highly	ADV
cana-6247	242	19	compressed	compressed	ADJ
cana-6247	242	20	latent	latent	NOUN
cana-6247	242	21	representation	representation	NOUN
cana-6247	242	22	,	,	PUNCT
cana-6247	242	23	where	where	SCONJ
cana-6247	242	24	the	the	DET
cana-6247	242	25	deblurring	deblurre	VERB
cana-6247	242	26	phase	phase	NOUN
cana-6247	242	27	integrates	integrate	VERB
cana-6247	242	28	condition	condition	NOUN
cana-6247	242	29	-	-	PUNCT
cana-6247	242	30	adaptive	adaptive	ADJ
cana-6247	242	31	priors	prior	NOUN
cana-6247	242	32	at	at	ADP
cana-6247	242	33	various	various	ADJ
cana-6247	242	34	scales	scale	NOUN
cana-6247	242	35	.	.	PUNCT
cana-6247	243	1	while	while	SCONJ
cana-6247	243	2	other	other	ADJ
cana-6247	243	3	merging	merging	NOUN
cana-6247	243	4	integrations	integration	NOUN
cana-6247	243	5	will	will	AUX
cana-6247	243	6	generate	generate	VERB
cana-6247	243	7	with	with	ADP
cana-6247	243	8	adaptation	adaptation	NOUN
cana-6247	243	9	to	to	PART
cana-6247	243	10	appropriate	appropriate	VERB
cana-6247	243	11	denoising	denoising	NOUN
cana-6247	243	12	priors	prior	NOUN
cana-6247	243	13	,	,	PUNCT
cana-6247	243	14	the	the	DET
cana-6247	243	15	hierarchical	hierarchical	ADJ
cana-6247	243	16	integration	integration	NOUN
cana-6247	243	17	facilitates	facilitate	VERB
cana-6247	243	18	a	a	DET
cana-6247	243	19	more	more	ADV
cana-6247	243	20	effective	effective	ADJ
cana-6247	243	21	generalization	generalization	NOUN
cana-6247	243	22	and	and	CCONJ
cana-6247	243	23	efficiency	efficiency	NOUN
cana-6247	243	24	across	across	ADP
cana-6247	243	25	complex	complex	ADJ
cana-6247	243	26	blurry	blurry	ADJ
cana-6247	243	27	scenarios	scenario	NOUN
cana-6247	243	28	since	since	SCONJ
cana-6247	243	29	the	the	DET
cana-6247	243	30	priors	prior	NOUN
cana-6247	243	31	and	and	CCONJ
cana-6247	243	32	thus	thus	ADV
cana-6247	243	33	noise	noise	NOUN
cana-6247	243	34	process	process	NOUN
cana-6247	243	35	are	be	AUX
cana-6247	243	36	conditioned	condition	VERB
cana-6247	243	37	explicitly	explicitly	ADV
cana-6247	243	38	on	on	ADP
cana-6247	243	39	the	the	DET
cana-6247	243	40	input	input	NOUN
cana-6247	243	41	blur	blur	NOUN
cana-6247	243	42	characteristics	characteristic	NOUN
cana-6247	243	43	.	.	PUNCT
cana-6247	244	1	[	[	X
cana-6247	244	2	5	5	NUM
cana-6247	244	3	]	]	PUNCT
cana-6247	244	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	244	5	988	988	NUM
cana-6247	244	6	communications	communication	NOUN
cana-6247	244	7	on	on	ADP
cana-6247	244	8	applied	apply	VERB
cana-6247	244	9	nonlinear	nonlinear	ADJ
cana-6247	244	10	analysis	analysis	NOUN
cana-6247	244	11	issn	issn	NOUN
cana-6247	244	12	:	:	PUNCT
cana-6247	244	13	1074	1074	NUM
cana-6247	244	14	-	-	PUNCT
cana-6247	244	15	133x	133x	NUM
cana-6247	244	16	vol	vol	NOUN
cana-6247	244	17	.	.	PROPN
cana-6247	245	1	31	31	NUM
cana-6247	246	1	no	no	NOUN
cana-6247	246	2	.	.	PUNCT
cana-6247	247	1	7s	7	NOUN
cana-6247	247	2	(	(	PUNCT
cana-6247	247	3	2024	2024	NUM
cana-6247	247	4	)	)	PUNCT
cana-6247	247	5	in	in	ADP
cana-6247	247	6	sum	sum	NOUN
cana-6247	247	7	.	.	PUNCT
cana-6247	248	1	conditional	conditional	ADJ
cana-6247	248	2	noise	noise	NOUN
cana-6247	248	3	injection	injection	NOUN
cana-6247	248	4	for	for	ADP
cana-6247	248	5	image	image	NOUN
cana-6247	248	6	deblurring	deblurring	NOUN
cana-6247	248	7	can	can	AUX
cana-6247	248	8	use	use	VERB
cana-6247	248	9	specific	specific	ADJ
cana-6247	248	10	information	information	NOUN
cana-6247	248	11	about	about	ADP
cana-6247	248	12	the	the	DET
cana-6247	248	13	input	input	NOUN
cana-6247	248	14	through	through	ADP
cana-6247	248	15	one	one	NUM
cana-6247	248	16	or	or	CCONJ
cana-6247	248	17	more	more	ADV
cana-6247	248	18	estimated	estimate	VERB
cana-6247	248	19	blur	blur	NOUN
cana-6247	248	20	kernels	kernel	NOUN
cana-6247	248	21	,	,	PUNCT
cana-6247	248	22	degradation	degradation	NOUN
cana-6247	248	23	maps	map	NOUN
cana-6247	248	24	,	,	PUNCT
cana-6247	248	25	or	or	CCONJ
cana-6247	248	26	edge	edge	VERB
cana-6247	248	27	information	information	NOUN
cana-6247	248	28	to	to	PART
cana-6247	248	29	dynamically	dynamically	ADV
cana-6247	248	30	bypass	bypass	VERB
cana-6247	248	31	,	,	PUNCT
cana-6247	248	32	modulate	modulate	ADJ
cana-6247	248	33	,	,	PUNCT
cana-6247	248	34	or	or	CCONJ
cana-6247	248	35	control	control	NOUN
cana-6247	248	36	noise	noise	NOUN
cana-6247	248	37	injection	injection	NOUN
cana-6247	248	38	and	and	CCONJ
cana-6247	248	39	carry	carry	VERB
cana-6247	248	40	out	out	ADP
cana-6247	248	41	restoration	restoration	NOUN
cana-6247	248	42	on	on	ADP
cana-6247	248	43	image	image	NOUN
cana-6247	248	44	zones	zone	NOUN
cana-6247	248	45	with	with	ADP
cana-6247	248	46	the	the	DET
cana-6247	248	47	worst	bad	ADJ
cana-6247	248	48	blur	blur	PROPN
cana-6247	248	49	.	.	PUNCT
cana-6247	248	50	cold	cold	ADJ
cana-6247	248	51	diffusion	diffusion	NOUN
cana-6247	248	52	is	be	AUX
cana-6247	248	53	a	a	DET
cana-6247	248	54	more	more	ADV
cana-6247	248	55	generalized	generalized	ADJ
cana-6247	248	56	version	version	NOUN
cana-6247	248	57	of	of	ADP
cana-6247	248	58	the	the	DET
cana-6247	248	59	standard	standard	ADJ
cana-6247	248	60	forward	forward	ADJ
cana-6247	248	61	diffusion	diffusion	NOUN
cana-6247	248	62	process	process	NOUN
cana-6247	248	63	,	,	PUNCT
cana-6247	248	64	where	where	SCONJ
cana-6247	248	65	deterministic	deterministic	ADJ
cana-6247	248	66	degradations	degradation	NOUN
cana-6247	248	67	like	like	ADP
cana-6247	248	68	blur	blur	NOUN
cana-6247	248	69	are	be	AUX
cana-6247	248	70	considered	consider	VERB
cana-6247	248	71	for	for	ADP
cana-6247	248	72	better	well	ADJ
cana-6247	248	73	restoration	restoration	NOUN
cana-6247	248	74	of	of	ADP
cana-6247	248	75	image	image	NOUN
cana-6247	248	76	-	-	PUNCT
cana-6247	248	77	specific	specific	ADJ
cana-6247	248	78	degradations	degradation	NOUN
cana-6247	248	79	.	.	PUNCT
cana-6247	249	1	diffblur	diffblur	PROPN
cana-6247	249	2	performs	perform	VERB
cana-6247	249	3	pixelwise	pixelwise	ADJ
cana-6247	249	4	blur	blur	NOUN
cana-6247	249	5	modeling	modeling	NOUN
cana-6247	249	6	variably	variably	ADV
cana-6247	249	7	through	through	ADP
cana-6247	249	8	a	a	DET
cana-6247	249	9	latent	latent	ADJ
cana-6247	249	10	kernel	kernel	NOUN
cana-6247	249	11	prediction	prediction	NOUN
cana-6247	249	12	network	network	NOUN
cana-6247	249	13	to	to	PART
cana-6247	249	14	introduce	introduce	VERB
cana-6247	249	15	noise	noise	NOUN
cana-6247	249	16	imperfectly	imperfectly	ADV
cana-6247	249	17	under	under	ADP
cana-6247	249	18	a	a	DET
cana-6247	249	19	certain	certain	ADJ
cana-6247	249	20	condition	condition	NOUN
cana-6247	249	21	while	while	SCONJ
cana-6247	249	22	preserving	preserve	VERB
cana-6247	249	23	sharp	sharp	ADJ
cana-6247	249	24	structure	structure	NOUN
cana-6247	249	25	and	and	CCONJ
cana-6247	249	26	restoring	restore	VERB
cana-6247	249	27	fine	fine	ADJ
cana-6247	249	28	details	detail	NOUN
cana-6247	249	29	[	[	X
cana-6247	249	30	16	16	NUM
cana-6247	249	31	]	]	PUNCT
cana-6247	249	32	.	.	PUNCT
cana-6247	250	1	the	the	DET
cana-6247	250	2	posteriors	posterior	NOUN
cana-6247	250	3	,	,	PUNCT
cana-6247	250	4	like	like	ADP
cana-6247	250	5	diffpir	diffpir	NOUN
cana-6247	250	6	,	,	PUNCT
cana-6247	250	7	can	can	AUX
cana-6247	250	8	use	use	VERB
cana-6247	250	9	degradation	degradation	NOUN
cana-6247	250	10	priors	prior	NOUN
cana-6247	250	11	from	from	ADP
cana-6247	250	12	the	the	DET
cana-6247	250	13	initial	initial	ADJ
cana-6247	250	14	noise	noise	NOUN
cana-6247	250	15	scheduling	scheduling	NOUN
cana-6247	250	16	step	step	NOUN
cana-6247	250	17	up	up	ADP
cana-6247	250	18	to	to	ADP
cana-6247	250	19	the	the	DET
cana-6247	250	20	denoising	denoise	VERB
cana-6247	250	21	step	step	NOUN
cana-6247	250	22	to	to	PART
cana-6247	250	23	enhance	enhance	VERB
cana-6247	250	24	robustness	robustness	NOUN
cana-6247	250	25	and	and	CCONJ
cana-6247	250	26	perceptual	perceptual	ADJ
cana-6247	250	27	quality	quality	NOUN
cana-6247	250	28	[	[	X
cana-6247	250	29	8	8	NUM
cana-6247	250	30	]	]	PUNCT
cana-6247	250	31	.	.	PUNCT
cana-6247	251	1	difface	difface	PROPN
cana-6247	251	2	is	be	AUX
cana-6247	251	3	aimed	aim	VERB
cana-6247	251	4	toward	toward	ADP
cana-6247	251	5	blind	blind	ADJ
cana-6247	251	6	-	-	PUNCT
cana-6247	251	7	aware	aware	ADJ
cana-6247	251	8	restoration	restoration	NOUN
cana-6247	251	9	of	of	ADP
cana-6247	251	10	face	face	NOUN
cana-6247	251	11	images	image	NOUN
cana-6247	251	12	by	by	ADP
cana-6247	251	13	degraded	degraded	ADJ
cana-6247	251	14	input	input	NOUN
cana-6247	251	15	information	information	NOUN
cana-6247	251	16	and	and	CCONJ
cana-6247	251	17	has	have	AUX
cana-6247	251	18	been	be	AUX
cana-6247	251	19	demonstrated	demonstrate	VERB
cana-6247	251	20	to	to	PART
cana-6247	251	21	generalize	generalize	VERB
cana-6247	251	22	well	well	ADV
cana-6247	251	23	to	to	ADP
cana-6247	251	24	unseen	unseen	ADJ
cana-6247	251	25	or	or	CCONJ
cana-6247	251	26	hard	hard	ADJ
cana-6247	251	27	blur	blur	NOUN
cana-6247	251	28	in	in	ADP
cana-6247	251	29	testing	testing	NOUN
cana-6247	251	30	[	[	X
cana-6247	251	31	49	49	NUM
cana-6247	251	32	]	]	PUNCT
cana-6247	252	1	[	[	X
cana-6247	252	2	5	5	NUM
cana-6247	252	3	]	]	PUNCT
cana-6247	252	4	.	.	PUNCT
cana-6247	253	1	an	an	DET
cana-6247	253	2	elaborate	elaborate	ADJ
cana-6247	253	3	situation	situation	NOUN
cana-6247	253	4	having	have	VERB
cana-6247	253	5	diffbir	diffbir	NOUN
cana-6247	253	6	and	and	CCONJ
cana-6247	253	7	hi	hi	NOUN
cana-6247	253	8	-	-	NOUN
cana-6247	253	9	diff	diff	NOUN
cana-6247	253	10	comprises	comprise	VERB
cana-6247	253	11	the	the	DET
cana-6247	253	12	adaptive	adaptive	ADJ
cana-6247	253	13	prior	prior	ADJ
cana-6247	253	14	condition	condition	NOUN
cana-6247	253	15	with	with	ADP
cana-6247	253	16	multi	multi	ADJ
cana-6247	253	17	-	-	NOUN
cana-6247	253	18	scale	scale	ADJ
cana-6247	253	19	or	or	CCONJ
cana-6247	253	20	hierarchical	hierarchical	ADJ
cana-6247	253	21	latent	latent	NOUN
cana-6247	253	22	-	-	PUNCT
cana-6247	253	23	diffusion	diffusion	NOUN
cana-6247	253	24	set	set	NOUN
cana-6247	253	25	-	-	PUNCT
cana-6247	253	26	ups	up	NOUN
cana-6247	253	27	for	for	ADP
cana-6247	253	28	an	an	DET
cana-6247	253	29	efficient	efficient	ADJ
cana-6247	253	30	,	,	PUNCT
cana-6247	253	31	scalable	scalable	ADJ
cana-6247	253	32	image	image	NOUN
cana-6247	253	33	restoration	restoration	NOUN
cana-6247	253	34	method	method	NOUN
cana-6247	253	35	for	for	ADP
cana-6247	253	36	many	many	ADJ
cana-6247	253	37	categories	category	NOUN
cana-6247	253	38	of	of	ADP
cana-6247	253	39	blur	blur	PROPN
cana-6247	253	40	,	,	PUNCT
cana-6247	253	41	even	even	ADV
cana-6247	253	42	those	those	PRON
cana-6247	253	43	spatially	spatially	ADV
cana-6247	253	44	varying	vary	VERB
cana-6247	253	45	[	[	X
cana-6247	253	46	19	19	NUM
cana-6247	253	47	]	]	PUNCT
cana-6247	253	48	.	.	PUNCT
cana-6247	254	1	all	all	PRON
cana-6247	254	2	taken	take	VERB
cana-6247	254	3	together	together	ADV
cana-6247	254	4	,	,	PUNCT
cana-6247	254	5	such	such	ADJ
cana-6247	254	6	work	work	NOUN
cana-6247	254	7	suggests	suggest	VERB
cana-6247	254	8	that	that	SCONJ
cana-6247	254	9	such	such	DET
cana-6247	254	10	a	a	DET
cana-6247	254	11	dependence	dependence	NOUN
cana-6247	254	12	on	on	ADP
cana-6247	254	13	data	datum	NOUN
cana-6247	254	14	and	and	CCONJ
cana-6247	254	15	spatial	spatial	ADJ
cana-6247	254	16	awareness	awareness	NOUN
cana-6247	254	17	to	to	PART
cana-6247	254	18	noise	noise	VERB
cana-6247	254	19	injection	injection	NOUN
cana-6247	254	20	via	via	ADP
cana-6247	254	21	cni	cni	PROPN
cana-6247	254	22	is	be	AUX
cana-6247	254	23	central	central	ADJ
cana-6247	254	24	to	to	ADP
cana-6247	254	25	successful	successful	ADJ
cana-6247	254	26	deblurring	deblurring	NOUN
cana-6247	254	27	,	,	PUNCT
cana-6247	254	28	more	more	ADV
cana-6247	254	29	so	so	ADV
cana-6247	254	30	toward	toward	ADP
cana-6247	254	31	a	a	DET
cana-6247	254	32	sharp	sharp	ADJ
cana-6247	254	33	,	,	PUNCT
cana-6247	254	34	artifact	artifact	NOUN
cana-6247	254	35	-	-	PUNCT
cana-6247	254	36	free	free	ADJ
cana-6247	254	37	,	,	PUNCT
cana-6247	254	38	and	and	CCONJ
cana-6247	254	39	more	more	ADV
cana-6247	254	40	faithful	faithful	ADJ
cana-6247	254	41	and	and	CCONJ
cana-6247	254	42	generalized	generalized	ADJ
cana-6247	254	43	reconstruction	reconstruction	NOUN
cana-6247	254	44	of	of	ADP
cana-6247	254	45	the	the	DET
cana-6247	254	46	blur	blur	NOUN
cana-6247	254	47	introduced	introduce	VERB
cana-6247	254	48	by	by	ADP
cana-6247	254	49	the	the	DET
cana-6247	254	50	world	world	NOUN
cana-6247	254	51	in	in	ADP
cana-6247	254	52	real	real	ADJ
cana-6247	254	53	-	-	PUNCT
cana-6247	254	54	life	life	NOUN
cana-6247	254	55	scenarios	scenario	NOUN
cana-6247	254	56	.	.	PUNCT
cana-6247	255	1	3	3	NUM
cana-6247	255	2	learned	learn	VERB
cana-6247	255	3	noise	noise	NOUN
cana-6247	255	4	a	a	DET
cana-6247	255	5	learnt	learnt	ADJ
cana-6247	255	6	noise	noise	NOUN
cana-6247	255	7	diffusion	diffusion	NOUN
cana-6247	255	8	model	model	NOUN
cana-6247	255	9	is	be	AUX
cana-6247	255	10	a	a	DET
cana-6247	255	11	subset	subset	NOUN
cana-6247	255	12	of	of	ADP
cana-6247	255	13	the	the	DET
cana-6247	255	14	diffusion	diffusion	NOUN
cana-6247	255	15	model	model	NOUN
cana-6247	255	16	family	family	NOUN
cana-6247	255	17	in	in	ADP
cana-6247	255	18	which	which	PRON
cana-6247	255	19	the	the	DET
cana-6247	255	20	noise	noise	NOUN
cana-6247	255	21	process	process	NOUN
cana-6247	255	22	is	be	AUX
cana-6247	255	23	parameterized	parameterized	ADJ
cana-6247	255	24	and	and	CCONJ
cana-6247	255	25	modified	modify	VERB
cana-6247	255	26	—	—	PUNCT
cana-6247	255	27	learned	learn	VERB
cana-6247	255	28	directly	directly	ADV
cana-6247	255	29	from	from	ADP
cana-6247	255	30	data	datum	NOUN
cana-6247	255	31	and	and	CCONJ
cana-6247	255	32	task	task	NOUN
cana-6247	255	33	context	context	NOUN
cana-6247	255	34	during	during	ADP
cana-6247	255	35	model	model	NOUN
cana-6247	255	36	training	training	NOUN
cana-6247	255	37	—	—	PUNCT
cana-6247	255	38	instead	instead	ADV
cana-6247	255	39	of	of	ADP
cana-6247	255	40	being	be	AUX
cana-6247	255	41	fixed	fix	VERB
cana-6247	255	42	(	(	PUNCT
cana-6247	255	43	e.g.	e.g.	ADV
cana-6247	255	44	,	,	PUNCT
cana-6247	255	45	conventional	conventional	ADJ
cana-6247	255	46	isotropic	isotropic	ADJ
cana-6247	255	47	gaussian	gaussian	NOUN
cana-6247	255	48	at	at	ADP
cana-6247	255	49	every	every	DET
cana-6247	255	50	timestep	timestep	NOUN
cana-6247	255	51	)	)	PUNCT
cana-6247	256	1	[	[	X
cana-6247	256	2	35	35	NUM
cana-6247	256	3	]	]	X
cana-6247	257	1	[	[	X
cana-6247	257	2	51	51	NUM
cana-6247	257	3	]	]	PUNCT
cana-6247	257	4	.	.	PUNCT
cana-6247	258	1	diffusion	diffusion	NOUN
cana-6247	258	2	models	model	NOUN
cana-6247	258	3	can	can	AUX
cana-6247	258	4	now	now	ADV
cana-6247	258	5	inject	inject	VERB
cana-6247	258	6	,	,	PUNCT
cana-6247	258	7	modulate	modulate	ADJ
cana-6247	258	8	,	,	PUNCT
cana-6247	258	9	and	and	CCONJ
cana-6247	258	10	remove	remove	VERB
cana-6247	258	11	noise	noise	NOUN
cana-6247	258	12	in	in	ADP
cana-6247	258	13	a	a	DET
cana-6247	258	14	way	way	NOUN
cana-6247	258	15	that	that	PRON
cana-6247	258	16	is	be	AUX
cana-6247	258	17	dynamically	dynamically	ADV
cana-6247	258	18	matched	match	VERB
cana-6247	258	19	to	to	ADP
cana-6247	258	20	the	the	DET
cana-6247	258	21	content	content	NOUN
cana-6247	258	22	and	and	CCONJ
cana-6247	258	23	context	context	NOUN
cana-6247	258	24	of	of	ADP
cana-6247	258	25	the	the	DET
cana-6247	258	26	image	image	NOUN
cana-6247	258	27	being	be	AUX
cana-6247	258	28	restored	restore	VERB
cana-6247	258	29	,	,	PUNCT
cana-6247	258	30	synthesized	synthesize	VERB
cana-6247	258	31	,	,	PUNCT
cana-6247	258	32	or	or	CCONJ
cana-6247	258	33	upsampled	upsample	VERB
cana-6247	258	34	[	[	X
cana-6247	258	35	17	17	NUM
cana-6247	258	36	]	]	X
cana-6247	259	1	[	[	X
cana-6247	259	2	44	44	NUM
cana-6247	259	3	]	]	PUNCT
cana-6247	259	4	.	.	PUNCT
cana-6247	260	1	this	this	DET
cana-6247	260	2	approach	approach	NOUN
cana-6247	260	3	theoretically	theoretically	ADV
cana-6247	260	4	and	and	CCONJ
cana-6247	260	5	mathematically	mathematically	ADV
cana-6247	260	6	removes	remove	VERB
cana-6247	260	7	the	the	DET
cana-6247	260	8	strong	strong	ADJ
cana-6247	260	9	simplifying	simplifying	NOUN
cana-6247	260	10	assumption	assumption	NOUN
cana-6247	260	11	that	that	SCONJ
cana-6247	260	12	noise	noise	NOUN
cana-6247	260	13	must	must	AUX
cana-6247	260	14	be	be	AUX
cana-6247	260	15	static	static	ADJ
cana-6247	260	16	and	and	CCONJ
cana-6247	260	17	content	content	NOUN
cana-6247	260	18	-	-	ADJ
cana-6247	260	19	agnostic	agnostic	ADJ
cana-6247	260	20	.	.	PUNCT
cana-6247	261	1	the	the	DET
cana-6247	261	2	forward	forward	ADJ
cana-6247	261	3	process	process	NOUN
cana-6247	261	4	in	in	ADP
cana-6247	261	5	classical	classical	ADJ
cana-6247	261	6	diffusion	diffusion	NOUN
cana-6247	261	7	models	model	NOUN
cana-6247	261	8	applies	apply	VERB
cana-6247	261	9	predetermined	predetermine	VERB
cana-6247	261	10	,	,	PUNCT
cana-6247	261	11	typically	typically	ADV
cana-6247	261	12	white	white	ADJ
cana-6247	261	13	gaussian	gaussian	NOUN
cana-6247	261	14	noise	noise	NOUN
cana-6247	261	15	incrementally	incrementally	ADV
cana-6247	261	16	at	at	ADP
cana-6247	261	17	each	each	DET
cana-6247	261	18	timestep	timestep	NOUN
cana-6247	261	19	to	to	PART
cana-6247	261	20	turn	turn	VERB
cana-6247	261	21	a	a	DET
cana-6247	261	22	data	data	NOUN
cana-6247	261	23	sample	sample	NOUN
cana-6247	261	24	(	(	PUNCT
cana-6247	261	25	such	such	DET
cana-6247	261	26	an	an	DET
cana-6247	261	27	image	image	NOUN
cana-6247	261	28	)	)	PUNCT
cana-6247	261	29	into	into	ADP
cana-6247	261	30	pure	pure	ADJ
cana-6247	261	31	noise	noise	NOUN
cana-6247	261	32	.	.	PUNCT
cana-6247	262	1	given	give	VERB
cana-6247	262	2	that	that	SCONJ
cana-6247	262	3	it	it	PRON
cana-6247	262	4	treats	treat	VERB
cana-6247	262	5	all	all	DET
cana-6247	262	6	pixels	pixel	NOUN
cana-6247	262	7	,	,	PUNCT
cana-6247	262	8	spatial	spatial	ADJ
cana-6247	262	9	regions	region	NOUN
cana-6247	262	10	,	,	PUNCT
cana-6247	262	11	and	and	CCONJ
cana-6247	262	12	image	image	NOUN
cana-6247	262	13	contexts	context	NOUN
cana-6247	262	14	as	as	ADP
cana-6247	262	15	equally	equally	ADV
cana-6247	262	16	uncertain	uncertain	ADJ
cana-6247	262	17	and	and	CCONJ
cana-6247	262	18	challenging	challenging	ADJ
cana-6247	262	19	to	to	PART
cana-6247	262	20	reconstruct	reconstruct	VERB
cana-6247	262	21	,	,	PUNCT
cana-6247	262	22	this	this	DET
cana-6247	262	23	standardization	standardization	NOUN
cana-6247	262	24	may	may	AUX
cana-6247	262	25	be	be	AUX
cana-6247	262	26	less	less	ADJ
cana-6247	262	27	-	-	PUNCT
cana-6247	262	28	than	than	ADP
cana-6247	262	29	-	-	PUNCT
cana-6247	262	30	ideal	ideal	ADJ
cana-6247	262	31	,	,	PUNCT
cana-6247	262	32	particularly	particularly	ADV
cana-6247	262	33	for	for	ADP
cana-6247	262	34	tasks	task	NOUN
cana-6247	262	35	like	like	ADP
cana-6247	262	36	super	super	NOUN
cana-6247	262	37	-	-	NOUN
cana-6247	262	38	resolution	resolution	NOUN
cana-6247	262	39	,	,	PUNCT
cana-6247	262	40	inpainting	inpainting	NOUN
cana-6247	262	41	,	,	PUNCT
cana-6247	262	42	or	or	CCONJ
cana-6247	262	43	restoration	restoration	NOUN
cana-6247	262	44	.	.	PUNCT
cana-6247	263	1	however	however	ADV
cana-6247	263	2	,	,	PUNCT
cana-6247	263	3	it	it	PRON
cana-6247	263	4	does	do	AUX
cana-6247	263	5	make	make	VERB
cana-6247	263	6	training	training	NOUN
cana-6247	263	7	and	and	CCONJ
cana-6247	263	8	theoretical	theoretical	ADJ
cana-6247	263	9	analysis	analysis	NOUN
cana-6247	263	10	easier	easy	ADJ
cana-6247	263	11	.	.	PUNCT
cana-6247	264	1	this	this	DET
cana-6247	264	2	limitation	limitation	NOUN
cana-6247	264	3	is	be	AUX
cana-6247	264	4	addressed	address	VERB
cana-6247	264	5	by	by	ADP
cana-6247	264	6	the	the	DET
cana-6247	264	7	idea	idea	NOUN
cana-6247	264	8	of	of	ADP
cana-6247	264	9	learned	learn	VERB
cana-6247	264	10	noise	noise	NOUN
cana-6247	264	11	,	,	PUNCT
cana-6247	264	12	which	which	PRON
cana-6247	264	13	parameterizes	parameterize	VERB
cana-6247	264	14	the	the	DET
cana-6247	264	15	forward	forward	ADJ
cana-6247	264	16	noise	noise	NOUN
cana-6247	264	17	process	process	NOUN
cana-6247	264	18	as	as	ADP
cana-6247	264	19	a	a	DET
cana-6247	264	20	learnable	learnable	ADJ
cana-6247	264	21	function	function	NOUN
cana-6247	264	22	as	as	SCONJ
cana-6247	264	23	opposed	oppose	VERB
cana-6247	264	24	to	to	ADP
cana-6247	264	25	treating	treat	VERB
cana-6247	264	26	it	it	PRON
cana-6247	264	27	as	as	ADP
cana-6247	264	28	immutable	immutable	ADJ
cana-6247	264	29	[	[	X
cana-6247	264	30	35	35	NUM
cana-6247	264	31	]	]	PUNCT
cana-6247	264	32	.	.	PUNCT
cana-6247	265	1	the	the	DET
cana-6247	265	2	input	input	NOUN
cana-6247	265	3	itself	itself	PRON
cana-6247	265	4	,	,	PUNCT
cana-6247	265	5	uncertainty	uncertainty	NOUN
cana-6247	265	6	maps	map	NOUN
cana-6247	265	7	,	,	PUNCT
cana-6247	265	8	semantic	semantic	ADJ
cana-6247	265	9	labels	label	NOUN
cana-6247	265	10	,	,	PUNCT
cana-6247	265	11	structural	structural	ADJ
cana-6247	265	12	cues	cue	NOUN
cana-6247	265	13	,	,	PUNCT
cana-6247	265	14	degradation	degradation	NOUN
cana-6247	265	15	levels	level	NOUN
cana-6247	265	16	,	,	PUNCT
cana-6247	265	17	or	or	CCONJ
cana-6247	265	18	any	any	DET
cana-6247	265	19	other	other	ADJ
cana-6247	265	20	domainrelevant	domainrelevant	ADJ
cana-6247	265	21	signal	signal	NOUN
cana-6247	265	22	may	may	AUX
cana-6247	265	23	all	all	PRON
cana-6247	265	24	influence	influence	VERB
cana-6247	265	25	this	this	DET
cana-6247	265	26	function	function	NOUN
cana-6247	265	27	[	[	X
cana-6247	265	28	41	41	NUM
cana-6247	265	29	]	]	X
cana-6247	266	1	[	[	X
cana-6247	266	2	44	44	NUM
cana-6247	266	3	]	]	PUNCT
cana-6247	266	4	.	.	PUNCT
cana-6247	267	1	as	as	ADP
cana-6247	267	2	a	a	DET
cana-6247	267	3	result	result	NOUN
cana-6247	267	4	,	,	PUNCT
cana-6247	267	5	the	the	DET
cana-6247	267	6	model	model	NOUN
cana-6247	267	7	can	can	AUX
cana-6247	267	8	"	"	PUNCT
cana-6247	267	9	inject	inject	VERB
cana-6247	267	10	"	"	PUNCT
cana-6247	267	11	more	more	ADJ
cana-6247	267	12	randomness	randomness	NOUN
cana-6247	267	13	where	where	SCONJ
cana-6247	267	14	the	the	DET
cana-6247	267	15	mapping	mapping	NOUN
cana-6247	267	16	from	from	ADP
cana-6247	267	17	degraded	degrade	VERB
cana-6247	267	18	to	to	ADP
cana-6247	267	19	high	high	ADJ
cana-6247	267	20	-	-	PUNCT
cana-6247	267	21	fidelity	fidelity	NOUN
cana-6247	267	22	content	content	NOUN
cana-6247	267	23	is	be	AUX
cana-6247	267	24	unclear	unclear	ADJ
cana-6247	267	25	or	or	CCONJ
cana-6247	267	26	underdetermined	underdetermined	ADJ
cana-6247	267	27	,	,	PUNCT
cana-6247	267	28	and	and	CCONJ
cana-6247	267	29	less	less	ADV
cana-6247	267	30	where	where	SCONJ
cana-6247	267	31	context	context	NOUN
cana-6247	267	32	constraints	constraint	NOUN
cana-6247	267	33	are	be	AUX
cana-6247	267	34	tight	tight	ADJ
cana-6247	267	35	or	or	CCONJ
cana-6247	267	36	features	feature	NOUN
cana-6247	267	37	are	be	AUX
cana-6247	267	38	well	well	ADV
cana-6247	267	39	-	-	PUNCT
cana-6247	267	40	predicted	predict	VERB
cana-6247	267	41	[	[	X
cana-6247	267	42	17	17	NUM
cana-6247	267	43	]	]	PUNCT
cana-6247	267	44	.	.	PUNCT
cana-6247	268	1	noise	noise	NOUN
cana-6247	268	2	can	can	AUX
cana-6247	268	3	also	also	ADV
cana-6247	268	4	vary	vary	VERB
cana-6247	268	5	in	in	ADP
cana-6247	268	6	space	space	NOUN
cana-6247	268	7	,	,	PUNCT
cana-6247	268	8	time	time	NOUN
cana-6247	268	9	,	,	PUNCT
cana-6247	268	10	and	and	CCONJ
cana-6247	268	11	context	context	NOUN
cana-6247	268	12	.	.	PUNCT
cana-6247	269	1	as	as	SCONJ
cana-6247	269	2	the	the	DET
cana-6247	269	3	denoising	denoising	NOUN
cana-6247	269	4	procedure	procedure	NOUN
cana-6247	269	5	is	be	AUX
cana-6247	269	6	adjusted	adjust	VERB
cana-6247	269	7	to	to	ADP
cana-6247	269	8	the	the	DET
cana-6247	269	9	precise	precise	ADJ
cana-6247	269	10	geometry	geometry	NOUN
cana-6247	269	11	and	and	CCONJ
cana-6247	269	12	statistics	statistic	NOUN
cana-6247	269	13	of	of	ADP
cana-6247	269	14	the	the	DET
cana-6247	269	15	data	datum	NOUN
cana-6247	269	16	,	,	PUNCT
cana-6247	269	17	this	this	PRON
cana-6247	269	18	results	result	VERB
cana-6247	269	19	in	in	ADP
cana-6247	269	20	increased	increase	VERB
cana-6247	269	21	model	model	NOUN
cana-6247	269	22	flexibility	flexibility	NOUN
cana-6247	269	23	,	,	PUNCT
cana-6247	269	24	better	well	ADJ
cana-6247	269	25	uncertainty	uncertainty	NOUN
cana-6247	269	26	modeling	modeling	NOUN
cana-6247	269	27	,	,	PUNCT
cana-6247	269	28	and	and	CCONJ
cana-6247	269	29	enhanced	enhanced	ADJ
cana-6247	269	30	restoration	restoration	NOUN
cana-6247	269	31	performance	performance	NOUN
cana-6247	269	32	.	.	PUNCT
cana-6247	270	1	the	the	DET
cana-6247	270	2	standard	standard	ADJ
cana-6247	270	3	diffusion	diffusion	NOUN
cana-6247	270	4	forward	forward	ADJ
cana-6247	270	5	process	process	NOUN
cana-6247	270	6	is	be	AUX
cana-6247	270	7	defined	define	VERB
cana-6247	270	8	as	as	ADP
cana-6247	270	9	a	a	DET
cana-6247	270	10	markov	markov	NOUN
cana-6247	270	11	chain	chain	NOUN
cana-6247	270	12	,	,	PUNCT
cana-6247	270	13	as	as	SCONJ
cana-6247	270	14	shown	show	VERB
cana-6247	270	15	in	in	ADP
cana-6247	270	16	eq	eq	ADP
cana-6247	270	17	.	.	PUNCT
cana-6247	271	1	(	(	PUNCT
cana-6247	271	2	1	1	X
cana-6247	271	3	)	)	PUNCT
cana-6247	271	4	where	where	SCONJ
cana-6247	271	5	βt	βt	NOUN
cana-6247	271	6	is	be	AUX
cana-6247	271	7	the	the	DET
cana-6247	271	8	(	(	PUNCT
cana-6247	271	9	typically	typically	ADV
cana-6247	271	10	hand	hand	NOUN
cana-6247	271	11	-	-	PUNCT
cana-6247	271	12	designed	design	VERB
cana-6247	271	13	)	)	PUNCT
cana-6247	271	14	noise	noise	NOUN
cana-6247	271	15	variance	variance	NOUN
cana-6247	271	16	schedule	schedule	NOUN
cana-6247	271	17	,	,	PUNCT
cana-6247	271	18	and	and	CCONJ
cana-6247	271	19	i	i	PRON
cana-6247	271	20	is	be	AUX
cana-6247	271	21	the	the	DET
cana-6247	271	22	identity	identity	NOUN
cana-6247	271	23	matrix	matrix	NOUN
cana-6247	271	24	,	,	PUNCT
cana-6247	271	25	yielding	yield	VERB
cana-6247	271	26	isotropic	isotropic	ADJ
cana-6247	271	27	gaussian	gaussian	NOUN
cana-6247	271	28	noise	noise	NOUN
cana-6247	271	29	.	.	PUNCT
cana-6247	272	1	the	the	DET
cana-6247	272	2	cumulative	cumulative	ADJ
cana-6247	272	3	process	process	NOUN
cana-6247	272	4	over	over	ADP
cana-6247	272	5	t	t	NOUN
cana-6247	272	6	steps	step	NOUN
cana-6247	272	7	can	can	AUX
cana-6247	272	8	be	be	AUX
cana-6247	272	9	re	re	VERB
cana-6247	272	10	-	-	VERB
cana-6247	272	11	expressed	express	VERB
cana-6247	272	12	in	in	ADP
cana-6247	272	13	closed	closed	ADJ
cana-6247	272	14	form	form	NOUN
cana-6247	272	15	shown	show	VERB
cana-6247	272	16	in	in	ADP
cana-6247	272	17	eq	eq	ADP
cana-6247	272	18	.	.	PUNCT
cana-6247	273	1	(	(	PUNCT
cana-6247	273	2	2	2	NUM
cana-6247	273	3	)	)	PUNCT
cana-6247	273	4	with	with	ADP
cana-6247	273	5	ᾱt	ᾱt	VERB
cana-6247	273	6	=	=	SYM
cana-6247	273	7	∏t	∏t	NOUN
cana-6247	273	8	s=1(1−	s=1(1−	PROPN
cana-6247	273	9	βs	βs	NUM
cana-6247	273	10	)	)	PUNCT
cana-6247	273	11	.	.	PUNCT
cana-6247	274	1	learned	learn	VERB
cana-6247	274	2	noise	noise	NOUN
cana-6247	274	3	diffusion	diffusion	NOUN
cana-6247	274	4	models	model	NOUN
cana-6247	274	5	generalize	generalize	VERB
cana-6247	274	6	this	this	DET
cana-6247	274	7	framework	framework	NOUN
cana-6247	274	8	by	by	ADP
cana-6247	274	9	making	make	VERB
cana-6247	274	10	both	both	CCONJ
cana-6247	274	11	the	the	DET
cana-6247	274	12	variance	variance	NOUN
cana-6247	274	13	schedule	schedule	NOUN
cana-6247	274	14	and/or	and/or	CCONJ
cana-6247	274	15	the	the	DET
cana-6247	274	16	noise	noise	NOUN
cana-6247	274	17	sampling	sampling	NOUN
cana-6247	274	18	process	process	NOUN
cana-6247	274	19	themselves	themselves	PRON
cana-6247	274	20	functions	function	NOUN
cana-6247	274	21	of	of	ADP
cana-6247	274	22	the	the	DET
cana-6247	274	23	input	input	NOUN
cana-6247	274	24	,	,	PUNCT
cana-6247	274	25	conditioning	condition	VERB
cana-6247	274	26	context	context	NOUN
cana-6247	274	27	,	,	PUNCT
cana-6247	274	28	or	or	CCONJ
cana-6247	274	29	even	even	ADV
cana-6247	274	30	learnable	learnable	ADJ
cana-6247	274	31	parameters	parameter	NOUN
cana-6247	274	32	.	.	PUNCT
cana-6247	275	1	mathematically	mathematically	ADV
cana-6247	275	2	,	,	PUNCT
cana-6247	275	3	the	the	DET
cana-6247	275	4	forward	forward	ADJ
cana-6247	275	5	process	process	NOUN
cana-6247	275	6	becomes	become	VERB
cana-6247	275	7	:	:	PUNCT
cana-6247	275	8	q(xt	q(xt	NOUN
cana-6247	275	9	|	|	ADP
cana-6247	275	10	xt−1	xt−1	PROPN
cana-6247	275	11	,	,	PUNCT
cana-6247	275	12	c	c	NOUN
cana-6247	275	13	)	)	PUNCT
cana-6247	275	14	=	=	SYM
cana-6247	276	1	n	n	CCONJ
cana-6247	276	2	(	(	PUNCT
cana-6247	276	3	xt	xt	X
cana-6247	276	4	;	;	PUNCT
cana-6247	276	5	√	√	PROPN
cana-6247	276	6	1−	1−	NUM
cana-6247	276	7	βt(c)xt−1	βt(c)xt−1	NOUN
cana-6247	276	8	,	,	PUNCT
cana-6247	276	9	βt(c)σ(c	βt(c)σ(c	NUM
cana-6247	276	10	)	)	PUNCT
cana-6247	276	11	)	)	PUNCT
cana-6247	277	1	(	(	PUNCT
cana-6247	277	2	3	3	X
cana-6247	277	3	)	)	PUNCT
cana-6247	277	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	277	5	989	989	NUM
cana-6247	277	6	communications	communication	NOUN
cana-6247	277	7	on	on	ADP
cana-6247	277	8	applied	apply	VERB
cana-6247	277	9	nonlinear	nonlinear	ADJ
cana-6247	277	10	analysis	analysis	NOUN
cana-6247	277	11	issn	issn	NOUN
cana-6247	277	12	:	:	PUNCT
cana-6247	277	13	1074	1074	NUM
cana-6247	277	14	-	-	PUNCT
cana-6247	277	15	133x	133x	NUM
cana-6247	277	16	vol	vol	NOUN
cana-6247	277	17	.	.	PROPN
cana-6247	278	1	31	31	NUM
cana-6247	279	1	no	no	NOUN
cana-6247	279	2	.	.	PUNCT
cana-6247	280	1	7s	7	NOUN
cana-6247	280	2	(	(	PUNCT
cana-6247	280	3	2024	2024	NUM
cana-6247	280	4	)	)	PUNCT
cana-6247	280	5	or	or	CCONJ
cana-6247	280	6	,	,	PUNCT
cana-6247	280	7	more	more	ADV
cana-6247	280	8	flexibly	flexibly	ADV
cana-6247	280	9	,	,	PUNCT
cana-6247	280	10	xt	xt	X
cana-6247	280	11	=	=	SYM
cana-6247	281	1	√	√	ADP
cana-6247	281	2	ᾱt(c)x0	ᾱt(c)x0	PROPN
cana-6247	281	3	+	+	CCONJ
cana-6247	281	4	√	√	PROPN
cana-6247	281	5	1−	1−	NUM
cana-6247	281	6	ᾱt(c	ᾱt(c	VERB
cana-6247	281	7	)	)	PUNCT
cana-6247	281	8	z(c	z(c	PROPN
cana-6247	281	9	)	)	PUNCT
cana-6247	281	10	(	(	PUNCT
cana-6247	281	11	4	4	X
cana-6247	281	12	)	)	PUNCT
cana-6247	281	13	where	where	SCONJ
cana-6247	281	14	:	:	PUNCT
cana-6247	281	15	•	•	NUM
cana-6247	281	16	c	c	NOUN
cana-6247	281	17	is	be	AUX
cana-6247	281	18	any	any	DET
cana-6247	281	19	conditioning	conditioning	NOUN
cana-6247	281	20	variable	variable	NOUN
cana-6247	281	21	(	(	PUNCT
cana-6247	281	22	e.g.	e.g.	ADV
cana-6247	281	23	,	,	PUNCT
cana-6247	281	24	lr	lr	NOUN
cana-6247	281	25	image	image	NOUN
cana-6247	281	26	for	for	ADP
cana-6247	281	27	sr	sr	PROPN
cana-6247	281	28	,	,	PUNCT
cana-6247	281	29	uncertainty	uncertainty	NOUN
cana-6247	281	30	maps	map	NOUN
cana-6247	281	31	,	,	PUNCT
cana-6247	281	32	semantic	semantic	ADJ
cana-6247	281	33	masks	mask	NOUN
cana-6247	281	34	)	)	PUNCT
cana-6247	281	35	.	.	PUNCT
cana-6247	282	1	•	•	NUM
cana-6247	282	2	βt(c	βt(c	NUM
cana-6247	282	3	)	)	PUNCT
cana-6247	282	4	,	,	PUNCT
cana-6247	282	5	σ(c	σ(c	PROPN
cana-6247	282	6	)	)	PUNCT
cana-6247	282	7	,	,	PUNCT
cana-6247	282	8	and	and	CCONJ
cana-6247	282	9	even	even	ADV
cana-6247	282	10	z(c	z(c	NOUN
cana-6247	282	11	)	)	PUNCT
cana-6247	282	12	are	be	AUX
cana-6247	282	13	now	now	ADV
cana-6247	282	14	functions	function	NOUN
cana-6247	282	15	learned	learn	VERB
cana-6247	282	16	from	from	ADP
cana-6247	282	17	data	datum	NOUN
cana-6247	282	18	,	,	PUNCT
cana-6247	282	19	possibly	possibly	ADV
cana-6247	282	20	via	via	ADP
cana-6247	282	21	neural	neural	ADJ
cana-6247	282	22	networks	network	NOUN
cana-6247	282	23	.	.	PUNCT
cana-6247	283	1	•	•	NUM
cana-6247	283	2	in	in	ADP
cana-6247	283	3	the	the	DET
cana-6247	283	4	most	most	ADV
cana-6247	283	5	general	general	ADJ
cana-6247	283	6	case	case	NOUN
cana-6247	283	7	,	,	PUNCT
cana-6247	283	8	the	the	DET
cana-6247	283	9	noise	noise	NOUN
cana-6247	283	10	z	z	PROPN
cana-6247	283	11	itself	itself	PRON
cana-6247	283	12	may	may	AUX
cana-6247	283	13	not	not	PART
cana-6247	283	14	be	be	AUX
cana-6247	283	15	gaussian	gaussian	ADJ
cana-6247	283	16	,	,	PUNCT
cana-6247	283	17	but	but	CCONJ
cana-6247	283	18	sampled	sample	VERB
cana-6247	283	19	from	from	ADP
cana-6247	283	20	a	a	DET
cana-6247	283	21	learned	learn	VERB
cana-6247	283	22	distribution	distribution	NOUN
cana-6247	283	23	tailored	tailor	VERB
cana-6247	283	24	to	to	ADP
cana-6247	283	25	the	the	DET
cana-6247	283	26	application	application	NOUN
cana-6247	283	27	’s	’s	PART
cana-6247	283	28	requirements	requirement	NOUN
cana-6247	283	29	.	.	PUNCT
cana-6247	284	1	in	in	ADP
cana-6247	284	2	structure	structure	NOUN
cana-6247	284	3	-	-	PUNCT
cana-6247	284	4	aware	aware	ADJ
cana-6247	284	5	super	super	NOUN
cana-6247	284	6	-	-	NOUN
cana-6247	284	7	resolution	resolution	NOUN
cana-6247	284	8	,	,	PUNCT
cana-6247	284	9	for	for	ADP
cana-6247	284	10	instance	instance	NOUN
cana-6247	284	11	,	,	PUNCT
cana-6247	284	12	the	the	DET
cana-6247	284	13	variance	variance	NOUN
cana-6247	284	14	βt	βt	NOUN
cana-6247	284	15	can	can	AUX
cana-6247	284	16	be	be	AUX
cana-6247	284	17	kept	keep	VERB
cana-6247	284	18	higher	high	ADJ
cana-6247	284	19	for	for	ADP
cana-6247	284	20	flat	flat	ADJ
cana-6247	284	21	or	or	CCONJ
cana-6247	284	22	ambiguous	ambiguous	ADJ
cana-6247	284	23	regions	region	NOUN
cana-6247	284	24	to	to	PART
cana-6247	284	25	inject	inject	VERB
cana-6247	284	26	diversity	diversity	NOUN
cana-6247	284	27	and/or	and/or	CCONJ
cana-6247	284	28	lower	low	ADJ
cana-6247	284	29	for	for	ADP
cana-6247	284	30	well	well	ADV
cana-6247	284	31	-	-	PUNCT
cana-6247	284	32	constrained	constrain	VERB
cana-6247	284	33	ones	one	NOUN
cana-6247	284	34	;	;	PUNCT
cana-6247	284	35	in	in	ADP
cana-6247	284	36	semantic	semantic	ADJ
cana-6247	284	37	-	-	PUNCT
cana-6247	284	38	aware	aware	ADJ
cana-6247	284	39	models	model	NOUN
cana-6247	284	40	,	,	PUNCT
cana-6247	284	41	the	the	DET
cana-6247	284	42	noise	noise	NOUN
cana-6247	284	43	can	can	AUX
cana-6247	284	44	be	be	AUX
cana-6247	284	45	scheduled	schedule	VERB
cana-6247	284	46	independently	independently	ADV
cana-6247	284	47	for	for	SCONJ
cana-6247	284	48	any	any	DET
cana-6247	284	49	image	image	NOUN
cana-6247	284	50	region	region	NOUN
cana-6247	284	51	to	to	PART
cana-6247	284	52	further	far	ADV
cana-6247	284	53	optimize	optimize	VERB
cana-6247	284	54	content	content	NOUN
cana-6247	284	55	recovery	recovery	NOUN
cana-6247	284	56	.	.	PUNCT
cana-6247	285	1	some	some	DET
cana-6247	285	2	models	model	NOUN
cana-6247	285	3	go	go	VERB
cana-6247	285	4	beyond	beyond	ADP
cana-6247	285	5	this	this	PRON
cana-6247	285	6	by	by	ADP
cana-6247	285	7	learning	learn	VERB
cana-6247	285	8	mixture	mixture	NOUN
cana-6247	285	9	-	-	PUNCT
cana-6247	285	10	of	of	ADP
cana-6247	285	11	-	-	PUNCT
cana-6247	285	12	gaussian	gaussian	NOUN
cana-6247	285	13	or	or	CCONJ
cana-6247	285	14	even	even	ADV
cana-6247	285	15	non	non	ADJ
cana-6247	285	16	-	-	ADJ
cana-6247	285	17	gaussian	gaussian	ADJ
cana-6247	285	18	noise	noise	NOUN
cana-6247	285	19	distributions	distribution	NOUN
cana-6247	285	20	for	for	ADP
cana-6247	285	21	greater	great	ADJ
cana-6247	285	22	generative	generative	ADJ
cana-6247	285	23	expressiveness	expressiveness	NOUN
cana-6247	285	24	.	.	PUNCT
cana-6247	286	1	backwards	backwards	ADV
cana-6247	286	2	(	(	PUNCT
cana-6247	286	3	or	or	CCONJ
cana-6247	286	4	reverse	reverse	VERB
cana-6247	286	5	)	)	PUNCT
cana-6247	286	6	diffusion	diffusion	NOUN
cana-6247	286	7	process	process	NOUN
cana-6247	286	8	,	,	PUNCT
cana-6247	286	9	learns	learn	VERB
cana-6247	286	10	to	to	PART
cana-6247	286	11	filter	filter	VERB
cana-6247	286	12	out	out	ADP
cana-6247	286	13	this	this	DET
cana-6247	286	14	learned	learn	VERB
cana-6247	286	15	context	context	NOUN
cana-6247	286	16	-	-	PUNCT
cana-6247	286	17	dependent	dependent	ADJ
cana-6247	286	18	noise	noise	NOUN
cana-6247	286	19	,	,	PUNCT
cana-6247	286	20	thus	thus	ADV
cana-6247	286	21	obtaining	obtain	VERB
cana-6247	286	22	high	high	ADJ
cana-6247	286	23	-	-	PUNCT
cana-6247	286	24	fidelity	fidelity	NOUN
cana-6247	286	25	or	or	CCONJ
cana-6247	286	26	plausible	plausible	ADJ
cana-6247	286	27	candidates	candidate	NOUN
cana-6247	286	28	that	that	PRON
cana-6247	286	29	are	be	AUX
cana-6247	286	30	consistent	consistent	ADJ
cana-6247	286	31	with	with	ADP
cana-6247	286	32	the	the	DET
cana-6247	286	33	context	context	NOUN
cana-6247	286	34	(	(	PUNCT
cana-6247	286	35	e.g.	e.g.	ADV
cana-6247	286	36	,	,	PUNCT
cana-6247	286	37	lr	lr	NOUN
cana-6247	286	38	image	image	NOUN
cana-6247	286	39	,	,	PUNCT
cana-6247	286	40	semantic	semantic	ADJ
cana-6247	286	41	segmentation	segmentation	NOUN
cana-6247	286	42	,	,	PUNCT
cana-6247	286	43	mask	mask	NOUN
cana-6247	286	44	)	)	PUNCT
cana-6247	286	45	and	and	CCONJ
cana-6247	286	46	statistically	statistically	ADV
cana-6247	286	47	matched	match	VERB
cana-6247	286	48	with	with	ADP
cana-6247	286	49	the	the	DET
cana-6247	286	50	trained	train	VERB
cana-6247	286	51	data	datum	NOUN
cana-6247	286	52	distribution	distribution	NOUN
cana-6247	286	53	.	.	PUNCT
cana-6247	287	1	figure	figure	VERB
cana-6247	287	2	4	4	NUM
cana-6247	287	3	:	:	PUNCT
cana-6247	287	4	forward	forward	ADJ
cana-6247	287	5	diffusion	diffusion	NOUN
cana-6247	287	6	process	process	NOUN
cana-6247	287	7	in	in	ADP
cana-6247	287	8	learned	learn	VERB
cana-6247	287	9	noise	noise	NOUN
cana-6247	287	10	type	type	NOUN
cana-6247	287	11	diffusion	diffusion	NOUN
cana-6247	287	12	model	model	NOUN
cana-6247	287	13	3.1	3.1	NUM
cana-6247	287	14	super	super	ADJ
cana-6247	287	15	resolution	resolution	NOUN
cana-6247	287	16	(	(	PUNCT
cana-6247	287	17	sr	sr	PROPN
cana-6247	287	18	)	)	PUNCT
cana-6247	287	19	image	image	VERB
cana-6247	287	20	super	super	NOUN
cana-6247	287	21	-	-	NOUN
cana-6247	287	22	resolution	resolution	NOUN
cana-6247	287	23	(	(	PUNCT
cana-6247	287	24	sr	sr	PROPN
cana-6247	287	25	)	)	PUNCT
cana-6247	287	26	is	be	AUX
cana-6247	287	27	the	the	DET
cana-6247	287	28	recovery	recovery	NOUN
cana-6247	287	29	of	of	ADP
cana-6247	287	30	a	a	DET
cana-6247	287	31	high	high	ADJ
cana-6247	287	32	-	-	PUNCT
cana-6247	287	33	resolution	resolution	NOUN
cana-6247	287	34	(	(	PUNCT
cana-6247	287	35	hr	hr	NOUN
cana-6247	287	36	)	)	PUNCT
cana-6247	287	37	image	image	NOUN
cana-6247	287	38	from	from	ADP
cana-6247	287	39	a	a	DET
cana-6247	287	40	low	low	ADJ
cana-6247	287	41	-	-	PUNCT
cana-6247	287	42	resolution	resolution	NOUN
cana-6247	287	43	(	(	PUNCT
cana-6247	287	44	lr	lr	NOUN
cana-6247	287	45	)	)	PUNCT
cana-6247	287	46	counterpart	counterpart	NOUN
cana-6247	287	47	.	.	PUNCT
cana-6247	288	1	sr	sr	PROPN
cana-6247	288	2	is	be	AUX
cana-6247	288	3	an	an	DET
cana-6247	288	4	important	important	ADJ
cana-6247	288	5	operation	operation	NOUN
cana-6247	288	6	in	in	ADP
cana-6247	288	7	many	many	ADJ
cana-6247	288	8	areas	area	NOUN
cana-6247	288	9	,	,	PUNCT
cana-6247	288	10	including	include	VERB
cana-6247	288	11	medical	medical	ADJ
cana-6247	288	12	imaging	imaging	NOUN
cana-6247	288	13	,	,	PUNCT
cana-6247	288	14	remote	remote	ADJ
cana-6247	288	15	sensing	sensing	NOUN
cana-6247	288	16	,	,	PUNCT
cana-6247	288	17	scientific	scientific	ADJ
cana-6247	288	18	visualization	visualization	NOUN
cana-6247	288	19	,	,	PUNCT
cana-6247	288	20	security	security	NOUN
cana-6247	288	21	video	video	NOUN
cana-6247	288	22	forensics	forensic	NOUN
cana-6247	288	23	,	,	PUNCT
cana-6247	288	24	and	and	CCONJ
cana-6247	288	25	consumer	consumer	NOUN
cana-6247	288	26	photo	photo	NOUN
cana-6247	288	27	upscaling	upscaling	NOUN
cana-6247	288	28	.	.	PUNCT
cana-6247	289	1	the	the	DET
cana-6247	289	2	most	most	ADV
cana-6247	289	3	challenging	challenging	ADJ
cana-6247	289	4	part	part	NOUN
cana-6247	289	5	is	be	AUX
cana-6247	289	6	restoring	restore	VERB
cana-6247	289	7	high	high	ADJ
cana-6247	289	8	-	-	PUNCT
cana-6247	289	9	frequency	frequency	NOUN
cana-6247	289	10	details	detail	NOUN
cana-6247	289	11	that	that	PRON
cana-6247	289	12	are	be	AUX
cana-6247	289	13	lost	lose	VERB
cana-6247	289	14	from	from	ADP
cana-6247	289	15	downsampling	downsample	VERB
cana-6247	289	16	or	or	CCONJ
cana-6247	289	17	sensor	sensor	NOUN
cana-6247	289	18	resolution	resolution	NOUN
cana-6247	289	19	,	,	PUNCT
cana-6247	289	20	with	with	ADP
cana-6247	289	21	the	the	DET
cana-6247	289	22	guarantee	guarantee	NOUN
cana-6247	289	23	of	of	ADP
cana-6247	289	24	both	both	CCONJ
cana-6247	289	25	sharpness	sharpness	ADJ
cana-6247	289	26	and	and	CCONJ
cana-6247	289	27	perceptual	perceptual	ADJ
cana-6247	289	28	coherence	coherence	NOUN
cana-6247	289	29	in	in	ADP
cana-6247	289	30	the	the	DET
cana-6247	289	31	constructed	construct	VERB
cana-6247	289	32	hr	hr	NOUN
cana-6247	289	33	image	image	NOUN
cana-6247	289	34	.	.	PUNCT
cana-6247	290	1	diffusion	diffusion	NOUN
cana-6247	290	2	models	model	NOUN
cana-6247	290	3	have	have	AUX
cana-6247	290	4	become	become	VERB
cana-6247	290	5	a	a	DET
cana-6247	290	6	preferred	preferred	ADJ
cana-6247	290	7	architecture	architecture	NOUN
cana-6247	290	8	for	for	ADP
cana-6247	290	9	sr	sr	PROPN
cana-6247	290	10	,	,	PUNCT
cana-6247	290	11	capitalizing	capitalize	VERB
cana-6247	290	12	on	on	ADP
cana-6247	290	13	their	their	PRON
cana-6247	290	14	success	success	NOUN
cana-6247	290	15	in	in	ADP
cana-6247	290	16	generative	generative	ADJ
cana-6247	290	17	modeling	modeling	NOUN
cana-6247	290	18	and	and	CCONJ
cana-6247	290	19	image	image	NOUN
cana-6247	290	20	restoration	restoration	NOUN
cana-6247	290	21	.	.	PUNCT
cana-6247	291	1	they	they	PRON
cana-6247	291	2	iteratively	iteratively	ADV
cana-6247	291	3	modify	modify	VERB
cana-6247	291	4	the	the	DET
cana-6247	291	5	noisy	noisy	ADJ
cana-6247	291	6	hr	hr	NOUN
cana-6247	291	7	candidate	candidate	NOUN
cana-6247	291	8	,	,	PUNCT
cana-6247	291	9	often	often	ADV
cana-6247	291	10	guided	guide	VERB
cana-6247	291	11	by	by	ADP
cana-6247	291	12	the	the	DET
cana-6247	291	13	lr	lr	NOUN
cana-6247	291	14	image	image	NOUN
cana-6247	291	15	,	,	PUNCT
cana-6247	291	16	so	so	SCONJ
cana-6247	291	17	as	as	SCONJ
cana-6247	291	18	to	to	PART
cana-6247	291	19	induce	induce	VERB
cana-6247	291	20	textures	texture	NOUN
cana-6247	291	21	,	,	PUNCT
cana-6247	291	22	structures	structure	NOUN
cana-6247	291	23	,	,	PUNCT
cana-6247	291	24	and	and	CCONJ
cana-6247	291	25	plausible	plausible	ADJ
cana-6247	291	26	details	detail	NOUN
cana-6247	291	27	that	that	PRON
cana-6247	291	28	are	be	AUX
cana-6247	291	29	not	not	PART
cana-6247	291	30	present	present	ADJ
cana-6247	291	31	in	in	ADP
cana-6247	291	32	the	the	DET
cana-6247	291	33	input	input	NOUN
cana-6247	291	34	image	image	NOUN
cana-6247	291	35	.	.	PUNCT
cana-6247	292	1	early	early	ADJ
cana-6247	292	2	diffusion	diffusion	NOUN
cana-6247	292	3	models	model	NOUN
cana-6247	292	4	performed	perform	VERB
cana-6247	292	5	a	a	DET
cana-6247	292	6	fixed	fix	VERB
cana-6247	292	7	-	-	PUNCT
cana-6247	292	8	noise	noise	NOUN
cana-6247	292	9	schedule	schedule	NOUN
cana-6247	292	10	for	for	ADP
cana-6247	292	11	every	every	DET
cana-6247	292	12	image	image	NOUN
cana-6247	292	13	and	and	CCONJ
cana-6247	292	14	timestep	timestep	NOUN
cana-6247	292	15	—	—	PUNCT
cana-6247	292	16	often	often	ADV
cana-6247	292	17	just	just	ADV
cana-6247	292	18	isotropic	isotropic	ADJ
cana-6247	292	19	gaussian	gaussian	ADJ
cana-6247	292	20	noise	noise	NOUN
cana-6247	292	21	.	.	PUNCT
cana-6247	293	1	however	however	ADV
cana-6247	293	2	,	,	PUNCT
cana-6247	293	3	conditioning	condition	VERB
cana-6247	293	4	the	the	DET
cana-6247	293	5	noise	noise	NOUN
cana-6247	293	6	process	process	NOUN
cana-6247	293	7	on	on	ADP
cana-6247	293	8	the	the	DET
cana-6247	293	9	lr	lr	NOUN
cana-6247	293	10	input	input	NOUN
cana-6247	293	11	,	,	PUNCT
cana-6247	293	12	on	on	ADP
cana-6247	293	13	uncertainty	uncertainty	NOUN
cana-6247	293	14	estimates	estimate	NOUN
cana-6247	293	15	,	,	PUNCT
cana-6247	293	16	or	or	CCONJ
cana-6247	293	17	on	on	ADP
cana-6247	293	18	semantic	semantic	ADJ
cana-6247	293	19	cue	cue	NOUN
cana-6247	293	20	has	have	AUX
cana-6247	293	21	proven	prove	VERB
cana-6247	293	22	to	to	PART
cana-6247	293	23	greatly	greatly	ADV
cana-6247	293	24	improve	improve	VERB
cana-6247	293	25	quality	quality	NOUN
cana-6247	293	26	and	and	CCONJ
cana-6247	293	27	efficiency	efficiency	NOUN
cana-6247	293	28	.	.	PUNCT
cana-6247	294	1	learned	learn	VERB
cana-6247	294	2	noise	noise	NOUN
cana-6247	294	3	refers	refer	VERB
cana-6247	294	4	to	to	ADP
cana-6247	294	5	the	the	DET
cana-6247	294	6	approach	approach	NOUN
cana-6247	294	7	in	in	ADP
cana-6247	294	8	which	which	PRON
cana-6247	294	9	the	the	DET
cana-6247	294	10	quantity	quantity	NOUN
cana-6247	294	11	,	,	PUNCT
cana-6247	294	12	distribution	distribution	NOUN
cana-6247	294	13	,	,	PUNCT
cana-6247	294	14	or	or	CCONJ
cana-6247	294	15	style	style	NOUN
cana-6247	294	16	of	of	ADP
cana-6247	294	17	noise	noise	NOUN
cana-6247	294	18	injected	inject	VERB
cana-6247	294	19	during	during	ADP
cana-6247	294	20	the	the	DET
cana-6247	294	21	diffusion	diffusion	NOUN
cana-6247	294	22	process	process	NOUN
cana-6247	294	23	itself	itself	PRON
cana-6247	294	24	,	,	PUNCT
cana-6247	294	25	is	be	AUX
cana-6247	294	26	learned	learn	VERB
cana-6247	294	27	from	from	ADP
cana-6247	294	28	data	datum	NOUN
cana-6247	294	29	,	,	PUNCT
cana-6247	294	30	usually	usually	ADV
cana-6247	294	31	made	make	VERB
cana-6247	294	32	a	a	DET
cana-6247	294	33	function	function	NOUN
cana-6247	294	34	of	of	ADP
cana-6247	294	35	the	the	DET
cana-6247	294	36	particular	particular	ADJ
cana-6247	294	37	lr	lr	NOUN
cana-6247	294	38	guidance	guidance	NOUN
cana-6247	294	39	,	,	PUNCT
cana-6247	294	40	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	294	41	990	990	NUM
cana-6247	294	42	communications	communication	NOUN
cana-6247	294	43	on	on	ADP
cana-6247	294	44	applied	apply	VERB
cana-6247	294	45	nonlinear	nonlinear	ADJ
cana-6247	294	46	analysis	analysis	NOUN
cana-6247	294	47	issn	issn	NOUN
cana-6247	294	48	:	:	PUNCT
cana-6247	294	49	1074	1074	NUM
cana-6247	294	50	-	-	PUNCT
cana-6247	294	51	133x	133x	NUM
cana-6247	294	52	vol	vol	NOUN
cana-6247	294	53	.	.	PROPN
cana-6247	295	1	31	31	NUM
cana-6247	296	1	no	no	NOUN
cana-6247	296	2	.	.	PUNCT
cana-6247	297	1	7s	7	NOUN
cana-6247	297	2	(	(	PUNCT
cana-6247	297	3	2024	2024	NUM
cana-6247	297	4	)	)	PUNCT
cana-6247	297	5	region	region	NOUN
cana-6247	297	6	content	content	NOUN
cana-6247	297	7	,	,	PUNCT
cana-6247	297	8	or	or	CCONJ
cana-6247	297	9	semantic	semantic	ADJ
cana-6247	297	10	priors	prior	NOUN
cana-6247	297	11	.	.	PUNCT
cana-6247	298	1	applying	apply	VERB
cana-6247	298	2	noise	noise	NOUN
cana-6247	298	3	can	can	AUX
cana-6247	298	4	serve	serve	VERB
cana-6247	298	5	for	for	ADP
cana-6247	298	6	restoration	restoration	NOUN
cana-6247	298	7	or	or	CCONJ
cana-6247	298	8	editing	editing	NOUN
cana-6247	298	9	functions	function	NOUN
cana-6247	298	10	.	.	PUNCT
cana-6247	299	1	rather	rather	ADV
cana-6247	299	2	than	than	ADP
cana-6247	299	3	applying	apply	VERB
cana-6247	299	4	generic	generic	ADJ
cana-6247	299	5	or	or	CCONJ
cana-6247	299	6	uniform	uniform	ADJ
cana-6247	299	7	schedules	schedule	NOUN
cana-6247	299	8	,	,	PUNCT
cana-6247	299	9	the	the	DET
cana-6247	299	10	noise	noise	NOUN
cana-6247	299	11	processes	process	NOUN
cana-6247	299	12	can	can	AUX
cana-6247	299	13	allocate	allocate	VERB
cana-6247	299	14	selective	selective	ADJ
cana-6247	299	15	model	model	NOUN
cana-6247	299	16	capacity	capacity	NOUN
cana-6247	299	17	it	it	PRON
cana-6247	299	18	adds	add	VERB
cana-6247	299	19	less	less	ADJ
cana-6247	299	20	noise	noise	NOUN
cana-6247	299	21	to	to	ADP
cana-6247	299	22	regions	region	NOUN
cana-6247	299	23	that	that	SCONJ
cana-6247	299	24	it	it	PRON
cana-6247	299	25	can	can	AUX
cana-6247	299	26	reliably	reliably	ADV
cana-6247	299	27	predict	predict	VERB
cana-6247	299	28	(	(	PUNCT
cana-6247	299	29	e.g.	e.g.	ADV
cana-6247	299	30	,	,	PUNCT
cana-6247	299	31	smooth	smooth	ADJ
cana-6247	299	32	surfaces	surface	NOUN
cana-6247	299	33	)	)	PUNCT
cana-6247	299	34	and	and	CCONJ
cana-6247	299	35	more	more	ADJ
cana-6247	299	36	noise	noise	NOUN
cana-6247	299	37	to	to	ADP
cana-6247	299	38	areas	area	NOUN
cana-6247	299	39	that	that	PRON
cana-6247	299	40	remain	remain	VERB
cana-6247	299	41	ambiguous	ambiguous	ADJ
cana-6247	299	42	or	or	CCONJ
cana-6247	299	43	underdetermined	underdetermined	ADJ
cana-6247	299	44	(	(	PUNCT
cana-6247	299	45	e.g.	e.g.	ADV
cana-6247	299	46	,	,	PUNCT
cana-6247	299	47	textured	textured	ADJ
cana-6247	299	48	edges	edge	NOUN
cana-6247	299	49	,	,	PUNCT
cana-6247	299	50	fine	fine	ADJ
cana-6247	299	51	details	detail	NOUN
cana-6247	299	52	)	)	PUNCT
cana-6247	299	53	.	.	PUNCT
cana-6247	300	1	this	this	DET
cana-6247	300	2	selective	selective	ADJ
cana-6247	300	3	addition	addition	NOUN
cana-6247	300	4	of	of	ADP
cana-6247	300	5	noise	noise	NOUN
cana-6247	300	6	helps	help	VERB
cana-6247	300	7	to	to	PART
cana-6247	300	8	boost	boost	VERB
cana-6247	300	9	the	the	DET
cana-6247	300	10	diversity	diversity	NOUN
cana-6247	300	11	and	and	CCONJ
cana-6247	300	12	realism	realism	NOUN
cana-6247	300	13	in	in	ADP
cana-6247	300	14	high	high	ADJ
cana-6247	300	15	-	-	PUNCT
cana-6247	300	16	frequency	frequency	NOUN
cana-6247	300	17	areas	area	NOUN
cana-6247	300	18	,	,	PUNCT
cana-6247	300	19	gain	gain	VERB
cana-6247	300	20	consistent	consistent	ADJ
cana-6247	300	21	restoration	restoration	NOUN
cana-6247	300	22	behaviour	behaviour	NOUN
cana-6247	300	23	on	on	ADP
cana-6247	300	24	semantic	semantic	ADJ
cana-6247	300	25	boundaries	boundary	NOUN
cana-6247	300	26	,	,	PUNCT
cana-6247	300	27	and	and	CCONJ
cana-6247	300	28	allow	allow	VERB
cana-6247	300	29	faster	fast	ADJ
cana-6247	300	30	convergence	convergence	NOUN
cana-6247	300	31	during	during	ADP
cana-6247	300	32	inference	inference	NOUN
cana-6247	300	33	.	.	PUNCT
cana-6247	301	1	for	for	ADP
cana-6247	301	2	example	example	NOUN
cana-6247	301	3	,	,	PUNCT
cana-6247	301	4	in	in	ADP
cana-6247	301	5	ecdp	ecdp	PROPN
cana-6247	301	6	[	[	PUNCT
cana-6247	301	7	47	47	NUM
cana-6247	301	8	]	]	PUNCT
cana-6247	301	9	,	,	PUNCT
cana-6247	301	10	the	the	DET
cana-6247	301	11	noise	noise	NOUN
cana-6247	301	12	schedule	schedule	NOUN
cana-6247	301	13	is	be	AUX
cana-6247	301	14	modulated	modulate	VERB
cana-6247	301	15	depending	depend	VERB
cana-6247	301	16	on	on	ADP
cana-6247	301	17	the	the	DET
cana-6247	301	18	probability	probability	NOUN
cana-6247	301	19	flow	flow	NOUN
cana-6247	301	20	given	give	VERB
cana-6247	301	21	by	by	ADP
cana-6247	301	22	the	the	DET
cana-6247	301	23	lr	lr	NOUN
cana-6247	301	24	structure	structure	NOUN
cana-6247	301	25	,	,	PUNCT
cana-6247	301	26	thereby	thereby	ADV
cana-6247	301	27	avoiding	avoid	VERB
cana-6247	301	28	unnecessary	unnecessary	ADJ
cana-6247	301	29	updates	update	NOUN
cana-6247	301	30	and	and	CCONJ
cana-6247	301	31	finishing	finish	VERB
cana-6247	301	32	with	with	ADP
cana-6247	301	33	high	high	ADJ
cana-6247	301	34	-	-	PUNCT
cana-6247	301	35	quality	quality	NOUN
cana-6247	301	36	hr	hr	NOUN
cana-6247	301	37	results	result	NOUN
cana-6247	301	38	in	in	ADP
cana-6247	301	39	fewer	few	ADJ
cana-6247	301	40	denoising	denoising	NOUN
cana-6247	301	41	steps	step	NOUN
cana-6247	301	42	.	.	PUNCT
cana-6247	302	1	in	in	ADP
cana-6247	302	2	the	the	DET
cana-6247	302	3	uncertainty	uncertainty	NOUN
cana-6247	302	4	-	-	PUNCT
cana-6247	302	5	guided	guide	VERB
cana-6247	302	6	perturbation	perturbation	NOUN
cana-6247	302	7	framework	framework	NOUN
cana-6247	302	8	,	,	PUNCT
cana-6247	302	9	the	the	DET
cana-6247	302	10	noise	noise	NOUN
cana-6247	302	11	scale	scale	NOUN
cana-6247	302	12	is	be	AUX
cana-6247	302	13	,	,	PUNCT
cana-6247	302	14	in	in	ADP
cana-6247	302	15	fact	fact	NOUN
cana-6247	302	16	,	,	PUNCT
cana-6247	302	17	inversely	inversely	ADV
cana-6247	302	18	proportional	proportional	ADJ
cana-6247	302	19	to	to	ADP
cana-6247	302	20	uncertainty	uncertainty	NOUN
cana-6247	302	21	maps	map	NOUN
cana-6247	302	22	generated	generate	VERB
cana-6247	302	23	through	through	ADP
cana-6247	302	24	the	the	DET
cana-6247	302	25	lr	lr	PROPN
cana-6247	302	26	-	-	PUNCT
cana-6247	302	27	to	to	ADP
cana-6247	302	28	-	-	PUNCT
cana-6247	302	29	hr	hr	NOUN
cana-6247	302	30	mapping	mapping	NOUN
cana-6247	302	31	so	so	SCONJ
cana-6247	302	32	that	that	SCONJ
cana-6247	302	33	only	only	ADV
cana-6247	302	34	the	the	DET
cana-6247	302	35	hard	hard	ADJ
cana-6247	302	36	-	-	PUNCT
cana-6247	302	37	to	to	AUX
cana-6247	302	38	-	-	PUNCT
cana-6247	302	39	predict	predict	VERB
cana-6247	302	40	regions	region	NOUN
cana-6247	302	41	are	be	AUX
cana-6247	302	42	perturbed	perturb	VERB
cana-6247	302	43	.	.	PUNCT
cana-6247	303	1	this	this	DET
cana-6247	303	2	results	result	VERB
cana-6247	303	3	in	in	ADP
cana-6247	303	4	both	both	CCONJ
cana-6247	303	5	more	more	ADV
cana-6247	303	6	convincing	convincing	ADJ
cana-6247	303	7	and	and	CCONJ
cana-6247	303	8	sharper	sharp	ADJ
cana-6247	303	9	hr	hr	NOUN
cana-6247	303	10	reconstructions	reconstruction	NOUN
cana-6247	303	11	.	.	PUNCT
cana-6247	304	1	3.1.1	3.1.1	NUM
cana-6247	304	2	evidence	evidence	NOUN
cana-6247	304	3	from	from	ADP
cana-6247	304	4	recent	recent	ADJ
cana-6247	304	5	research	research	NOUN
cana-6247	304	6	recent	recent	ADJ
cana-6247	304	7	works	work	NOUN
cana-6247	304	8	provide	provide	VERB
cana-6247	304	9	compelling	compelling	ADJ
cana-6247	304	10	empirical	empirical	ADJ
cana-6247	304	11	and	and	CCONJ
cana-6247	304	12	architectural	architectural	ADJ
cana-6247	304	13	support	support	NOUN
cana-6247	304	14	for	for	ADP
cana-6247	304	15	the	the	DET
cana-6247	304	16	superiority	superiority	NOUN
cana-6247	304	17	of	of	ADP
cana-6247	304	18	learned	learn	VERB
cana-6247	304	19	noise	noise	NOUN
cana-6247	304	20	in	in	ADP
cana-6247	304	21	diffusion	diffusion	NOUN
cana-6247	304	22	-	-	PUNCT
cana-6247	304	23	based	base	VERB
cana-6247	304	24	super	super	NOUN
cana-6247	304	25	-	-	NOUN
cana-6247	304	26	resolution	resolution	NOUN
cana-6247	304	27	.	.	PUNCT
cana-6247	305	1	•	•	NUM
cana-6247	305	2	efficient	efficient	ADJ
cana-6247	305	3	conditional	conditional	ADJ
cana-6247	305	4	diffusion	diffusion	NOUN
cana-6247	305	5	model	model	NOUN
cana-6247	305	6	with	with	ADP
cana-6247	305	7	probability	probability	NOUN
cana-6247	305	8	flow	flow	NOUN
cana-6247	305	9	sampling	sample	VERB
cana-6247	305	10	for	for	ADP
cana-6247	305	11	image	image	NOUN
cana-6247	305	12	super	super	NOUN
cana-6247	305	13	-	-	NOUN
cana-6247	305	14	resolution	resolution	NOUN
cana-6247	305	15	(	(	PUNCT
cana-6247	305	16	ecdp	ecdp	PROPN
cana-6247	305	17	)	)	PUNCT
cana-6247	305	18	introduced	introduce	VERB
cana-6247	305	19	a	a	DET
cana-6247	305	20	conditional	conditional	ADJ
cana-6247	305	21	sr	sr	PROPN
cana-6247	305	22	framework	framework	NOUN
cana-6247	305	23	that	that	PRON
cana-6247	305	24	tries	try	VERB
cana-6247	305	25	to	to	PART
cana-6247	305	26	dynamically	dynamically	ADV
cana-6247	305	27	adjust	adjust	VERB
cana-6247	305	28	the	the	DET
cana-6247	305	29	noise	noise	NOUN
cana-6247	305	30	schedule	schedule	NOUN
cana-6247	305	31	on	on	ADP
cana-6247	305	32	the	the	DET
cana-6247	305	33	lr	lr	NOUN
cana-6247	305	34	so	so	SCONJ
cana-6247	305	35	that	that	SCONJ
cana-6247	305	36	the	the	DET
cana-6247	305	37	noise	noise	NOUN
cana-6247	305	38	schedule	schedule	NOUN
cana-6247	305	39	depends	depend	VERB
cana-6247	305	40	on	on	ADP
cana-6247	305	41	it	it	PRON
cana-6247	305	42	and	and	CCONJ
cana-6247	305	43	on	on	ADP
cana-6247	305	44	structural	structural	ADJ
cana-6247	305	45	cues	cue	NOUN
cana-6247	305	46	.	.	PUNCT
cana-6247	306	1	hence	hence	ADV
cana-6247	306	2	,	,	PUNCT
cana-6247	306	3	very	very	ADV
cana-6247	306	4	little	little	ADJ
cana-6247	306	5	noise	noise	NOUN
cana-6247	306	6	needs	need	NOUN
cana-6247	306	7	to	to	PART
cana-6247	306	8	be	be	AUX
cana-6247	306	9	injected	inject	VERB
cana-6247	306	10	into	into	ADP
cana-6247	306	11	those	those	DET
cana-6247	306	12	areas	area	NOUN
cana-6247	306	13	that	that	PRON
cana-6247	306	14	can	can	AUX
cana-6247	306	15	already	already	ADV
cana-6247	306	16	be	be	AUX
cana-6247	306	17	predicted	predict	VERB
cana-6247	306	18	with	with	ADP
cana-6247	306	19	a	a	DET
cana-6247	306	20	high	high	ADJ
cana-6247	306	21	degree	degree	NOUN
cana-6247	306	22	of	of	ADP
cana-6247	306	23	confidence	confidence	NOUN
cana-6247	306	24	,	,	PUNCT
cana-6247	306	25	and	and	CCONJ
cana-6247	306	26	maximum	maximum	ADJ
cana-6247	306	27	generative	generative	ADJ
cana-6247	306	28	effort	effort	NOUN
cana-6247	306	29	is	be	AUX
cana-6247	306	30	applied	apply	VERB
cana-6247	306	31	where	where	SCONJ
cana-6247	306	32	lr	lr	PRON
cana-6247	306	33	evidence	evidence	NOUN
cana-6247	306	34	is	be	AUX
cana-6247	306	35	weak	weak	ADJ
cana-6247	306	36	,	,	PUNCT
cana-6247	306	37	resulting	result	VERB
cana-6247	306	38	in	in	ADP
cana-6247	306	39	sharper	sharp	ADJ
cana-6247	306	40	hr	hr	NOUN
cana-6247	306	41	outputs	output	NOUN
cana-6247	306	42	via	via	ADP
cana-6247	306	43	fewer	few	ADJ
cana-6247	306	44	sampling	sampling	NOUN
cana-6247	306	45	steps	step	NOUN
cana-6247	306	46	.	.	PUNCT
cana-6247	307	1	[	[	X
cana-6247	307	2	47	47	NUM
cana-6247	307	3	]	]	SYM
cana-6247	307	4	•	•	NUM
cana-6247	307	5	cdpmsr	cdpmsr	NOUN
cana-6247	307	6	(	(	PUNCT
cana-6247	307	7	conditional	conditional	ADJ
cana-6247	307	8	diffusion	diffusion	NOUN
cana-6247	307	9	probabilistic	probabilistic	ADJ
cana-6247	307	10	models	model	NOUN
cana-6247	307	11	for	for	ADP
cana-6247	307	12	single	single	ADJ
cana-6247	307	13	image	image	NOUN
cana-6247	307	14	super	super	NOUN
cana-6247	307	15	-	-	NOUN
cana-6247	307	16	resolution	resolution	NOUN
cana-6247	307	17	)	)	PUNCT
cana-6247	307	18	and	and	CCONJ
cana-6247	307	19	acdmsr	acdmsr	PROPN
cana-6247	307	20	(	(	PUNCT
cana-6247	307	21	accelerated	accelerate	VERB
cana-6247	307	22	conditional	conditional	ADJ
cana-6247	307	23	diffusion	diffusion	NOUN
cana-6247	307	24	models	model	NOUN
cana-6247	307	25	for	for	ADP
cana-6247	307	26	super	super	NOUN
cana-6247	307	27	-	-	NOUN
cana-6247	307	28	resolution	resolution	NOUN
cana-6247	307	29	)	)	PUNCT
cana-6247	307	30	,	,	PUNCT
cana-6247	307	31	further	far	ADV
cana-6247	307	32	demonstrated	demonstrate	VERB
cana-6247	307	33	that	that	SCONJ
cana-6247	307	34	conditioning	condition	VERB
cana-6247	307	35	both	both	CCONJ
cana-6247	307	36	the	the	DET
cana-6247	307	37	forward	forward	ADJ
cana-6247	307	38	diffusion	diffusion	NOUN
cana-6247	307	39	process	process	NOUN
cana-6247	307	40	and	and	CCONJ
cana-6247	307	41	reverse	reverse	ADJ
cana-6247	307	42	diffusion	diffusion	NOUN
cana-6247	307	43	process	process	NOUN
cana-6247	307	44	on	on	ADP
cana-6247	307	45	the	the	DET
cana-6247	307	46	lr	lr	NOUN
cana-6247	307	47	image	image	NOUN
cana-6247	307	48	improves	improve	VERB
cana-6247	307	49	restoration	restoration	NOUN
cana-6247	307	50	performance	performance	NOUN
cana-6247	307	51	.	.	PUNCT
cana-6247	308	1	especially	especially	ADV
cana-6247	308	2	in	in	ADP
cana-6247	308	3	the	the	DET
cana-6247	308	4	real	real	ADJ
cana-6247	308	5	world	world	NOUN
cana-6247	308	6	,	,	PUNCT
cana-6247	308	7	a	a	DET
cana-6247	308	8	restoration	restoration	NOUN
cana-6247	308	9	algorithm	algorithm	NOUN
cana-6247	308	10	is	be	AUX
cana-6247	308	11	needed	need	VERB
cana-6247	308	12	to	to	PART
cana-6247	308	13	condition	condition	VERB
cana-6247	308	14	on	on	ADP
cana-6247	308	15	the	the	DET
cana-6247	308	16	lr	lr	NOUN
cana-6247	308	17	image	image	NOUN
cana-6247	308	18	,	,	PUNCT
cana-6247	308	19	where	where	SCONJ
cana-6247	308	20	the	the	DET
cana-6247	308	21	appropriate	appropriate	ADJ
cana-6247	308	22	noise	noise	NOUN
cana-6247	308	23	level	level	NOUN
cana-6247	308	24	is	be	AUX
cana-6247	308	25	different	different	ADJ
cana-6247	308	26	for	for	ADP
cana-6247	308	27	each	each	DET
cana-6247	308	28	semantic	semantic	ADJ
cana-6247	308	29	region	region	NOUN
cana-6247	308	30	and	and	CCONJ
cana-6247	308	31	texture	texture	NOUN
cana-6247	308	32	in	in	ADP
cana-6247	308	33	the	the	DET
cana-6247	308	34	image	image	NOUN
cana-6247	308	35	.	.	PUNCT
cana-6247	309	1	[	[	X
cana-6247	309	2	27	27	NUM
cana-6247	309	3	]	]	X
cana-6247	309	4	[	[	X
cana-6247	309	5	24	24	NUM
cana-6247	309	6	]	]	SYM
cana-6247	309	7	•	•	NUM
cana-6247	309	8	latent	latent	PROPN
cana-6247	309	9	space	space	NOUN
cana-6247	309	10	super	super	NOUN
cana-6247	309	11	-	-	NOUN
cana-6247	309	12	resolution	resolution	NOUN
cana-6247	309	13	for	for	ADP
cana-6247	309	14	higher	high	ADJ
cana-6247	309	15	-	-	PUNCT
cana-6247	309	16	resolution	resolution	NOUN
cana-6247	309	17	image	image	NOUN
cana-6247	309	18	generation	generation	NOUN
cana-6247	309	19	with	with	ADP
cana-6247	309	20	diffusion	diffusion	NOUN
cana-6247	309	21	models	model	NOUN
cana-6247	309	22	operates	operate	VERB
cana-6247	309	23	in	in	ADP
cana-6247	309	24	learned	learn	VERB
cana-6247	309	25	feature	feature	NOUN
cana-6247	309	26	space	space	NOUN
cana-6247	309	27	,	,	PUNCT
cana-6247	309	28	where	where	SCONJ
cana-6247	309	29	the	the	DET
cana-6247	309	30	injected	inject	VERB
cana-6247	309	31	noise	noise	NOUN
cana-6247	309	32	is	be	AUX
cana-6247	309	33	modulated	modulate	VERB
cana-6247	309	34	by	by	ADP
cana-6247	309	35	the	the	DET
cana-6247	309	36	semantic	semantic	ADJ
cana-6247	309	37	structure	structure	NOUN
cana-6247	309	38	,	,	PUNCT
cana-6247	309	39	allowing	allow	VERB
cana-6247	309	40	for	for	ADP
cana-6247	309	41	fast	fast	ADJ
cana-6247	309	42	sr	sr	PROPN
cana-6247	309	43	inference	inference	NOUN
cana-6247	309	44	and	and	CCONJ
cana-6247	309	45	high	high	ADJ
cana-6247	309	46	-	-	PUNCT
cana-6247	309	47	resolution	resolution	NOUN
cana-6247	309	48	restorations	restoration	NOUN
cana-6247	309	49	at	at	ADP
cana-6247	309	50	minimal	minimal	ADJ
cana-6247	309	51	memory	memory	NOUN
cana-6247	309	52	cost	cost	NOUN
cana-6247	309	53	.	.	PUNCT
cana-6247	310	1	[	[	X
cana-6247	310	2	13	13	NUM
cana-6247	310	3	]	]	SYM
cana-6247	310	4	•	•	NUM
cana-6247	310	5	resdiff	resdiff	NOUN
cana-6247	310	6	:	:	PUNCT
cana-6247	310	7	combining	combine	VERB
cana-6247	310	8	cnn	cnn	PROPN
cana-6247	310	9	and	and	CCONJ
cana-6247	310	10	diffusion	diffusion	NOUN
cana-6247	310	11	model	model	NOUN
cana-6247	310	12	for	for	ADP
cana-6247	310	13	image	image	NOUN
cana-6247	310	14	super	super	ADJ
cana-6247	310	15	-	-	NOUN
cana-6247	310	16	resolution	resolution	NOUN
cana-6247	310	17	stresses	stress	VERB
cana-6247	310	18	the	the	DET
cana-6247	310	19	benefits	benefit	NOUN
cana-6247	310	20	of	of	ADP
cana-6247	310	21	learned	learn	VERB
cana-6247	310	22	noise	noise	NOUN
cana-6247	310	23	processes	process	NOUN
cana-6247	310	24	with	with	ADP
cana-6247	310	25	cnn	cnn	PROPN
cana-6247	310	26	-	-	PUNCT
cana-6247	310	27	based	base	VERB
cana-6247	310	28	residuals	residual	NOUN
cana-6247	310	29	,	,	PUNCT
cana-6247	310	30	where	where	SCONJ
cana-6247	310	31	noise	noise	NOUN
cana-6247	310	32	is	be	AUX
cana-6247	310	33	injected	inject	VERB
cana-6247	310	34	guided	guide	VERB
cana-6247	310	35	by	by	ADP
cana-6247	310	36	the	the	DET
cana-6247	310	37	lr	lr	PROPN
cana-6247	310	38	context	context	NOUN
cana-6247	310	39	and	and	CCONJ
cana-6247	310	40	the	the	DET
cana-6247	310	41	residual	residual	ADJ
cana-6247	310	42	feature	feature	NOUN
cana-6247	310	43	maps	map	NOUN
cana-6247	310	44	.	.	PUNCT
cana-6247	311	1	this	this	PRON
cana-6247	311	2	allows	allow	VERB
cana-6247	311	3	for	for	ADP
cana-6247	311	4	efficient	efficient	ADJ
cana-6247	311	5	hr	hr	NOUN
cana-6247	311	6	reconstruction	reconstruction	NOUN
cana-6247	311	7	using	use	VERB
cana-6247	311	8	the	the	DET
cana-6247	311	9	wellpositioned	wellpositioned	ADJ
cana-6247	311	10	residuals	residual	NOUN
cana-6247	311	11	with	with	ADP
cana-6247	311	12	highly	highly	ADV
cana-6247	311	13	contextual	contextual	ADJ
cana-6247	311	14	noise	noise	NOUN
cana-6247	311	15	injected	inject	VERB
cana-6247	311	16	.	.	PUNCT
cana-6247	312	1	[	[	X
cana-6247	312	2	33	33	NUM
cana-6247	312	3	]	]	SYM
cana-6247	312	4	•	•	NUM
cana-6247	312	5	uncertainty	uncertainty	NOUN
cana-6247	312	6	-	-	PUNCT
cana-6247	312	7	guided	guide	VERB
cana-6247	312	8	perturbation	perturbation	NOUN
cana-6247	312	9	for	for	ADP
cana-6247	312	10	image	image	NOUN
cana-6247	312	11	super	super	ADJ
cana-6247	312	12	-	-	ADJ
cana-6247	312	13	resolution	resolution	ADJ
cana-6247	312	14	diffusion	diffusion	NOUN
cana-6247	312	15	model	model	NOUN
cana-6247	312	16	and	and	CCONJ
cana-6247	312	17	seesr	seesr	NOUN
cana-6247	312	18	:	:	PUNCT
cana-6247	312	19	towards	towards	ADP
cana-6247	312	20	semantics	semantic	NOUN
cana-6247	312	21	-	-	PUNCT
cana-6247	312	22	aware	aware	ADJ
cana-6247	312	23	real	real	ADJ
cana-6247	312	24	-	-	PUNCT
cana-6247	312	25	world	world	NOUN
cana-6247	312	26	image	image	NOUN
cana-6247	312	27	super	super	ADJ
cana-6247	312	28	-	-	ADJ
cana-6247	312	29	resolution	resolution	ADJ
cana-6247	312	30	introduced	introduce	VERB
cana-6247	312	31	models	model	NOUN
cana-6247	312	32	that	that	PRON
cana-6247	312	33	provide	provide	VERB
cana-6247	312	34	estimates	estimate	NOUN
cana-6247	312	35	of	of	ADP
cana-6247	312	36	region	region	NOUN
cana-6247	312	37	-	-	PUNCT
cana-6247	312	38	wise	wise	ADJ
cana-6247	312	39	uncertainty	uncertainty	NOUN
cana-6247	312	40	,	,	PUNCT
cana-6247	312	41	or	or	CCONJ
cana-6247	312	42	semantic	semantic	ADJ
cana-6247	312	43	segmentation	segmentation	NOUN
cana-6247	312	44	maps	map	NOUN
cana-6247	312	45	,	,	PUNCT
cana-6247	312	46	and	and	CCONJ
cana-6247	312	47	would	would	AUX
cana-6247	312	48	then	then	ADV
cana-6247	312	49	modify	modify	VERB
cana-6247	312	50	the	the	DET
cana-6247	312	51	noise	noise	NOUN
cana-6247	312	52	injection	injection	NOUN
cana-6247	312	53	to	to	PART
cana-6247	312	54	incorporate	incorporate	VERB
cana-6247	312	55	the	the	DET
cana-6247	312	56	region	region	NOUN
cana-6247	312	57	-	-	PUNCT
cana-6247	312	58	wise	wise	ADJ
cana-6247	312	59	uncertainty	uncertainty	NOUN
cana-6247	312	60	.	.	PUNCT
cana-6247	313	1	these	these	DET
cana-6247	313	2	approaches	approach	NOUN
cana-6247	313	3	can	can	AUX
cana-6247	313	4	produce	produce	VERB
cana-6247	313	5	highly	highly	ADV
cana-6247	313	6	photorealistic	photorealistic	ADJ
cana-6247	313	7	and	and	CCONJ
cana-6247	313	8	structure	structure	NOUN
cana-6247	313	9	-	-	PUNCT
cana-6247	313	10	preserving	preserve	VERB
cana-6247	313	11	sr	sr	PROPN
cana-6247	313	12	results	result	NOUN
cana-6247	313	13	,	,	PUNCT
cana-6247	313	14	which	which	PRON
cana-6247	313	15	are	be	AUX
cana-6247	313	16	crucial	crucial	ADJ
cana-6247	313	17	for	for	ADP
cana-6247	313	18	medical	medical	ADJ
cana-6247	313	19	,	,	PUNCT
cana-6247	313	20	remote	remote	ADJ
cana-6247	313	21	sensing	sensing	NOUN
cana-6247	313	22	or	or	CCONJ
cana-6247	313	23	out	out	ADV
cana-6247	313	24	-	-	PUNCT
cana-6247	313	25	of	of	ADP
cana-6247	313	26	-	-	PUNCT
cana-6247	313	27	distribution	distribution	NOUN
cana-6247	313	28	images	image	NOUN
cana-6247	313	29	.	.	PUNCT
cana-6247	314	1	[	[	X
cana-6247	314	2	41	41	NUM
cana-6247	314	3	]	]	X
cana-6247	314	4	[	[	X
cana-6247	314	5	52	52	NUM
cana-6247	314	6	]	]	PUNCT
cana-6247	314	7	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	314	8	991	991	NUM
cana-6247	314	9	communications	communication	NOUN
cana-6247	314	10	on	on	ADP
cana-6247	314	11	applied	apply	VERB
cana-6247	314	12	nonlinear	nonlinear	ADJ
cana-6247	314	13	analysis	analysis	NOUN
cana-6247	314	14	issn	issn	NOUN
cana-6247	314	15	:	:	PUNCT
cana-6247	314	16	1074	1074	NUM
cana-6247	314	17	-	-	PUNCT
cana-6247	314	18	133x	133x	NUM
cana-6247	314	19	vol	vol	NOUN
cana-6247	314	20	.	.	PROPN
cana-6247	315	1	31	31	NUM
cana-6247	316	1	no	no	NOUN
cana-6247	316	2	.	.	PUNCT
cana-6247	317	1	7s	7	NOUN
cana-6247	317	2	(	(	PUNCT
cana-6247	317	3	2024	2024	NUM
cana-6247	317	4	)	)	PUNCT
cana-6247	317	5	•	•	NUM
cana-6247	317	6	supresdiffgan	supresdiffgan	NOUN
cana-6247	317	7	fuses	fuse	NOUN
cana-6247	317	8	diffusion	diffusion	NOUN
cana-6247	317	9	and	and	CCONJ
cana-6247	317	10	adversarial	adversarial	ADJ
cana-6247	317	11	learning	learning	NOUN
cana-6247	317	12	through	through	ADP
cana-6247	317	13	adaptive	adaptive	ADJ
cana-6247	317	14	,	,	PUNCT
cana-6247	317	15	task	task	NOUN
cana-6247	317	16	-	-	PUNCT
cana-6247	317	17	driven	drive	VERB
cana-6247	317	18	noise	noise	NOUN
cana-6247	317	19	modulations	modulation	NOUN
cana-6247	317	20	used	use	VERB
cana-6247	317	21	to	to	PART
cana-6247	317	22	stabilize	stabilize	VERB
cana-6247	317	23	training	training	NOUN
cana-6247	317	24	and	and	CCONJ
cana-6247	317	25	enhance	enhance	VERB
cana-6247	317	26	both	both	DET
cana-6247	317	27	sharpness	sharpness	NOUN
cana-6247	317	28	and	and	CCONJ
cana-6247	317	29	realism	realism	NOUN
cana-6247	317	30	in	in	ADP
cana-6247	317	31	the	the	DET
cana-6247	317	32	output	output	NOUN
cana-6247	317	33	generation	generation	NOUN
cana-6247	317	34	of	of	ADP
cana-6247	317	35	hr	hr	NOUN
cana-6247	317	36	,	,	PUNCT
cana-6247	317	37	with	with	ADP
cana-6247	317	38	more	more	ADJ
cana-6247	317	39	speed	speed	NOUN
cana-6247	317	40	and	and	CCONJ
cana-6247	317	41	quality	quality	NOUN
cana-6247	317	42	than	than	ADP
cana-6247	317	43	gan	gan	NOUN
cana-6247	317	44	or	or	CCONJ
cana-6247	317	45	diffusion	diffusion	NOUN
cana-6247	317	46	-	-	PUNCT
cana-6247	317	47	only	only	ADJ
cana-6247	317	48	baselines	baseline	NOUN
cana-6247	317	49	.	.	PUNCT
cana-6247	318	1	[	[	X
cana-6247	318	2	17	17	NUM
cana-6247	318	3	]	]	SYM
cana-6247	318	4	•	•	PRON
cana-6247	318	5	multi	multi	NOUN
cana-6247	318	6	-	-	NOUN
cana-6247	318	7	stage	stage	NOUN
cana-6247	318	8	or	or	CCONJ
cana-6247	318	9	cascaded	cascade	VERB
cana-6247	318	10	diffusion	diffusion	NOUN
cana-6247	318	11	is	be	AUX
cana-6247	318	12	utilized	utilize	VERB
cana-6247	318	13	in	in	ADP
cana-6247	318	14	relay	relay	NOUN
cana-6247	318	15	diffusion	diffusion	NOUN
cana-6247	318	16	and	and	CCONJ
cana-6247	318	17	multi	multi	ADJ
cana-6247	318	18	-	-	ADJ
cana-6247	318	19	scale	scale	ADJ
cana-6247	318	20	adversarial	adversarial	ADJ
cana-6247	318	21	diffusion	diffusion	NOUN
cana-6247	318	22	network	network	NOUN
cana-6247	318	23	,	,	PUNCT
cana-6247	318	24	so	so	ADV
cana-6247	318	25	noise	noise	NOUN
cana-6247	318	26	and	and	CCONJ
cana-6247	318	27	denoising	denoising	NOUN
cana-6247	318	28	are	be	AUX
cana-6247	318	29	learned	learn	VERB
cana-6247	318	30	and	and	CCONJ
cana-6247	318	31	distributed	distribute	VERB
cana-6247	318	32	hierarchically	hierarchically	ADV
cana-6247	318	33	across	across	ADP
cana-6247	318	34	resolutions	resolution	NOUN
cana-6247	318	35	.	.	PUNCT
cana-6247	319	1	this	this	PRON
cana-6247	319	2	allows	allow	VERB
cana-6247	319	3	for	for	ADP
cana-6247	319	4	fine	fine	ADJ
cana-6247	319	5	-	-	PUNCT
cana-6247	319	6	detail	detail	NOUN
cana-6247	319	7	recovery	recovery	NOUN
cana-6247	319	8	and	and	CCONJ
cana-6247	319	9	a	a	DET
cana-6247	319	10	global	global	ADJ
cana-6247	319	11	consistency	consistency	NOUN
cana-6247	319	12	which	which	PRON
cana-6247	319	13	is	be	AUX
cana-6247	319	14	important	important	ADJ
cana-6247	319	15	for	for	ADP
cana-6247	319	16	very	very	ADV
cana-6247	319	17	high	high	ADJ
cana-6247	319	18	resolution	resolution	NOUN
cana-6247	319	19	and	and	CCONJ
cana-6247	319	20	real	real	ADJ
cana-6247	319	21	-	-	PUNCT
cana-6247	319	22	world	world	NOUN
cana-6247	319	23	sr	sr	PROPN
cana-6247	319	24	tasks	task	NOUN
cana-6247	319	25	.	.	PUNCT
cana-6247	320	1	[	[	X
cana-6247	320	2	37	37	NUM
cana-6247	320	3	]	]	X
cana-6247	320	4	[	[	X
cana-6247	320	5	34	34	NUM
cana-6247	320	6	]	]	SYM
cana-6247	320	7	•	•	NUM
cana-6247	320	8	partdiff	partdiff	NOUN
cana-6247	320	9	partitions	partition	VERB
cana-6247	320	10	the	the	DET
cana-6247	320	11	noise	noise	NOUN
cana-6247	320	12	process	process	NOUN
cana-6247	320	13	across	across	ADP
cana-6247	320	14	meaningful	meaningful	ADJ
cana-6247	320	15	segments	segment	NOUN
cana-6247	320	16	,	,	PUNCT
cana-6247	320	17	focusing	focus	VERB
cana-6247	320	18	generative	generative	ADJ
cana-6247	320	19	capacity	capacity	NOUN
cana-6247	320	20	on	on	ADP
cana-6247	320	21	ambiguous	ambiguous	ADJ
cana-6247	320	22	or	or	CCONJ
cana-6247	320	23	underdefined	underdefined	ADJ
cana-6247	320	24	regions	region	NOUN
cana-6247	320	25	,	,	PUNCT
cana-6247	320	26	while	while	SCONJ
cana-6247	320	27	xpsr	xpsr	NOUN
cana-6247	320	28	(	(	PUNCT
cana-6247	320	29	cross	cross	ADJ
cana-6247	320	30	-	-	ADJ
cana-6247	320	31	modal	modal	ADJ
cana-6247	320	32	priors	prior	NOUN
cana-6247	320	33	for	for	ADP
cana-6247	320	34	diffusion	diffusion	NOUN
cana-6247	320	35	-	-	PUNCT
cana-6247	320	36	based	base	VERB
cana-6247	320	37	sr	sr	PROPN
cana-6247	320	38	)	)	PUNCT
cana-6247	320	39	uses	use	VERB
cana-6247	320	40	external	external	ADJ
cana-6247	320	41	priors	prior	NOUN
cana-6247	320	42	and	and	CCONJ
cana-6247	320	43	semantic	semantic	ADJ
cana-6247	320	44	information	information	NOUN
cana-6247	320	45	to	to	PART
cana-6247	320	46	drive	drive	VERB
cana-6247	320	47	noise	noise	NOUN
cana-6247	320	48	scheduling	scheduling	NOUN
cana-6247	320	49	and	and	CCONJ
cana-6247	320	50	enhance	enhance	VERB
cana-6247	320	51	cross	cross	ADJ
cana-6247	320	52	-	-	ADJ
cana-6247	320	53	domain	domain	ADJ
cana-6247	320	54	sr	sr	PROPN
cana-6247	320	55	fidelity	fidelity	PROPN
cana-6247	320	56	.	.	PUNCT
cana-6247	321	1	[	[	X
cana-6247	321	2	53	53	NUM
cana-6247	321	3	]	]	X
cana-6247	322	1	[	[	X
cana-6247	322	2	29	29	NUM
cana-6247	322	3	]	]	PUNCT
cana-6247	322	4	in	in	ADP
cana-6247	322	5	sum	sum	NOUN
cana-6247	322	6	.	.	PUNCT
cana-6247	323	1	learned	learn	VERB
cana-6247	323	2	noise	noise	NOUN
cana-6247	323	3	diffusion	diffusion	NOUN
cana-6247	323	4	models	model	NOUN
cana-6247	323	5	stand	stand	VERB
cana-6247	323	6	out	out	ADP
cana-6247	323	7	by	by	ADP
cana-6247	323	8	utilizing	utilize	VERB
cana-6247	323	9	a	a	DET
cana-6247	323	10	trainable	trainable	ADJ
cana-6247	323	11	noise	noise	NOUN
cana-6247	323	12	process	process	NOUN
cana-6247	323	13	that	that	PRON
cana-6247	323	14	is	be	AUX
cana-6247	323	15	informed	inform	VERB
cana-6247	323	16	by	by	ADP
cana-6247	323	17	not	not	PART
cana-6247	323	18	just	just	ADV
cana-6247	323	19	the	the	DET
cana-6247	323	20	diffusion	diffusion	NOUN
cana-6247	323	21	time	time	NOUN
cana-6247	323	22	step	step	NOUN
cana-6247	323	23	,	,	PUNCT
cana-6247	323	24	but	but	CCONJ
cana-6247	323	25	also	also	ADV
cana-6247	323	26	the	the	DET
cana-6247	323	27	content	content	NOUN
cana-6247	323	28	of	of	ADP
cana-6247	323	29	the	the	DET
cana-6247	323	30	underlying	underlie	VERB
cana-6247	323	31	image	image	NOUN
cana-6247	323	32	,	,	PUNCT
cana-6247	323	33	its	its	PRON
cana-6247	323	34	semantics	semantic	NOUN
cana-6247	323	35	,	,	PUNCT
cana-6247	323	36	and	and	CCONJ
cana-6247	323	37	the	the	DET
cana-6247	323	38	uncertainty	uncertainty	NOUN
cana-6247	323	39	in	in	ADP
cana-6247	323	40	each	each	DET
cana-6247	323	41	region	region	NOUN
cana-6247	323	42	.	.	PUNCT
cana-6247	324	1	this	this	PRON
cana-6247	324	2	enables	enable	VERB
cana-6247	324	3	the	the	DET
cana-6247	324	4	model	model	NOUN
cana-6247	324	5	to	to	PART
cana-6247	324	6	apply	apply	VERB
cana-6247	324	7	varying	vary	VERB
cana-6247	324	8	amounts	amount	NOUN
cana-6247	324	9	of	of	ADP
cana-6247	324	10	noise	noise	NOUN
cana-6247	324	11	in	in	ADP
cana-6247	324	12	different	different	ADJ
cana-6247	324	13	regions	region	NOUN
cana-6247	324	14	and	and	CCONJ
cana-6247	324	15	contexts	contexts	NOUN
cana-6247	324	16	,	,	PUNCT
cana-6247	324	17	allowing	allow	VERB
cana-6247	324	18	it	it	PRON
cana-6247	324	19	to	to	PART
cana-6247	324	20	utilize	utilize	VERB
cana-6247	324	21	its	its	PRON
cana-6247	324	22	generative	generative	ADJ
cana-6247	324	23	ability	ability	NOUN
cana-6247	324	24	where	where	SCONJ
cana-6247	324	25	uncertainty	uncertainty	NOUN
cana-6247	324	26	or	or	CCONJ
cana-6247	324	27	ambiguity	ambiguity	NOUN
cana-6247	324	28	is	be	AUX
cana-6247	324	29	greatest	great	ADJ
cana-6247	324	30	,	,	PUNCT
cana-6247	324	31	and	and	CCONJ
cana-6247	324	32	preserve	preserve	VERB
cana-6247	324	33	structure	structure	NOUN
cana-6247	324	34	and	and	CCONJ
cana-6247	324	35	detail	detail	NOUN
cana-6247	324	36	where	where	SCONJ
cana-6247	324	37	the	the	DET
cana-6247	324	38	context	context	NOUN
cana-6247	324	39	is	be	AUX
cana-6247	324	40	most	most	ADV
cana-6247	324	41	defined	define	VERB
cana-6247	324	42	.	.	PUNCT
cana-6247	325	1	the	the	DET
cana-6247	325	2	benefits	benefit	NOUN
cana-6247	325	3	of	of	ADP
cana-6247	325	4	learned	learn	VERB
cana-6247	325	5	noise	noise	NOUN
cana-6247	325	6	are	be	AUX
cana-6247	325	7	highly	highly	ADV
cana-6247	325	8	adaptive	adaptive	ADJ
cana-6247	325	9	restoration	restoration	NOUN
cana-6247	325	10	,	,	PUNCT
cana-6247	325	11	which	which	PRON
cana-6247	325	12	allows	allow	VERB
cana-6247	325	13	the	the	DET
cana-6247	325	14	model	model	NOUN
cana-6247	325	15	to	to	PART
cana-6247	325	16	restore	restore	VERB
cana-6247	325	17	textures	texture	NOUN
cana-6247	325	18	,	,	PUNCT
cana-6247	325	19	fine	fine	ADJ
cana-6247	325	20	details	detail	NOUN
cana-6247	325	21	,	,	PUNCT
cana-6247	325	22	and	and	CCONJ
cana-6247	325	23	semantic	semantic	ADJ
cana-6247	325	24	boundaries	boundary	NOUN
cana-6247	325	25	in	in	ADP
cana-6247	325	26	any	any	DET
cana-6247	325	27	region	region	NOUN
cana-6247	325	28	by	by	ADP
cana-6247	325	29	varying	vary	VERB
cana-6247	325	30	the	the	DET
cana-6247	325	31	noise	noise	NOUN
cana-6247	325	32	schedule	schedule	NOUN
cana-6247	325	33	it	it	PRON
cana-6247	325	34	applies	apply	VERB
cana-6247	325	35	to	to	ADP
cana-6247	325	36	each	each	DET
cana-6247	325	37	part	part	NOUN
cana-6247	325	38	of	of	ADP
cana-6247	325	39	the	the	DET
cana-6247	325	40	image	image	NOUN
cana-6247	325	41	,	,	PUNCT
cana-6247	325	42	depending	depend	VERB
cana-6247	325	43	on	on	ADP
cana-6247	325	44	the	the	DET
cana-6247	325	45	requirements	requirement	NOUN
cana-6247	325	46	of	of	ADP
cana-6247	325	47	that	that	DET
cana-6247	325	48	part	part	NOUN
cana-6247	325	49	of	of	ADP
cana-6247	325	50	the	the	DET
cana-6247	325	51	image	image	NOUN
cana-6247	325	52	.	.	PUNCT
cana-6247	326	1	this	this	DET
cana-6247	326	2	adaptivity	adaptivity	NOUN
cana-6247	326	3	is	be	AUX
cana-6247	326	4	particularly	particularly	ADV
cana-6247	326	5	useful	useful	ADJ
cana-6247	326	6	in	in	ADP
cana-6247	326	7	complex	complex	ADJ
cana-6247	326	8	tasks	task	NOUN
cana-6247	326	9	such	such	ADJ
cana-6247	326	10	as	as	ADP
cana-6247	326	11	image	image	NOUN
cana-6247	326	12	super	super	NOUN
cana-6247	326	13	-	-	NOUN
cana-6247	326	14	resolution	resolution	NOUN
cana-6247	326	15	and	and	CCONJ
cana-6247	326	16	when	when	SCONJ
cana-6247	326	17	degradation	degradation	NOUN
cana-6247	326	18	is	be	AUX
cana-6247	326	19	spatially	spatially	ADV
cana-6247	326	20	inconsistent	inconsistent	ADJ
cana-6247	326	21	.	.	PUNCT
cana-6247	327	1	this	this	DET
cana-6247	327	2	flexibility	flexibility	NOUN
cana-6247	327	3	also	also	ADV
cana-6247	327	4	allows	allow	VERB
cana-6247	327	5	the	the	DET
cana-6247	327	6	learned	learn	VERB
cana-6247	327	7	noise	noise	NOUN
cana-6247	327	8	model	model	NOUN
cana-6247	327	9	to	to	PART
cana-6247	327	10	extract	extract	VERB
cana-6247	327	11	and	and	CCONJ
cana-6247	327	12	leverage	leverage	VERB
cana-6247	327	13	uncertainty	uncertainty	NOUN
cana-6247	327	14	maps	map	NOUN
cana-6247	327	15	,	,	PUNCT
cana-6247	327	16	semantic	semantic	ADJ
cana-6247	327	17	priors	prior	NOUN
cana-6247	327	18	,	,	PUNCT
cana-6247	327	19	or	or	CCONJ
cana-6247	327	20	degradation	degradation	NOUN
cana-6247	327	21	cues	cue	NOUN
cana-6247	327	22	,	,	PUNCT
cana-6247	327	23	allowing	allow	VERB
cana-6247	327	24	for	for	ADP
cana-6247	327	25	sharper	sharp	ADJ
cana-6247	327	26	and	and	CCONJ
cana-6247	327	27	more	more	ADV
cana-6247	327	28	perceptually	perceptually	ADV
cana-6247	327	29	realistic	realistic	ADJ
cana-6247	327	30	reconstructions	reconstruction	NOUN
cana-6247	327	31	,	,	PUNCT
cana-6247	327	32	while	while	SCONJ
cana-6247	327	33	avoiding	avoid	VERB
cana-6247	327	34	the	the	DET
cana-6247	327	35	smoothing	smoothing	NOUN
cana-6247	327	36	and	and	CCONJ
cana-6247	327	37	artifacts	artifact	NOUN
cana-6247	327	38	common	common	ADJ
cana-6247	327	39	in	in	ADP
cana-6247	327	40	static	static	ADJ
cana-6247	327	41	-	-	PUNCT
cana-6247	327	42	noise	noise	NOUN
cana-6247	327	43	models	model	NOUN
cana-6247	327	44	.	.	PUNCT
cana-6247	328	1	nevertheless	nevertheless	ADV
cana-6247	328	2	,	,	PUNCT
cana-6247	328	3	it	it	PRON
cana-6247	328	4	is	be	AUX
cana-6247	328	5	crucial	crucial	ADJ
cana-6247	328	6	to	to	PART
cana-6247	328	7	acknowledge	acknowledge	VERB
cana-6247	328	8	the	the	DET
cana-6247	328	9	limitations	limitation	NOUN
cana-6247	328	10	associated	associate	VERB
cana-6247	328	11	with	with	ADP
cana-6247	328	12	this	this	DET
cana-6247	328	13	power	power	NOUN
cana-6247	328	14	;	;	PUNCT
cana-6247	328	15	learned	learn	VERB
cana-6247	328	16	noise	noise	NOUN
cana-6247	328	17	is	be	AUX
cana-6247	328	18	substantially	substantially	ADV
cana-6247	328	19	more	more	ADV
cana-6247	328	20	complex	complex	ADJ
cana-6247	328	21	,	,	PUNCT
cana-6247	328	22	more	more	ADV
cana-6247	328	23	computationally	computationally	ADV
cana-6247	328	24	intensive	intensive	ADJ
cana-6247	328	25	,	,	PUNCT
cana-6247	328	26	and	and	CCONJ
cana-6247	328	27	requires	require	VERB
cana-6247	328	28	a	a	DET
cana-6247	328	29	significant	significant	ADJ
cana-6247	328	30	amount	amount	NOUN
cana-6247	328	31	of	of	ADP
cana-6247	328	32	high	high	ADJ
cana-6247	328	33	-	-	PUNCT
cana-6247	328	34	quality	quality	NOUN
cana-6247	328	35	data	datum	NOUN
cana-6247	328	36	and	and	CCONJ
cana-6247	328	37	careful	careful	ADJ
cana-6247	328	38	fine	fine	ADJ
cana-6247	328	39	-	-	PUNCT
cana-6247	328	40	tuning	tuning	NOUN
cana-6247	328	41	of	of	ADP
cana-6247	328	42	the	the	DET
cana-6247	328	43	conditioning	conditioning	NOUN
cana-6247	328	44	pathways	pathway	NOUN
cana-6247	328	45	.	.	PUNCT
cana-6247	329	1	the	the	DET
cana-6247	329	2	specific	specific	ADJ
cana-6247	329	3	conditioning	conditioning	NOUN
cana-6247	329	4	pathways	pathway	NOUN
cana-6247	329	5	and	and	CCONJ
cana-6247	329	6	network	network	NOUN
cana-6247	329	7	architecture	architecture	NOUN
cana-6247	329	8	also	also	ADV
cana-6247	329	9	influence	influence	VERB
cana-6247	329	10	model	model	NOUN
cana-6247	329	11	performance	performance	NOUN
cana-6247	329	12	and	and	CCONJ
cana-6247	329	13	generalization	generalization	NOUN
cana-6247	329	14	,	,	PUNCT
cana-6247	329	15	making	make	VERB
cana-6247	329	16	the	the	DET
cana-6247	329	17	overall	overall	ADJ
cana-6247	329	18	performance	performance	NOUN
cana-6247	329	19	of	of	ADP
cana-6247	329	20	learned	learn	VERB
cana-6247	329	21	noise	noise	NOUN
cana-6247	329	22	more	more	ADV
cana-6247	329	23	sensitive	sensitive	ADJ
cana-6247	329	24	to	to	ADP
cana-6247	329	25	these	these	DET
cana-6247	329	26	design	design	NOUN
cana-6247	329	27	choices	choice	NOUN
cana-6247	329	28	.	.	PUNCT
cana-6247	330	1	taking	take	VERB
cana-6247	330	2	everything	everything	PRON
cana-6247	330	3	into	into	ADP
cana-6247	330	4	account	account	NOUN
cana-6247	330	5	,	,	PUNCT
cana-6247	330	6	learned	learn	VERB
cana-6247	330	7	noise	noise	NOUN
cana-6247	330	8	adds	add	VERB
cana-6247	330	9	flexibility	flexibility	NOUN
cana-6247	330	10	and	and	CCONJ
cana-6247	330	11	strength	strength	NOUN
cana-6247	330	12	to	to	ADP
cana-6247	330	13	diffusion	diffusion	NOUN
cana-6247	330	14	models	model	NOUN
cana-6247	330	15	,	,	PUNCT
cana-6247	330	16	but	but	CCONJ
cana-6247	330	17	introduces	introduce	VERB
cana-6247	330	18	complexities	complexity	NOUN
cana-6247	330	19	during	during	ADP
cana-6247	330	20	training	training	NOUN
cana-6247	330	21	and	and	CCONJ
cana-6247	330	22	inference	inference	NOUN
cana-6247	330	23	that	that	PRON
cana-6247	330	24	may	may	AUX
cana-6247	330	25	require	require	VERB
cana-6247	330	26	significant	significant	ADJ
cana-6247	330	27	engineering	engineering	NOUN
cana-6247	330	28	and	and	CCONJ
cana-6247	330	29	computational	computational	ADJ
cana-6247	330	30	efforts	effort	NOUN
cana-6247	330	31	.	.	PUNCT
cana-6247	331	1	4	4	NUM
cana-6247	331	2	poison	poison	NOUN
cana-6247	331	3	/	/	SYM
cana-6247	331	4	signal	signal	NOUN
cana-6247	331	5	dependent	dependent	ADJ
cana-6247	331	6	noise	noise	NOUN
cana-6247	331	7	models	model	NOUN
cana-6247	331	8	for	for	ADP
cana-6247	331	9	diffusion	diffusion	NOUN
cana-6247	331	10	arising	arise	VERB
cana-6247	331	11	out	out	ADP
cana-6247	331	12	of	of	ADP
cana-6247	331	13	poisson	poisson	PROPN
cana-6247	331	14	mechanics	mechanic	NOUN
cana-6247	331	15	/	/	SYM
cana-6247	331	16	signal	signal	NOUN
cana-6247	331	17	noise	noise	NOUN
cana-6247	331	18	form	form	NOUN
cana-6247	331	19	the	the	DET
cana-6247	331	20	primary	primary	ADJ
cana-6247	331	21	stage	stage	NOUN
cana-6247	331	22	of	of	ADP
cana-6247	331	23	advancement	advancement	NOUN
cana-6247	331	24	over	over	ADP
cana-6247	331	25	classical	classical	ADJ
cana-6247	331	26	diffusion	diffusion	NOUN
cana-6247	331	27	approaches	approach	NOUN
cana-6247	331	28	for	for	ADP
cana-6247	331	29	image	image	NOUN
cana-6247	331	30	scenarios	scenario	NOUN
cana-6247	331	31	dominated	dominate	VERB
cana-6247	331	32	by	by	ADP
cana-6247	331	33	signal	signal	NOUN
cana-6247	331	34	-	-	PUNCT
cana-6247	331	35	dependent	dependent	ADJ
cana-6247	331	36	,	,	PUNCT
cana-6247	331	37	discrete	discrete	ADJ
cana-6247	331	38	noise	noise	NOUN
cana-6247	331	39	,	,	PUNCT
cana-6247	331	40	such	such	ADJ
cana-6247	331	41	as	as	ADP
cana-6247	331	42	low	low	ADJ
cana-6247	331	43	-	-	PUNCT
cana-6247	331	44	light	light	NOUN
cana-6247	331	45	photography	photography	NOUN
cana-6247	331	46	,	,	PUNCT
cana-6247	331	47	medical	medical	ADJ
cana-6247	331	48	imaging	imaging	NOUN
cana-6247	331	49	via	via	ADP
cana-6247	331	50	photon	photon	NOUN
cana-6247	331	51	counting	counting	NOUN
cana-6247	331	52	,	,	PUNCT
cana-6247	331	53	or	or	CCONJ
cana-6247	331	54	defect	defect	VERB
cana-6247	331	55	detection	detection	NOUN
cana-6247	331	56	in	in	ADP
cana-6247	331	57	industrial	industrial	ADJ
cana-6247	331	58	applications	application	NOUN
cana-6247	331	59	.	.	PUNCT
cana-6247	332	1	each	each	DET
cana-6247	332	2	diffusion	diffusion	NOUN
cana-6247	332	3	step	step	NOUN
cana-6247	332	4	,	,	PUNCT
cana-6247	332	5	instead	instead	ADV
cana-6247	332	6	,	,	PUNCT
cana-6247	332	7	is	be	AUX
cana-6247	332	8	mathematically	mathematically	ADV
cana-6247	332	9	defined	define	VERB
cana-6247	332	10	with	with	ADP
cana-6247	332	11	a	a	DET
cana-6247	332	12	poisson	poisson	NOUN
cana-6247	332	13	kernel	kernel	NOUN
cana-6247	332	14	as	as	ADP
cana-6247	332	15	its	its	PRON
cana-6247	332	16	forward	forward	ADJ
cana-6247	332	17	corruption	corruption	NOUN
cana-6247	332	18	process	process	NOUN
cana-6247	332	19	,	,	PUNCT
cana-6247	332	20	rather	rather	ADV
cana-6247	332	21	than	than	ADP
cana-6247	332	22	adding	add	VERB
cana-6247	332	23	gaussian	gaussian	ADJ
cana-6247	332	24	noise	noise	NOUN
cana-6247	332	25	of	of	ADP
cana-6247	332	26	fixed	fix	VERB
cana-6247	332	27	variance	variance	NOUN
cana-6247	332	28	.	.	PUNCT
cana-6247	333	1	that	that	PRON
cana-6247	333	2	is	be	AUX
cana-6247	333	3	,	,	PUNCT
cana-6247	333	4	for	for	ADP
cana-6247	333	5	every	every	DET
cana-6247	333	6	pixel	pixel	NOUN
cana-6247	333	7	or	or	CCONJ
cana-6247	333	8	signal	signal	NOUN
cana-6247	333	9	element	element	NOUN
cana-6247	333	10	,	,	PUNCT
cana-6247	333	11	the	the	DET
cana-6247	333	12	corrupted	corrupted	ADJ
cana-6247	333	13	observation	observation	NOUN
cana-6247	333	14	zγ	zγ	X
cana-6247	333	15	at	at	ADP
cana-6247	333	16	diffusion	diffusion	NOUN
cana-6247	333	17	level	level	NOUN
cana-6247	333	18	γ	γ	NOUN
cana-6247	333	19	is	be	AUX
cana-6247	333	20	sampled	sample	VERB
cana-6247	333	21	from	from	ADP
cana-6247	333	22	a	a	DET
cana-6247	333	23	poisson	poisson	NOUN
cana-6247	333	24	distribution	distribution	NOUN
cana-6247	333	25	with	with	ADP
cana-6247	333	26	mean	mean	ADJ
cana-6247	333	27	γx	γx	NOUN
cana-6247	333	28	,	,	PUNCT
cana-6247	333	29	directly	directly	ADV
cana-6247	333	30	mirroring	mirror	VERB
cana-6247	333	31	the	the	DET
cana-6247	333	32	stochastic	stochastic	ADJ
cana-6247	333	33	arrival	arrival	NOUN
cana-6247	333	34	of	of	ADP
cana-6247	333	35	photons	photon	NOUN
cana-6247	333	36	or	or	CCONJ
cana-6247	333	37	counts	count	VERB
cana-6247	333	38	observable	observable	ADJ
cana-6247	333	39	in	in	ADP
cana-6247	333	40	real	real	ADJ
cana-6247	333	41	sensors	sensor	NOUN
cana-6247	333	42	:	:	PUNCT
cana-6247	333	43	p	p	X
cana-6247	333	44	(	(	PUNCT
cana-6247	333	45	zγ	zγ	PROPN
cana-6247	333	46	|	|	ADV
cana-6247	333	47	x	x	NOUN
cana-6247	333	48	)	)	PUNCT
cana-6247	333	49	=	=	PUNCT
cana-6247	334	1	(	(	PUNCT
cana-6247	334	2	γx)zγe−γx	γx)zγe−γx	VERB
cana-6247	334	3	zγ	zγ	X
cana-6247	334	4	!	!	PUNCT
cana-6247	335	1	z	z	X
cana-6247	335	2	,	,	PUNCT
cana-6247	335	3	γ	γ	PROPN
cana-6247	335	4	∈	∈	PROPN
cana-6247	335	5	{	{	PUNCT
cana-6247	335	6	0	0	NUM
cana-6247	335	7	,	,	PUNCT
cana-6247	335	8	1	1	NUM
cana-6247	335	9	,	,	PUNCT
cana-6247	335	10	2	2	NUM
cana-6247	335	11	,	,	PUNCT
cana-6247	335	12	.	.	PUNCT
cana-6247	335	13	.	.	PUNCT
cana-6247	335	14	.	.	PUNCT
cana-6247	335	15	}	}	PUNCT
cana-6247	336	1	(	(	PUNCT
cana-6247	336	2	5	5	X
cana-6247	336	3	)	)	PUNCT
cana-6247	337	1	[	[	X
cana-6247	337	2	10	10	NUM
cana-6247	337	3	]	]	X
cana-6247	338	1	[	[	X
cana-6247	338	2	30	30	NUM
cana-6247	338	3	]	]	PUNCT
cana-6247	338	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	338	5	992	992	NUM
cana-6247	338	6	communications	communication	NOUN
cana-6247	338	7	on	on	ADP
cana-6247	338	8	applied	apply	VERB
cana-6247	338	9	nonlinear	nonlinear	ADJ
cana-6247	338	10	analysis	analysis	NOUN
cana-6247	338	11	issn	issn	NOUN
cana-6247	338	12	:	:	PUNCT
cana-6247	338	13	1074	1074	NUM
cana-6247	338	14	-	-	PUNCT
cana-6247	338	15	133x	133x	NUM
cana-6247	338	16	vol	vol	NOUN
cana-6247	338	17	.	.	PROPN
cana-6247	339	1	31	31	NUM
cana-6247	340	1	no	no	NOUN
cana-6247	340	2	.	.	PUNCT
cana-6247	341	1	7s	7	NOUN
cana-6247	341	2	(	(	PUNCT
cana-6247	341	3	2024	2024	NUM
cana-6247	341	4	)	)	PUNCT
cana-6247	341	5	here	here	ADV
cana-6247	341	6	,	,	PUNCT
cana-6247	341	7	noise	noise	NOUN
cana-6247	341	8	mean	mean	NOUN
cana-6247	341	9	and	and	CCONJ
cana-6247	341	10	variance	variance	NOUN
cana-6247	341	11	increase	increase	NOUN
cana-6247	341	12	with	with	ADP
cana-6247	341	13	the	the	DET
cana-6247	341	14	signal	signal	ADJ
cana-6247	341	15	intensity	intensity	NOUN
cana-6247	341	16	,	,	PUNCT
cana-6247	341	17	mirroring	mirror	VERB
cana-6247	341	18	the	the	DET
cana-6247	341	19	empirical	empirical	ADJ
cana-6247	341	20	scenario	scenario	NOUN
cana-6247	341	21	of	of	ADP
cana-6247	341	22	photon	photon	NOUN
cana-6247	341	23	noise	noise	NOUN
cana-6247	341	24	where	where	SCONJ
cana-6247	341	25	brighter	bright	ADJ
cana-6247	341	26	regions	region	NOUN
cana-6247	341	27	contain	contain	VERB
cana-6247	341	28	more	more	ADJ
cana-6247	341	29	information	information	NOUN
cana-6247	341	30	but	but	CCONJ
cana-6247	341	31	are	be	AUX
cana-6247	341	32	also	also	ADV
cana-6247	341	33	more	more	ADV
cana-6247	341	34	variable	variable	ADJ
cana-6247	341	35	.	.	PUNCT
cana-6247	342	1	with	with	ADP
cana-6247	342	2	an	an	DET
cana-6247	342	3	increasing	increase	VERB
cana-6247	342	4	diffusion	diffusion	NOUN
cana-6247	342	5	-	-	PUNCT
cana-6247	342	6	step	step	NOUN
cana-6247	342	7	sequence	sequence	NOUN
cana-6247	342	8	γ	γ	PROPN
cana-6247	342	9	any	any	DET
cana-6247	342	10	initial	initial	ADJ
cana-6247	342	11	image	image	NOUN
cana-6247	342	12	transforms	transform	VERB
cana-6247	342	13	into	into	ADP
cana-6247	342	14	a	a	DET
cana-6247	342	15	thoroughly	thoroughly	ADV
cana-6247	342	16	randomized	randomize	VERB
cana-6247	342	17	,	,	PUNCT
cana-6247	342	18	poisson	poisson	NOUN
cana-6247	342	19	-	-	PUNCT
cana-6247	342	20	distributed	distribute	VERB
cana-6247	342	21	array	array	NOUN
cana-6247	342	22	.	.	PUNCT
cana-6247	343	1	contrarily	contrarily	ADV
cana-6247	343	2	,	,	PUNCT
cana-6247	343	3	in	in	ADP
cana-6247	343	4	the	the	DET
cana-6247	343	5	context	context	NOUN
cana-6247	343	6	of	of	ADP
cana-6247	343	7	gaussian	gaussian	ADJ
cana-6247	343	8	corruption	corruption	NOUN
cana-6247	343	9	,	,	PUNCT
cana-6247	343	10	noise	noise	NOUN
cana-6247	343	11	is	be	AUX
cana-6247	343	12	uniform	uniform	ADJ
cana-6247	343	13	and	and	CCONJ
cana-6247	343	14	agnostic	agnostic	ADJ
cana-6247	343	15	of	of	ADP
cana-6247	343	16	the	the	DET
cana-6247	343	17	signal	signal	NOUN
cana-6247	343	18	,	,	PUNCT
cana-6247	343	19	whereas	whereas	SCONJ
cana-6247	343	20	poisson	poisson	NOUN
cana-6247	343	21	diffusion	diffusion	NOUN
cana-6247	343	22	ensures	ensure	VERB
cana-6247	343	23	each	each	DET
cana-6247	343	24	location	location	NOUN
cana-6247	343	25	’s	’s	PART
cana-6247	343	26	uncertainty	uncertainty	NOUN
cana-6247	343	27	is	be	AUX
cana-6247	343	28	aware	aware	ADJ
cana-6247	343	29	of	of	ADP
cana-6247	343	30	and	and	CCONJ
cana-6247	343	31	physically	physically	ADV
cana-6247	343	32	meaningful	meaningful	ADJ
cana-6247	343	33	concerning	concern	VERB
cana-6247	343	34	the	the	DET
cana-6247	343	35	signal	signal	NOUN
cana-6247	343	36	.	.	PUNCT
cana-6247	344	1	from	from	ADP
cana-6247	344	2	a	a	DET
cana-6247	344	3	mathematical	mathematical	ADJ
cana-6247	344	4	perspective	perspective	NOUN
cana-6247	344	5	,	,	PUNCT
cana-6247	344	6	the	the	DET
cana-6247	344	7	forward	forward	ADJ
cana-6247	344	8	diffusion	diffusion	NOUN
cana-6247	344	9	kernel	kernel	NOUN
cana-6247	344	10	at	at	ADP
cana-6247	344	11	an	an	DET
cana-6247	344	12	intermediate	intermediate	ADJ
cana-6247	344	13	step	step	NOUN
cana-6247	344	14	is	be	AUX
cana-6247	344	15	given	give	VERB
cana-6247	344	16	by	by	ADP
cana-6247	344	17	q(zt	q(zt	PROPN
cana-6247	344	18	|	|	ADV
cana-6247	344	19	zt−1	zt−1	NOUN
cana-6247	344	20	,	,	PUNCT
cana-6247	344	21	x	x	NOUN
cana-6247	344	22	)	)	PUNCT
cana-6247	344	23	=	=	SYM
cana-6247	344	24	poisson(λt(x	poisson(λt(x	NOUN
cana-6247	344	25	)	)	PUNCT
cana-6247	344	26	)	)	PUNCT
cana-6247	344	27	where	where	SCONJ
cana-6247	344	28	the	the	DET
cana-6247	344	29	schedule	schedule	NOUN
cana-6247	344	30	λt(x	λt(x	PUNCT
cana-6247	344	31	)	)	PUNCT
cana-6247	344	32	(	(	PUNCT
cana-6247	344	33	x	x	X
cana-6247	344	34	)	)	PUNCT
cana-6247	344	35	increases	increase	VERB
cana-6247	344	36	monotonically	monotonically	ADV
cana-6247	344	37	so	so	SCONJ
cana-6247	344	38	as	as	SCONJ
cana-6247	344	39	to	to	PART
cana-6247	344	40	randomize	randomize	VERB
cana-6247	344	41	the	the	DET
cana-6247	344	42	image	image	NOUN
cana-6247	344	43	.	.	PUNCT
cana-6247	345	1	note	note	VERB
cana-6247	345	2	the	the	DET
cana-6247	345	3	main	main	ADJ
cana-6247	345	4	aspect	aspect	NOUN
cana-6247	345	5	that	that	PRON
cana-6247	345	6	distinguishes	distinguish	VERB
cana-6247	345	7	this	this	DET
cana-6247	345	8	process	process	NOUN
cana-6247	345	9	,	,	PUNCT
cana-6247	345	10	the	the	DET
cana-6247	345	11	noise	noise	NOUN
cana-6247	345	12	is	be	AUX
cana-6247	345	13	not	not	PART
cana-6247	345	14	additive	additive	ADJ
cana-6247	345	15	and	and	CCONJ
cana-6247	345	16	is	be	AUX
cana-6247	345	17	data	datum	NOUN
cana-6247	345	18	-	-	PUNCT
cana-6247	345	19	dependent	dependent	ADJ
cana-6247	345	20	:	:	PUNCT
cana-6247	345	21	image	image	NOUN
cana-6247	345	22	regions	region	NOUN
cana-6247	345	23	with	with	ADP
cana-6247	345	24	weak	weak	ADJ
cana-6247	345	25	signal	signal	NOUN
cana-6247	345	26	vanish	vanish	VERB
cana-6247	345	27	in	in	ADP
cana-6247	345	28	the	the	DET
cana-6247	345	29	noise	noise	NOUN
cana-6247	345	30	(	(	PUNCT
cana-6247	345	31	just	just	ADV
cana-6247	345	32	like	like	INTJ
cana-6247	345	33	in	in	ADP
cana-6247	345	34	actual	actual	ADJ
cana-6247	345	35	photon	photon	NOUN
cana-6247	345	36	-	-	PUNCT
cana-6247	345	37	starved	starve	VERB
cana-6247	345	38	sensors	sensor	NOUN
cana-6247	345	39	)	)	PUNCT
cana-6247	345	40	,	,	PUNCT
cana-6247	345	41	whereas	whereas	SCONJ
cana-6247	345	42	bright	bright	ADJ
cana-6247	345	43	,	,	PUNCT
cana-6247	345	44	more	more	ADV
cana-6247	345	45	textured	textured	ADJ
cana-6247	345	46	regions	region	NOUN
cana-6247	345	47	gather	gather	VERB
cana-6247	345	48	more	more	ADJ
cana-6247	345	49	noise	noise	NOUN
cana-6247	345	50	energy	energy	NOUN
cana-6247	345	51	-	-	PUNCT
cana-6247	345	52	mimicking	mimic	VERB
cana-6247	345	53	a	a	DET
cana-6247	345	54	more	more	ADV
cana-6247	345	55	realistic	realistic	ADJ
cana-6247	345	56	model	model	NOUN
cana-6247	345	57	of	of	ADP
cana-6247	345	58	degraded	degraded	ADJ
cana-6247	345	59	image	image	NOUN
cana-6247	345	60	.	.	PUNCT
cana-6247	346	1	reverse	reverse	ADJ
cana-6247	346	2	diffusion	diffusion	NOUN
cana-6247	346	3	in	in	ADP
cana-6247	346	4	these	these	DET
cana-6247	346	5	models	model	NOUN
cana-6247	346	6	thus	thus	ADV
cana-6247	346	7	involves	involve	VERB
cana-6247	346	8	learning	learn	VERB
cana-6247	346	9	a	a	DET
cana-6247	346	10	generative	generative	ADJ
cana-6247	346	11	mapping	mapping	NOUN
cana-6247	346	12	p(x	p(x	PROPN
cana-6247	346	13	|	|	NOUN
cana-6247	346	14	zt	zt	PROPN
cana-6247	346	15	)	)	PUNCT
cana-6247	346	16	that	that	PRON
cana-6247	346	17	reconstructs	reconstruct	VERB
cana-6247	346	18	the	the	DET
cana-6247	346	19	original	original	ADJ
cana-6247	346	20	,	,	PUNCT
cana-6247	346	21	clean	clean	ADJ
cana-6247	346	22	image	image	NOUN
cana-6247	346	23	x	x	PUNCT
cana-6247	346	24	from	from	ADP
cana-6247	346	25	noisy	noisy	ADJ
cana-6247	346	26	observations	observation	NOUN
cana-6247	347	1	zt	zt	PROPN
cana-6247	347	2	.	.	PUNCT
cana-6247	348	1	this	this	PRON
cana-6247	348	2	is	be	AUX
cana-6247	348	3	often	often	ADV
cana-6247	348	4	operationalized	operationalize	VERB
cana-6247	348	5	as	as	ADP
cana-6247	348	6	a	a	DET
cana-6247	348	7	neural	neural	ADJ
cana-6247	348	8	network	network	NOUN
cana-6247	348	9	trained	train	VERB
cana-6247	348	10	to	to	PART
cana-6247	348	11	minimize	minimize	VERB
cana-6247	348	12	a	a	DET
cana-6247	348	13	poisson	poisson	NOUN
cana-6247	348	14	negative	negative	ADJ
cana-6247	348	15	log	log	NOUN
cana-6247	348	16	-	-	PUNCT
cana-6247	348	17	likelihood	likelihood	NOUN
cana-6247	348	18	loss	loss	NOUN
cana-6247	348	19	lpoisson	lpoisson	NOUN
cana-6247	348	20	=	=	PUNCT
cana-6247	348	21	−	−	PROPN
cana-6247	349	1	∑	∑	PROPN
cana-6247	349	2	i	i	PRON
cana-6247	349	3	(	(	PUNCT
cana-6247	349	4	zi	zi	NOUN
cana-6247	349	5	log	log	VERB
cana-6247	349	6	x̂i	x̂i	PROPN
cana-6247	349	7	−	−	PROPN
cana-6247	349	8	x̂i	x̂i	NUM
cana-6247	350	1	−	−	NOUN
cana-6247	350	2	log(zi	log(zi	NOUN
cana-6247	350	3	!	!	PUNCT
cana-6247	350	4	)	)	PUNCT
cana-6247	350	5	)	)	PUNCT
cana-6247	350	6	which	which	PRON
cana-6247	350	7	is	be	AUX
cana-6247	350	8	both	both	PRON
cana-6247	350	9	physically	physically	ADV
cana-6247	350	10	grounded	ground	VERB
cana-6247	350	11	and	and	CCONJ
cana-6247	350	12	provides	provide	VERB
cana-6247	350	13	an	an	DET
cana-6247	350	14	exact	exact	ADJ
cana-6247	350	15	likelihood	likelihood	NOUN
cana-6247	350	16	term	term	NOUN
cana-6247	350	17	for	for	ADP
cana-6247	350	18	image	image	NOUN
cana-6247	350	19	counts	count	NOUN
cana-6247	350	20	or	or	CCONJ
cana-6247	350	21	intensities	intensity	NOUN
cana-6247	350	22	[	[	X
cana-6247	350	23	10	10	NUM
cana-6247	350	24	]	]	PUNCT
cana-6247	350	25	.	.	PUNCT
cana-6247	351	1	some	some	DET
cana-6247	351	2	poisson	poisson	PROPN
cana-6247	351	3	/	/	SYM
cana-6247	351	4	signal	signal	NOUN
cana-6247	351	5	diffusion	diffusion	NOUN
cana-6247	351	6	models	model	NOUN
cana-6247	351	7	are	be	AUX
cana-6247	351	8	placed	place	VERB
cana-6247	351	9	in	in	ADP
cana-6247	351	10	continuous	continuous	ADJ
cana-6247	351	11	time	time	NOUN
cana-6247	351	12	or	or	CCONJ
cana-6247	351	13	high	high	ADJ
cana-6247	351	14	-	-	PUNCT
cana-6247	351	15	dimensional	dimensional	ADJ
cana-6247	351	16	space	space	NOUN
cana-6247	351	17	(	(	PUNCT
cana-6247	351	18	as	as	ADP
cana-6247	351	19	with	with	ADP
cana-6247	351	20	poisson	poisson	NOUN
cana-6247	351	21	flow	flow	NOUN
cana-6247	351	22	generative	generative	ADJ
cana-6247	351	23	models	model	NOUN
cana-6247	351	24	)	)	PUNCT
cana-6247	351	25	,	,	PUNCT
cana-6247	351	26	wherein	wherein	SCONJ
cana-6247	351	27	data	datum	NOUN
cana-6247	351	28	evolves	evolve	VERB
cana-6247	351	29	along	along	ADP
cana-6247	351	30	the	the	DET
cana-6247	351	31	vector	vector	NOUN
cana-6247	351	32	field	field	NOUN
cana-6247	351	33	solution	solution	NOUN
cana-6247	351	34	of	of	ADP
cana-6247	351	35	the	the	DET
cana-6247	351	36	poisson	poisson	NOUN
cana-6247	351	37	equationjust	equationjust	ADV
cana-6247	351	38	like	like	ADP
cana-6247	351	39	how	how	SCONJ
cana-6247	351	40	physical	physical	ADJ
cana-6247	351	41	fields	field	NOUN
cana-6247	351	42	propagate	propagate	VERB
cana-6247	351	43	.	.	PUNCT
cana-6247	352	1	the	the	DET
cana-6247	352	2	forward	forward	ADJ
cana-6247	352	3	trajectory	trajectory	NOUN
cana-6247	352	4	follows	follow	VERB
cana-6247	352	5	stochastic	stochastic	NOUN
cana-6247	352	6	,	,	PUNCT
cana-6247	352	7	data	data	NOUN
cana-6247	352	8	-	-	PUNCT
cana-6247	352	9	dependent	dependent	ADJ
cana-6247	352	10	progressions	progression	NOUN
cana-6247	352	11	prescribed	prescribe	VERB
cana-6247	352	12	on	on	ADP
cana-6247	352	13	the	the	DET
cana-6247	352	14	image	image	NOUN
cana-6247	352	15	intensity	intensity	NOUN
cana-6247	352	16	,	,	PUNCT
cana-6247	352	17	while	while	SCONJ
cana-6247	352	18	the	the	DET
cana-6247	352	19	reverse	reverse	NOUN
cana-6247	352	20	tries	try	VERB
cana-6247	352	21	to	to	PART
cana-6247	352	22	bring	bring	VERB
cana-6247	352	23	data	datum	NOUN
cana-6247	352	24	to	to	ADP
cana-6247	352	25	the	the	DET
cana-6247	352	26	original	original	ADJ
cana-6247	352	27	manifold	manifold	NOUN
cana-6247	352	28	.	.	PUNCT
cana-6247	353	1	[	[	X
cana-6247	353	2	21	21	NUM
cana-6247	353	3	]	]	PUNCT
cana-6247	353	4	the	the	DET
cana-6247	353	5	advantages	advantage	NOUN
cana-6247	353	6	are	be	AUX
cana-6247	353	7	quite	quite	ADV
cana-6247	353	8	clear	clear	ADJ
cana-6247	353	9	:	:	PUNCT
cana-6247	353	10	signal	signal	NOUN
cana-6247	353	11	-	-	PUNCT
cana-6247	353	12	based	base	VERB
cana-6247	353	13	systems	system	NOUN
cana-6247	353	14	are	be	AUX
cana-6247	353	15	inherently	inherently	ADV
cana-6247	353	16	able	able	ADJ
cana-6247	353	17	to	to	PART
cana-6247	353	18	adapt	adapt	VERB
cana-6247	353	19	to	to	ADP
cana-6247	353	20	scenarios	scenario	NOUN
cana-6247	353	21	where	where	SCONJ
cana-6247	353	22	noise	noise	NOUN
cana-6247	353	23	statistics	statistic	NOUN
cana-6247	353	24	are	be	AUX
cana-6247	353	25	signal	signal	NOUN
cana-6247	353	26	-	-	PUNCT
cana-6247	353	27	dependent	dependent	ADJ
cana-6247	353	28	and	and	CCONJ
cana-6247	353	29	discrete	discrete	ADJ
cana-6247	353	30	,	,	PUNCT
cana-6247	353	31	unlike	unlike	ADP
cana-6247	353	32	gaussian	gaussian	NOUN
cana-6247	353	33	-	-	PUNCT
cana-6247	353	34	based	base	VERB
cana-6247	353	35	models	model	NOUN
cana-6247	353	36	,	,	PUNCT
cana-6247	353	37	which	which	PRON
cana-6247	353	38	have	have	VERB
cana-6247	353	39	the	the	DET
cana-6247	353	40	tendency	tendency	NOUN
cana-6247	353	41	of	of	ADP
cana-6247	353	42	oversmoothing	oversmoothe	VERB
cana-6247	353	43	textures	texture	NOUN
cana-6247	353	44	or	or	CCONJ
cana-6247	353	45	creating	create	VERB
cana-6247	353	46	artifacts	artifact	NOUN
cana-6247	353	47	in	in	ADP
cana-6247	353	48	the	the	DET
cana-6247	353	49	presence	presence	NOUN
cana-6247	353	50	of	of	ADP
cana-6247	353	51	quantization	quantization	NOUN
cana-6247	353	52	,	,	PUNCT
cana-6247	353	53	degradation	degradation	NOUN
cana-6247	353	54	,	,	PUNCT
cana-6247	353	55	or	or	CCONJ
cana-6247	353	56	spartan	spartan	ADJ
cana-6247	353	57	photon	photon	NOUN
cana-6247	353	58	counts	count	VERB
cana-6247	353	59	[	[	X
cana-6247	353	60	30	30	NUM
cana-6247	353	61	]	]	X
cana-6247	354	1	[	[	X
cana-6247	354	2	46	46	NUM
cana-6247	354	3	]	]	PUNCT
cana-6247	354	4	.	.	PUNCT
cana-6247	355	1	poisson	poisson	NOUN
cana-6247	355	2	-	-	PUNCT
cana-6247	355	3	guided	guide	VERB
cana-6247	355	4	models	model	NOUN
cana-6247	355	5	preserve	preserve	VERB
cana-6247	355	6	edges	edge	NOUN
cana-6247	355	7	and	and	CCONJ
cana-6247	355	8	fine	fine	ADJ
cana-6247	355	9	details	detail	NOUN
cana-6247	355	10	mainly	mainly	ADV
cana-6247	355	11	because	because	SCONJ
cana-6247	355	12	the	the	DET
cana-6247	355	13	noise	noise	NOUN
cana-6247	355	14	adaptation	adaptation	NOUN
cana-6247	355	15	is	be	AUX
cana-6247	355	16	strongest	strong	ADJ
cana-6247	355	17	in	in	ADP
cana-6247	355	18	flat	flat	ADJ
cana-6247	355	19	or	or	CCONJ
cana-6247	355	20	low	low	ADJ
cana-6247	355	21	-	-	PUNCT
cana-6247	355	22	signal	signal	NOUN
cana-6247	355	23	areas	area	NOUN
cana-6247	355	24	,	,	PUNCT
cana-6247	355	25	while	while	SCONJ
cana-6247	355	26	being	be	AUX
cana-6247	355	27	numerically	numerically	ADV
cana-6247	355	28	more	more	ADV
cana-6247	355	29	stable	stable	ADJ
cana-6247	355	30	and	and	CCONJ
cana-6247	355	31	robust	robust	ADJ
cana-6247	355	32	regarding	regard	VERB
cana-6247	355	33	training	training	NOUN
cana-6247	355	34	and	and	CCONJ
cana-6247	355	35	inference	inference	NOUN
cana-6247	355	36	,	,	PUNCT
cana-6247	355	37	precisely	precisely	ADV
cana-6247	355	38	due	due	ADJ
cana-6247	355	39	to	to	ADP
cana-6247	355	40	an	an	DET
cana-6247	355	41	exact	exact	ADJ
cana-6247	355	42	match	match	NOUN
cana-6247	355	43	with	with	ADP
cana-6247	355	44	the	the	DET
cana-6247	355	45	expected	expect	VERB
cana-6247	355	46	data	data	NOUN
cana-6247	355	47	likelihoods	likelihood	NOUN
cana-6247	355	48	.	.	PUNCT
cana-6247	356	1	these	these	DET
cana-6247	356	2	models	model	NOUN
cana-6247	356	3	are	be	AUX
cana-6247	356	4	thus	thus	ADV
cana-6247	356	5	the	the	DET
cana-6247	356	6	best	good	ADJ
cana-6247	356	7	fit	fit	NOUN
cana-6247	356	8	towards	towards	ADP
cana-6247	356	9	reconstructing	reconstruct	VERB
cana-6247	356	10	images	image	NOUN
cana-6247	356	11	from	from	ADP
cana-6247	356	12	photon	photon	NOUN
cana-6247	356	13	-	-	PUNCT
cana-6247	356	14	limited	limit	VERB
cana-6247	356	15	sensors	sensor	NOUN
cana-6247	356	16	,	,	PUNCT
cana-6247	356	17	denoising	denoise	VERB
cana-6247	356	18	in	in	ADP
cana-6247	356	19	medical	medical	ADJ
cana-6247	356	20	and	and	CCONJ
cana-6247	356	21	scientific	scientific	ADJ
cana-6247	356	22	spaces	space	NOUN
cana-6247	356	23	,	,	PUNCT
cana-6247	356	24	as	as	ADV
cana-6247	356	25	well	well	ADV
cana-6247	356	26	as	as	ADP
cana-6247	356	27	enhancing	enhance	VERB
cana-6247	356	28	highly	highly	ADV
cana-6247	356	29	-	-	PUNCT
cana-6247	356	30	degraded	degrade	VERB
cana-6247	356	31	images	image	NOUN
cana-6247	356	32	towards	towards	ADP
cana-6247	356	33	defect	defect	ADJ
cana-6247	356	34	detection	detection	NOUN
cana-6247	356	35	with	with	ADP
cana-6247	356	36	severely	severely	ADV
cana-6247	356	37	low	low	ADJ
cana-6247	356	38	snr	snr	NOUN
cana-6247	356	39	.	.	PUNCT
cana-6247	357	1	[	[	X
cana-6247	357	2	55	55	NUM
cana-6247	357	3	]	]	X
cana-6247	357	4	[	[	X
cana-6247	357	5	44	44	NUM
cana-6247	357	6	]	]	SYM
cana-6247	357	7	innovations	innovation	NOUN
cana-6247	357	8	like	like	ADP
cana-6247	357	9	information	information	NOUN
cana-6247	357	10	-	-	PUNCT
cana-6247	357	11	theoretic	theoretic	NOUN
cana-6247	357	12	discrete	discrete	NOUN
cana-6247	357	13	poisson	poisson	NOUN
cana-6247	357	14	diffusion	diffusion	NOUN
cana-6247	357	15	(	(	PUNCT
cana-6247	357	16	itdpdm	itdpdm	NOUN
cana-6247	357	17	)	)	PUNCT
cana-6247	357	18	eliminate	eliminate	VERB
cana-6247	357	19	the	the	DET
cana-6247	357	20	need	need	NOUN
cana-6247	357	21	for	for	ADP
cana-6247	357	22	variational	variational	ADJ
cana-6247	357	23	approximations	approximation	NOUN
cana-6247	357	24	.	.	PUNCT
cana-6247	358	1	they	they	PRON
cana-6247	358	2	use	use	VERB
cana-6247	358	3	exact	exact	ADJ
cana-6247	358	4	poisson	poisson	NOUN
cana-6247	358	5	-	-	PUNCT
cana-6247	358	6	based	base	VERB
cana-6247	358	7	likelihoods	likelihood	NOUN
cana-6247	358	8	to	to	PART
cana-6247	358	9	reinforce	reinforce	VERB
cana-6247	358	10	statistics	statistic	NOUN
cana-6247	358	11	during	during	ADP
cana-6247	358	12	training	training	NOUN
cana-6247	358	13	and	and	CCONJ
cana-6247	358	14	inference	inference	NOUN
cana-6247	358	15	,	,	PUNCT
cana-6247	358	16	which	which	PRON
cana-6247	358	17	enhances	enhance	VERB
cana-6247	358	18	reliability	reliability	NOUN
cana-6247	358	19	.	.	PUNCT
cana-6247	359	1	signal	signal	ADJ
cana-6247	359	2	diffusion	diffusion	NOUN
cana-6247	359	3	formulations	formulation	NOUN
cana-6247	359	4	,	,	PUNCT
cana-6247	359	5	particularly	particularly	ADV
cana-6247	359	6	those	those	PRON
cana-6247	359	7	drawn	draw	VERB
cana-6247	359	8	from	from	ADP
cana-6247	359	9	poisson	poisson	PROPN
cana-6247	359	10	field	field	NOUN
cana-6247	359	11	equations	equation	NOUN
cana-6247	359	12	,	,	PUNCT
cana-6247	359	13	provide	provide	VERB
cana-6247	359	14	both	both	DET
cana-6247	359	15	physical	physical	ADJ
cana-6247	359	16	consistency	consistency	NOUN
cana-6247	359	17	and	and	CCONJ
cana-6247	359	18	generative	generative	ADJ
cana-6247	359	19	flexibility	flexibility	NOUN
cana-6247	359	20	.	.	PUNCT
cana-6247	360	1	the	the	DET
cana-6247	360	2	forward	forward	ADJ
cana-6247	360	3	process	process	NOUN
cana-6247	360	4	is	be	AUX
cana-6247	360	5	no	no	ADV
cana-6247	360	6	longer	long	ADV
cana-6247	360	7	uniform	uniform	ADJ
cana-6247	360	8	or	or	CCONJ
cana-6247	360	9	random	random	ADJ
cana-6247	360	10	;	;	PUNCT
cana-6247	360	11	it	it	PRON
cana-6247	360	12	is	be	AUX
cana-6247	360	13	controlled	control	VERB
cana-6247	360	14	by	by	ADP
cana-6247	360	15	field	field	NOUN
cana-6247	360	16	theory	theory	NOUN
cana-6247	360	17	,	,	PUNCT
cana-6247	360	18	leading	lead	VERB
cana-6247	360	19	to	to	ADP
cana-6247	360	20	better	well	ADJ
cana-6247	360	21	convergence	convergence	NOUN
cana-6247	360	22	and	and	CCONJ
cana-6247	360	23	realism	realism	NOUN
cana-6247	360	24	.	.	PUNCT
cana-6247	361	1	[	[	X
cana-6247	361	2	18	18	NUM
cana-6247	361	3	]	]	X
cana-6247	361	4	[	[	X
cana-6247	361	5	45	45	NUM
cana-6247	361	6	]	]	PUNCT
cana-6247	361	7	in	in	ADP
cana-6247	361	8	practical	practical	ADJ
cana-6247	361	9	terms	term	NOUN
cana-6247	361	10	,	,	PUNCT
cana-6247	361	11	this	this	PRON
cana-6247	361	12	means	mean	VERB
cana-6247	361	13	that	that	SCONJ
cana-6247	361	14	poisson	poisson	NOUN
cana-6247	361	15	/	/	SYM
cana-6247	361	16	signal	signal	NOUN
cana-6247	361	17	diffusion	diffusion	NOUN
cana-6247	361	18	frameworks	framework	NOUN
cana-6247	361	19	work	work	VERB
cana-6247	361	20	better	well	ADV
cana-6247	361	21	than	than	ADP
cana-6247	361	22	gaussian	gaussian	ADJ
cana-6247	361	23	models	model	NOUN
cana-6247	361	24	when	when	SCONJ
cana-6247	361	25	imaging	imaging	NOUN
cana-6247	361	26	is	be	AUX
cana-6247	361	27	limited	limit	VERB
cana-6247	361	28	by	by	ADP
cana-6247	361	29	photon	photon	NOUN
cana-6247	361	30	count	count	NOUN
cana-6247	361	31	or	or	CCONJ
cana-6247	361	32	sensor	sensor	NOUN
cana-6247	361	33	physics	physic	NOUN
cana-6247	361	34	.	.	PUNCT
cana-6247	362	1	they	they	PRON
cana-6247	362	2	offer	offer	VERB
cana-6247	362	3	sharper	sharp	ADJ
cana-6247	362	4	restoration	restoration	NOUN
cana-6247	362	5	,	,	PUNCT
cana-6247	362	6	introduce	introduce	VERB
cana-6247	362	7	fewer	few	ADJ
cana-6247	362	8	artifacts	artifact	NOUN
cana-6247	362	9	,	,	PUNCT
cana-6247	362	10	recover	recover	VERB
cana-6247	362	11	more	more	ADV
cana-6247	362	12	faithfully	faithfully	ADV
cana-6247	362	13	in	in	ADP
cana-6247	362	14	low	low	ADJ
cana-6247	362	15	-	-	PUNCT
cana-6247	362	16	light	light	ADJ
cana-6247	362	17	,	,	PUNCT
cana-6247	362	18	medical	medical	ADJ
cana-6247	362	19	,	,	PUNCT
cana-6247	362	20	or	or	CCONJ
cana-6247	362	21	quantized	quantize	VERB
cana-6247	362	22	situations	situation	NOUN
cana-6247	362	23	,	,	PUNCT
cana-6247	362	24	and	and	CCONJ
cana-6247	362	25	are	be	AUX
cana-6247	362	26	more	more	ADV
cana-6247	362	27	resilient	resilient	ADJ
cana-6247	362	28	to	to	ADP
cana-6247	362	29	the	the	DET
cana-6247	362	30	modeling	modeling	NOUN
cana-6247	362	31	and	and	CCONJ
cana-6247	362	32	dequantization	dequantization	NOUN
cana-6247	362	33	artifacts	artifact	NOUN
cana-6247	362	34	that	that	PRON
cana-6247	362	35	often	often	ADV
cana-6247	362	36	affect	affect	VERB
cana-6247	362	37	classical	classical	ADJ
cana-6247	362	38	models	model	NOUN
cana-6247	362	39	.	.	PUNCT
cana-6247	363	1	the	the	DET
cana-6247	363	2	solid	solid	ADJ
cana-6247	363	3	math	math	NOUN
cana-6247	363	4	and	and	CCONJ
cana-6247	363	5	real	real	ADJ
cana-6247	363	6	sensor	sensor	NOUN
cana-6247	363	7	data	datum	NOUN
cana-6247	363	8	backing	back	VERB
cana-6247	363	9	these	these	DET
cana-6247	363	10	techniques	technique	NOUN
cana-6247	363	11	make	make	VERB
cana-6247	363	12	them	they	PRON
cana-6247	363	13	the	the	DET
cana-6247	363	14	best	good	ADJ
cana-6247	363	15	option	option	NOUN
cana-6247	363	16	for	for	ADP
cana-6247	363	17	the	the	DET
cana-6247	363	18	next	next	ADJ
cana-6247	363	19	generation	generation	NOUN
cana-6247	363	20	of	of	ADP
cana-6247	363	21	image	image	NOUN
cana-6247	363	22	improvement	improvement	NOUN
cana-6247	363	23	,	,	PUNCT
cana-6247	363	24	restoration	restoration	NOUN
cana-6247	363	25	,	,	PUNCT
cana-6247	363	26	and	and	CCONJ
cana-6247	363	27	scientific	scientific	ADJ
cana-6247	363	28	imaging	imaging	NOUN
cana-6247	363	29	systems	system	NOUN
cana-6247	363	30	.	.	PUNCT
cana-6247	364	1	4.1	4.1	NUM
cana-6247	364	2	low	low	ADJ
cana-6247	364	3	light	light	ADJ
cana-6247	364	4	image	image	NOUN
cana-6247	364	5	enhancement	enhancement	NOUN
cana-6247	364	6	poisson	poisson	NOUN
cana-6247	364	7	/	/	SYM
cana-6247	364	8	signal	signal	NOUN
cana-6247	364	9	-	-	PUNCT
cana-6247	364	10	dependent	dependent	ADJ
cana-6247	364	11	noise	noise	NOUN
cana-6247	364	12	diffusion	diffusion	NOUN
cana-6247	364	13	models	model	NOUN
cana-6247	364	14	are	be	AUX
cana-6247	364	15	excellent	excellent	ADJ
cana-6247	364	16	for	for	ADP
cana-6247	364	17	low	low	ADJ
cana-6247	364	18	-	-	PUNCT
cana-6247	364	19	light	light	NOUN
cana-6247	364	20	enhancement	enhancement	NOUN
cana-6247	364	21	as	as	SCONJ
cana-6247	364	22	they	they	PRON
cana-6247	364	23	"	"	PUNCT
cana-6247	364	24	naturally	naturally	ADV
cana-6247	364	25	"	"	PUNCT
cana-6247	364	26	model	model	NOUN
cana-6247	364	27	the	the	DET
cana-6247	364	28	heteroscedastic	heteroscedastic	ADJ
cana-6247	364	29	,	,	PUNCT
cana-6247	364	30	signal	signal	NOUN
cana-6247	364	31	-	-	PUNCT
cana-6247	364	32	dependent	dependent	ADJ
cana-6247	364	33	nature	nature	NOUN
cana-6247	364	34	of	of	ADP
cana-6247	364	35	photon	photon	NOUN
cana-6247	364	36	noise	noise	NOUN
cana-6247	364	37	under	under	ADP
cana-6247	364	38	these	these	DET
cana-6247	364	39	conditions	condition	NOUN
cana-6247	364	40	.	.	PUNCT
cana-6247	365	1	as	as	ADP
cana-6247	365	2	a	a	DET
cana-6247	365	3	result	result	NOUN
cana-6247	365	4	,	,	PUNCT
cana-6247	365	5	they	they	PRON
cana-6247	365	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	365	7	993	993	NUM
cana-6247	365	8	communications	communication	NOUN
cana-6247	365	9	on	on	ADP
cana-6247	365	10	applied	apply	VERB
cana-6247	365	11	nonlinear	nonlinear	ADJ
cana-6247	365	12	analysis	analysis	NOUN
cana-6247	365	13	issn	issn	NOUN
cana-6247	365	14	:	:	PUNCT
cana-6247	365	15	1074	1074	NUM
cana-6247	365	16	-	-	PUNCT
cana-6247	365	17	133x	133x	NUM
cana-6247	365	18	vol	vol	NOUN
cana-6247	365	19	.	.	PROPN
cana-6247	366	1	31	31	NUM
cana-6247	367	1	no	no	NOUN
cana-6247	367	2	.	.	PUNCT
cana-6247	368	1	7s	7	NOUN
cana-6247	368	2	(	(	PUNCT
cana-6247	368	3	2024	2024	NUM
cana-6247	368	4	)	)	PUNCT
cana-6247	368	5	figure	figure	NOUN
cana-6247	368	6	5	5	NUM
cana-6247	368	7	:	:	PUNCT
cana-6247	368	8	forward	forward	ADJ
cana-6247	368	9	diffusion	diffusion	NOUN
cana-6247	368	10	process	process	NOUN
cana-6247	368	11	in	in	ADP
cana-6247	368	12	poison	poison	NOUN
cana-6247	368	13	/	/	SYM
cana-6247	368	14	signal	signal	NOUN
cana-6247	368	15	dependent	dependent	ADJ
cana-6247	368	16	noise	noise	NOUN
cana-6247	368	17	type	type	NOUN
cana-6247	368	18	diffusion	diffusion	NOUN
cana-6247	368	19	model	model	NOUN
cana-6247	368	20	afford	afford	VERB
cana-6247	368	21	accurate	accurate	ADJ
cana-6247	368	22	noise	noise	NOUN
cana-6247	368	23	modeling	modeling	NOUN
cana-6247	368	24	with	with	ADP
cana-6247	368	25	appropriately	appropriately	ADV
cana-6247	368	26	low	low	ADJ
cana-6247	368	27	noise	noise	NOUN
cana-6247	368	28	variance	variance	NOUN
cana-6247	368	29	for	for	ADP
cana-6247	368	30	darker	dark	ADJ
cana-6247	368	31	,	,	PUNCT
cana-6247	368	32	lower	low	ADJ
cana-6247	368	33	photon	photon	NOUN
cana-6247	368	34	regions	region	NOUN
cana-6247	368	35	,	,	PUNCT
cana-6247	368	36	and	and	CCONJ
cana-6247	368	37	with	with	ADP
cana-6247	368	38	proportionately	proportionately	ADV
cana-6247	368	39	more	more	ADJ
cana-6247	368	40	noise	noise	NOUN
cana-6247	368	41	for	for	ADP
cana-6247	368	42	brighter	bright	ADJ
cana-6247	368	43	,	,	PUNCT
cana-6247	368	44	higher	high	ADJ
cana-6247	368	45	photon	photon	NOUN
cana-6247	368	46	count	count	NOUN
cana-6247	368	47	regions	region	NOUN
cana-6247	368	48	,	,	PUNCT
cana-6247	368	49	giving	give	VERB
cana-6247	368	50	restoration	restoration	NOUN
cana-6247	368	51	of	of	ADP
cana-6247	368	52	both	both	DET
cana-6247	368	53	light	light	ADJ
cana-6247	368	54	highlights	highlight	NOUN
cana-6247	368	55	and	and	CCONJ
cana-6247	368	56	subtle	subtle	ADJ
cana-6247	368	57	shadows	shadow	NOUN
cana-6247	368	58	[	[	X
cana-6247	368	59	30	30	NUM
cana-6247	368	60	]	]	PUNCT
cana-6247	368	61	.	.	PUNCT
cana-6247	369	1	furthermore	furthermore	ADV
cana-6247	369	2	,	,	PUNCT
cana-6247	369	3	unlike	unlike	ADP
cana-6247	369	4	gaussian	gaussian	NOUN
cana-6247	369	5	-	-	PUNCT
cana-6247	369	6	based	base	VERB
cana-6247	369	7	diffusion	diffusion	NOUN
cana-6247	369	8	,	,	PUNCT
cana-6247	369	9	which	which	PRON
cana-6247	369	10	frequently	frequently	ADV
cana-6247	369	11	calls	call	VERB
cana-6247	369	12	for	for	ADP
cana-6247	369	13	approximations	approximation	NOUN
cana-6247	369	14	or	or	CCONJ
cana-6247	369	15	variance	variance	NOUN
cana-6247	369	16	stabilisation	stabilisation	NOUN
cana-6247	369	17	transforms	transform	VERB
cana-6247	369	18	,	,	PUNCT
cana-6247	369	19	the	the	DET
cana-6247	369	20	precise	precise	ADJ
cana-6247	369	21	poisson	poisson	NOUN
cana-6247	369	22	likelihood	likelihood	NOUN
cana-6247	369	23	formulations	formulation	NOUN
cana-6247	369	24	used	use	VERB
cana-6247	369	25	to	to	PART
cana-6247	369	26	train	train	VERB
cana-6247	369	27	these	these	DET
cana-6247	369	28	models	model	NOUN
cana-6247	369	29	allow	allow	VERB
cana-6247	369	30	for	for	ADP
cana-6247	369	31	stable	stable	ADJ
cana-6247	369	32	,	,	PUNCT
cana-6247	369	33	theoretically	theoretically	ADV
cana-6247	369	34	-	-	PUNCT
cana-6247	369	35	grounded	ground	VERB
cana-6247	369	36	optimisation	optimisation	NOUN
cana-6247	369	37	that	that	PRON
cana-6247	369	38	is	be	AUX
cana-6247	369	39	more	more	ADV
cana-6247	369	40	appropriate	appropriate	ADJ
cana-6247	369	41	for	for	ADP
cana-6247	369	42	discrete	discrete	ADJ
cana-6247	369	43	photon	photon	NOUN
cana-6247	369	44	count	count	NOUN
cana-6247	369	45	data	datum	NOUN
cana-6247	369	46	[	[	X
cana-6247	369	47	30	30	NUM
cana-6247	369	48	]	]	PUNCT
cana-6247	369	49	.	.	PUNCT
cana-6247	370	1	this	this	PRON
cana-6247	370	2	makes	make	VERB
cana-6247	370	3	models	model	NOUN
cana-6247	370	4	more	more	ADV
cana-6247	370	5	resilient	resilient	ADJ
cana-6247	370	6	to	to	AUX
cana-6247	370	7	domain	domain	VERB
cana-6247	370	8	shifts	shift	NOUN
cana-6247	370	9	and	and	CCONJ
cana-6247	370	10	improves	improve	VERB
cana-6247	370	11	generalisation	generalisation	NOUN
cana-6247	370	12	to	to	ADP
cana-6247	370	13	real	real	ADJ
cana-6247	370	14	-	-	PUNCT
cana-6247	370	15	world	world	NOUN
cana-6247	370	16	low	low	ADJ
cana-6247	370	17	-	-	PUNCT
cana-6247	370	18	light	light	NOUN
cana-6247	370	19	photos	photo	NOUN
cana-6247	370	20	,	,	PUNCT
cana-6247	370	21	which	which	PRON
cana-6247	370	22	differ	differ	VERB
cana-6247	370	23	greatly	greatly	ADV
cana-6247	370	24	in	in	ADP
cana-6247	370	25	sensor	sensor	NOUN
cana-6247	370	26	characteristics	characteristic	NOUN
cana-6247	370	27	and	and	CCONJ
cana-6247	370	28	illumination	illumination	NOUN
cana-6247	370	29	levels	level	NOUN
cana-6247	370	30	.	.	PUNCT
cana-6247	371	1	by	by	ADP
cana-6247	371	2	using	use	VERB
cana-6247	371	3	reaction	reaction	NOUN
cana-6247	371	4	-	-	PUNCT
cana-6247	371	5	diffusion	diffusion	NOUN
cana-6247	371	6	dynamics	dynamic	NOUN
cana-6247	371	7	in	in	ADP
cana-6247	371	8	conjunction	conjunction	NOUN
cana-6247	371	9	with	with	ADP
cana-6247	371	10	poisson	poisson	PROPN
cana-6247	371	11	noise	noise	NOUN
cana-6247	371	12	modeling	modeling	NOUN
cana-6247	371	13	,	,	PUNCT
cana-6247	371	14	models	model	NOUN
cana-6247	371	15	that	that	PRON
cana-6247	371	16	employ	employ	VERB
cana-6247	371	17	signal	signal	NOUN
cana-6247	371	18	-	-	PUNCT
cana-6247	371	19	dependent	dependent	ADJ
cana-6247	371	20	noise	noise	NOUN
cana-6247	371	21	can	can	AUX
cana-6247	371	22	also	also	ADV
cana-6247	371	23	improve	improve	VERB
cana-6247	371	24	underlying	underlie	VERB
cana-6247	371	25	image	image	NOUN
cana-6247	371	26	structures	structure	NOUN
cana-6247	371	27	while	while	SCONJ
cana-6247	371	28	suppressing	suppress	VERB
cana-6247	371	29	noise	noise	NOUN
cana-6247	371	30	.	.	PUNCT
cana-6247	372	1	this	this	DET
cana-6247	372	2	combined	combined	ADJ
cana-6247	372	3	effect	effect	NOUN
cana-6247	372	4	smoothes	smooth	VERB
cana-6247	372	5	noise	noise	NOUN
cana-6247	372	6	in	in	ADP
cana-6247	372	7	flat	flat	ADJ
cana-6247	372	8	areas	area	NOUN
cana-6247	372	9	and	and	CCONJ
cana-6247	372	10	enhances	enhance	VERB
cana-6247	372	11	the	the	DET
cana-6247	372	12	recovery	recovery	NOUN
cana-6247	372	13	of	of	ADP
cana-6247	372	14	high	high	ADJ
cana-6247	372	15	-	-	PUNCT
cana-6247	372	16	frequency	frequency	NOUN
cana-6247	372	17	details	detail	NOUN
cana-6247	372	18	[	[	X
cana-6247	372	19	1	1	X
cana-6247	372	20	]	]	X
cana-6247	372	21	[	[	X
cana-6247	372	22	12	12	NUM
cana-6247	372	23	]	]	PUNCT
cana-6247	372	24	.	.	PUNCT
cana-6247	373	1	thus	thus	ADV
cana-6247	373	2	,	,	PUNCT
cana-6247	373	3	in	in	ADP
cana-6247	373	4	addition	addition	NOUN
cana-6247	373	5	to	to	ADP
cana-6247	373	6	having	have	VERB
cana-6247	373	7	physical	physical	ADJ
cana-6247	373	8	significance	significance	NOUN
cana-6247	373	9	,	,	PUNCT
cana-6247	373	10	the	the	DET
cana-6247	373	11	noise	noise	NOUN
cana-6247	373	12	process	process	NOUN
cana-6247	373	13	serves	serve	VERB
cana-6247	373	14	as	as	ADP
cana-6247	373	15	a	a	DET
cana-6247	373	16	learnt	learnt	NOUN
cana-6247	373	17	prior	prior	ADV
cana-6247	373	18	that	that	SCONJ
cana-6247	373	19	aids	aid	NOUN
cana-6247	373	20	in	in	ADP
cana-6247	373	21	adaptively	adaptively	ADV
cana-6247	373	22	separating	separate	VERB
cana-6247	373	23	noise	noise	NOUN
cana-6247	373	24	from	from	ADP
cana-6247	373	25	the	the	DET
cana-6247	373	26	real	real	ADJ
cana-6247	373	27	signal	signal	NOUN
cana-6247	373	28	.	.	PUNCT
cana-6247	374	1	4.1.1	4.1.1	NUM
cana-6247	374	2	evidence	evidence	NOUN
cana-6247	374	3	from	from	ADP
cana-6247	374	4	research	research	NOUN
cana-6247	374	5	•	•	NUM
cana-6247	374	6	poisson	poisson	NOUN
cana-6247	374	7	-	-	PUNCT
cana-6247	374	8	guided	guide	VERB
cana-6247	374	9	decomposition	decomposition	NOUN
cana-6247	374	10	network	network	NOUN
cana-6247	374	11	for	for	ADP
cana-6247	374	12	extreme	extreme	ADJ
cana-6247	374	13	low	low	ADJ
cana-6247	374	14	-	-	PUNCT
cana-6247	374	15	light	light	NOUN
cana-6247	374	16	image	image	NOUN
cana-6247	374	17	enhancement	enhancement	NOUN
cana-6247	374	18	showed	show	VERB
cana-6247	374	19	that	that	SCONJ
cana-6247	374	20	by	by	ADP
cana-6247	374	21	directly	directly	ADV
cana-6247	374	22	modeling	model	VERB
cana-6247	374	23	signal	signal	NOUN
cana-6247	374	24	-	-	PUNCT
cana-6247	374	25	dependent	dependent	ADJ
cana-6247	374	26	poisson	poisson	NOUN
cana-6247	374	27	noise	noise	NOUN
cana-6247	374	28	,	,	PUNCT
cana-6247	374	29	the	the	DET
cana-6247	374	30	method	method	NOUN
cana-6247	374	31	could	could	AUX
cana-6247	374	32	better	well	ADV
cana-6247	374	33	separate	separate	VERB
cana-6247	374	34	real	real	ADJ
cana-6247	374	35	image	image	NOUN
cana-6247	374	36	structure	structure	NOUN
cana-6247	374	37	from	from	ADP
cana-6247	374	38	noise	noise	NOUN
cana-6247	374	39	,	,	PUNCT
cana-6247	374	40	even	even	ADV
cana-6247	374	41	in	in	ADP
cana-6247	374	42	very	very	ADV
cana-6247	374	43	low	low	ADJ
cana-6247	374	44	light	light	NOUN
cana-6247	374	45	.	.	PUNCT
cana-6247	375	1	by	by	ADP
cana-6247	375	2	adjusting	adjust	VERB
cana-6247	375	3	the	the	DET
cana-6247	375	4	diffusion	diffusion	NOUN
cana-6247	375	5	and	and	CCONJ
cana-6247	375	6	enhancement	enhancement	NOUN
cana-6247	375	7	process	process	NOUN
cana-6247	375	8	to	to	PART
cana-6247	375	9	fit	fit	VERB
cana-6247	375	10	the	the	DET
cana-6247	375	11	actual	actual	ADJ
cana-6247	375	12	noise	noise	NOUN
cana-6247	375	13	distribution	distribution	NOUN
cana-6247	375	14	in	in	ADP
cana-6247	375	15	low	low	ADJ
cana-6247	375	16	-	-	PUNCT
cana-6247	375	17	light	light	NOUN
cana-6247	375	18	sensor	sensor	NOUN
cana-6247	375	19	readings	reading	NOUN
cana-6247	375	20	,	,	PUNCT
cana-6247	375	21	the	the	DET
cana-6247	375	22	model	model	NOUN
cana-6247	375	23	significantly	significantly	ADV
cana-6247	375	24	improved	improve	VERB
cana-6247	375	25	the	the	DET
cana-6247	375	26	preservation	preservation	NOUN
cana-6247	375	27	of	of	ADP
cana-6247	375	28	fine	fine	ADJ
cana-6247	375	29	edges	edge	NOUN
cana-6247	375	30	and	and	CCONJ
cana-6247	375	31	complex	complex	ADJ
cana-6247	375	32	textures	texture	NOUN
cana-6247	375	33	that	that	PRON
cana-6247	375	34	traditional	traditional	ADJ
cana-6247	375	35	gaussian	gaussian	NOUN
cana-6247	375	36	-	-	PUNCT
cana-6247	375	37	based	base	VERB
cana-6247	375	38	enhancement	enhancement	NOUN
cana-6247	375	39	would	would	AUX
cana-6247	375	40	often	often	ADV
cana-6247	375	41	smooth	smooth	VERB
cana-6247	375	42	out	out	ADP
cana-6247	375	43	or	or	CCONJ
cana-6247	375	44	lose	lose	VERB
cana-6247	375	45	.	.	PUNCT
cana-6247	376	1	this	this	DET
cana-6247	376	2	approach	approach	NOUN
cana-6247	376	3	created	create	VERB
cana-6247	376	4	visual	visual	ADJ
cana-6247	376	5	results	result	NOUN
cana-6247	376	6	with	with	ADP
cana-6247	376	7	clear	clear	ADJ
cana-6247	376	8	boundaries	boundary	NOUN
cana-6247	376	9	and	and	CCONJ
cana-6247	376	10	detailed	detailed	ADJ
cana-6247	376	11	surfaces	surface	NOUN
cana-6247	376	12	,	,	PUNCT
cana-6247	376	13	especially	especially	ADV
cana-6247	376	14	in	in	ADP
cana-6247	376	15	difficult	difficult	ADJ
cana-6247	376	16	,	,	PUNCT
cana-6247	376	17	dark	dark	ADJ
cana-6247	376	18	environments	environment	NOUN
cana-6247	376	19	.	.	PUNCT
cana-6247	377	1	[	[	X
cana-6247	377	2	30	30	NUM
cana-6247	377	3	]	]	SYM
cana-6247	377	4	•	•	NUM
cana-6247	377	5	photon	photon	NOUN
cana-6247	377	6	-	-	PUNCT
cana-6247	377	7	counting	count	VERB
cana-6247	377	8	ct	ct	NOUN
cana-6247	377	9	denoising	denoising	NOUN
cana-6247	377	10	using	use	VERB
cana-6247	377	11	unsupervised	unsupervised	ADJ
cana-6247	377	12	poisson	poisson	NOUN
cana-6247	377	13	flow	flow	NOUN
cana-6247	377	14	generative	generative	ADJ
cana-6247	377	15	models	model	NOUN
cana-6247	377	16	advanced	advance	VERB
cana-6247	377	17	the	the	DET
cana-6247	377	18	field	field	NOUN
cana-6247	377	19	by	by	ADP
cana-6247	377	20	demonstrating	demonstrate	VERB
cana-6247	377	21	that	that	SCONJ
cana-6247	377	22	generative	generative	ADJ
cana-6247	377	23	models	model	NOUN
cana-6247	377	24	designed	design	VERB
cana-6247	377	25	for	for	ADP
cana-6247	377	26	poisson	poisson	NOUN
cana-6247	377	27	noise	noise	NOUN
cana-6247	377	28	can	can	AUX
cana-6247	377	29	deliver	deliver	VERB
cana-6247	377	30	topnotch	topnotch	NOUN
cana-6247	377	31	denoising	denoising	NOUN
cana-6247	377	32	results	result	NOUN
cana-6247	377	33	in	in	ADP
cana-6247	377	34	photon	photon	NOUN
cana-6247	377	35	-	-	PUNCT
cana-6247	377	36	limited	limit	VERB
cana-6247	377	37	ct	ct	PROPN
cana-6247	377	38	scans	scan	NOUN
cana-6247	377	39	.	.	PUNCT
cana-6247	378	1	in	in	ADP
cana-6247	378	2	these	these	DET
cana-6247	378	3	scans	scan	NOUN
cana-6247	378	4	,	,	PUNCT
cana-6247	378	5	each	each	DET
cana-6247	378	6	pixel	pixel	PROPN
cana-6247	378	7	count	count	PROPN
cana-6247	378	8	directly	directly	ADV
cana-6247	378	9	reflects	reflect	VERB
cana-6247	378	10	the	the	DET
cana-6247	378	11	sparse	sparse	ADJ
cana-6247	378	12	measurement	measurement	NOUN
cana-6247	378	13	of	of	ADP
cana-6247	378	14	incoming	incoming	ADJ
cana-6247	378	15	photons	photon	NOUN
cana-6247	378	16	.	.	PUNCT
cana-6247	379	1	importantly	importantly	ADV
cana-6247	379	2	,	,	PUNCT
cana-6247	379	3	their	their	PRON
cana-6247	379	4	method	method	NOUN
cana-6247	379	5	functioned	function	VERB
cana-6247	379	6	well	well	ADV
cana-6247	379	7	even	even	ADV
cana-6247	379	8	without	without	ADP
cana-6247	379	9	paired	pair	VERB
cana-6247	379	10	clean	clean	ADJ
cana-6247	379	11	and	and	CCONJ
cana-6247	379	12	noisy	noisy	ADJ
cana-6247	379	13	data	datum	NOUN
cana-6247	379	14	for	for	ADP
cana-6247	379	15	guidance	guidance	NOUN
cana-6247	379	16	.	.	PUNCT
cana-6247	380	1	this	this	DET
cana-6247	380	2	unsupervised	unsupervised	ADJ
cana-6247	380	3	approach	approach	NOUN
cana-6247	380	4	is	be	AUX
cana-6247	380	5	especially	especially	ADV
cana-6247	380	6	relevant	relevant	ADJ
cana-6247	380	7	for	for	ADP
cana-6247	380	8	lowlight	lowlight	ADJ
cana-6247	380	9	photography	photography	NOUN
cana-6247	380	10	,	,	PUNCT
cana-6247	380	11	as	as	SCONJ
cana-6247	380	12	it	it	PRON
cana-6247	380	13	shows	show	VERB
cana-6247	380	14	that	that	SCONJ
cana-6247	380	15	effective	effective	ADJ
cana-6247	380	16	signal	signal	NOUN
cana-6247	380	17	-	-	PUNCT
cana-6247	380	18	aware	aware	ADJ
cana-6247	380	19	noise	noise	NOUN
cana-6247	380	20	modeling	modeling	NOUN
cana-6247	380	21	allows	allow	VERB
cana-6247	380	22	for	for	ADP
cana-6247	380	23	strong	strong	ADJ
cana-6247	380	24	restoration	restoration	NOUN
cana-6247	380	25	in	in	ADP
cana-6247	380	26	data	data	NOUN
cana-6247	380	27	-	-	PUNCT
cana-6247	380	28	scarce	scarce	NOUN
cana-6247	380	29	or	or	CCONJ
cana-6247	380	30	challenging	challenging	ADJ
cana-6247	380	31	imaging	imaging	NOUN
cana-6247	380	32	conditions	condition	NOUN
cana-6247	380	33	.	.	PUNCT
cana-6247	381	1	this	this	DET
cana-6247	381	2	research	research	NOUN
cana-6247	381	3	connects	connect	VERB
cana-6247	381	4	simulation	simulation	NOUN
cana-6247	381	5	with	with	ADP
cana-6247	381	6	real	real	ADJ
cana-6247	381	7	-	-	PUNCT
cana-6247	381	8	world	world	NOUN
cana-6247	381	9	sensor	sensor	NOUN
cana-6247	381	10	data	datum	NOUN
cana-6247	381	11	.	.	PUNCT
cana-6247	382	1	[	[	X
cana-6247	382	2	10	10	NUM
cana-6247	382	3	]	]	SYM
cana-6247	382	4	•	•	NUM
cana-6247	382	5	noise	noise	NOUN
cana-6247	382	6	synthesis	synthesis	NOUN
cana-6247	382	7	for	for	ADP
cana-6247	382	8	low	low	ADJ
cana-6247	382	9	-	-	PUNCT
cana-6247	382	10	light	light	NOUN
cana-6247	382	11	image	image	NOUN
cana-6247	382	12	denoising	denoise	VERB
cana-6247	382	13	with	with	ADP
cana-6247	382	14	diffusion	diffusion	NOUN
cana-6247	382	15	models	model	NOUN
cana-6247	382	16	showed	show	VERB
cana-6247	382	17	the	the	DET
cana-6247	382	18	benefits	benefit	NOUN
cana-6247	382	19	of	of	ADP
cana-6247	382	20	simulating	simulate	VERB
cana-6247	382	21	poisson	poisson	NOUN
cana-6247	382	22	noise	noise	NOUN
cana-6247	382	23	during	during	ADP
cana-6247	382	24	the	the	DET
cana-6247	382	25	training	training	NOUN
cana-6247	382	26	of	of	ADP
cana-6247	382	27	diffusion	diffusion	NOUN
cana-6247	382	28	models	model	NOUN
cana-6247	382	29	for	for	ADP
cana-6247	382	30	denoising	denoise	VERB
cana-6247	382	31	.	.	PUNCT
cana-6247	383	1	by	by	ADP
cana-6247	383	2	creating	create	VERB
cana-6247	383	3	artificial	artificial	ADJ
cana-6247	383	4	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	383	5	994	994	NUM
cana-6247	383	6	communications	communication	NOUN
cana-6247	383	7	on	on	ADP
cana-6247	383	8	applied	apply	VERB
cana-6247	383	9	nonlinear	nonlinear	ADJ
cana-6247	383	10	analysis	analysis	NOUN
cana-6247	383	11	issn	issn	NOUN
cana-6247	383	12	:	:	PUNCT
cana-6247	383	13	1074	1074	NUM
cana-6247	383	14	-	-	PUNCT
cana-6247	383	15	133x	133x	NUM
cana-6247	383	16	vol	vol	NOUN
cana-6247	383	17	.	.	PROPN
cana-6247	383	18	31	31	NUM
cana-6247	384	1	no	no	NOUN
cana-6247	384	2	.	.	PUNCT
cana-6247	385	1	7s	7	NOUN
cana-6247	385	2	(	(	PUNCT
cana-6247	385	3	2024	2024	NUM
cana-6247	385	4	)	)	PUNCT
cana-6247	385	5	noisy	noisy	ADJ
cana-6247	385	6	images	image	NOUN
cana-6247	385	7	that	that	PRON
cana-6247	385	8	closely	closely	ADV
cana-6247	385	9	resembled	resemble	VERB
cana-6247	385	10	actual	actual	ADJ
cana-6247	385	11	sensor	sensor	NOUN
cana-6247	385	12	noise	noise	NOUN
cana-6247	385	13	from	from	ADP
cana-6247	385	14	low	low	ADJ
cana-6247	385	15	-	-	PUNCT
cana-6247	385	16	light	light	NOUN
cana-6247	385	17	cameras	camera	NOUN
cana-6247	385	18	,	,	PUNCT
cana-6247	385	19	they	they	PRON
cana-6247	385	20	trained	train	VERB
cana-6247	385	21	denoisers	denoiser	NOUN
cana-6247	385	22	that	that	PRON
cana-6247	385	23	yielded	yield	VERB
cana-6247	385	24	more	more	ADV
cana-6247	385	25	realistic	realistic	ADJ
cana-6247	385	26	,	,	PUNCT
cana-6247	385	27	artifact	artifact	ADJ
cana-6247	385	28	-	-	PUNCT
cana-6247	385	29	free	free	ADJ
cana-6247	385	30	results	result	NOUN
cana-6247	385	31	.	.	PUNCT
cana-6247	386	1	their	their	PRON
cana-6247	386	2	research	research	NOUN
cana-6247	386	3	demonstrated	demonstrate	VERB
cana-6247	386	4	that	that	SCONJ
cana-6247	386	5	models	model	NOUN
cana-6247	386	6	adjusted	adjust	VERB
cana-6247	386	7	with	with	ADP
cana-6247	386	8	signal	signal	NOUN
cana-6247	386	9	-	-	PUNCT
cana-6247	386	10	dependent	dependent	ADJ
cana-6247	386	11	noise	noise	NOUN
cana-6247	386	12	synthesis	synthesis	NOUN
cana-6247	386	13	perform	perform	VERB
cana-6247	386	14	better	well	ADJ
cana-6247	386	15	than	than	ADP
cana-6247	386	16	standard	standard	ADJ
cana-6247	386	17	diffusion	diffusion	NOUN
cana-6247	386	18	models	model	NOUN
cana-6247	386	19	,	,	PUNCT
cana-6247	386	20	which	which	PRON
cana-6247	386	21	often	often	ADV
cana-6247	386	22	cause	cause	VERB
cana-6247	386	23	oversmoothing	oversmoothe	VERB
cana-6247	386	24	,	,	PUNCT
cana-6247	386	25	leave	leave	VERB
cana-6247	386	26	residual	residual	ADJ
cana-6247	386	27	artifacts	artifact	NOUN
cana-6247	386	28	,	,	PUNCT
cana-6247	386	29	or	or	CCONJ
cana-6247	386	30	struggle	struggle	VERB
cana-6247	386	31	to	to	PART
cana-6247	386	32	adapt	adapt	VERB
cana-6247	386	33	in	in	ADP
cana-6247	386	34	low	low	ADJ
cana-6247	386	35	-	-	PUNCT
cana-6247	386	36	light	light	NOUN
cana-6247	386	37	conditions	condition	NOUN
cana-6247	386	38	.	.	PUNCT
cana-6247	387	1	[	[	X
cana-6247	387	2	20	20	NUM
cana-6247	387	3	]	]	SYM
cana-6247	387	4	•	•	NUM
cana-6247	387	5	diff	diff	NOUN
cana-6247	387	6	-	-	PUNCT
cana-6247	387	7	retinex	retinex	NOUN
cana-6247	387	8	:	:	PUNCT
cana-6247	387	9	rethinking	rethink	VERB
cana-6247	387	10	low	low	ADJ
cana-6247	387	11	-	-	PUNCT
cana-6247	387	12	light	light	NOUN
cana-6247	387	13	image	image	NOUN
cana-6247	387	14	enhancement	enhancement	NOUN
cana-6247	387	15	with	with	ADP
cana-6247	387	16	a	a	DET
cana-6247	387	17	generative	generative	ADJ
cana-6247	387	18	diffusion	diffusion	NOUN
cana-6247	387	19	model	model	NOUN
cana-6247	387	20	and	and	CCONJ
cana-6247	387	21	difflle	difflle	NOUN
cana-6247	387	22	:	:	PUNCT
cana-6247	387	23	diffusion	diffusion	NOUN
cana-6247	387	24	-	-	PUNCT
cana-6247	387	25	guided	guide	VERB
cana-6247	387	26	domain	domain	NOUN
cana-6247	387	27	calibration	calibration	NOUN
cana-6247	387	28	for	for	ADP
cana-6247	387	29	unsupervised	unsupervised	ADJ
cana-6247	387	30	low	low	ADJ
cana-6247	387	31	-	-	PUNCT
cana-6247	387	32	light	light	NOUN
cana-6247	387	33	image	image	NOUN
cana-6247	387	34	enhancement	enhancement	NOUN
cana-6247	387	35	both	both	PRON
cana-6247	387	36	showed	show	VERB
cana-6247	387	37	that	that	SCONJ
cana-6247	387	38	poisson	poisson	NOUN
cana-6247	387	39	-	-	PUNCT
cana-6247	387	40	based	base	VERB
cana-6247	387	41	diffusion	diffusion	NOUN
cana-6247	387	42	modeling	modeling	NOUN
cana-6247	387	43	is	be	AUX
cana-6247	387	44	essential	essential	ADJ
cana-6247	387	45	for	for	ADP
cana-6247	387	46	overall	overall	ADJ
cana-6247	387	47	lowlight	lowlight	ADJ
cana-6247	387	48	recovery	recovery	NOUN
cana-6247	387	49	.	.	PUNCT
cana-6247	388	1	diff	diff	NOUN
cana-6247	388	2	-	-	PUNCT
cana-6247	388	3	retinex	retinex	PROPN
cana-6247	388	4	proved	prove	VERB
cana-6247	388	5	that	that	SCONJ
cana-6247	388	6	using	use	VERB
cana-6247	388	7	poisson	poisson	NOUN
cana-6247	388	8	noise	noise	NOUN
cana-6247	388	9	not	not	PART
cana-6247	388	10	only	only	ADV
cana-6247	388	11	improved	improve	VERB
cana-6247	388	12	the	the	DET
cana-6247	388	13	reconstruction	reconstruction	NOUN
cana-6247	388	14	of	of	ADP
cana-6247	388	15	lost	lost	ADJ
cana-6247	388	16	brightness	brightness	NOUN
cana-6247	388	17	and	and	CCONJ
cana-6247	388	18	detail	detail	NOUN
cana-6247	388	19	but	but	CCONJ
cana-6247	388	20	also	also	ADV
cana-6247	388	21	allowed	allow	VERB
cana-6247	388	22	for	for	ADP
cana-6247	388	23	a	a	DET
cana-6247	388	24	more	more	ADV
cana-6247	388	25	accurate	accurate	ADJ
cana-6247	388	26	restoration	restoration	NOUN
cana-6247	388	27	of	of	ADP
cana-6247	388	28	dynamic	dynamic	ADJ
cana-6247	388	29	range	range	NOUN
cana-6247	388	30	.	.	PUNCT
cana-6247	389	1	this	this	PRON
cana-6247	389	2	ensured	ensure	VERB
cana-6247	389	3	that	that	SCONJ
cana-6247	389	4	highlights	highlight	NOUN
cana-6247	389	5	and	and	CCONJ
cana-6247	389	6	shadows	shadow	NOUN
cana-6247	389	7	stayed	stay	VERB
cana-6247	389	8	balanced	balanced	ADJ
cana-6247	389	9	and	and	CCONJ
cana-6247	389	10	visually	visually	ADV
cana-6247	389	11	realistic	realistic	ADJ
cana-6247	389	12	.	.	PUNCT
cana-6247	390	1	difflle	difflle	NOUN
cana-6247	390	2	built	build	VERB
cana-6247	390	3	on	on	ADP
cana-6247	390	4	these	these	DET
cana-6247	390	5	ideas	idea	NOUN
cana-6247	390	6	,	,	PUNCT
cana-6247	390	7	demonstrating	demonstrate	VERB
cana-6247	390	8	that	that	SCONJ
cana-6247	390	9	domain	domain	NOUN
cana-6247	390	10	calibration	calibration	NOUN
cana-6247	390	11	techniques	technique	NOUN
cana-6247	390	12	based	base	VERB
cana-6247	390	13	on	on	ADP
cana-6247	390	14	signal	signal	NOUN
cana-6247	390	15	-	-	PUNCT
cana-6247	390	16	dependent	dependent	ADJ
cana-6247	390	17	diffusion	diffusion	NOUN
cana-6247	390	18	models	model	NOUN
cana-6247	390	19	could	could	AUX
cana-6247	390	20	effectively	effectively	ADV
cana-6247	390	21	transfer	transfer	VERB
cana-6247	390	22	enhancement	enhancement	NOUN
cana-6247	390	23	skills	skill	NOUN
cana-6247	390	24	across	across	ADP
cana-6247	390	25	very	very	ADV
cana-6247	390	26	different	different	ADJ
cana-6247	390	27	domains	domain	NOUN
cana-6247	390	28	and	and	CCONJ
cana-6247	390	29	device	device	NOUN
cana-6247	390	30	types	type	NOUN
cana-6247	390	31	.	.	PUNCT
cana-6247	391	1	this	this	PRON
cana-6247	391	2	led	lead	VERB
cana-6247	391	3	to	to	ADP
cana-6247	391	4	images	image	NOUN
cana-6247	391	5	with	with	ADP
cana-6247	391	6	better	well	ADJ
cana-6247	391	7	perceptual	perceptual	ADJ
cana-6247	391	8	quality	quality	NOUN
cana-6247	391	9	,	,	PUNCT
cana-6247	391	10	natural	natural	ADJ
cana-6247	391	11	light	light	NOUN
cana-6247	391	12	balance	balance	NOUN
cana-6247	391	13	,	,	PUNCT
cana-6247	391	14	and	and	CCONJ
cana-6247	391	15	adaptability	adaptability	NOUN
cana-6247	391	16	to	to	ADP
cana-6247	391	17	various	various	ADJ
cana-6247	391	18	lighting	lighting	NOUN
cana-6247	391	19	conditions	condition	NOUN
cana-6247	391	20	.	.	PUNCT
cana-6247	392	1	these	these	DET
cana-6247	392	2	results	result	NOUN
cana-6247	392	3	are	be	AUX
cana-6247	392	4	hard	hard	ADJ
cana-6247	392	5	to	to	PART
cana-6247	392	6	achieve	achieve	VERB
cana-6247	392	7	with	with	ADP
cana-6247	392	8	static	static	ADJ
cana-6247	392	9	,	,	PUNCT
cana-6247	392	10	gaussian	gaussian	ADJ
cana-6247	392	11	-	-	PUNCT
cana-6247	392	12	noise	noise	NOUN
cana-6247	392	13	-	-	PUNCT
cana-6247	392	14	based	base	VERB
cana-6247	392	15	methods	method	NOUN
cana-6247	392	16	.	.	PUNCT
cana-6247	393	1	[	[	X
cana-6247	393	2	46	46	NUM
cana-6247	393	3	]	]	X
cana-6247	393	4	[	[	X
cana-6247	393	5	44	44	NUM
cana-6247	393	6	]	]	PUNCT
cana-6247	393	7	in	in	ADP
cana-6247	393	8	sum	sum	NOUN
cana-6247	393	9	.	.	PUNCT
cana-6247	394	1	poisson	poisson	NOUN
cana-6247	394	2	and	and	CCONJ
cana-6247	394	3	signal	signal	NOUN
cana-6247	394	4	-	-	PUNCT
cana-6247	394	5	dependent	dependent	ADJ
cana-6247	394	6	diffusion	diffusion	NOUN
cana-6247	394	7	models	model	NOUN
cana-6247	394	8	provide	provide	VERB
cana-6247	394	9	major	major	ADJ
cana-6247	394	10	advantages	advantage	NOUN
cana-6247	394	11	for	for	ADP
cana-6247	394	12	low	low	ADJ
cana-6247	394	13	-	-	PUNCT
cana-6247	394	14	light	light	NOUN
cana-6247	394	15	image	image	NOUN
cana-6247	394	16	enhancement	enhancement	NOUN
cana-6247	394	17	and	and	CCONJ
cana-6247	394	18	for	for	ADP
cana-6247	394	19	tasks	task	NOUN
cana-6247	394	20	that	that	PRON
cana-6247	394	21	are	be	AUX
cana-6247	394	22	also	also	ADV
cana-6247	394	23	sensitive	sensitive	ADJ
cana-6247	394	24	to	to	ADP
cana-6247	394	25	the	the	DET
cana-6247	394	26	photon	photon	NOUN
cana-6247	394	27	statistics	statistic	NOUN
cana-6247	394	28	,	,	PUNCT
cana-6247	394	29	such	such	ADJ
cana-6247	394	30	as	as	ADP
cana-6247	394	31	in	in	ADP
cana-6247	394	32	medical	medical	ADJ
cana-6247	394	33	and	and	CCONJ
cana-6247	394	34	scientific	scientific	ADJ
cana-6247	394	35	imaging	imaging	NOUN
cana-6247	394	36	applications	application	NOUN
cana-6247	394	37	.	.	PUNCT
cana-6247	395	1	diffusion	diffusion	NOUN
cana-6247	395	2	models	model	NOUN
cana-6247	395	3	leverage	leverage	NOUN
cana-6247	395	4	noise	noise	NOUN
cana-6247	395	5	as	as	ADP
cana-6247	395	6	a	a	DET
cana-6247	395	7	function	function	NOUN
cana-6247	395	8	of	of	ADP
cana-6247	395	9	local	local	ADJ
cana-6247	395	10	signal	signal	NOUN
cana-6247	395	11	intensity	intensity	NOUN
cana-6247	395	12	,	,	PUNCT
cana-6247	395	13	rather	rather	ADV
cana-6247	395	14	than	than	ADP
cana-6247	395	15	fixed	fix	VERB
cana-6247	395	16	or	or	CCONJ
cana-6247	395	17	gaussian	gaussian	ADJ
cana-6247	395	18	noise	noise	NOUN
cana-6247	395	19	processes	process	NOUN
cana-6247	395	20	;	;	PUNCT
cana-6247	395	21	this	this	PRON
cana-6247	395	22	helps	help	VERB
cana-6247	395	23	to	to	PART
cana-6247	395	24	reflect	reflect	VERB
cana-6247	395	25	the	the	DET
cana-6247	395	26	physical	physical	ADJ
cana-6247	395	27	reality	reality	NOUN
cana-6247	395	28	of	of	ADP
cana-6247	395	29	the	the	DET
cana-6247	395	30	photon	photon	NOUN
cana-6247	395	31	-	-	PUNCT
cana-6247	395	32	limited	limit	VERB
cana-6247	395	33	sensor	sensor	NOUN
cana-6247	395	34	,	,	PUNCT
cana-6247	395	35	and	and	CCONJ
cana-6247	395	36	allows	allow	VERB
cana-6247	395	37	for	for	ADP
cana-6247	395	38	custom	custom	NOUN
cana-6247	395	39	denoising	denoising	NOUN
cana-6247	395	40	for	for	ADP
cana-6247	395	41	recovery	recovery	NOUN
cana-6247	395	42	of	of	ADP
cana-6247	395	43	shadows	shadow	NOUN
cana-6247	395	44	and	and	CCONJ
cana-6247	395	45	highlights	highlight	NOUN
cana-6247	395	46	from	from	ADP
cana-6247	395	47	the	the	DET
cana-6247	395	48	image	image	NOUN
cana-6247	395	49	,	,	PUNCT
cana-6247	395	50	corrupted	corrupt	VERB
cana-6247	395	51	by	by	ADP
cana-6247	395	52	noise	noise	NOUN
cana-6247	395	53	.	.	PUNCT
cana-6247	396	1	using	use	VERB
cana-6247	396	2	a	a	DET
cana-6247	396	3	signalaware	signalaware	ADJ
cana-6247	396	4	strategy	strategy	NOUN
cana-6247	396	5	is	be	AUX
cana-6247	396	6	a	a	DET
cana-6247	396	7	good	good	ADJ
cana-6247	396	8	method	method	NOUN
cana-6247	396	9	to	to	PART
cana-6247	396	10	enhance	enhance	VERB
cana-6247	396	11	the	the	DET
cana-6247	396	12	textural	textural	ADJ
cana-6247	396	13	content	content	NOUN
cana-6247	396	14	of	of	ADP
cana-6247	396	15	low	low	ADJ
cana-6247	396	16	-	-	PUNCT
cana-6247	396	17	light	light	NOUN
cana-6247	396	18	imagery	imagery	NOUN
cana-6247	396	19	while	while	SCONJ
cana-6247	396	20	maintaining	maintain	VERB
cana-6247	396	21	local	local	ADJ
cana-6247	396	22	edge	edge	NOUN
cana-6247	396	23	fidelity	fidelity	NOUN
cana-6247	396	24	and	and	CCONJ
cana-6247	396	25	a	a	DET
cana-6247	396	26	natural	natural	ADJ
cana-6247	396	27	balance	balance	NOUN
cana-6247	396	28	of	of	ADP
cana-6247	396	29	global	global	ADJ
cana-6247	396	30	lightness	lightness	NOUN
cana-6247	396	31	,	,	PUNCT
cana-6247	396	32	where	where	SCONJ
cana-6247	396	33	one	one	PRON
cana-6247	396	34	has	have	VERB
cana-6247	396	35	fewer	few	ADJ
cana-6247	396	36	artifacts	artifact	NOUN
cana-6247	396	37	or	or	CCONJ
cana-6247	396	38	saturation	saturation	NOUN
cana-6247	396	39	and	and	CCONJ
cana-6247	396	40	a	a	DET
cana-6247	396	41	higher	high	ADJ
cana-6247	396	42	dynamic	dynamic	ADJ
cana-6247	396	43	range	range	NOUN
cana-6247	396	44	than	than	ADP
cana-6247	396	45	the	the	DET
cana-6247	396	46	existing	exist	VERB
cana-6247	396	47	methods	method	NOUN
cana-6247	396	48	.	.	PUNCT
cana-6247	397	1	in	in	ADP
cana-6247	397	2	addition	addition	NOUN
cana-6247	397	3	,	,	PUNCT
cana-6247	397	4	exact	exact	ADJ
cana-6247	397	5	poisson	poisson	NOUN
cana-6247	397	6	-	-	PUNCT
cana-6247	397	7	likelihood	likelihood	NOUN
cana-6247	397	8	training	training	NOUN
cana-6247	397	9	allows	allow	VERB
cana-6247	397	10	for	for	ADP
cana-6247	397	11	handling	handle	VERB
cana-6247	397	12	discrete	discrete	ADJ
cana-6247	397	13	count	count	NOUN
cana-6247	397	14	data	datum	NOUN
cana-6247	397	15	in	in	ADP
cana-6247	397	16	a	a	DET
cana-6247	397	17	robust	robust	ADJ
cana-6247	397	18	mathematical	mathematical	ADJ
cana-6247	397	19	framework	framework	NOUN
cana-6247	397	20	,	,	PUNCT
cana-6247	397	21	which	which	PRON
cana-6247	397	22	helps	help	VERB
cana-6247	397	23	with	with	ADP
cana-6247	397	24	resilience	resilience	NOUN
cana-6247	397	25	to	to	ADP
cana-6247	397	26	different	different	ADJ
cana-6247	397	27	types	type	NOUN
cana-6247	397	28	of	of	ADP
cana-6247	397	29	sensors	sensor	NOUN
cana-6247	397	30	or	or	CCONJ
cana-6247	397	31	lighting	lighting	NOUN
cana-6247	397	32	in	in	ADP
cana-6247	397	33	the	the	DET
cana-6247	397	34	scenes	scene	NOUN
cana-6247	397	35	.	.	PUNCT
cana-6247	398	1	this	this	DET
cana-6247	398	2	option	option	NOUN
cana-6247	398	3	is	be	AUX
cana-6247	398	4	not	not	PART
cana-6247	398	5	without	without	ADP
cana-6247	398	6	some	some	DET
cana-6247	398	7	limitations	limitation	NOUN
cana-6247	398	8	,	,	PUNCT
cana-6247	398	9	as	as	SCONJ
cana-6247	398	10	all	all	DET
cana-6247	398	11	these	these	DET
cana-6247	398	12	models	model	NOUN
cana-6247	398	13	depend	depend	VERB
cana-6247	398	14	on	on	ADP
cana-6247	398	15	accurate	accurate	ADJ
cana-6247	398	16	signal	signal	NOUN
cana-6247	398	17	statistics	statistic	NOUN
cana-6247	398	18	,	,	PUNCT
cana-6247	398	19	and	and	CCONJ
cana-6247	398	20	they	they	PRON
cana-6247	398	21	are	be	AUX
cana-6247	398	22	also	also	ADV
cana-6247	398	23	less	less	ADV
cana-6247	398	24	generalizable	generalizable	ADJ
cana-6247	398	25	to	to	ADP
cana-6247	398	26	all	all	DET
cana-6247	398	27	tasks	task	NOUN
cana-6247	398	28	if	if	SCONJ
cana-6247	398	29	the	the	DET
cana-6247	398	30	source	source	NOUN
cana-6247	398	31	of	of	ADP
cana-6247	398	32	noise	noise	NOUN
cana-6247	398	33	in	in	ADP
cana-6247	398	34	images	image	NOUN
cana-6247	398	35	is	be	AUX
cana-6247	398	36	not	not	PART
cana-6247	398	37	intrinsically	intrinsically	ADV
cana-6247	398	38	referred	refer	VERB
cana-6247	398	39	to	to	ADP
cana-6247	398	40	as	as	ADP
cana-6247	398	41	signal	signal	NOUN
cana-6247	398	42	-	-	PUNCT
cana-6247	398	43	dependent	dependent	ADJ
cana-6247	398	44	.	.	PUNCT
cana-6247	399	1	they	they	PRON
cana-6247	399	2	may	may	AUX
cana-6247	399	3	also	also	ADV
cana-6247	399	4	require	require	VERB
cana-6247	399	5	some	some	DET
cana-6247	399	6	aspects	aspect	NOUN
cana-6247	399	7	of	of	ADP
cana-6247	399	8	domain	domain	NOUN
cana-6247	399	9	knowledge	knowledge	NOUN
cana-6247	399	10	when	when	SCONJ
cana-6247	399	11	implemented	implement	VERB
cana-6247	399	12	for	for	ADP
cana-6247	399	13	optimal	optimal	ADJ
cana-6247	399	14	denoising	denoising	NOUN
cana-6247	399	15	performance	performance	NOUN
cana-6247	399	16	,	,	PUNCT
cana-6247	399	17	as	as	SCONJ
cana-6247	399	18	this	this	PRON
cana-6247	399	19	may	may	AUX
cana-6247	399	20	be	be	AUX
cana-6247	399	21	less	less	ADV
cana-6247	399	22	extensible	extensible	ADJ
cana-6247	399	23	for	for	ADP
cana-6247	399	24	other	other	ADJ
cana-6247	399	25	types	type	NOUN
cana-6247	399	26	of	of	ADP
cana-6247	399	27	image	image	NOUN
cana-6247	399	28	restoration	restoration	NOUN
cana-6247	399	29	or	or	CCONJ
cana-6247	399	30	enhancement	enhancement	NOUN
cana-6247	399	31	applications	application	NOUN
cana-6247	399	32	.	.	PUNCT
cana-6247	400	1	conclusion	conclusion	NOUN
cana-6247	400	2	this	this	DET
cana-6247	400	3	review	review	NOUN
cana-6247	400	4	provides	provide	VERB
cana-6247	400	5	a	a	DET
cana-6247	400	6	noise	noise	NOUN
cana-6247	400	7	-	-	PUNCT
cana-6247	400	8	based	base	VERB
cana-6247	400	9	framework	framework	NOUN
cana-6247	400	10	for	for	ADP
cana-6247	400	11	classifying	classify	VERB
cana-6247	400	12	image	image	NOUN
cana-6247	400	13	enhancement	enhancement	NOUN
cana-6247	400	14	diffusion	diffusion	NOUN
cana-6247	400	15	models	model	NOUN
cana-6247	400	16	.	.	PUNCT
cana-6247	401	1	a	a	DET
cana-6247	401	2	taxonomy	taxonomy	NOUN
cana-6247	401	3	of	of	ADP
cana-6247	401	4	diffusion	diffusion	NOUN
cana-6247	401	5	models	model	NOUN
cana-6247	401	6	based	base	VERB
cana-6247	401	7	on	on	ADP
cana-6247	401	8	noise	noise	NOUN
cana-6247	401	9	mechanisms	mechanism	NOUN
cana-6247	401	10	enables	enable	VERB
cana-6247	401	11	the	the	DET
cana-6247	401	12	classification	classification	NOUN
cana-6247	401	13	of	of	ADP
cana-6247	401	14	models	model	NOUN
cana-6247	401	15	into	into	ADP
cana-6247	401	16	four	four	NUM
cana-6247	401	17	categories	category	NOUN
cana-6247	401	18	:	:	PUNCT
cana-6247	401	19	additive	additive	ADJ
cana-6247	401	20	gaussian	gaussian	NOUN
cana-6247	401	21	noise	noise	NOUN
cana-6247	401	22	(	(	PUNCT
cana-6247	401	23	agn	agn	NOUN
cana-6247	401	24	)	)	PUNCT
cana-6247	401	25	,	,	PUNCT
cana-6247	401	26	conditional	conditional	ADJ
cana-6247	401	27	noise	noise	NOUN
cana-6247	401	28	injection	injection	NOUN
cana-6247	401	29	(	(	PUNCT
cana-6247	401	30	cni	cni	PROPN
cana-6247	401	31	)	)	PUNCT
cana-6247	401	32	,	,	PUNCT
cana-6247	401	33	learned	learn	VERB
cana-6247	401	34	noise	noise	NOUN
cana-6247	401	35	,	,	PUNCT
cana-6247	401	36	and	and	CCONJ
cana-6247	401	37	poisson	poisson	PROPN
cana-6247	401	38	/	/	SYM
cana-6247	401	39	signaldependent	signaldependent	PROPN
cana-6247	401	40	.	.	PUNCT
cana-6247	402	1	the	the	DET
cana-6247	402	2	agn	agn	PROPN
cana-6247	402	3	model	model	NOUN
cana-6247	402	4	remains	remain	VERB
cana-6247	402	5	the	the	DET
cana-6247	402	6	state	state	NOUN
cana-6247	402	7	-	-	PUNCT
cana-6247	402	8	of	of	ADP
cana-6247	402	9	-	-	PUNCT
cana-6247	402	10	the	the	DET
cana-6247	402	11	-	-	PUNCT
cana-6247	402	12	art	art	NOUN
cana-6247	402	13	for	for	ADP
cana-6247	402	14	restoration	restoration	NOUN
cana-6247	402	15	since	since	SCONJ
cana-6247	402	16	agn	agn	PROPN
cana-6247	402	17	is	be	AUX
cana-6247	402	18	easily	easily	ADV
cana-6247	402	19	scalable	scalable	ADJ
cana-6247	402	20	across	across	ADP
cana-6247	402	21	dimensions	dimension	NOUN
cana-6247	402	22	while	while	SCONJ
cana-6247	402	23	remaining	remain	VERB
cana-6247	402	24	robust	robust	ADJ
cana-6247	402	25	and	and	CCONJ
cana-6247	402	26	tractable	tractable	ADJ
cana-6247	402	27	.	.	PUNCT
cana-6247	403	1	the	the	DET
cana-6247	403	2	cni	cni	PROPN
cana-6247	403	3	models	model	NOUN
cana-6247	403	4	introduce	introduce	VERB
cana-6247	403	5	critical	critical	ADJ
cana-6247	403	6	advancements	advancement	NOUN
cana-6247	403	7	toward	toward	ADP
cana-6247	403	8	flexible	flexible	ADJ
cana-6247	403	9	and	and	CCONJ
cana-6247	403	10	region	region	NOUN
cana-6247	403	11	-	-	PUNCT
cana-6247	403	12	aware	aware	ADJ
cana-6247	403	13	editing	editing	NOUN
cana-6247	403	14	,	,	PUNCT
cana-6247	403	15	improving	improve	VERB
cana-6247	403	16	boundary	boundary	ADJ
cana-6247	403	17	fidelity	fidelity	NOUN
cana-6247	403	18	across	across	ADP
cana-6247	403	19	a	a	DET
cana-6247	403	20	variety	variety	NOUN
cana-6247	403	21	of	of	ADP
cana-6247	403	22	image	image	NOUN
cana-6247	403	23	enhancement	enhancement	NOUN
cana-6247	403	24	tasks	task	NOUN
cana-6247	403	25	.	.	PUNCT
cana-6247	404	1	the	the	DET
cana-6247	404	2	learned	learn	VERB
cana-6247	404	3	noise	noise	NOUN
cana-6247	404	4	models	model	NOUN
cana-6247	404	5	prioritize	prioritize	VERB
cana-6247	404	6	flexible	flexible	ADJ
cana-6247	404	7	applicability	applicability	NOUN
cana-6247	404	8	and	and	CCONJ
cana-6247	404	9	perceptual	perceptual	ADJ
cana-6247	404	10	quality	quality	NOUN
cana-6247	404	11	in	in	ADP
cana-6247	404	12	difficult	difficult	ADJ
cana-6247	404	13	tasks	task	NOUN
cana-6247	404	14	such	such	ADJ
cana-6247	404	15	as	as	ADP
cana-6247	404	16	super	super	NOUN
cana-6247	404	17	-	-	NOUN
cana-6247	404	18	resolution	resolution	NOUN
cana-6247	404	19	,	,	PUNCT
cana-6247	404	20	using	use	VERB
cana-6247	404	21	noise	noise	NOUN
cana-6247	404	22	characteristics	characteristic	NOUN
cana-6247	404	23	that	that	PRON
cana-6247	404	24	are	be	AUX
cana-6247	404	25	tailored	tailor	VERB
cana-6247	404	26	based	base	VERB
cana-6247	404	27	on	on	ADP
cana-6247	404	28	the	the	DET
cana-6247	404	29	means	mean	NOUN
cana-6247	404	30	of	of	ADP
cana-6247	404	31	uncertainty	uncertainty	NOUN
cana-6247	404	32	and	and	CCONJ
cana-6247	404	33	task	task	NOUN
cana-6247	404	34	context	context	NOUN
cana-6247	404	35	.	.	PUNCT
cana-6247	405	1	the	the	DET
cana-6247	405	2	poisson	poisson	NOUN
cana-6247	405	3	/	/	SYM
cana-6247	405	4	signal	signal	NOUN
cana-6247	405	5	-	-	PUNCT
cana-6247	405	6	dependent	dependent	ADJ
cana-6247	405	7	models	model	NOUN
cana-6247	405	8	abstract	abstract	VERB
cana-6247	405	9	significant	significant	ADJ
cana-6247	405	10	benefits	benefit	NOUN
cana-6247	405	11	over	over	ADP
cana-6247	405	12	low	low	ADJ
cana-6247	405	13	-	-	PUNCT
cana-6247	405	14	light	light	NOUN
cana-6247	405	15	and	and	CCONJ
cana-6247	405	16	sensor	sensor	NOUN
cana-6247	405	17	-	-	PUNCT
cana-6247	405	18	limited	limit	VERB
cana-6247	405	19	images	image	NOUN
cana-6247	405	20	by	by	ADP
cana-6247	405	21	utilizing	utilize	VERB
cana-6247	405	22	noise	noise	NOUN
cana-6247	405	23	that	that	PRON
cana-6247	405	24	fits	fit	VERB
cana-6247	405	25	real	real	ADJ
cana-6247	405	26	-	-	PUNCT
cana-6247	405	27	world	world	NOUN
cana-6247	405	28	optical	optical	ADJ
cana-6247	405	29	sensor	sensor	NOUN
cana-6247	405	30	patterns	pattern	NOUN
cana-6247	405	31	.	.	PUNCT
cana-6247	406	1	taken	take	VERB
cana-6247	406	2	independently	independently	ADV
cana-6247	406	3	but	but	CCONJ
cana-6247	406	4	in	in	ADP
cana-6247	406	5	concert	concert	NOUN
cana-6247	406	6	,	,	PUNCT
cana-6247	406	7	the	the	DET
cana-6247	406	8	field	field	NOUN
cana-6247	406	9	continues	continue	VERB
cana-6247	406	10	to	to	PART
cana-6247	406	11	advance	advance	VERB
cana-6247	406	12	in	in	ADP
cana-6247	406	13	all	all	DET
cana-6247	406	14	classes	class	NOUN
cana-6247	406	15	of	of	ADP
cana-6247	406	16	noise	noise	NOUN
cana-6247	406	17	methods	method	NOUN
cana-6247	406	18	whereby	whereby	SCONJ
cana-6247	406	19	scheduling	scheduling	NOUN
cana-6247	406	20	of	of	ADP
cana-6247	406	21	noise	noise	NOUN
cana-6247	406	22	function	function	NOUN
cana-6247	406	23	and	and	CCONJ
cana-6247	406	24	conditioning	conditioning	NOUN
cana-6247	406	25	provides	provide	VERB
cana-6247	406	26	improvements	improvement	NOUN
cana-6247	406	27	in	in	ADP
cana-6247	406	28	quality	quality	NOUN
cana-6247	406	29	,	,	PUNCT
cana-6247	406	30	speed	speed	NOUN
cana-6247	406	31	,	,	PUNCT
cana-6247	406	32	and	and	CCONJ
cana-6247	406	33	generalization	generalization	NOUN
cana-6247	406	34	outlines	outline	VERB
cana-6247	406	35	the	the	DET
cana-6247	406	36	progress	progress	NOUN
cana-6247	406	37	made	make	VERB
cana-6247	406	38	and	and	CCONJ
cana-6247	406	39	the	the	DET
cana-6247	406	40	importance	importance	NOUN
cana-6247	406	41	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	406	42	995	995	NUM
cana-6247	406	43	communications	communication	NOUN
cana-6247	406	44	on	on	ADP
cana-6247	406	45	applied	apply	VERB
cana-6247	406	46	nonlinear	nonlinear	ADJ
cana-6247	406	47	analysis	analysis	NOUN
cana-6247	406	48	issn	issn	NOUN
cana-6247	406	49	:	:	PUNCT
cana-6247	406	50	1074	1074	NUM
cana-6247	406	51	-	-	PUNCT
cana-6247	406	52	133x	133x	NUM
cana-6247	406	53	vol	vol	NOUN
cana-6247	406	54	.	.	PROPN
cana-6247	407	1	31	31	NUM
cana-6247	408	1	no	no	NOUN
cana-6247	408	2	.	.	PUNCT
cana-6247	409	1	7s	7	NOUN
cana-6247	409	2	(	(	PUNCT
cana-6247	409	3	2024	2024	NUM
cana-6247	409	4	)	)	PUNCT
cana-6247	409	5	figure	figure	NOUN
cana-6247	409	6	6	6	NUM
cana-6247	409	7	:	:	PUNCT
cana-6247	409	8	comparison	comparison	NOUN
cana-6247	409	9	table	table	NOUN
cana-6247	409	10	of	of	ADP
cana-6247	409	11	continuing	continue	VERB
cana-6247	409	12	to	to	PART
cana-6247	409	13	develop	develop	VERB
cana-6247	409	14	task	task	NOUN
cana-6247	409	15	-	-	PUNCT
cana-6247	409	16	relevant	relevant	ADJ
cana-6247	409	17	noise	noise	NOUN
cana-6247	409	18	methods	method	NOUN
cana-6247	409	19	to	to	PART
cana-6247	409	20	improve	improve	VERB
cana-6247	409	21	practical	practical	ADJ
cana-6247	409	22	and	and	CCONJ
cana-6247	409	23	robust	robust	ADJ
cana-6247	409	24	image	image	NOUN
cana-6247	409	25	enhancement	enhancement	NOUN
cana-6247	409	26	outcomes	outcome	NOUN
cana-6247	410	1	[	[	X
cana-6247	410	2	23	23	NUM
cana-6247	410	3	]	]	PUNCT
cana-6247	410	4	references	reference	NOUN
cana-6247	410	5	[	[	X
cana-6247	410	6	1	1	NUM
cana-6247	410	7	]	]	PUNCT
cana-6247	410	8	a	a	DET
cana-6247	410	9	abirami	abirami	NOUN
cana-6247	410	10	,	,	PUNCT
cana-6247	410	11	p	p	PROPN
cana-6247	410	12	prakash	prakash	PROPN
cana-6247	410	13	,	,	PUNCT
cana-6247	410	14	and	and	CCONJ
cana-6247	410	15	y	y	PROPN
cana-6247	410	16	-	-	PROPN
cana-6247	410	17	k	k	PROPN
cana-6247	410	18	ma	ma	PROPN
cana-6247	410	19	.	.	PROPN
cana-6247	410	20	variable	variable	ADJ
cana-6247	410	21	-	-	PUNCT
cana-6247	410	22	order	order	NOUN
cana-6247	410	23	fractional	fractional	ADJ
cana-6247	410	24	diffusion	diffusion	NOUN
cana-6247	410	25	model	model	NOUN
cana-6247	410	26	-	-	PUNCT
cana-6247	410	27	based	base	VERB
cana-6247	410	28	medical	medical	ADJ
cana-6247	410	29	image	image	NOUN
cana-6247	410	30	denoising	denoising	PROPN
cana-6247	410	31	.	.	PUNCT
cana-6247	411	1	math	math	NOUN
cana-6247	411	2	.	.	PUNCT
cana-6247	412	1	probl	probl	PROPN
cana-6247	412	2	.	.	PUNCT
cana-6247	413	1	eng	eng	PROPN
cana-6247	413	2	.	.	PROPN
cana-6247	413	3	,	,	PUNCT
cana-6247	413	4	2021:1–10	2021:1–10	NOUN
cana-6247	413	5	,	,	PUNCT
cana-6247	413	6	december	december	PROPN
cana-6247	413	7	2021	2021	NUM
cana-6247	413	8	.	.	PUNCT
cana-6247	414	1	[	[	X
cana-6247	414	2	2	2	NUM
cana-6247	414	3	]	]	PUNCT
cana-6247	414	4	noureddine	noureddine	ADP
cana-6247	414	5	alaa	alaa	PROPN
cana-6247	414	6	.	.	PUNCT
cana-6247	415	1	generic	generic	ADJ
cana-6247	415	2	reaction	reaction	NOUN
cana-6247	415	3	-	-	PUNCT
cana-6247	415	4	diffusion	diffusion	NOUN
cana-6247	415	5	model	model	NOUN
cana-6247	415	6	with	with	ADP
cana-6247	415	7	application	application	NOUN
cana-6247	415	8	to	to	ADP
cana-6247	415	9	image	image	NOUN
cana-6247	415	10	restoration	restoration	NOUN
cana-6247	415	11	and	and	CCONJ
cana-6247	415	12	enhancement	enhancement	NOUN
cana-6247	415	13	.	.	PUNCT
cana-6247	416	1	unknown	unknown	ADJ
cana-6247	416	2	,	,	PUNCT
cana-6247	416	3	2018	2018	NUM
cana-6247	416	4	.	.	PUNCT
cana-6247	417	1	[	[	X
cana-6247	417	2	3	3	NUM
cana-6247	417	3	]	]	PUNCT
cana-6247	417	4	author1firstname	author1firstname	NOUN
cana-6247	417	5	author1lastname	author1lastname	NOUN
cana-6247	417	6	and	and	CCONJ
cana-6247	417	7	author2firstname	author2firstname	NUM
cana-6247	417	8	author2lastname	author2lastname	PROPN
cana-6247	417	9	.	.	PUNCT
cana-6247	418	1	finite	finite	ADJ
cana-6247	418	2	difference	difference	NOUN
cana-6247	418	3	and	and	CCONJ
cana-6247	418	4	legendre	legendre	PROPN
cana-6247	418	5	spectral	spectral	PROPN
cana-6247	418	6	method	method	NOUN
cana-6247	418	7	for	for	ADP
cana-6247	418	8	a	a	DET
cana-6247	418	9	time	time	NOUN
cana-6247	418	10	-	-	PUNCT
cana-6247	418	11	fractional	fractional	ADJ
cana-6247	418	12	diffusion	diffusion	NOUN
cana-6247	418	13	-	-	PUNCT
cana-6247	418	14	convection	convection	NOUN
cana-6247	418	15	equation	equation	NOUN
cana-6247	418	16	for	for	ADP
cana-6247	418	17	image	image	NOUN
cana-6247	418	18	restoration	restoration	NOUN
cana-6247	418	19	.	.	PUNCT
cana-6247	419	1	journal	journal	NOUN
cana-6247	419	2	name	name	NOUN
cana-6247	419	3	,	,	PUNCT
cana-6247	419	4	volume	volume	NOUN
cana-6247	419	5	number(issue	number(issue	PROPN
cana-6247	419	6	number):page	number):page	NOUN
cana-6247	419	7	range	range	NOUN
cana-6247	419	8	,	,	PUNCT
cana-6247	419	9	2025	2025	NUM
cana-6247	419	10	.	.	PUNCT
cana-6247	420	1	[	[	X
cana-6247	420	2	4	4	NUM
cana-6247	420	3	]	]	X
cana-6247	420	4	arpit	arpit	PROPN
cana-6247	420	5	bansal	bansal	PROPN
cana-6247	420	6	,	,	PUNCT
cana-6247	420	7	eitan	eitan	PROPN
cana-6247	420	8	borgnia	borgnia	PROPN
cana-6247	420	9	,	,	PUNCT
cana-6247	420	10	hong	hong	PROPN
cana-6247	420	11	-	-	PUNCT
cana-6247	420	12	min	min	PROPN
cana-6247	420	13	chu	chu	PROPN
cana-6247	420	14	,	,	PUNCT
cana-6247	420	15	jie	jie	PROPN
cana-6247	420	16	s	s	PROPN
cana-6247	420	17	li	li	PROPN
cana-6247	420	18	,	,	PUNCT
cana-6247	420	19	hamid	hamid	PROPN
cana-6247	420	20	kazemi	kazemi	PROPN
cana-6247	420	21	,	,	PUNCT
cana-6247	420	22	furong	furong	PROPN
cana-6247	420	23	huang	huang	PROPN
cana-6247	420	24	,	,	PUNCT
cana-6247	420	25	micah	micah	PROPN
cana-6247	420	26	goldblum	goldblum	PROPN
cana-6247	420	27	,	,	PUNCT
cana-6247	420	28	jonas	jonas	PROPN
cana-6247	420	29	geiping	geiping	NOUN
cana-6247	420	30	,	,	PUNCT
cana-6247	420	31	and	and	CCONJ
cana-6247	420	32	tom	tom	PROPN
cana-6247	420	33	goldstein	goldstein	PROPN
cana-6247	420	34	.	.	PUNCT
cana-6247	420	35	cold	cold	PROPN
cana-6247	420	36	diffusion	diffusion	NOUN
cana-6247	420	37	:	:	PUNCT
cana-6247	420	38	inverting	invert	VERB
cana-6247	420	39	arbitrary	arbitrary	ADJ
cana-6247	420	40	image	image	NOUN
cana-6247	420	41	transforms	transform	VERB
cana-6247	420	42	without	without	ADP
cana-6247	420	43	noise	noise	NOUN
cana-6247	420	44	.	.	PUNCT
cana-6247	421	1	august	august	PROPN
cana-6247	421	2	2022	2022	NUM
cana-6247	421	3	.	.	PUNCT
cana-6247	422	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	422	2	996	996	NUM
cana-6247	422	3	communications	communication	NOUN
cana-6247	422	4	on	on	ADP
cana-6247	422	5	applied	apply	VERB
cana-6247	422	6	nonlinear	nonlinear	ADJ
cana-6247	422	7	analysis	analysis	NOUN
cana-6247	422	8	issn	issn	NOUN
cana-6247	422	9	:	:	PUNCT
cana-6247	422	10	1074	1074	NUM
cana-6247	422	11	-	-	PUNCT
cana-6247	422	12	133x	133x	NUM
cana-6247	422	13	vol	vol	NOUN
cana-6247	422	14	.	.	PROPN
cana-6247	423	1	31	31	NUM
cana-6247	424	1	no	no	NOUN
cana-6247	424	2	.	.	PUNCT
cana-6247	425	1	7s	7	NOUN
cana-6247	425	2	(	(	PUNCT
cana-6247	425	3	2024	2024	NUM
cana-6247	425	4	)	)	PUNCT
cana-6247	426	1	[	[	X
cana-6247	426	2	5	5	X
cana-6247	426	3	]	]	PUNCT
cana-6247	426	4	zheng	zheng	PROPN
cana-6247	426	5	chen	chen	PROPN
cana-6247	426	6	,	,	PUNCT
cana-6247	426	7	yulun	yulun	PROPN
cana-6247	426	8	zhang	zhang	PROPN
cana-6247	426	9	,	,	PUNCT
cana-6247	426	10	ding	ding	PROPN
cana-6247	426	11	liu	liu	PROPN
cana-6247	426	12	,	,	PUNCT
cana-6247	426	13	bin	bin	PROPN
cana-6247	426	14	xia	xia	PROPN
cana-6247	426	15	,	,	PUNCT
cana-6247	426	16	jinjin	jinjin	PROPN
cana-6247	426	17	gu	gu	PROPN
cana-6247	426	18	,	,	PUNCT
cana-6247	426	19	linghe	linghe	PROPN
cana-6247	426	20	kong	kong	PROPN
cana-6247	426	21	,	,	PUNCT
cana-6247	426	22	and	and	CCONJ
cana-6247	426	23	xin	xin	PROPN
cana-6247	426	24	yuan	yuan	PROPN
cana-6247	426	25	.	.	PUNCT
cana-6247	427	1	hierarchical	hierarchical	ADJ
cana-6247	427	2	integration	integration	NOUN
cana-6247	427	3	diffusion	diffusion	NOUN
cana-6247	427	4	model	model	NOUN
cana-6247	427	5	for	for	ADP
cana-6247	427	6	realistic	realistic	ADJ
cana-6247	427	7	image	image	NOUN
cana-6247	427	8	deblurring	deblurring	NOUN
cana-6247	427	9	.	.	PUNCT
cana-6247	428	1	2023	2023	NUM
cana-6247	428	2	.	.	PUNCT
cana-6247	429	1	[	[	X
cana-6247	429	2	6	6	NUM
cana-6247	429	3	]	]	PUNCT
cana-6247	429	4	ciprian	ciprian	ADJ
cana-6247	429	5	corneanu	corneanu	NOUN
cana-6247	429	6	,	,	PUNCT
cana-6247	429	7	raghudeep	raghudeep	NOUN
cana-6247	429	8	gadde	gadde	NOUN
cana-6247	429	9	,	,	PUNCT
cana-6247	429	10	and	and	CCONJ
cana-6247	429	11	aleix	aleix	VERB
cana-6247	429	12	m	m	PROPN
cana-6247	429	13	martinez	martinez	PROPN
cana-6247	429	14	.	.	PUNCT
cana-6247	430	1	latentpaint	latentpaint	NOUN
cana-6247	430	2	:	:	PUNCT
cana-6247	430	3	image	image	NOUN
cana-6247	430	4	inpainting	inpainte	VERB
cana-6247	430	5	in	in	ADP
cana-6247	430	6	latent	latent	ADJ
cana-6247	430	7	space	space	NOUN
cana-6247	430	8	with	with	ADP
cana-6247	430	9	diffusion	diffusion	NOUN
cana-6247	430	10	models	model	NOUN
cana-6247	430	11	.	.	PUNCT
cana-6247	431	1	in	in	ADP
cana-6247	431	2	2024	2024	NUM
cana-6247	431	3	ieee	ieee	NOUN
cana-6247	431	4	/	/	SYM
cana-6247	431	5	cvf	cvf	NOUN
cana-6247	431	6	winter	winter	NOUN
cana-6247	431	7	conference	conference	NOUN
cana-6247	431	8	on	on	ADP
cana-6247	431	9	applications	application	NOUN
cana-6247	431	10	of	of	ADP
cana-6247	431	11	computer	computer	NOUN
cana-6247	431	12	vision	vision	NOUN
cana-6247	431	13	(	(	PUNCT
cana-6247	431	14	wacv	wacv	NOUN
cana-6247	431	15	)	)	PUNCT
cana-6247	431	16	,	,	PUNCT
cana-6247	431	17	pages	page	NOUN
cana-6247	431	18	4322–4331	4322–4331	NUM
cana-6247	431	19	.	.	PUNCT
cana-6247	432	1	ieee	ieee	PROPN
cana-6247	432	2	,	,	PUNCT
cana-6247	432	3	january	january	PROPN
cana-6247	432	4	2024	2024	NUM
cana-6247	432	5	.	.	PUNCT
cana-6247	433	1	[	[	X
cana-6247	433	2	7	7	X
cana-6247	433	3	]	]	PUNCT
cana-6247	433	4	m.	m.	PROPN
cana-6247	433	5	b.	b.	PROPN
cana-6247	433	6	gharsallah	gharsallah	PROPN
cana-6247	433	7	and	and	CCONJ
cana-6247	433	8	e.	e.	PROPN
cana-6247	433	9	b.	b.	PROPN
cana-6247	433	10	braiek	braiek	PROPN
cana-6247	433	11	.	.	PUNCT
cana-6247	434	1	new	new	ADJ
cana-6247	434	2	anisotropic	anisotropic	NOUN
cana-6247	434	3	diffusion	diffusion	NOUN
cana-6247	434	4	method	method	NOUN
cana-6247	434	5	to	to	PART
cana-6247	434	6	improve	improve	VERB
cana-6247	434	7	radiographic	radiographic	ADJ
cana-6247	434	8	image	image	NOUN
cana-6247	434	9	quality	quality	NOUN
cana-6247	434	10	.	.	PUNCT
cana-6247	435	1	kuwait	kuwait	PROPN
cana-6247	435	2	journal	journal	PROPN
cana-6247	435	3	of	of	ADP
cana-6247	435	4	science	science	NOUN
cana-6247	435	5	,	,	PUNCT
cana-6247	435	6	2017	2017	NUM
cana-6247	435	7	.	.	PUNCT
cana-6247	436	1	[	[	X
cana-6247	436	2	8	8	NUM
cana-6247	436	3	]	]	PUNCT
cana-6247	436	4	vidit	vidit	PROPN
cana-6247	436	5	goel	goel	PROPN
cana-6247	436	6	,	,	PUNCT
cana-6247	436	7	elia	elia	PROPN
cana-6247	436	8	peruzzo	peruzzo	PROPN
cana-6247	436	9	,	,	PUNCT
cana-6247	436	10	yifan	yifan	PROPN
cana-6247	436	11	jiang	jiang	PROPN
cana-6247	436	12	,	,	PUNCT
cana-6247	436	13	dejia	dejia	PROPN
cana-6247	436	14	xu	xu	PROPN
cana-6247	436	15	,	,	PUNCT
cana-6247	436	16	nicu	nicu	PROPN
cana-6247	436	17	sebe	sebe	PROPN
cana-6247	436	18	,	,	PUNCT
cana-6247	436	19	trevor	trevor	PROPN
cana-6247	436	20	darrell	darrell	PROPN
cana-6247	436	21	,	,	PUNCT
cana-6247	436	22	zhangyang	zhangyang	PROPN
cana-6247	436	23	wang	wang	PROPN
cana-6247	436	24	,	,	PUNCT
cana-6247	436	25	and	and	CCONJ
cana-6247	436	26	humphrey	humphrey	PROPN
cana-6247	436	27	shi	shi	PROPN
cana-6247	436	28	.	.	PUNCT
cana-6247	437	1	pair	pair	NOUN
cana-6247	437	2	-	-	PUNCT
cana-6247	437	3	diffusion	diffusion	NOUN
cana-6247	437	4	:	:	PUNCT
cana-6247	437	5	object	object	NOUN
cana-6247	437	6	-	-	PUNCT
cana-6247	437	7	level	level	NOUN
cana-6247	437	8	image	image	NOUN
cana-6247	437	9	editing	edit	VERB
cana-6247	437	10	with	with	ADP
cana-6247	437	11	structure	structure	NOUN
cana-6247	437	12	-	-	PUNCT
cana-6247	437	13	and	and	CCONJ
cana-6247	437	14	-	-	PUNCT
cana-6247	437	15	appearance	appearance	NOUN
cana-6247	437	16	paired	pair	VERB
cana-6247	437	17	diffusion	diffusion	NOUN
cana-6247	437	18	models	model	NOUN
cana-6247	437	19	.	.	PUNCT
cana-6247	438	1	arxiv	arxiv	PROPN
cana-6247	438	2	preprint	preprint	NOUN
cana-6247	438	3	arxiv:2303.17546	arxiv:2303.17546	VERB
cana-6247	438	4	,	,	PUNCT
cana-6247	438	5	2023	2023	NUM
cana-6247	438	6	.	.	PUNCT
cana-6247	439	1	[	[	X
cana-6247	439	2	9	9	NUM
cana-6247	439	3	]	]	PUNCT
cana-6247	439	4	asya	asya	PROPN
cana-6247	439	5	grechka	grechka	PROPN
cana-6247	439	6	,	,	PUNCT
cana-6247	439	7	guillaume	guillaume	PROPN
cana-6247	439	8	couairon	couairon	PROPN
cana-6247	439	9	,	,	PUNCT
cana-6247	439	10	and	and	CCONJ
cana-6247	439	11	matthieu	matthieu	PROPN
cana-6247	439	12	cord	cord	NOUN
cana-6247	439	13	.	.	PUNCT
cana-6247	440	1	gradpaint	gradpaint	NOUN
cana-6247	440	2	:	:	PUNCT
cana-6247	440	3	gradient	gradient	NOUN
cana-6247	440	4	-	-	PUNCT
cana-6247	440	5	guided	guide	VERB
cana-6247	440	6	inpainting	inpainte	VERB
cana-6247	440	7	with	with	ADP
cana-6247	440	8	diffusion	diffusion	NOUN
cana-6247	440	9	models	model	NOUN
cana-6247	440	10	.	.	PUNCT
cana-6247	441	1	2023	2023	NUM
cana-6247	441	2	.	.	PUNCT
cana-6247	442	1	[	[	X
cana-6247	442	2	10	10	NUM
cana-6247	442	3	]	]	X
cana-6247	442	4	dennis	dennis	PROPN
cana-6247	442	5	hein	hein	PROPN
cana-6247	442	6	,	,	PUNCT
cana-6247	442	7	staffan	staffan	PROPN
cana-6247	442	8	holmin	holmin	NOUN
cana-6247	442	9	,	,	PUNCT
cana-6247	442	10	timothy	timothy	PROPN
cana-6247	442	11	p.	p.	PROPN
cana-6247	442	12	szczykutowicz	szczykutowicz	PROPN
cana-6247	442	13	,	,	PUNCT
cana-6247	442	14	mats	mat	VERB
cana-6247	442	15	danielsson	danielsson	PROPN
cana-6247	442	16	,	,	PUNCT
cana-6247	442	17	and	and	CCONJ
cana-6247	442	18	mats	mats	PROPN
cana-6247	442	19	persson	persson	PROPN
cana-6247	442	20	.	.	PUNCT
cana-6247	442	21	noise	noise	NOUN
cana-6247	442	22	suppression	suppression	NOUN
cana-6247	442	23	in	in	ADP
cana-6247	442	24	photon	photon	NOUN
cana-6247	442	25	-	-	PUNCT
cana-6247	442	26	counting	counting	NOUN
cana-6247	442	27	computed	compute	VERB
cana-6247	442	28	tomography	tomography	NOUN
cana-6247	442	29	using	use	VERB
cana-6247	442	30	unsupervised	unsupervised	ADJ
cana-6247	442	31	poisson	poisson	NOUN
cana-6247	442	32	flow	flow	NOUN
cana-6247	442	33	generative	generative	ADJ
cana-6247	442	34	models	model	NOUN
cana-6247	442	35	.	.	PUNCT
cana-6247	443	1	visual	visual	ADJ
cana-6247	443	2	computing	computing	NOUN
cana-6247	443	3	for	for	ADP
cana-6247	443	4	industry	industry	NOUN
cana-6247	443	5	,	,	PUNCT
cana-6247	443	6	biomedicine	biomedicine	NOUN
cana-6247	443	7	and	and	CCONJ
cana-6247	443	8	art	art	NOUN
cana-6247	443	9	,	,	PUNCT
cana-6247	443	10	7(1):24	7(1):24	NUM
cana-6247	443	11	,	,	PUNCT
cana-6247	443	12	2024	2024	NUM
cana-6247	443	13	.	.	PUNCT
cana-6247	444	1	[	[	X
cana-6247	444	2	11	11	NUM
cana-6247	444	3	]	]	X
cana-6247	444	4	jonathan	jonathan	PROPN
cana-6247	444	5	ho	ho	PROPN
cana-6247	444	6	,	,	PUNCT
cana-6247	444	7	ajay	ajay	PROPN
cana-6247	444	8	jain	jain	PROPN
cana-6247	444	9	,	,	PUNCT
cana-6247	444	10	and	and	CCONJ
cana-6247	444	11	pieter	pieter	NOUN
cana-6247	444	12	abbeel	abbeel	NOUN
cana-6247	444	13	.	.	PUNCT
cana-6247	445	1	denoising	denoise	VERB
cana-6247	445	2	diffusion	diffusion	NOUN
cana-6247	445	3	probabilistic	probabilistic	ADJ
cana-6247	445	4	models	model	NOUN
cana-6247	445	5	.	.	PUNCT
cana-6247	446	1	2020	2020	NUM
cana-6247	446	2	.	.	PUNCT
cana-6247	447	1	[	[	X
cana-6247	447	2	12	12	NUM
cana-6247	447	3	]	]	PUNCT
cana-6247	447	4	jinhui	jinhui	PROPN
cana-6247	447	5	hou	hou	PROPN
cana-6247	447	6	,	,	PUNCT
cana-6247	447	7	zhiyu	zhiyu	PROPN
cana-6247	447	8	zhu	zhu	PROPN
cana-6247	447	9	,	,	PUNCT
cana-6247	447	10	junhui	junhui	PROPN
cana-6247	447	11	hou	hou	PROPN
cana-6247	447	12	,	,	PUNCT
cana-6247	447	13	hui	hui	PROPN
cana-6247	447	14	liu	liu	PROPN
cana-6247	447	15	,	,	PUNCT
cana-6247	447	16	huanqiang	huanqiang	PROPN
cana-6247	447	17	zeng	zeng	PROPN
cana-6247	447	18	,	,	PUNCT
cana-6247	447	19	and	and	CCONJ
cana-6247	447	20	hui	hui	PROPN
cana-6247	447	21	yuan	yuan	PROPN
cana-6247	447	22	.	.	PUNCT
cana-6247	448	1	global	global	ADJ
cana-6247	448	2	structure	structure	NOUN
cana-6247	448	3	-	-	PUNCT
cana-6247	448	4	aware	aware	ADJ
cana-6247	448	5	diffusion	diffusion	NOUN
cana-6247	448	6	process	process	NOUN
cana-6247	448	7	for	for	ADP
cana-6247	448	8	low	low	ADJ
cana-6247	448	9	-	-	PUNCT
cana-6247	448	10	light	light	ADJ
cana-6247	448	11	image	image	NOUN
cana-6247	448	12	enhancement	enhancement	NOUN
cana-6247	448	13	.	.	PUNCT
cana-6247	449	1	october	october	PROPN
cana-6247	449	2	2023	2023	NUM
cana-6247	449	3	.	.	PUNCT
cana-6247	450	1	[	[	X
cana-6247	450	2	13	13	NUM
cana-6247	450	3	]	]	PUNCT
cana-6247	450	4	jinho	jinho	PROPN
cana-6247	450	5	jeong	jeong	PROPN
cana-6247	450	6	,	,	PUNCT
cana-6247	450	7	sangmin	sangmin	PROPN
cana-6247	450	8	han	han	PROPN
cana-6247	450	9	,	,	PUNCT
cana-6247	450	10	jinwoo	jinwoo	PROPN
cana-6247	450	11	kim	kim	PROPN
cana-6247	450	12	,	,	PUNCT
cana-6247	450	13	and	and	CCONJ
cana-6247	450	14	seon	seon	PROPN
cana-6247	450	15	joo	joo	PROPN
cana-6247	450	16	kim	kim	PROPN
cana-6247	450	17	.	.	PUNCT
cana-6247	451	1	latent	latent	PROPN
cana-6247	451	2	space	space	NOUN
cana-6247	451	3	super	super	NOUN
cana-6247	451	4	-	-	NOUN
cana-6247	451	5	resolution	resolution	NOUN
cana-6247	451	6	for	for	ADP
cana-6247	451	7	higher	high	ADJ
cana-6247	451	8	resolution	resolution	NOUN
cana-6247	451	9	image	image	NOUN
cana-6247	451	10	generation	generation	NOUN
cana-6247	451	11	with	with	ADP
cana-6247	451	12	diffusion	diffusion	NOUN
cana-6247	451	13	models	model	NOUN
cana-6247	451	14	.	.	PUNCT
cana-6247	452	1	march	march	PROPN
cana-6247	452	2	2025	2025	NUM
cana-6247	452	3	.	.	PUNCT
cana-6247	453	1	[	[	X
cana-6247	453	2	14	14	NUM
cana-6247	453	3	]	]	X
cana-6247	453	4	hai	hai	PROPN
cana-6247	453	5	jiang	jiang	PROPN
cana-6247	453	6	,	,	PUNCT
cana-6247	453	7	ao	ao	PROPN
cana-6247	453	8	luo	luo	PROPN
cana-6247	453	9	,	,	PUNCT
cana-6247	453	10	xiaohong	xiaohong	PROPN
cana-6247	453	11	liu	liu	PROPN
cana-6247	453	12	,	,	PUNCT
cana-6247	453	13	songchen	songchen	PROPN
cana-6247	453	14	han	han	PROPN
cana-6247	453	15	,	,	PUNCT
cana-6247	453	16	and	and	CCONJ
cana-6247	453	17	shuaicheng	shuaicheng	PROPN
cana-6247	453	18	liu	liu	PROPN
cana-6247	454	1	.	.	PUNCT
cana-6247	454	2	lightendiffusion	lightendiffusion	PROPN
cana-6247	454	3	:	:	PUNCT
cana-6247	454	4	unsupervised	unsupervised	ADJ
cana-6247	454	5	low	low	ADJ
cana-6247	454	6	-	-	PUNCT
cana-6247	454	7	light	light	NOUN
cana-6247	454	8	image	image	NOUN
cana-6247	454	9	enhancement	enhancement	NOUN
cana-6247	454	10	with	with	ADP
cana-6247	454	11	latent	latent	ADJ
cana-6247	454	12	-	-	PUNCT
cana-6247	454	13	retinex	retinex	ADJ
cana-6247	454	14	diffusion	diffusion	NOUN
cana-6247	454	15	models	model	NOUN
cana-6247	454	16	.	.	PUNCT
cana-6247	455	1	july	july	PROPN
cana-6247	455	2	2024	2024	NUM
cana-6247	455	3	.	.	PUNCT
cana-6247	456	1	[	[	X
cana-6247	456	2	15	15	NUM
cana-6247	456	3	]	]	X
cana-6247	456	4	bahjat	bahjat	PROPN
cana-6247	456	5	kawar	kawar	PROPN
cana-6247	456	6	,	,	PUNCT
cana-6247	456	7	jiaming	jiame	VERB
cana-6247	456	8	song	song	NOUN
cana-6247	456	9	,	,	PUNCT
cana-6247	456	10	stefano	stefano	PROPN
cana-6247	456	11	ermon	ermon	NOUN
cana-6247	456	12	,	,	PUNCT
cana-6247	456	13	and	and	CCONJ
cana-6247	456	14	michael	michael	PROPN
cana-6247	456	15	elad	elad	PROPN
cana-6247	456	16	.	.	PUNCT
cana-6247	457	1	jpeg	jpeg	PROPN
cana-6247	457	2	artifact	artifact	NOUN
cana-6247	457	3	correction	correction	NOUN
cana-6247	457	4	using	use	VERB
cana-6247	457	5	denoising	denoising	NOUN
cana-6247	457	6	diffusion	diffusion	NOUN
cana-6247	457	7	restoration	restoration	NOUN
cana-6247	457	8	models	model	NOUN
cana-6247	457	9	.	.	PUNCT
cana-6247	458	1	2022	2022	NUM
cana-6247	458	2	.	.	PUNCT
cana-6247	459	1	[	[	X
cana-6247	459	2	16	16	NUM
cana-6247	459	3	]	]	PUNCT
cana-6247	459	4	lingshun	lingshun	PROPN
cana-6247	459	5	kong	kong	PROPN
cana-6247	459	6	,	,	PUNCT
cana-6247	459	7	jiawei	jiawei	PROPN
cana-6247	459	8	zhang	zhang	PROPN
cana-6247	459	9	,	,	PUNCT
cana-6247	459	10	dongqing	dongqe	VERB
cana-6247	459	11	zou	zou	PRON
cana-6247	459	12	,	,	PUNCT
cana-6247	459	13	jimmy	jimmy	PROPN
cana-6247	459	14	ren	ren	PROPN
cana-6247	459	15	,	,	PUNCT
cana-6247	459	16	xiaohe	xiaohe	PROPN
cana-6247	459	17	wu	wu	PROPN
cana-6247	459	18	,	,	PUNCT
cana-6247	459	19	jiangxin	jiangxin	PROPN
cana-6247	459	20	dong	dong	PROPN
cana-6247	459	21	,	,	PUNCT
cana-6247	459	22	and	and	CCONJ
cana-6247	459	23	jinshan	jinshan	PROPN
cana-6247	459	24	pan	pan	PROPN
cana-6247	459	25	.	.	PROPN
cana-6247	460	1	deblurdiff	deblurdiff	PROPN
cana-6247	460	2	:	:	PUNCT
cana-6247	460	3	real	real	ADJ
cana-6247	460	4	-	-	PUNCT
cana-6247	460	5	world	world	NOUN
cana-6247	460	6	image	image	NOUN
cana-6247	460	7	deblurring	deblurre	VERB
cana-6247	460	8	with	with	ADP
cana-6247	460	9	generative	generative	ADJ
cana-6247	460	10	diffusion	diffusion	NOUN
cana-6247	460	11	models	model	NOUN
cana-6247	460	12	.	.	PUNCT
cana-6247	461	1	arxiv	arxiv	PROPN
cana-6247	461	2	preprint	preprint	PROPN
cana-6247	461	3	arxiv:2502.03810	arxiv:2502.03810	NOUN
cana-6247	461	4	,	,	PUNCT
cana-6247	461	5	2025	2025	NUM
cana-6247	461	6	.	.	PUNCT
cana-6247	462	1	[	[	X
cana-6247	462	2	17	17	NUM
cana-6247	462	3	]	]	PUNCT
cana-6247	462	4	dawid	dawid	PROPN
cana-6247	462	5	kopeć	kopeć	PROPN
cana-6247	462	6	,	,	PUNCT
cana-6247	462	7	wojciech	wojciech	NOUN
cana-6247	462	8	kozłowski	kozłowski	NOUN
cana-6247	462	9	,	,	PUNCT
cana-6247	462	10	maciej	maciej	PROPN
cana-6247	462	11	wizerkaniuk	wizerkaniuk	NOUN
cana-6247	462	12	,	,	PUNCT
cana-6247	462	13	dawid	dawid	PROPN
cana-6247	462	14	krutul	krutul	PROPN
cana-6247	462	15	,	,	PUNCT
cana-6247	462	16	jan	jan	PROPN
cana-6247	462	17	kocoń	kocoń	PROPN
cana-6247	462	18	,	,	PUNCT
cana-6247	462	19	and	and	CCONJ
cana-6247	462	20	maciej	maciej	PROPN
cana-6247	462	21	zięba	zięba	PROPN
cana-6247	462	22	.	.	PUNCT
cana-6247	463	1	supresdiffgan	supresdiffgan	VERB
cana-6247	463	2	a	a	DET
cana-6247	463	3	new	new	ADJ
cana-6247	463	4	approach	approach	NOUN
cana-6247	463	5	for	for	ADP
cana-6247	463	6	the	the	DET
cana-6247	463	7	super	super	ADJ
cana-6247	463	8	-	-	ADJ
cana-6247	463	9	resolution	resolution	ADJ
cana-6247	463	10	task	task	NOUN
cana-6247	463	11	.	.	PUNCT
cana-6247	464	1	in	in	ADP
cana-6247	464	2	lecture	lecture	NOUN
cana-6247	464	3	notes	note	NOUN
cana-6247	464	4	in	in	ADP
cana-6247	464	5	computer	computer	NOUN
cana-6247	464	6	science	science	NOUN
cana-6247	464	7	,	,	PUNCT
cana-6247	464	8	lecture	lecture	NOUN
cana-6247	464	9	notes	note	NOUN
cana-6247	464	10	in	in	ADP
cana-6247	464	11	computer	computer	NOUN
cana-6247	464	12	science	science	NOUN
cana-6247	464	13	,	,	PUNCT
cana-6247	464	14	pages	page	NOUN
cana-6247	464	15	66–80	66–80	PROPN
cana-6247	464	16	.	.	PUNCT
cana-6247	465	1	springer	springer	PROPN
cana-6247	465	2	nature	nature	PROPN
cana-6247	465	3	switzerland	switzerland	PROPN
cana-6247	465	4	,	,	PUNCT
cana-6247	465	5	cham	cham	PROPN
cana-6247	465	6	,	,	PUNCT
cana-6247	465	7	2025	2025	NUM
cana-6247	465	8	.	.	PUNCT
cana-6247	466	1	[	[	X
cana-6247	466	2	18	18	NUM
cana-6247	466	3	]	]	PUNCT
cana-6247	466	4	yadong	yadong	PROPN
cana-6247	466	5	li	li	PROPN
cana-6247	466	6	,	,	PUNCT
cana-6247	466	7	jun	jun	PROPN
cana-6247	466	8	tian	tian	PROPN
cana-6247	466	9	,	,	PUNCT
cana-6247	466	10	yang	yang	PROPN
cana-6247	466	11	chen	chen	PROPN
cana-6247	466	12	,	,	PUNCT
cana-6247	466	13	hongdong	hongdong	PROPN
cana-6247	466	14	wang	wang	PROPN
cana-6247	466	15	,	,	PUNCT
cana-6247	466	16	hui	hui	PROPN
cana-6247	466	17	yan	yan	PROPN
cana-6247	466	18	,	,	PUNCT
cana-6247	466	19	yang	yang	PROPN
cana-6247	466	20	peng	peng	PROPN
cana-6247	466	21	,	,	PUNCT
cana-6247	466	22	and	and	CCONJ
cana-6247	466	23	tianjiao	tianjiao	VERB
cana-6247	466	24	wang	wang	PROPN
cana-6247	466	25	.	.	PUNCT
cana-6247	467	1	rw	rw	PROPN
cana-6247	467	2	-	-	PUNCT
cana-6247	467	3	dm	dm	NOUN
cana-6247	467	4	:	:	PUNCT
cana-6247	467	5	retinex	retinex	ADJ
cana-6247	467	6	and	and	CCONJ
cana-6247	467	7	wavelet	wavelet	NOUN
cana-6247	467	8	-	-	PUNCT
cana-6247	467	9	based	base	VERB
cana-6247	467	10	diffusion	diffusion	NOUN
cana-6247	467	11	model	model	NOUN
cana-6247	467	12	for	for	ADP
cana-6247	467	13	low	low	ADJ
cana-6247	467	14	-	-	PUNCT
cana-6247	467	15	light	light	NOUN
cana-6247	467	16	image	image	NOUN
cana-6247	467	17	enhancement	enhancement	NOUN
cana-6247	467	18	in	in	ADP
cana-6247	467	19	underground	underground	ADJ
cana-6247	467	20	coal	coal	NOUN
cana-6247	467	21	mines	mine	NOUN
cana-6247	467	22	.	.	PUNCT
cana-6247	468	1	complex	complex	ADJ
cana-6247	468	2	intell	intell	PROPN
cana-6247	468	3	.	.	PUNCT
cana-6247	469	1	syst	syst	PROPN
cana-6247	469	2	.	.	PROPN
cana-6247	469	3	,	,	PUNCT
cana-6247	469	4	11(8	11(8	PROPN
cana-6247	469	5	)	)	PUNCT
cana-6247	469	6	,	,	PUNCT
cana-6247	469	7	august	august	PROPN
cana-6247	469	8	2025	2025	NUM
cana-6247	469	9	.	.	PUNCT
cana-6247	470	1	[	[	X
cana-6247	470	2	19	19	NUM
cana-6247	470	3	]	]	X
cana-6247	470	4	xinqi	xinqi	PROPN
cana-6247	470	5	lin	lin	PROPN
cana-6247	470	6	,	,	PUNCT
cana-6247	470	7	jingwen	jingwen	VERB
cana-6247	470	8	he	he	PRON
cana-6247	470	9	,	,	PUNCT
cana-6247	470	10	ziyan	ziyan	PROPN
cana-6247	470	11	chen	chen	PROPN
cana-6247	470	12	,	,	PUNCT
cana-6247	470	13	zhaoyang	zhaoyang	PROPN
cana-6247	470	14	lyu	lyu	NOUN
cana-6247	470	15	,	,	PUNCT
cana-6247	470	16	bo	bo	PROPN
cana-6247	470	17	dai	dai	PROPN
cana-6247	470	18	,	,	PUNCT
cana-6247	470	19	fanghua	fanghua	PROPN
cana-6247	470	20	yu	yu	PROPN
cana-6247	470	21	,	,	PUNCT
cana-6247	470	22	yu	yu	PROPN
cana-6247	470	23	qiao	qiao	PROPN
cana-6247	470	24	,	,	PUNCT
cana-6247	470	25	wanli	wanli	PROPN
cana-6247	470	26	ouyang	ouyang	PROPN
cana-6247	470	27	,	,	PUNCT
cana-6247	470	28	and	and	CCONJ
cana-6247	470	29	chao	chao	PROPN
cana-6247	470	30	dong	dong	PROPN
cana-6247	470	31	.	.	PUNCT
cana-6247	471	1	diffbir	diffbir	PROPN
cana-6247	471	2	:	:	PUNCT
cana-6247	471	3	toward	toward	ADP
cana-6247	471	4	blind	blind	ADJ
cana-6247	471	5	image	image	NOUN
cana-6247	471	6	restoration	restoration	NOUN
cana-6247	471	7	with	with	ADP
cana-6247	471	8	generative	generative	ADJ
cana-6247	471	9	diffusion	diffusion	NOUN
cana-6247	471	10	prior	prior	ADV
cana-6247	471	11	.	.	PUNCT
cana-6247	472	1	in	in	ADP
cana-6247	472	2	lecture	lecture	NOUN
cana-6247	472	3	notes	note	NOUN
cana-6247	472	4	in	in	ADP
cana-6247	472	5	computer	computer	NOUN
cana-6247	472	6	science	science	NOUN
cana-6247	472	7	,	,	PUNCT
cana-6247	472	8	lecture	lecture	NOUN
cana-6247	472	9	notes	note	NOUN
cana-6247	472	10	in	in	ADP
cana-6247	472	11	computer	computer	NOUN
cana-6247	472	12	science	science	NOUN
cana-6247	472	13	,	,	PUNCT
cana-6247	472	14	pages	page	NOUN
cana-6247	472	15	430–448	430–448	NUM
cana-6247	472	16	.	.	PUNCT
cana-6247	472	17	springer	springer	PROPN
cana-6247	472	18	nature	nature	PROPN
cana-6247	472	19	switzerland	switzerland	PROPN
cana-6247	472	20	,	,	PUNCT
cana-6247	472	21	cham	cham	PROPN
cana-6247	472	22	,	,	PUNCT
cana-6247	472	23	2025	2025	NUM
cana-6247	472	24	.	.	PUNCT
cana-6247	473	1	[	[	X
cana-6247	473	2	20	20	NUM
cana-6247	473	3	]	]	PUNCT
cana-6247	473	4	liying	liye	VERB
cana-6247	473	5	lu	lu	PROPN
cana-6247	473	6	,	,	PUNCT
cana-6247	473	7	raphael	raphael	PROPN
cana-6247	473	8	achddou	achddou	PROPN
cana-6247	473	9	,	,	PUNCT
cana-6247	473	10	and	and	CCONJ
cana-6247	473	11	sabine	sabine	PROPN
cana-6247	473	12	susstrunk	susstrunk	PROPN
cana-6247	473	13	.	.	PUNCT
cana-6247	474	1	dark	dark	PROPN
cana-6247	474	2	noise	noise	PROPN
cana-6247	474	3	diffusion	diffusion	NOUN
cana-6247	474	4	:	:	PUNCT
cana-6247	474	5	noise	noise	NOUN
cana-6247	474	6	synthesis	synthesis	NOUN
cana-6247	474	7	for	for	ADP
cana-6247	474	8	low	low	ADJ
cana-6247	474	9	-	-	PUNCT
cana-6247	474	10	light	light	NOUN
cana-6247	474	11	image	image	NOUN
cana-6247	474	12	denoising	denoising	NOUN
cana-6247	474	13	.	.	PUNCT
cana-6247	475	1	ieee	ieee	PROPN
cana-6247	475	2	trans	trans	PROPN
cana-6247	475	3	.	.	PUNCT
cana-6247	475	4	pattern	pattern	PROPN
cana-6247	475	5	anal	anal	PROPN
cana-6247	475	6	.	.	PUNCT
cana-6247	476	1	mach	mach	PROPN
cana-6247	476	2	.	.	PUNCT
cana-6247	477	1	intell	intell	PROPN
cana-6247	477	2	.	.	PUNCT
cana-6247	478	1	,	,	PUNCT
cana-6247	478	2	pp	pp	ADV
cana-6247	478	3	,	,	PUNCT
cana-6247	478	4	august	august	PROPN
cana-6247	478	5	2025	2025	NUM
cana-6247	478	6	.	.	PUNCT
cana-6247	479	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	479	2	997	997	NUM
cana-6247	479	3	communications	communication	NOUN
cana-6247	479	4	on	on	ADP
cana-6247	479	5	applied	apply	VERB
cana-6247	479	6	nonlinear	nonlinear	ADJ
cana-6247	479	7	analysis	analysis	NOUN
cana-6247	479	8	issn	issn	NOUN
cana-6247	479	9	:	:	PUNCT
cana-6247	479	10	1074	1074	NUM
cana-6247	479	11	-	-	PUNCT
cana-6247	479	12	133x	133x	NUM
cana-6247	479	13	vol	vol	NOUN
cana-6247	479	14	.	.	PROPN
cana-6247	480	1	31	31	NUM
cana-6247	481	1	no	no	NOUN
cana-6247	481	2	.	.	PUNCT
cana-6247	482	1	7s	7	NOUN
cana-6247	482	2	(	(	PUNCT
cana-6247	482	3	2024	2024	NUM
cana-6247	482	4	)	)	PUNCT
cana-6247	483	1	[	[	X
cana-6247	483	2	21	21	NUM
cana-6247	483	3	]	]	X
cana-6247	483	4	liying	liye	VERB
cana-6247	483	5	lu	lu	PROPN
cana-6247	483	6	,	,	PUNCT
cana-6247	483	7	raphael	raphael	PROPN
cana-6247	483	8	achddou	achddou	PROPN
cana-6247	483	9	,	,	PUNCT
cana-6247	483	10	and	and	CCONJ
cana-6247	483	11	sabine	sabine	PROPN
cana-6247	483	12	susstrunk	susstrunk	PROPN
cana-6247	483	13	.	.	PUNCT
cana-6247	484	1	dark	dark	PROPN
cana-6247	484	2	noise	noise	PROPN
cana-6247	484	3	diffusion	diffusion	NOUN
cana-6247	484	4	:	:	PUNCT
cana-6247	484	5	noise	noise	NOUN
cana-6247	484	6	synthesis	synthesis	NOUN
cana-6247	484	7	for	for	ADP
cana-6247	484	8	low	low	ADJ
cana-6247	484	9	-	-	PUNCT
cana-6247	484	10	light	light	NOUN
cana-6247	484	11	image	image	NOUN
cana-6247	484	12	denoising	denoising	NOUN
cana-6247	484	13	.	.	PUNCT
cana-6247	485	1	ieee	ieee	PROPN
cana-6247	485	2	trans	trans	PROPN
cana-6247	485	3	.	.	PUNCT
cana-6247	485	4	pattern	pattern	PROPN
cana-6247	485	5	anal	anal	PROPN
cana-6247	485	6	.	.	PUNCT
cana-6247	486	1	mach	mach	PROPN
cana-6247	486	2	.	.	PUNCT
cana-6247	487	1	intell	intell	PROPN
cana-6247	487	2	.	.	PUNCT
cana-6247	488	1	,	,	PUNCT
cana-6247	488	2	pp	pp	ADV
cana-6247	488	3	,	,	PUNCT
cana-6247	488	4	august	august	PROPN
cana-6247	488	5	2025	2025	NUM
cana-6247	488	6	.	.	PUNCT
cana-6247	489	1	[	[	X
cana-6247	489	2	22	22	NUM
cana-6247	489	3	]	]	X
cana-6247	489	4	andreas	andreas	PROPN
cana-6247	489	5	lugmayr	lugmayr	PROPN
cana-6247	489	6	,	,	PUNCT
cana-6247	489	7	martin	martin	PROPN
cana-6247	489	8	danelljan	danelljan	PROPN
cana-6247	489	9	,	,	PUNCT
cana-6247	489	10	andres	andres	PROPN
cana-6247	489	11	romero	romero	PROPN
cana-6247	489	12	,	,	PUNCT
cana-6247	489	13	fisher	fisher	PROPN
cana-6247	489	14	yu	yu	PROPN
cana-6247	489	15	,	,	PUNCT
cana-6247	489	16	radu	radu	PROPN
cana-6247	489	17	timofte	timofte	NOUN
cana-6247	489	18	,	,	PUNCT
cana-6247	489	19	and	and	CCONJ
cana-6247	489	20	luc	luc	PROPN
cana-6247	489	21	van	van	PROPN
cana-6247	489	22	gool	gool	PROPN
cana-6247	489	23	.	.	PUNCT
cana-6247	490	1	repaint	repaint	NOUN
cana-6247	490	2	:	:	PUNCT
cana-6247	490	3	inpainting	inpainte	VERB
cana-6247	490	4	using	use	VERB
cana-6247	490	5	denoising	denoising	NOUN
cana-6247	490	6	diffusion	diffusion	NOUN
cana-6247	490	7	probabilistic	probabilistic	ADJ
cana-6247	490	8	models	model	NOUN
cana-6247	490	9	.	.	PUNCT
cana-6247	491	1	in	in	ADP
cana-6247	491	2	2022	2022	NUM
cana-6247	491	3	ieee	ieee	NOUN
cana-6247	491	4	/	/	SYM
cana-6247	491	5	cvf	cvf	NOUN
cana-6247	491	6	conference	conference	NOUN
cana-6247	491	7	on	on	ADP
cana-6247	491	8	computer	computer	NOUN
cana-6247	491	9	vision	vision	NOUN
cana-6247	491	10	and	and	CCONJ
cana-6247	491	11	pattern	pattern	NOUN
cana-6247	491	12	recognition	recognition	NOUN
cana-6247	491	13	(	(	PUNCT
cana-6247	491	14	cvpr	cvpr	NOUN
cana-6247	491	15	)	)	PUNCT
cana-6247	491	16	,	,	PUNCT
cana-6247	491	17	pages	page	NOUN
cana-6247	491	18	11451–11461	11451–11461	NUM
cana-6247	491	19	.	.	PUNCT
cana-6247	492	1	ieee	ieee	PROPN
cana-6247	492	2	,	,	PUNCT
cana-6247	492	3	june	june	PROPN
cana-6247	492	4	2022	2022	NUM
cana-6247	492	5	.	.	PUNCT
cana-6247	493	1	[	[	X
cana-6247	493	2	23	23	NUM
cana-6247	493	3	]	]	X
cana-6247	493	4	zhaoyang	zhaoyang	PROPN
cana-6247	493	5	lyu	lyu	NOUN
cana-6247	493	6	,	,	PUNCT
cana-6247	493	7	xudong	xudong	PROPN
cana-6247	493	8	xu	xu	PROPN
cana-6247	493	9	,	,	PUNCT
cana-6247	493	10	ceyuan	ceyuan	PROPN
cana-6247	493	11	yang	yang	PROPN
cana-6247	493	12	,	,	PUNCT
cana-6247	493	13	dahua	dahua	PROPN
cana-6247	493	14	lin	lin	PROPN
cana-6247	493	15	,	,	PUNCT
cana-6247	493	16	and	and	CCONJ
cana-6247	493	17	bo	bo	PROPN
cana-6247	493	18	dai	dai	PROPN
cana-6247	493	19	.	.	PUNCT
cana-6247	493	20	accelerating	accelerate	VERB
cana-6247	493	21	diffusion	diffusion	NOUN
cana-6247	493	22	models	model	NOUN
cana-6247	493	23	via	via	ADP
cana-6247	493	24	early	early	ADJ
cana-6247	493	25	stop	stop	NOUN
cana-6247	493	26	of	of	ADP
cana-6247	493	27	the	the	DET
cana-6247	493	28	diffusion	diffusion	NOUN
cana-6247	493	29	process	process	NOUN
cana-6247	493	30	.	.	PUNCT
cana-6247	494	1	2022	2022	NUM
cana-6247	494	2	.	.	PUNCT
cana-6247	495	1	[	[	X
cana-6247	495	2	24	24	NUM
cana-6247	495	3	]	]	X
cana-6247	495	4	zhaoyang	zhaoyang	PROPN
cana-6247	495	5	lyu	lyu	NOUN
cana-6247	495	6	,	,	PUNCT
cana-6247	495	7	xudong	xudong	PROPN
cana-6247	495	8	xu	xu	PROPN
cana-6247	495	9	,	,	PUNCT
cana-6247	495	10	ceyuan	ceyuan	PROPN
cana-6247	495	11	yang	yang	PROPN
cana-6247	495	12	,	,	PUNCT
cana-6247	495	13	dahua	dahua	PROPN
cana-6247	495	14	lin	lin	PROPN
cana-6247	495	15	,	,	PUNCT
cana-6247	495	16	and	and	CCONJ
cana-6247	495	17	bo	bo	PROPN
cana-6247	495	18	dai	dai	PROPN
cana-6247	495	19	.	.	PUNCT
cana-6247	495	20	accelerating	accelerate	VERB
cana-6247	495	21	diffusion	diffusion	NOUN
cana-6247	495	22	models	model	NOUN
cana-6247	495	23	via	via	ADP
cana-6247	495	24	early	early	ADJ
cana-6247	495	25	stop	stop	NOUN
cana-6247	495	26	of	of	ADP
cana-6247	495	27	the	the	DET
cana-6247	495	28	diffusion	diffusion	NOUN
cana-6247	495	29	process	process	NOUN
cana-6247	495	30	.	.	PUNCT
cana-6247	496	1	2022	2022	NUM
cana-6247	496	2	.	.	PUNCT
cana-6247	497	1	[	[	X
cana-6247	497	2	25	25	NUM
cana-6247	497	3	]	]	PUNCT
cana-6247	497	4	chenlin	chenlin	PROPN
cana-6247	497	5	meng	meng	PROPN
cana-6247	497	6	,	,	PUNCT
cana-6247	497	7	yutong	yutong	PROPN
cana-6247	497	8	he	he	PRON
cana-6247	497	9	,	,	PUNCT
cana-6247	497	10	yang	yang	PROPN
cana-6247	497	11	song	song	PROPN
cana-6247	497	12	,	,	PUNCT
cana-6247	497	13	jiaming	jiame	VERB
cana-6247	497	14	song	song	NOUN
cana-6247	497	15	,	,	PUNCT
cana-6247	497	16	jiajun	jiajun	PROPN
cana-6247	497	17	wu	wu	PROPN
cana-6247	497	18	,	,	PUNCT
cana-6247	497	19	jun	jun	PROPN
cana-6247	497	20	-	-	PUNCT
cana-6247	497	21	yan	yan	PROPN
cana-6247	497	22	zhu	zhu	PROPN
cana-6247	497	23	,	,	PUNCT
cana-6247	497	24	and	and	CCONJ
cana-6247	497	25	stefano	stefano	PROPN
cana-6247	497	26	ermon	ermon	NOUN
cana-6247	497	27	.	.	PUNCT
cana-6247	498	1	sdedit	sdedit	NOUN
cana-6247	498	2	:	:	PUNCT
cana-6247	498	3	guided	guide	VERB
cana-6247	498	4	image	image	NOUN
cana-6247	498	5	synthesis	synthesis	NOUN
cana-6247	498	6	and	and	CCONJ
cana-6247	498	7	editing	edit	VERB
cana-6247	498	8	with	with	ADP
cana-6247	498	9	stochastic	stochastic	ADJ
cana-6247	498	10	differential	differential	ADJ
cana-6247	498	11	equations	equation	NOUN
cana-6247	498	12	.	.	PUNCT
cana-6247	499	1	august	august	PROPN
cana-6247	499	2	2021	2021	NUM
cana-6247	499	3	.	.	PUNCT
cana-6247	500	1	[	[	X
cana-6247	500	2	26	26	NUM
cana-6247	500	3	]	]	PUNCT
cana-6247	500	4	axi	axi	PROPN
cana-6247	500	5	niu	niu	PROPN
cana-6247	500	6	,	,	PUNCT
cana-6247	500	7	trung	trung	PROPN
cana-6247	500	8	x	x	PROPN
cana-6247	500	9	pham	pham	PROPN
cana-6247	500	10	,	,	PUNCT
cana-6247	500	11	kang	kang	PROPN
cana-6247	500	12	zhang	zhang	PROPN
cana-6247	500	13	,	,	PUNCT
cana-6247	500	14	jinqiu	jinqiu	PROPN
cana-6247	500	15	sun	sun	PROPN
cana-6247	500	16	,	,	PUNCT
cana-6247	500	17	yu	yu	PROPN
cana-6247	500	18	zhu	zhu	PROPN
cana-6247	500	19	,	,	PUNCT
cana-6247	500	20	qingsen	qingsen	PROPN
cana-6247	500	21	yan	yan	PROPN
cana-6247	500	22	,	,	PUNCT
cana-6247	500	23	in	in	ADP
cana-6247	500	24	so	so	ADV
cana-6247	500	25	kweon	kweon	PROPN
cana-6247	500	26	,	,	PUNCT
cana-6247	500	27	and	and	CCONJ
cana-6247	500	28	yanning	yanne	VERB
cana-6247	500	29	zhang	zhang	PROPN
cana-6247	500	30	.	.	PUNCT
cana-6247	501	1	acdmsr	acdmsr	PROPN
cana-6247	501	2	:	:	PUNCT
cana-6247	501	3	accelerated	accelerate	VERB
cana-6247	501	4	conditional	conditional	ADJ
cana-6247	501	5	diffusion	diffusion	NOUN
cana-6247	501	6	models	model	NOUN
cana-6247	501	7	for	for	ADP
cana-6247	501	8	single	single	ADJ
cana-6247	501	9	image	image	NOUN
cana-6247	501	10	super	super	NOUN
cana-6247	501	11	-	-	NOUN
cana-6247	501	12	resolution	resolution	NOUN
cana-6247	501	13	.	.	PUNCT
cana-6247	502	1	ieee	ieee	PROPN
cana-6247	502	2	trans	trans	PROPN
cana-6247	502	3	.	.	PUNCT
cana-6247	503	1	on	on	ADP
cana-6247	503	2	broadcast	broadcast	NOUN
cana-6247	503	3	.	.	PUNCT
cana-6247	504	1	,	,	PUNCT
cana-6247	504	2	70(2):492–504	70(2):492–504	NUM
cana-6247	504	3	,	,	PUNCT
cana-6247	504	4	june	june	PROPN
cana-6247	504	5	2024	2024	NUM
cana-6247	504	6	.	.	PUNCT
cana-6247	505	1	[	[	X
cana-6247	505	2	27	27	NUM
cana-6247	505	3	]	]	X
cana-6247	505	4	axi	axi	PROPN
cana-6247	505	5	niu	niu	PROPN
cana-6247	505	6	,	,	PUNCT
cana-6247	505	7	kang	kang	PROPN
cana-6247	505	8	zhang	zhang	PROPN
cana-6247	505	9	,	,	PUNCT
cana-6247	505	10	trung	trung	PROPN
cana-6247	505	11	x	x	PROPN
cana-6247	505	12	pham	pham	PROPN
cana-6247	505	13	,	,	PUNCT
cana-6247	505	14	jinqiu	jinqiu	PROPN
cana-6247	505	15	sun	sun	PROPN
cana-6247	505	16	,	,	PUNCT
cana-6247	505	17	yu	yu	PROPN
cana-6247	505	18	zhu	zhu	PROPN
cana-6247	505	19	,	,	PUNCT
cana-6247	505	20	in	in	ADP
cana-6247	505	21	so	so	ADV
cana-6247	505	22	kweon	kweon	PROPN
cana-6247	505	23	,	,	PUNCT
cana-6247	505	24	and	and	CCONJ
cana-6247	505	25	yanning	yanning	PROPN
cana-6247	505	26	zhang	zhang	PROPN
cana-6247	505	27	.	.	PUNCT
cana-6247	505	28	cdpmsr	cdpmsr	PROPN
cana-6247	505	29	:	:	PUNCT
cana-6247	505	30	conditional	conditional	ADJ
cana-6247	505	31	diffusion	diffusion	NOUN
cana-6247	505	32	probabilistic	probabilistic	ADJ
cana-6247	505	33	models	model	NOUN
cana-6247	505	34	for	for	ADP
cana-6247	505	35	single	single	ADJ
cana-6247	505	36	image	image	NOUN
cana-6247	505	37	super	super	NOUN
cana-6247	505	38	-	-	NOUN
cana-6247	505	39	resolution	resolution	NOUN
cana-6247	505	40	.	.	PUNCT
cana-6247	506	1	february	february	PROPN
cana-6247	506	2	2023	2023	NUM
cana-6247	506	3	.	.	PUNCT
cana-6247	507	1	[	[	X
cana-6247	507	2	28	28	NUM
cana-6247	507	3	]	]	PUNCT
cana-6247	507	4	ozan	ozan	PROPN
cana-6247	507	5	ozdenizci	ozdenizci	PROPN
cana-6247	507	6	and	and	CCONJ
cana-6247	507	7	robert	robert	PROPN
cana-6247	507	8	legenstein	legenstein	PROPN
cana-6247	507	9	.	.	PUNCT
cana-6247	508	1	restoring	restore	VERB
cana-6247	508	2	vision	vision	NOUN
cana-6247	508	3	in	in	ADP
cana-6247	508	4	adverse	adverse	ADJ
cana-6247	508	5	weather	weather	NOUN
cana-6247	508	6	conditions	condition	NOUN
cana-6247	508	7	with	with	ADP
cana-6247	508	8	patch	patch	NOUN
cana-6247	508	9	-	-	PUNCT
cana-6247	508	10	based	base	VERB
cana-6247	508	11	denoising	denoising	NOUN
cana-6247	508	12	diffusion	diffusion	NOUN
cana-6247	508	13	models	model	NOUN
cana-6247	508	14	.	.	PUNCT
cana-6247	509	1	ieee	ieee	PROPN
cana-6247	509	2	trans	trans	PROPN
cana-6247	509	3	.	.	PUNCT
cana-6247	509	4	pattern	pattern	PROPN
cana-6247	509	5	anal	anal	PROPN
cana-6247	509	6	.	.	PUNCT
cana-6247	510	1	mach	mach	PROPN
cana-6247	510	2	.	.	PUNCT
cana-6247	511	1	intell	intell	PROPN
cana-6247	511	2	.	.	PUNCT
cana-6247	511	3	,	,	PUNCT
cana-6247	512	1	45(8):10346–10357	45(8):10346–10357	NUM
cana-6247	512	2	,	,	PUNCT
cana-6247	512	3	august	august	PROPN
cana-6247	512	4	2023	2023	NUM
cana-6247	512	5	.	.	PUNCT
cana-6247	513	1	[	[	X
cana-6247	513	2	29	29	NUM
cana-6247	513	3	]	]	X
cana-6247	513	4	yunpeng	yunpeng	PROPN
cana-6247	513	5	qu	qu	PROPN
cana-6247	513	6	,	,	PUNCT
cana-6247	513	7	kun	kun	PROPN
cana-6247	513	8	yuan	yuan	PROPN
cana-6247	513	9	,	,	PUNCT
cana-6247	513	10	kai	kai	PROPN
cana-6247	513	11	zhao	zhao	PROPN
cana-6247	513	12	,	,	PUNCT
cana-6247	513	13	qizhi	qizhi	PROPN
cana-6247	513	14	xie	xie	PROPN
cana-6247	513	15	,	,	PUNCT
cana-6247	513	16	jinhua	jinhua	PROPN
cana-6247	513	17	hao	hao	PROPN
cana-6247	513	18	,	,	PUNCT
cana-6247	513	19	ming	ming	PROPN
cana-6247	513	20	sun	sun	PROPN
cana-6247	513	21	,	,	PUNCT
cana-6247	513	22	and	and	CCONJ
cana-6247	513	23	chao	chao	PROPN
cana-6247	513	24	zhou	zhou	PROPN
cana-6247	513	25	.	.	PUNCT
cana-6247	514	1	xpsr	xpsr	PROPN
cana-6247	514	2	:	:	PUNCT
cana-6247	514	3	cross	cross	VERB
cana-6247	514	4	modal	modal	ADJ
cana-6247	514	5	priors	prior	NOUN
cana-6247	514	6	for	for	ADP
cana-6247	514	7	diffusion	diffusion	NOUN
cana-6247	514	8	-	-	PUNCT
cana-6247	514	9	based	base	VERB
cana-6247	514	10	image	image	NOUN
cana-6247	514	11	super	super	NOUN
cana-6247	514	12	-	-	NOUN
cana-6247	514	13	resolution	resolution	NOUN
cana-6247	514	14	.	.	PUNCT
cana-6247	515	1	2024	2024	NUM
cana-6247	515	2	.	.	PUNCT
cana-6247	516	1	[	[	X
cana-6247	516	2	30	30	NUM
cana-6247	516	3	]	]	X
cana-6247	516	4	isha	isha	PROPN
cana-6247	516	5	rao	rao	PROPN
cana-6247	516	6	and	and	CCONJ
cana-6247	516	7	sanjay	sanjay	PROPN
cana-6247	516	8	ghosh	ghosh	PROPN
cana-6247	516	9	.	.	PUNCT
cana-6247	517	1	a	a	DET
cana-6247	517	2	poisson	poisson	NOUN
cana-6247	517	3	-	-	PUNCT
cana-6247	517	4	guided	guide	VERB
cana-6247	517	5	decomposition	decomposition	NOUN
cana-6247	517	6	network	network	NOUN
cana-6247	517	7	for	for	ADP
cana-6247	517	8	extreme	extreme	ADJ
cana-6247	517	9	low	low	ADJ
cana-6247	517	10	-	-	PUNCT
cana-6247	517	11	light	light	NOUN
cana-6247	517	12	image	image	NOUN
cana-6247	517	13	enhancement	enhancement	NOUN
cana-6247	517	14	.	.	PUNCT
cana-6247	518	1	june	june	PROPN
cana-6247	518	2	2025	2025	NUM
cana-6247	518	3	.	.	PUNCT
cana-6247	519	1	[	[	X
cana-6247	519	2	31	31	NUM
cana-6247	519	3	]	]	PUNCT
cana-6247	519	4	chitwan	chitwan	PROPN
cana-6247	519	5	saharia	saharia	PROPN
cana-6247	519	6	,	,	PUNCT
cana-6247	519	7	william	william	PROPN
cana-6247	519	8	chan	chan	PROPN
cana-6247	519	9	,	,	PUNCT
cana-6247	519	10	huiwen	huiwen	PROPN
cana-6247	519	11	chang	chang	PROPN
cana-6247	519	12	,	,	PUNCT
cana-6247	519	13	chris	chris	PROPN
cana-6247	519	14	a	a	DET
cana-6247	519	15	lee	lee	PROPN
cana-6247	519	16	,	,	PUNCT
cana-6247	519	17	jonathan	jonathan	PROPN
cana-6247	519	18	ho	ho	PROPN
cana-6247	519	19	,	,	PUNCT
cana-6247	519	20	tim	tim	PROPN
cana-6247	519	21	salimans	salimans	PROPN
cana-6247	519	22	,	,	PUNCT
cana-6247	519	23	david	david	PROPN
cana-6247	519	24	j	j	PROPN
cana-6247	519	25	fleet	fleet	PROPN
cana-6247	519	26	,	,	PUNCT
cana-6247	519	27	and	and	CCONJ
cana-6247	519	28	mohammad	mohammad	PROPN
cana-6247	519	29	norouzi	norouzi	PROPN
cana-6247	519	30	.	.	PUNCT
cana-6247	520	1	palette	palette	NOUN
cana-6247	520	2	:	:	PUNCT
cana-6247	520	3	image	image	NOUN
cana-6247	520	4	-	-	PUNCT
cana-6247	520	5	to	to	ADP
cana-6247	520	6	-	-	PUNCT
cana-6247	520	7	image	image	NOUN
cana-6247	520	8	diffusion	diffusion	NOUN
cana-6247	520	9	models	model	NOUN
cana-6247	520	10	.	.	PUNCT
cana-6247	521	1	november	november	PROPN
cana-6247	521	2	2021	2021	NUM
cana-6247	521	3	.	.	PUNCT
cana-6247	522	1	[	[	X
cana-6247	522	2	32	32	NUM
cana-6247	522	3	]	]	PUNCT
cana-6247	522	4	chitwan	chitwan	PROPN
cana-6247	522	5	saharia	saharia	PROPN
cana-6247	522	6	,	,	PUNCT
cana-6247	522	7	william	william	PROPN
cana-6247	522	8	fedus	fedus	PROPN
cana-6247	522	9	,	,	PUNCT
cana-6247	522	10	saurabh	saurabh	PROPN
cana-6247	522	11	khanna	khanna	PROPN
cana-6247	522	12	,	,	PUNCT
cana-6247	522	13	shariq	shariq	NOUN
cana-6247	522	14	iqbal	iqbal	PROPN
cana-6247	522	15	,	,	PUNCT
cana-6247	522	16	ben	ben	PROPN
cana-6247	522	17	wang	wang	PROPN
cana-6247	522	18	,	,	PUNCT
cana-6247	522	19	dmitry	dmitry	PROPN
cana-6247	522	20	kalenichenko	kalenichenko	PROPN
cana-6247	522	21	,	,	PUNCT
cana-6247	522	22	mohammad	mohammad	PROPN
cana-6247	522	23	norouzi	norouzi	PROPN
cana-6247	522	24	,	,	PUNCT
cana-6247	522	25	and	and	CCONJ
cana-6247	522	26	william	william	PROPN
cana-6247	522	27	t.	t.	PROPN
cana-6247	522	28	freeman	freeman	PROPN
cana-6247	522	29	.	.	PUNCT
cana-6247	523	1	palette	palette	NOUN
cana-6247	523	2	:	:	PUNCT
cana-6247	523	3	image	image	NOUN
cana-6247	523	4	-	-	PUNCT
cana-6247	523	5	to	to	ADP
cana-6247	523	6	-	-	PUNCT
cana-6247	523	7	image	image	NOUN
cana-6247	523	8	diffusion	diffusion	NOUN
cana-6247	523	9	models	model	NOUN
cana-6247	523	10	.	.	PUNCT
cana-6247	524	1	arxiv	arxiv	PROPN
cana-6247	524	2	preprint	preprint	PROPN
cana-6247	524	3	arxiv:2111.05826	arxiv:2111.05826	PROPN
cana-6247	524	4	,	,	PUNCT
cana-6247	524	5	2021	2021	NUM
cana-6247	524	6	.	.	PUNCT
cana-6247	525	1	[	[	X
cana-6247	525	2	33	33	NUM
cana-6247	525	3	]	]	PUNCT
cana-6247	525	4	shuyao	shuyao	PROPN
cana-6247	525	5	shang	shang	PROPN
cana-6247	525	6	,	,	PUNCT
cana-6247	525	7	zhengyang	zhengyang	PROPN
cana-6247	525	8	shan	shan	PROPN
cana-6247	525	9	,	,	PUNCT
cana-6247	525	10	guangxing	guangxe	VERB
cana-6247	525	11	liu	liu	PROPN
cana-6247	525	12	,	,	PUNCT
cana-6247	525	13	lunqian	lunqian	PROPN
cana-6247	525	14	wang	wang	PROPN
cana-6247	525	15	,	,	PUNCT
cana-6247	525	16	xinghua	xinghua	PROPN
cana-6247	525	17	wang	wang	PROPN
cana-6247	525	18	,	,	PUNCT
cana-6247	525	19	zekai	zekai	PROPN
cana-6247	525	20	zhang	zhang	PROPN
cana-6247	525	21	,	,	PUNCT
cana-6247	525	22	and	and	CCONJ
cana-6247	525	23	jinglin	jinglin	PROPN
cana-6247	525	24	zhang	zhang	PROPN
cana-6247	525	25	.	.	PUNCT
cana-6247	526	1	resdiff	resdiff	PROPN
cana-6247	526	2	:	:	PUNCT
cana-6247	526	3	combining	combine	VERB
cana-6247	526	4	cnn	cnn	PROPN
cana-6247	526	5	and	and	CCONJ
cana-6247	526	6	diffusion	diffusion	NOUN
cana-6247	526	7	model	model	NOUN
cana-6247	526	8	for	for	ADP
cana-6247	526	9	image	image	NOUN
cana-6247	526	10	super	super	NOUN
cana-6247	526	11	-	-	NOUN
cana-6247	526	12	resolution	resolution	NOUN
cana-6247	526	13	.	.	PUNCT
cana-6247	527	1	march	march	PROPN
cana-6247	527	2	2023	2023	NUM
cana-6247	527	3	.	.	PUNCT
cana-6247	528	1	[	[	X
cana-6247	528	2	34	34	NUM
cana-6247	528	3	]	]	SYM
cana-6247	528	4	yanli	yanli	PROPN
cana-6247	528	5	shi	shi	PROPN
cana-6247	528	6	,	,	PUNCT
cana-6247	528	7	xianhe	xianhe	PROPN
cana-6247	528	8	zhang	zhang	PROPN
cana-6247	528	9	,	,	PUNCT
cana-6247	528	10	yi	yi	PROPN
cana-6247	528	11	jia	jia	PROPN
cana-6247	528	12	,	,	PUNCT
cana-6247	528	13	and	and	CCONJ
cana-6247	528	14	jinxing	jinx	VERB
cana-6247	528	15	zhao	zhao	PROPN
cana-6247	528	16	.	.	PUNCT
cana-6247	529	1	multi	multi	ADJ
cana-6247	529	2	-	-	ADJ
cana-6247	529	3	scale	scale	ADJ
cana-6247	529	4	adversarial	adversarial	ADJ
cana-6247	529	5	diffusion	diffusion	NOUN
cana-6247	529	6	network	network	NOUN
cana-6247	529	7	for	for	ADP
cana-6247	529	8	image	image	NOUN
cana-6247	529	9	super	super	NOUN
cana-6247	529	10	-	-	NOUN
cana-6247	529	11	resolution	resolution	NOUN
cana-6247	529	12	.	.	PUNCT
cana-6247	530	1	sci	sci	PROPN
cana-6247	530	2	.	.	PROPN
cana-6247	530	3	rep	rep	PROPN
cana-6247	530	4	.	.	PROPN
cana-6247	530	5	,	,	PUNCT
cana-6247	530	6	15(1):11690	15(1):11690	NUM
cana-6247	530	7	,	,	PUNCT
cana-6247	530	8	april	april	PROPN
cana-6247	530	9	2025	2025	NUM
cana-6247	530	10	.	.	PUNCT
cana-6247	531	1	[	[	X
cana-6247	531	2	35	35	NUM
cana-6247	531	3	]	]	PUNCT
cana-6247	531	4	zhenning	zhenning	NOUN
cana-6247	531	5	shi	shi	PROPN
cana-6247	531	6	,	,	PUNCT
cana-6247	531	7	haoshuai	haoshuai	PROPN
cana-6247	531	8	zheng	zheng	PROPN
cana-6247	531	9	,	,	PUNCT
cana-6247	531	10	chen	chen	PROPN
cana-6247	531	11	xu	xu	PROPN
cana-6247	531	12	,	,	PUNCT
cana-6247	531	13	changsheng	changsheng	PROPN
cana-6247	531	14	dong	dong	PROPN
cana-6247	531	15	,	,	PUNCT
cana-6247	531	16	bin	bin	PROPN
cana-6247	531	17	pan	pan	PROPN
cana-6247	531	18	,	,	PUNCT
cana-6247	531	19	xueshuo	xueshuo	PROPN
cana-6247	531	20	xie	xie	PROPN
cana-6247	531	21	,	,	PUNCT
cana-6247	531	22	along	along	ADP
cana-6247	531	23	he	he	PRON
cana-6247	531	24	,	,	PUNCT
cana-6247	531	25	tao	tao	PROPN
cana-6247	531	26	li	li	PROPN
cana-6247	531	27	,	,	PUNCT
cana-6247	531	28	and	and	CCONJ
cana-6247	531	29	huazhu	huazhu	NOUN
cana-6247	531	30	fu	fu	PROPN
cana-6247	531	31	.	.	PUNCT
cana-6247	532	1	resfusion	resfusion	NOUN
cana-6247	532	2	:	:	PUNCT
cana-6247	532	3	denoising	denoise	VERB
cana-6247	532	4	diffusion	diffusion	NOUN
cana-6247	532	5	probabilistic	probabilistic	ADJ
cana-6247	532	6	models	model	NOUN
cana-6247	532	7	for	for	ADP
cana-6247	532	8	image	image	NOUN
cana-6247	532	9	restoration	restoration	NOUN
cana-6247	532	10	based	base	VERB
cana-6247	532	11	on	on	ADP
cana-6247	532	12	prior	prior	ADJ
cana-6247	532	13	residual	residual	ADJ
cana-6247	532	14	noise	noise	NOUN
cana-6247	532	15	.	.	PUNCT
cana-6247	533	1	november	november	PROPN
cana-6247	533	2	2023	2023	NUM
cana-6247	533	3	.	.	PUNCT
cana-6247	534	1	[	[	X
cana-6247	534	2	36	36	NUM
cana-6247	534	3	]	]	X
cana-6247	534	4	takahiro	takahiro	PROPN
cana-6247	534	5	shirakawa	shirakawa	PROPN
cana-6247	534	6	and	and	CCONJ
cana-6247	534	7	seiichi	seiichi	PROPN
cana-6247	534	8	uchida	uchida	PROPN
cana-6247	534	9	.	.	PUNCT
cana-6247	535	1	noisecollage	noisecollage	NOUN
cana-6247	535	2	:	:	PUNCT
cana-6247	535	3	a	a	DET
cana-6247	535	4	layout	layout	NOUN
cana-6247	535	5	-	-	PUNCT
cana-6247	535	6	aware	aware	ADJ
cana-6247	535	7	text	text	NOUN
cana-6247	535	8	-	-	PUNCT
cana-6247	535	9	to	to	ADP
cana-6247	535	10	-	-	PUNCT
cana-6247	535	11	image	image	NOUN
cana-6247	535	12	diffusion	diffusion	NOUN
cana-6247	535	13	model	model	NOUN
cana-6247	535	14	based	base	VERB
cana-6247	535	15	on	on	ADP
cana-6247	535	16	noise	noise	NOUN
cana-6247	535	17	cropping	cropping	NOUN
cana-6247	535	18	and	and	CCONJ
cana-6247	535	19	merging	merge	VERB
cana-6247	535	20	.	.	PUNCT
cana-6247	536	1	in	in	ADP
cana-6247	536	2	2024	2024	NUM
cana-6247	536	3	ieee	ieee	NOUN
cana-6247	536	4	/	/	SYM
cana-6247	536	5	cvf	cvf	NOUN
cana-6247	536	6	conference	conference	NOUN
cana-6247	536	7	on	on	ADP
cana-6247	536	8	computer	computer	NOUN
cana-6247	536	9	vision	vision	NOUN
cana-6247	536	10	and	and	CCONJ
cana-6247	536	11	pattern	pattern	NOUN
cana-6247	536	12	recognition	recognition	NOUN
cana-6247	536	13	(	(	PUNCT
cana-6247	536	14	cvpr	cvpr	NOUN
cana-6247	536	15	)	)	PUNCT
cana-6247	536	16	,	,	PUNCT
cana-6247	536	17	pages	page	NOUN
cana-6247	536	18	8921–8930	8921–8930	NUM
cana-6247	536	19	.	.	PUNCT
cana-6247	537	1	ieee	ieee	PROPN
cana-6247	537	2	,	,	PUNCT
cana-6247	537	3	june	june	PROPN
cana-6247	537	4	2024	2024	NUM
cana-6247	537	5	.	.	PUNCT
cana-6247	538	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	538	2	998	998	NUM
cana-6247	538	3	communications	communication	NOUN
cana-6247	538	4	on	on	ADP
cana-6247	538	5	applied	apply	VERB
cana-6247	538	6	nonlinear	nonlinear	ADJ
cana-6247	538	7	analysis	analysis	NOUN
cana-6247	538	8	issn	issn	NOUN
cana-6247	538	9	:	:	PUNCT
cana-6247	538	10	1074	1074	NUM
cana-6247	538	11	-	-	PUNCT
cana-6247	538	12	133x	133x	NUM
cana-6247	538	13	vol	vol	NOUN
cana-6247	538	14	.	.	PROPN
cana-6247	539	1	31	31	NUM
cana-6247	540	1	no	no	NOUN
cana-6247	540	2	.	.	PUNCT
cana-6247	541	1	7s	7	NOUN
cana-6247	541	2	(	(	PUNCT
cana-6247	541	3	2024	2024	NUM
cana-6247	541	4	)	)	PUNCT
cana-6247	542	1	[	[	X
cana-6247	542	2	37	37	NUM
cana-6247	542	3	]	]	PUNCT
cana-6247	542	4	jiayan	jiayan	PROPN
cana-6247	542	5	teng	teng	PROPN
cana-6247	542	6	,	,	PUNCT
cana-6247	542	7	wendi	wendi	PROPN
cana-6247	542	8	zheng	zheng	PROPN
cana-6247	542	9	,	,	PUNCT
cana-6247	542	10	ming	ming	PROPN
cana-6247	542	11	ding	ding	PROPN
cana-6247	542	12	,	,	PUNCT
cana-6247	542	13	wenyi	wenyi	PROPN
cana-6247	542	14	hong	hong	PROPN
cana-6247	542	15	,	,	PUNCT
cana-6247	542	16	jianqiao	jianqiao	NOUN
cana-6247	542	17	wangni	wangni	PROPN
cana-6247	542	18	,	,	PUNCT
cana-6247	542	19	zhuoyi	zhuoyi	PROPN
cana-6247	542	20	yang	yang	PROPN
cana-6247	542	21	,	,	PUNCT
cana-6247	542	22	and	and	CCONJ
cana-6247	542	23	jie	jie	PROPN
cana-6247	542	24	tang	tang	PROPN
cana-6247	542	25	.	.	PUNCT
cana-6247	543	1	relay	relay	NOUN
cana-6247	543	2	diffusion	diffusion	NOUN
cana-6247	543	3	:	:	PUNCT
cana-6247	543	4	unifying	unify	VERB
cana-6247	543	5	diffusion	diffusion	NOUN
cana-6247	543	6	process	process	NOUN
cana-6247	543	7	across	across	ADP
cana-6247	543	8	resolutions	resolution	NOUN
cana-6247	543	9	for	for	ADP
cana-6247	543	10	image	image	NOUN
cana-6247	543	11	synthesis	synthesis	NOUN
cana-6247	543	12	.	.	PUNCT
cana-6247	544	1	2023	2023	NUM
cana-6247	544	2	.	.	PUNCT
cana-6247	545	1	[	[	X
cana-6247	545	2	38	38	NUM
cana-6247	545	3	]	]	PUNCT
cana-6247	545	4	jinyi	jinyi	PROPN
cana-6247	545	5	wang	wang	PROPN
cana-6247	545	6	,	,	PUNCT
cana-6247	545	7	zhaoyang	zhaoyang	PROPN
cana-6247	545	8	lyu	lyu	PROPN
cana-6247	545	9	,	,	PUNCT
cana-6247	545	10	dahua	dahua	PROPN
cana-6247	545	11	lin	lin	PROPN
cana-6247	545	12	,	,	PUNCT
cana-6247	545	13	bo	bo	PROPN
cana-6247	545	14	dai	dai	PROPN
cana-6247	545	15	,	,	PUNCT
cana-6247	545	16	and	and	CCONJ
cana-6247	545	17	hongfei	hongfei	PROPN
cana-6247	545	18	fu	fu	PROPN
cana-6247	545	19	.	.	PUNCT
cana-6247	546	1	guided	guide	VERB
cana-6247	546	2	diffusion	diffusion	NOUN
cana-6247	546	3	model	model	NOUN
cana-6247	546	4	for	for	ADP
cana-6247	546	5	adversarial	adversarial	ADJ
cana-6247	546	6	purification	purification	NOUN
cana-6247	546	7	.	.	PUNCT
cana-6247	547	1	2022	2022	NUM
cana-6247	547	2	.	.	PUNCT
cana-6247	548	1	[	[	X
cana-6247	548	2	39	39	NUM
cana-6247	548	3	]	]	X
cana-6247	548	4	kai	kai	PROPN
cana-6247	548	5	wang	wang	PROPN
cana-6247	548	6	,	,	PUNCT
cana-6247	548	7	shaozhang	shaozhang	PROPN
cana-6247	548	8	niu	niu	PROPN
cana-6247	548	9	,	,	PUNCT
cana-6247	548	10	qixian	qixian	PROPN
cana-6247	548	11	hao	hao	PROPN
cana-6247	548	12	,	,	PUNCT
cana-6247	548	13	and	and	CCONJ
cana-6247	548	14	jiwei	jiwei	PROPN
cana-6247	548	15	zhang	zhang	PROPN
cana-6247	548	16	.	.	PUNCT
cana-6247	548	17	inpdiffusion	inpdiffusion	NOUN
cana-6247	548	18	:	:	PUNCT
cana-6247	548	19	image	image	NOUN
cana-6247	548	20	inpainting	inpainte	VERB
cana-6247	548	21	localization	localization	NOUN
cana-6247	548	22	via	via	ADP
cana-6247	548	23	conditional	conditional	ADJ
cana-6247	548	24	diffusion	diffusion	NOUN
cana-6247	548	25	models	model	NOUN
cana-6247	548	26	.	.	PUNCT
cana-6247	549	1	january	january	PROPN
cana-6247	549	2	2025	2025	NUM
cana-6247	549	3	.	.	PUNCT
cana-6247	550	1	[	[	X
cana-6247	550	2	40	40	NUM
cana-6247	550	3	]	]	PUNCT
cana-6247	550	4	na	na	ADP
cana-6247	550	5	wang	wang	PROPN
cana-6247	550	6	,	,	PUNCT
cana-6247	550	7	yu	yu	PROPN
cana-6247	550	8	shang	shang	PROPN
cana-6247	550	9	,	,	PUNCT
cana-6247	550	10	yang	yang	PROPN
cana-6247	550	11	chen	chen	PROPN
cana-6247	550	12	,	,	PUNCT
cana-6247	550	13	min	min	PROPN
cana-6247	550	14	yang	yang	PROPN
cana-6247	550	15	,	,	PUNCT
cana-6247	550	16	quan	quan	PROPN
cana-6247	550	17	zhang	zhang	PROPN
cana-6247	550	18	,	,	PUNCT
cana-6247	550	19	yi	yi	PROPN
cana-6247	550	20	liu	liu	PROPN
cana-6247	550	21	,	,	PUNCT
cana-6247	550	22	and	and	CCONJ
cana-6247	550	23	zhiguo	zhiguo	PROPN
cana-6247	550	24	gui	gui	PROPN
cana-6247	550	25	.	.	PUNCT
cana-6247	551	1	a	a	DET
cana-6247	551	2	hybrid	hybrid	ADJ
cana-6247	551	3	model	model	NOUN
cana-6247	551	4	for	for	ADP
cana-6247	551	5	image	image	NOUN
cana-6247	551	6	denoising	denoising	NOUN
cana-6247	551	7	combining	combine	VERB
cana-6247	551	8	modified	modify	VERB
cana-6247	551	9	isotropic	isotropic	NOUN
cana-6247	551	10	diffusion	diffusion	NOUN
cana-6247	551	11	model	model	NOUN
cana-6247	551	12	and	and	CCONJ
cana-6247	551	13	modified	modify	VERB
cana-6247	551	14	perona	perona	PROPN
cana-6247	551	15	-	-	PUNCT
cana-6247	551	16	malik	malik	PROPN
cana-6247	551	17	model	model	NOUN
cana-6247	551	18	.	.	PUNCT
cana-6247	552	1	ieee	ieee	NOUN
cana-6247	552	2	access	access	NOUN
cana-6247	552	3	,	,	PUNCT
cana-6247	552	4	6:33568–33582	6:33568–33582	PROPN
cana-6247	552	5	,	,	PUNCT
cana-6247	552	6	2018	2018	NUM
cana-6247	552	7	.	.	PUNCT
cana-6247	553	1	[	[	X
cana-6247	553	2	41	41	NUM
cana-6247	553	3	]	]	X
cana-6247	553	4	rongyuan	rongyuan	PROPN
cana-6247	553	5	wu	wu	PROPN
cana-6247	553	6	,	,	PUNCT
cana-6247	553	7	tao	tao	PROPN
cana-6247	553	8	yang	yang	PROPN
cana-6247	553	9	,	,	PUNCT
cana-6247	553	10	lingchen	lingchen	X
cana-6247	553	11	sun	sun	PROPN
cana-6247	553	12	,	,	PUNCT
cana-6247	553	13	zhengqiang	zhengqiang	PROPN
cana-6247	553	14	zhang	zhang	PROPN
cana-6247	553	15	,	,	PUNCT
cana-6247	553	16	shuai	shuai	PROPN
cana-6247	553	17	li	li	PROPN
cana-6247	553	18	,	,	PUNCT
cana-6247	553	19	and	and	CCONJ
cana-6247	553	20	lei	lei	PROPN
cana-6247	553	21	zhang	zhang	PROPN
cana-6247	553	22	.	.	PUNCT
cana-6247	554	1	seesr	seesr	PROPN
cana-6247	554	2	:	:	PUNCT
cana-6247	554	3	towards	towards	ADP
cana-6247	554	4	semantics	semantic	NOUN
cana-6247	554	5	-	-	PUNCT
cana-6247	554	6	aware	aware	ADJ
cana-6247	554	7	real	real	ADJ
cana-6247	554	8	-	-	PUNCT
cana-6247	554	9	world	world	NOUN
cana-6247	554	10	image	image	NOUN
cana-6247	554	11	super	super	NOUN
cana-6247	554	12	-	-	NOUN
cana-6247	554	13	resolution	resolution	NOUN
cana-6247	554	14	.	.	PUNCT
cana-6247	555	1	november	november	PROPN
cana-6247	555	2	2023	2023	NUM
cana-6247	555	3	.	.	PUNCT
cana-6247	556	1	[	[	X
cana-6247	556	2	42	42	NUM
cana-6247	556	3	]	]	X
cana-6247	556	4	bin	bin	PROPN
cana-6247	556	5	xia	xia	PROPN
cana-6247	556	6	,	,	PUNCT
cana-6247	556	7	yulun	yulun	PROPN
cana-6247	556	8	zhang	zhang	PROPN
cana-6247	556	9	,	,	PUNCT
cana-6247	556	10	shiyin	shiyin	PROPN
cana-6247	556	11	wang	wang	PROPN
cana-6247	556	12	,	,	PUNCT
cana-6247	556	13	yitong	yitong	PROPN
cana-6247	556	14	wang	wang	PROPN
cana-6247	556	15	,	,	PUNCT
cana-6247	556	16	xinglong	xinglong	PROPN
cana-6247	556	17	wu	wu	PROPN
cana-6247	556	18	,	,	PUNCT
cana-6247	556	19	yapeng	yapeng	PROPN
cana-6247	556	20	tian	tian	PROPN
cana-6247	556	21	,	,	PUNCT
cana-6247	556	22	wenming	wenme	VERB
cana-6247	556	23	yang	yang	PROPN
cana-6247	556	24	,	,	PUNCT
cana-6247	556	25	and	and	CCONJ
cana-6247	556	26	luc	luc	PROPN
cana-6247	556	27	van	van	PROPN
cana-6247	556	28	gool	gool	PROPN
cana-6247	556	29	.	.	PUNCT
cana-6247	557	1	diffir	diffir	PROPN
cana-6247	557	2	:	:	PUNCT
cana-6247	557	3	efficient	efficient	ADJ
cana-6247	557	4	diffusion	diffusion	NOUN
cana-6247	557	5	model	model	NOUN
cana-6247	557	6	for	for	ADP
cana-6247	557	7	image	image	NOUN
cana-6247	557	8	restoration	restoration	NOUN
cana-6247	557	9	.	.	PUNCT
cana-6247	558	1	iccv	iccv	PROPN
cana-6247	558	2	,	,	PUNCT
cana-6247	558	3	2023	2023	NUM
cana-6247	558	4	.	.	PUNCT
cana-6247	559	1	[	[	X
cana-6247	559	2	43	43	NUM
cana-6247	559	3	]	]	PUNCT
cana-6247	559	4	j.	j.	PROPN
cana-6247	559	5	xu	xu	PROPN
cana-6247	559	6	,	,	PUNCT
cana-6247	559	7	s.	s.	PROPN
cana-6247	559	8	motamed	motame	VERB
cana-6247	559	9	,	,	PUNCT
cana-6247	559	10	p.	p.	NOUN
cana-6247	559	11	vaddamanu	vaddamanu	PROPN
cana-6247	559	12	,	,	PUNCT
cana-6247	559	13	c.	c.	PROPN
cana-6247	559	14	wu	wu	PROPN
cana-6247	559	15	,	,	PUNCT
cana-6247	559	16	c.	c.	PROPN
cana-6247	559	17	häne	häne	PROPN
cana-6247	559	18	,	,	PUNCT
cana-6247	559	19	j.	j.	PROPN
cana-6247	559	20	bazin	bazin	PROPN
cana-6247	559	21	,	,	PUNCT
cana-6247	559	22	and	and	CCONJ
cana-6247	559	23	f.	f.	PROPN
cana-6247	559	24	la	la	PROPN
cana-6247	559	25	torre	torre	PROPN
cana-6247	559	26	.	.	PUNCT
cana-6247	560	1	personalized	personalized	ADJ
cana-6247	560	2	face	face	NOUN
cana-6247	560	3	inpainting	inpainte	VERB
cana-6247	560	4	with	with	ADP
cana-6247	560	5	diffusion	diffusion	NOUN
cana-6247	560	6	models	model	NOUN
cana-6247	560	7	by	by	ADP
cana-6247	560	8	parallel	parallel	ADJ
cana-6247	560	9	visual	visual	ADJ
cana-6247	560	10	attention	attention	NOUN
cana-6247	560	11	.	.	PUNCT
cana-6247	561	1	corr	corr	PROPN
cana-6247	561	2	,	,	PUNCT
cana-6247	561	3	abs/2312.03556	abs/2312.03556	X
cana-6247	561	4	,	,	PUNCT
cana-6247	561	5	2023	2023	NUM
cana-6247	561	6	.	.	PUNCT
cana-6247	562	1	[	[	X
cana-6247	562	2	44	44	NUM
cana-6247	562	3	]	]	SYM
cana-6247	562	4	shuzhou	shuzhou	PROPN
cana-6247	562	5	yang	yang	PROPN
cana-6247	562	6	,	,	PUNCT
cana-6247	562	7	xuanyu	xuanyu	PROPN
cana-6247	562	8	zhang	zhang	PROPN
cana-6247	562	9	,	,	PUNCT
cana-6247	562	10	yinhuai	yinhuai	PROPN
cana-6247	562	11	wang	wang	PROPN
cana-6247	562	12	,	,	PUNCT
cana-6247	562	13	jiwen	jiwen	PROPN
cana-6247	562	14	yu	yu	PROPN
cana-6247	562	15	,	,	PUNCT
cana-6247	562	16	yuhan	yuhan	PROPN
cana-6247	562	17	wang	wang	PROPN
cana-6247	562	18	,	,	PUNCT
cana-6247	562	19	and	and	CCONJ
cana-6247	562	20	jian	jian	PROPN
cana-6247	562	21	zhang	zhang	PROPN
cana-6247	562	22	.	.	PUNCT
cana-6247	563	1	difflle	difflle	PROPN
cana-6247	563	2	:	:	PUNCT
cana-6247	563	3	diffusion	diffusion	NOUN
cana-6247	563	4	-	-	PUNCT
cana-6247	563	5	guided	guide	VERB
cana-6247	563	6	domain	domain	NOUN
cana-6247	563	7	calibration	calibration	NOUN
cana-6247	563	8	for	for	ADP
cana-6247	563	9	unsupervised	unsupervised	ADJ
cana-6247	563	10	low	low	ADJ
cana-6247	563	11	-	-	PUNCT
cana-6247	563	12	light	light	NOUN
cana-6247	563	13	image	image	NOUN
cana-6247	563	14	enhancement	enhancement	NOUN
cana-6247	563	15	.	.	PUNCT
cana-6247	564	1	august	august	PROPN
cana-6247	564	2	2023	2023	NUM
cana-6247	564	3	.	.	PUNCT
cana-6247	565	1	[	[	X
cana-6247	565	2	45	45	NUM
cana-6247	565	3	]	]	PUNCT
cana-6247	565	4	chia	chia	NOUN
cana-6247	565	5	-	-	PUNCT
cana-6247	565	6	hung	hung	ADJ
cana-6247	565	7	yeh	yeh	PROPN
cana-6247	565	8	and	and	CCONJ
cana-6247	565	9	cheng	cheng	PROPN
cana-6247	565	10	-	-	PUNCT
cana-6247	565	11	yue	yue	PROPN
cana-6247	565	12	liou	liou	PROPN
cana-6247	565	13	.	.	PUNCT
cana-6247	565	14	diffusion	diffusion	NOUN
cana-6247	565	15	-	-	PUNCT
cana-6247	565	16	based	base	VERB
cana-6247	565	17	low	low	ADJ
cana-6247	565	18	-	-	PUNCT
cana-6247	565	19	light	light	NOUN
cana-6247	565	20	image	image	NOUN
cana-6247	565	21	enhancement	enhancement	NOUN
cana-6247	565	22	with	with	ADP
cana-6247	565	23	kolmogorov	kolmogorov	PROPN
cana-6247	565	24	arnold	arnold	PROPN
cana-6247	565	25	networks	networks	PROPN
cana-6247	565	26	(	(	PUNCT
cana-6247	565	27	kans	kans	PROPN
cana-6247	565	28	)	)	PUNCT
cana-6247	565	29	.	.	PUNCT
cana-6247	566	1	array	array	NOUN
cana-6247	566	2	(	(	PUNCT
cana-6247	566	3	n.	n.	PROPN
cana-6247	566	4	y.	y.	PROPN
cana-6247	566	5	)	)	PUNCT
cana-6247	566	6	,	,	PUNCT
cana-6247	566	7	27(100431):100431	27(100431):100431	NUM
cana-6247	566	8	,	,	PUNCT
cana-6247	566	9	september	september	PROPN
cana-6247	566	10	2025	2025	NUM
cana-6247	566	11	.	.	PUNCT
cana-6247	567	1	[	[	X
cana-6247	567	2	46	46	NUM
cana-6247	567	3	]	]	X
cana-6247	567	4	xunpeng	xunpeng	PROPN
cana-6247	567	5	yi	yi	PROPN
cana-6247	567	6	,	,	PUNCT
cana-6247	567	7	han	han	PROPN
cana-6247	567	8	xu	xu	PROPN
cana-6247	567	9	,	,	PUNCT
cana-6247	567	10	hao	hao	PROPN
cana-6247	567	11	zhang	zhang	PROPN
cana-6247	567	12	,	,	PUNCT
cana-6247	567	13	linfeng	linfeng	PROPN
cana-6247	567	14	tang	tang	PROPN
cana-6247	567	15	,	,	PUNCT
cana-6247	567	16	and	and	CCONJ
cana-6247	567	17	jiayi	jiayi	PROPN
cana-6247	567	18	ma	ma	PROPN
cana-6247	567	19	.	.	PROPN
cana-6247	567	20	diff	diff	PROPN
cana-6247	567	21	-	-	PUNCT
cana-6247	567	22	retinex	retinex	NOUN
cana-6247	567	23	:	:	PUNCT
cana-6247	567	24	rethinking	rethink	VERB
cana-6247	567	25	low	low	ADJ
cana-6247	567	26	-	-	PUNCT
cana-6247	567	27	light	light	NOUN
cana-6247	567	28	image	image	NOUN
cana-6247	567	29	enhancement	enhancement	NOUN
cana-6247	567	30	with	with	ADP
cana-6247	567	31	a	a	DET
cana-6247	567	32	generative	generative	ADJ
cana-6247	567	33	diffusion	diffusion	NOUN
cana-6247	567	34	model	model	NOUN
cana-6247	567	35	.	.	PUNCT
cana-6247	568	1	in	in	ADP
cana-6247	568	2	2023	2023	NUM
cana-6247	568	3	ieee	ieee	NOUN
cana-6247	568	4	/	/	SYM
cana-6247	568	5	cvf	cvf	NOUN
cana-6247	568	6	international	international	ADJ
cana-6247	568	7	conference	conference	NOUN
cana-6247	568	8	on	on	ADP
cana-6247	568	9	computer	computer	NOUN
cana-6247	568	10	vision	vision	NOUN
cana-6247	568	11	(	(	PUNCT
cana-6247	568	12	iccv	iccv	PROPN
cana-6247	568	13	)	)	PUNCT
cana-6247	568	14	,	,	PUNCT
cana-6247	568	15	pages	page	VERB
cana-6247	568	16	12268–12277	12268–12277	NUM
cana-6247	568	17	.	.	PUNCT
cana-6247	569	1	ieee	ieee	PROPN
cana-6247	569	2	,	,	PUNCT
cana-6247	569	3	october	october	PROPN
cana-6247	569	4	2023	2023	NUM
cana-6247	569	5	.	.	PUNCT
cana-6247	570	1	[	[	X
cana-6247	570	2	47	47	NUM
cana-6247	570	3	]	]	X
cana-6247	570	4	yutao	yutao	PROPN
cana-6247	570	5	yuan	yuan	PROPN
cana-6247	570	6	and	and	CCONJ
cana-6247	570	7	chun	chun	PROPN
cana-6247	570	8	yuan	yuan	PROPN
cana-6247	570	9	.	.	PUNCT
cana-6247	571	1	efficient	efficient	ADJ
cana-6247	571	2	conditional	conditional	ADJ
cana-6247	571	3	diffusion	diffusion	NOUN
cana-6247	571	4	model	model	NOUN
cana-6247	571	5	with	with	ADP
cana-6247	571	6	probability	probability	NOUN
cana-6247	571	7	flow	flow	NOUN
cana-6247	571	8	sampling	sample	VERB
cana-6247	571	9	for	for	ADP
cana-6247	571	10	image	image	NOUN
cana-6247	571	11	super	super	NOUN
cana-6247	571	12	-	-	NOUN
cana-6247	571	13	resolution	resolution	NOUN
cana-6247	571	14	.	.	PUNCT
cana-6247	572	1	april	april	PROPN
cana-6247	572	2	2024	2024	NUM
cana-6247	572	3	.	.	PUNCT
cana-6247	573	1	[	[	X
cana-6247	573	2	48	48	NUM
cana-6247	573	3	]	]	X
cana-6247	573	4	yufeng	yufeng	PROPN
cana-6247	573	5	yue	yue	PROPN
cana-6247	573	6	,	,	PUNCT
cana-6247	573	7	meng	meng	PROPN
cana-6247	573	8	yu	yu	PROPN
cana-6247	573	9	,	,	PUNCT
cana-6247	573	10	luojie	luojie	PROPN
cana-6247	573	11	yang	yang	PROPN
cana-6247	573	12	,	,	PUNCT
cana-6247	573	13	and	and	CCONJ
cana-6247	573	14	tong	tong	PROPN
cana-6247	573	15	liu	liu	PROPN
cana-6247	573	16	.	.	PUNCT
cana-6247	574	1	joint	joint	ADJ
cana-6247	574	2	conditional	conditional	ADJ
cana-6247	574	3	diffusion	diffusion	NOUN
cana-6247	574	4	model	model	NOUN
cana-6247	574	5	for	for	ADP
cana-6247	574	6	image	image	NOUN
cana-6247	574	7	restoration	restoration	NOUN
cana-6247	574	8	with	with	ADP
cana-6247	574	9	mixed	mixed	ADJ
cana-6247	574	10	degradations	degradation	NOUN
cana-6247	574	11	.	.	PUNCT
cana-6247	575	1	neurocomputing	neurocomputing	NOUN
cana-6247	575	2	,	,	PUNCT
cana-6247	575	3	626(129512):129512	626(129512):129512	NUM
cana-6247	575	4	,	,	PUNCT
cana-6247	575	5	april	april	PROPN
cana-6247	575	6	2025	2025	NUM
cana-6247	575	7	.	.	PUNCT
cana-6247	576	1	[	[	X
cana-6247	576	2	49	49	NUM
cana-6247	576	3	]	]	X
cana-6247	576	4	zongsheng	zongsheng	PROPN
cana-6247	576	5	yue	yue	PROPN
cana-6247	576	6	and	and	CCONJ
cana-6247	576	7	chen	chen	PROPN
cana-6247	576	8	change	change	PROPN
cana-6247	576	9	loy	loy	PROPN
cana-6247	576	10	.	.	PUNCT
cana-6247	577	1	difface	difface	PROPN
cana-6247	577	2	:	:	PUNCT
cana-6247	577	3	blind	blind	ADJ
cana-6247	577	4	face	face	NOUN
cana-6247	577	5	restoration	restoration	NOUN
cana-6247	577	6	with	with	ADP
cana-6247	577	7	diffused	diffused	ADJ
cana-6247	577	8	error	error	NOUN
cana-6247	577	9	contraction	contraction	NOUN
cana-6247	577	10	.	.	PUNCT
cana-6247	578	1	ieee	ieee	PROPN
cana-6247	578	2	trans	trans	PROPN
cana-6247	578	3	.	.	PUNCT
cana-6247	578	4	pattern	pattern	PROPN
cana-6247	578	5	anal	anal	PROPN
cana-6247	578	6	.	.	PUNCT
cana-6247	579	1	mach	mach	PROPN
cana-6247	579	2	.	.	PUNCT
cana-6247	580	1	intell	intell	PROPN
cana-6247	580	2	.	.	PUNCT
cana-6247	580	3	,	,	PUNCT
cana-6247	581	1	46(12):9991–10004	46(12):9991–10004	NUM
cana-6247	581	2	,	,	PUNCT
cana-6247	581	3	december	december	PROPN
cana-6247	581	4	2024	2024	NUM
cana-6247	581	5	.	.	PUNCT
cana-6247	582	1	[	[	X
cana-6247	582	2	50	50	NUM
cana-6247	582	3	]	]	PUNCT
cana-6247	582	4	zongsheng	zongsheng	PROPN
cana-6247	582	5	yue	yue	PROPN
cana-6247	582	6	,	,	PUNCT
cana-6247	582	7	jianyi	jianyi	PROPN
cana-6247	582	8	wang	wang	PROPN
cana-6247	582	9	,	,	PUNCT
cana-6247	582	10	and	and	CCONJ
cana-6247	582	11	chen	chen	PROPN
cana-6247	582	12	change	change	PROPN
cana-6247	582	13	loy	loy	PROPN
cana-6247	582	14	.	.	PUNCT
cana-6247	583	1	resshift	resshift	NOUN
cana-6247	583	2	:	:	PUNCT
cana-6247	583	3	efficient	efficient	ADJ
cana-6247	583	4	diffusion	diffusion	NOUN
cana-6247	583	5	model	model	NOUN
cana-6247	583	6	for	for	ADP
cana-6247	583	7	image	image	NOUN
cana-6247	583	8	super	super	NOUN
cana-6247	583	9	-	-	NOUN
cana-6247	583	10	resolution	resolution	NOUN
cana-6247	583	11	by	by	ADP
cana-6247	583	12	residual	residual	ADJ
cana-6247	583	13	shifting	shifting	NOUN
cana-6247	583	14	.	.	PUNCT
cana-6247	584	1	2023	2023	NUM
cana-6247	584	2	.	.	PUNCT
cana-6247	585	1	[	[	X
cana-6247	585	2	51	51	NUM
cana-6247	585	3	]	]	X
cana-6247	585	4	zongsheng	zongsheng	PROPN
cana-6247	585	5	yue	yue	PROPN
cana-6247	585	6	,	,	PUNCT
cana-6247	585	7	jianyi	jianyi	PROPN
cana-6247	585	8	wang	wang	PROPN
cana-6247	585	9	,	,	PUNCT
cana-6247	585	10	and	and	CCONJ
cana-6247	585	11	chen	chen	PROPN
cana-6247	585	12	change	change	PROPN
cana-6247	585	13	loy	loy	PROPN
cana-6247	585	14	.	.	PUNCT
cana-6247	586	1	efficient	efficient	ADJ
cana-6247	586	2	diffusion	diffusion	NOUN
cana-6247	586	3	model	model	NOUN
cana-6247	586	4	for	for	ADP
cana-6247	586	5	image	image	NOUN
cana-6247	586	6	restoration	restoration	NOUN
cana-6247	586	7	by	by	ADP
cana-6247	586	8	residual	residual	ADJ
cana-6247	586	9	shifting	shifting	NOUN
cana-6247	586	10	.	.	PUNCT
cana-6247	587	1	ieee	ieee	PROPN
cana-6247	587	2	trans	trans	PROPN
cana-6247	587	3	.	.	PUNCT
cana-6247	587	4	pattern	pattern	PROPN
cana-6247	587	5	anal	anal	PROPN
cana-6247	587	6	.	.	PUNCT
cana-6247	588	1	mach	mach	PROPN
cana-6247	588	2	.	.	PUNCT
cana-6247	589	1	intell	intell	PROPN
cana-6247	589	2	.	.	PUNCT
cana-6247	589	3	,	,	PUNCT
cana-6247	590	1	47(1):116–130	47(1):116–130	PROPN
cana-6247	590	2	,	,	PUNCT
cana-6247	590	3	january	january	PROPN
cana-6247	590	4	2025	2025	NUM
cana-6247	590	5	.	.	PUNCT
cana-6247	591	1	[	[	X
cana-6247	591	2	52	52	NUM
cana-6247	591	3	]	]	X
cana-6247	591	4	leheng	leheng	PROPN
cana-6247	591	5	zhang	zhang	PROPN
cana-6247	591	6	,	,	PUNCT
cana-6247	591	7	weiyi	weiyi	VERB
cana-6247	591	8	you	you	PRON
cana-6247	591	9	,	,	PUNCT
cana-6247	591	10	kexuan	kexuan	PROPN
cana-6247	591	11	shi	shi	PROPN
cana-6247	591	12	,	,	PUNCT
cana-6247	591	13	and	and	CCONJ
cana-6247	591	14	shuhang	shuhang	PROPN
cana-6247	591	15	gu	gu	PROPN
cana-6247	591	16	.	.	PUNCT
cana-6247	592	1	uncertainty	uncertainty	NOUN
cana-6247	592	2	-	-	PUNCT
cana-6247	592	3	guided	guide	VERB
cana-6247	592	4	perturbation	perturbation	NOUN
cana-6247	592	5	for	for	ADP
cana-6247	592	6	image	image	NOUN
cana-6247	592	7	super	super	ADJ
cana-6247	592	8	-	-	ADJ
cana-6247	592	9	resolution	resolution	ADJ
cana-6247	592	10	diffusion	diffusion	NOUN
cana-6247	592	11	model	model	NOUN
cana-6247	592	12	.	.	PUNCT
cana-6247	593	1	march	march	PROPN
cana-6247	593	2	2025	2025	NUM
cana-6247	593	3	.	.	PUNCT
cana-6247	594	1	[	[	X
cana-6247	594	2	53	53	NUM
cana-6247	594	3	]	]	X
cana-6247	594	4	kai	kai	PROPN
cana-6247	594	5	zhao	zhao	PROPN
cana-6247	594	6	,	,	PUNCT
cana-6247	594	7	alex	alex	PROPN
cana-6247	594	8	ling	ling	PROPN
cana-6247	594	9	yu	yu	PROPN
cana-6247	594	10	hung	hung	PROPN
cana-6247	594	11	,	,	PUNCT
cana-6247	594	12	kaifeng	kaifeng	PROPN
cana-6247	594	13	pang	pang	PROPN
cana-6247	594	14	,	,	PUNCT
cana-6247	594	15	haoxin	haoxin	PROPN
cana-6247	594	16	zheng	zheng	PROPN
cana-6247	594	17	,	,	PUNCT
cana-6247	594	18	and	and	CCONJ
cana-6247	594	19	kyunghyun	kyunghyun	VERB
cana-6247	594	20	sung	sung	PROPN
cana-6247	594	21	.	.	PUNCT
cana-6247	595	1	partdiff	partdiff	NOUN
cana-6247	595	2	:	:	PUNCT
cana-6247	595	3	image	image	NOUN
cana-6247	595	4	super	super	NOUN
cana-6247	595	5	-	-	NOUN
cana-6247	595	6	resolution	resolution	NOUN
cana-6247	595	7	with	with	ADP
cana-6247	595	8	partial	partial	ADJ
cana-6247	595	9	diffusion	diffusion	NOUN
cana-6247	595	10	models	model	NOUN
cana-6247	595	11	.	.	PUNCT
cana-6247	596	1	2023	2023	NUM
cana-6247	596	2	.	.	PUNCT
cana-6247	597	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	597	2	999	999	NUM
cana-6247	597	3	communications	communication	NOUN
cana-6247	597	4	on	on	ADP
cana-6247	597	5	applied	apply	VERB
cana-6247	597	6	nonlinear	nonlinear	ADJ
cana-6247	597	7	analysis	analysis	NOUN
cana-6247	597	8	issn	issn	NOUN
cana-6247	597	9	:	:	PUNCT
cana-6247	597	10	1074	1074	NUM
cana-6247	597	11	-	-	PUNCT
cana-6247	597	12	133x	133x	NUM
cana-6247	597	13	vol	vol	NOUN
cana-6247	597	14	.	.	PROPN
cana-6247	598	1	31	31	NUM
cana-6247	599	1	no	no	NOUN
cana-6247	599	2	.	.	PUNCT
cana-6247	600	1	7s	7	NOUN
cana-6247	600	2	(	(	PUNCT
cana-6247	600	3	2024	2024	NUM
cana-6247	600	4	)	)	PUNCT
cana-6247	601	1	[	[	X
cana-6247	601	2	54	54	NUM
cana-6247	601	3	]	]	PUNCT
cana-6247	601	4	zixiang	zixiang	PROPN
cana-6247	601	5	zhao	zhao	PROPN
cana-6247	601	6	,	,	PUNCT
cana-6247	601	7	haowen	haowen	NOUN
cana-6247	601	8	bai	bai	PROPN
cana-6247	601	9	,	,	PUNCT
cana-6247	601	10	yuanzhi	yuanzhi	PROPN
cana-6247	601	11	zhu	zhu	PROPN
cana-6247	601	12	,	,	PUNCT
cana-6247	601	13	jiangshe	jiangshe	PROPN
cana-6247	601	14	zhang	zhang	PROPN
cana-6247	601	15	,	,	PUNCT
cana-6247	601	16	shuang	shuang	PROPN
cana-6247	601	17	xu	xu	PROPN
cana-6247	601	18	,	,	PUNCT
cana-6247	601	19	yulun	yulun	PROPN
cana-6247	601	20	zhang	zhang	PROPN
cana-6247	601	21	,	,	PUNCT
cana-6247	601	22	kai	kai	PROPN
cana-6247	601	23	zhang	zhang	PROPN
cana-6247	601	24	,	,	PUNCT
cana-6247	601	25	deyu	deyu	PROPN
cana-6247	601	26	meng	meng	PROPN
cana-6247	601	27	,	,	PUNCT
cana-6247	601	28	radu	radu	PROPN
cana-6247	601	29	timofte	timofte	PROPN
cana-6247	601	30	,	,	PUNCT
cana-6247	601	31	and	and	CCONJ
cana-6247	601	32	luc	luc	PROPN
cana-6247	601	33	van	van	PROPN
cana-6247	601	34	gool	gool	PROPN
cana-6247	601	35	.	.	PUNCT
cana-6247	601	36	ddfm	ddfm	PROPN
cana-6247	601	37	:	:	PUNCT
cana-6247	601	38	denoising	denoise	VERB
cana-6247	601	39	diffusion	diffusion	NOUN
cana-6247	601	40	model	model	NOUN
cana-6247	601	41	for	for	ADP
cana-6247	601	42	multi	multi	ADJ
cana-6247	601	43	-	-	ADJ
cana-6247	601	44	modality	modality	ADJ
cana-6247	601	45	image	image	NOUN
cana-6247	601	46	fusion	fusion	NOUN
cana-6247	601	47	.	.	PUNCT
cana-6247	602	1	2023	2023	NUM
cana-6247	602	2	.	.	PUNCT
cana-6247	603	1	[	[	X
cana-6247	603	2	55	55	NUM
cana-6247	603	3	]	]	PUNCT
cana-6247	603	4	dewei	dewei	PROPN
cana-6247	603	5	zhou	zhou	PROPN
cana-6247	603	6	,	,	PUNCT
cana-6247	603	7	zongxin	zongxin	PROPN
cana-6247	603	8	yang	yang	PROPN
cana-6247	603	9	,	,	PUNCT
cana-6247	603	10	and	and	CCONJ
cana-6247	603	11	yi	yi	PROPN
cana-6247	603	12	yang	yang	PROPN
cana-6247	603	13	.	.	PUNCT
cana-6247	604	1	pyramid	pyramid	NOUN
cana-6247	604	2	diffusion	diffusion	NOUN
cana-6247	604	3	models	model	NOUN
cana-6247	604	4	for	for	ADP
cana-6247	604	5	low	low	ADJ
cana-6247	604	6	-	-	PUNCT
cana-6247	604	7	light	light	ADJ
cana-6247	604	8	image	image	NOUN
cana-6247	604	9	enhancement	enhancement	NOUN
cana-6247	604	10	.	.	PUNCT
cana-6247	605	1	in	in	ADP
cana-6247	605	2	proceedings	proceeding	NOUN
cana-6247	605	3	of	of	ADP
cana-6247	605	4	the	the	DET
cana-6247	605	5	thirty	thirty	NUM
cana-6247	605	6	-	-	PUNCT
cana-6247	605	7	second	second	NOUN
cana-6247	605	8	international	international	ADJ
cana-6247	605	9	joint	joint	ADJ
cana-6247	605	10	conference	conference	NOUN
cana-6247	605	11	on	on	ADP
cana-6247	605	12	artificial	artificial	ADJ
cana-6247	605	13	intelligence	intelligence	NOUN
cana-6247	605	14	,	,	PUNCT
cana-6247	605	15	pages	page	NOUN
cana-6247	605	16	1795	1795	NUM
cana-6247	605	17	–	–	PUNCT
cana-6247	605	18	1803	1803	NUM
cana-6247	605	19	,	,	PUNCT
cana-6247	605	20	california	california	PROPN
cana-6247	605	21	,	,	PUNCT
cana-6247	605	22	august	august	PROPN
cana-6247	605	23	2023	2023	NUM
cana-6247	605	24	.	.	PUNCT
cana-6247	606	1	international	international	ADJ
cana-6247	606	2	joint	joint	ADJ
cana-6247	606	3	conferences	conference	NOUN
cana-6247	606	4	on	on	ADP
cana-6247	606	5	artificial	artificial	ADJ
cana-6247	606	6	intelligence	intelligence	NOUN
cana-6247	606	7	organization	organization	NOUN
cana-6247	606	8	.	.	PUNCT
cana-6247	607	1	[	[	X
cana-6247	607	2	56	56	NUM
cana-6247	607	3	]	]	X
cana-6247	607	4	yuanzhi	yuanzhi	PROPN
cana-6247	607	5	zhu	zhu	PROPN
cana-6247	607	6	,	,	PUNCT
cana-6247	607	7	yutong	yutong	PROPN
cana-6247	607	8	li	li	PROPN
cana-6247	607	9	,	,	PUNCT
cana-6247	607	10	xing	xing	PROPN
cana-6247	607	11	guo	guo	PROPN
cana-6247	607	12	,	,	PUNCT
cana-6247	607	13	song	song	NOUN
cana-6247	607	14	he	he	PRON
cana-6247	607	15	,	,	PUNCT
cana-6247	607	16	and	and	CCONJ
cana-6247	607	17	shuhang	shuhang	PROPN
cana-6247	607	18	gu	gu	PROPN
cana-6247	607	19	.	.	PUNCT
cana-6247	608	1	denoising	denoise	VERB
cana-6247	608	2	diffusion	diffusion	NOUN
cana-6247	608	3	models	model	NOUN
cana-6247	608	4	for	for	ADP
cana-6247	608	5	plug	plug	VERB
cana-6247	608	6	-	-	PUNCT
cana-6247	608	7	and	and	CCONJ
cana-6247	608	8	play	play	VERB
cana-6247	608	9	image	image	NOUN
cana-6247	608	10	restoration	restoration	NOUN
cana-6247	608	11	.	.	PUNCT
cana-6247	609	1	in	in	ADP
cana-6247	609	2	proceedings	proceeding	NOUN
cana-6247	609	3	of	of	ADP
cana-6247	609	4	the	the	DET
cana-6247	609	5	ieee	ieee	NOUN
cana-6247	609	6	/	/	SYM
cana-6247	609	7	cvf	cvf	NOUN
cana-6247	609	8	conference	conference	NOUN
cana-6247	609	9	on	on	ADP
cana-6247	609	10	computer	computer	NOUN
cana-6247	609	11	vision	vision	NOUN
cana-6247	609	12	and	and	CCONJ
cana-6247	609	13	pattern	pattern	NOUN
cana-6247	609	14	recognition	recognition	NOUN
cana-6247	609	15	(	(	PUNCT
cana-6247	609	16	cvpr	cvpr	NOUN
cana-6247	609	17	)	)	PUNCT
cana-6247	609	18	workshops	workshop	NOUN
cana-6247	609	19	,	,	PUNCT
cana-6247	609	20	pages	page	NOUN
cana-6247	609	21	1220–1229	1220–1229	NUM
cana-6247	609	22	,	,	PUNCT
cana-6247	609	23	2023	2023	NUM
cana-6247	609	24	.	.	PUNCT
cana-6247	610	1	[	[	X
cana-6247	610	2	57	57	NUM
cana-6247	610	3	]	]	PUNCT
cana-6247	610	4	yuanzhi	yuanzhi	PROPN
cana-6247	610	5	zhu	zhu	PROPN
cana-6247	610	6	,	,	PUNCT
cana-6247	610	7	kai	kai	PROPN
cana-6247	610	8	zhang	zhang	PROPN
cana-6247	610	9	,	,	PUNCT
cana-6247	610	10	jingyun	jingyun	PROPN
cana-6247	610	11	liang	liang	PROPN
cana-6247	610	12	,	,	PUNCT
cana-6247	610	13	jiezhang	jiezhang	PROPN
cana-6247	610	14	cao	cao	PROPN
cana-6247	610	15	,	,	PUNCT
cana-6247	610	16	bihan	bihan	PROPN
cana-6247	610	17	wen	wen	PROPN
cana-6247	610	18	,	,	PUNCT
cana-6247	610	19	radu	radu	PROPN
cana-6247	610	20	timofte	timofte	NOUN
cana-6247	610	21	,	,	PUNCT
cana-6247	610	22	and	and	CCONJ
cana-6247	610	23	luc	luc	PROPN
cana-6247	610	24	van	van	PROPN
cana-6247	610	25	gool	gool	PROPN
cana-6247	610	26	.	.	PUNCT
cana-6247	611	1	denoising	denoise	VERB
cana-6247	611	2	diffusion	diffusion	NOUN
cana-6247	611	3	models	model	NOUN
cana-6247	611	4	for	for	ADP
cana-6247	611	5	plug	plug	NOUN
cana-6247	611	6	-	-	PUNCT
cana-6247	611	7	and	and	CCONJ
cana-6247	611	8	-	-	PUNCT
cana-6247	611	9	play	play	VERB
cana-6247	611	10	image	image	NOUN
cana-6247	611	11	restoration	restoration	NOUN
cana-6247	611	12	.	.	PUNCT
cana-6247	612	1	2023	2023	NUM
cana-6247	612	2	.	.	PUNCT
cana-6247	613	1	https://internationalpubls.com	https://internationalpubls.com	X
cana-6247	613	2	1000	1000	NUM
cana-6247	613	3	introduction	introduction	NOUN
cana-6247	613	4	proposed	propose	VERB
cana-6247	613	5	classification	classification	NOUN
cana-6247	613	6	additional	additional	ADJ
cana-6247	613	7	gaussian	gaussian	ADJ
cana-6247	613	8	noise	noise	NOUN
cana-6247	613	9	1.1	1.1	NUM
cana-6247	613	10	image	image	NOUN
cana-6247	613	11	restoration	restoration	NOUN
cana-6247	613	12	/	/	SYM
cana-6247	613	13	fusion	fusion	NOUN
cana-6247	613	14	summary	summary	NOUN
cana-6247	613	15	conditional	conditional	ADJ
cana-6247	613	16	noise	noise	NOUN
cana-6247	613	17	injection	injection	NOUN
cana-6247	613	18	(	(	PUNCT
cana-6247	613	19	cni	cni	NOUN
cana-6247	613	20	)	)	PUNCT
cana-6247	613	21	2.1	2.1	NUM
cana-6247	613	22	image	image	NOUN
cana-6247	613	23	inpainting	inpainte	VERB
cana-6247	613	24	/	/	SYM
cana-6247	613	25	outpainting	outpainte	VERB
cana-6247	613	26	2.1.1	2.1.1	NUM
cana-6247	613	27	evidence	evidence	NOUN
cana-6247	613	28	from	from	ADP
cana-6247	613	29	recent	recent	ADJ
cana-6247	613	30	research	research	NOUN
cana-6247	613	31	summary	summary	NOUN
cana-6247	613	32	2.1.1	2.1.1	NUM
cana-6247	613	33	deblurring	deblurre	VERB
cana-6247	613	34	2.1.1	2.1.1	NUM
cana-6247	613	35	evidence	evidence	NOUN
cana-6247	613	36	from	from	ADP
cana-6247	613	37	recent	recent	ADJ
cana-6247	613	38	research	research	NOUN
cana-6247	613	39	summary	summary	NOUN
cana-6247	613	40	learned	learn	VERB
cana-6247	613	41	noise	noise	NOUN
cana-6247	613	42	3.1	3.1	NUM
cana-6247	613	43	super	super	ADJ
cana-6247	613	44	resolution	resolution	NOUN
cana-6247	613	45	(	(	PUNCT
cana-6247	613	46	sr	sr	PROPN
cana-6247	613	47	)	)	PUNCT
cana-6247	613	48	2.1.1	2.1.1	NUM
cana-6247	613	49	evidence	evidence	NOUN
cana-6247	613	50	from	from	ADP
cana-6247	613	51	recent	recent	ADJ
cana-6247	613	52	research	research	NOUN
cana-6247	613	53	summary	summary	NOUN
cana-6247	613	54	poison	poison	NOUN
cana-6247	613	55	/	/	SYM
cana-6247	613	56	signal	signal	NOUN
cana-6247	613	57	dependent	dependent	ADJ
cana-6247	613	58	noise	noise	NOUN
cana-6247	613	59	4.1	4.1	NUM
cana-6247	613	60	low	low	ADJ
cana-6247	613	61	light	light	ADJ
cana-6247	613	62	image	image	NOUN
cana-6247	613	63	enhancement	enhancement	NOUN
cana-6247	613	64	4.1.1	4.1.1	NUM
cana-6247	613	65	evidence	evidence	NOUN
cana-6247	613	66	from	from	ADP
cana-6247	613	67	research	research	NOUN
cana-6247	613	68	summary	summary	NOUN