id	sid	tid	token	lemma	pos
ajst-12267	1	1	academic	academic	ADJ
ajst-12267	1	2	journal	journal	NOUN
ajst-12267	1	3	of	of	ADP
ajst-12267	1	4	science	science	NOUN
ajst-12267	1	5	and	and	CCONJ
ajst-12267	1	6	technology	technology	NOUN
ajst-12267	1	7	issn	issn	NOUN
ajst-12267	1	8	:	:	PUNCT
ajst-12267	1	9	2771	2771	NUM
ajst-12267	1	10	-	-	SYM
ajst-12267	1	11	3032	3032	NUM
ajst-12267	1	12	|	|	NOUN
ajst-12267	1	13	vol	vol	NOUN
ajst-12267	1	14	.	.	PROPN
ajst-12267	2	1	7	7	NUM
ajst-12267	2	2	,	,	PUNCT
ajst-12267	2	3	no	no	INTJ
ajst-12267	2	4	.	.	NOUN
ajst-12267	2	5	2	2	NUM
ajst-12267	2	6	,	,	PUNCT
ajst-12267	2	7	2023	2023	NUM
ajst-12267	2	8	183	183	NUM
ajst-12267	2	9	lung	lung	NOUN
ajst-12267	2	10	nodule	nodule	NOUN
ajst-12267	2	11	data	data	PROPN
ajst-12267	2	12	enhancement	enhancement	NOUN
ajst-12267	2	13	algorithm	algorithm	NOUN
ajst-12267	2	14	based	base	VERB
ajst-12267	2	15	on	on	ADP
ajst-12267	2	16	generative	generative	ADJ
ajst-12267	2	17	adversarial	adversarial	ADJ
ajst-12267	2	18	network	network	NOUN
ajst-12267	2	19	yunjiao	yunjiao	PROPN
ajst-12267	2	20	ma	ma	PROPN
ajst-12267	2	21	school	school	PROPN
ajst-12267	2	22	of	of	ADP
ajst-12267	2	23	information	information	NOUN
ajst-12267	2	24	,	,	PUNCT
ajst-12267	2	25	yunnan	yunnan	PROPN
ajst-12267	2	26	university	university	PROPN
ajst-12267	2	27	of	of	ADP
ajst-12267	2	28	finance	finance	NOUN
ajst-12267	2	29	and	and	CCONJ
ajst-12267	2	30	economics	economic	NOUN
ajst-12267	2	31	,	,	PUNCT
ajst-12267	2	32	kunming	kunming	PROPN
ajst-12267	2	33	yunnan	yunnan	PROPN
ajst-12267	2	34	,	,	PUNCT
ajst-12267	2	35	650221	650221	NUM
ajst-12267	2	36	,	,	PUNCT
ajst-12267	2	37	china	china	PROPN
ajst-12267	2	38	abstract	abstract	NOUN
ajst-12267	2	39	:	:	PUNCT
ajst-12267	2	40	due	due	ADP
ajst-12267	2	41	to	to	ADP
ajst-12267	2	42	the	the	DET
ajst-12267	2	43	strong	strong	ADJ
ajst-12267	2	44	privacy	privacy	NOUN
ajst-12267	2	45	of	of	ADP
ajst-12267	2	46	data	datum	NOUN
ajst-12267	2	47	and	and	CCONJ
ajst-12267	2	48	the	the	DET
ajst-12267	2	49	difficulty	difficulty	NOUN
ajst-12267	2	50	of	of	ADP
ajst-12267	2	51	annotation	annotation	NOUN
ajst-12267	2	52	,	,	PUNCT
ajst-12267	2	53	the	the	DET
ajst-12267	2	54	amount	amount	NOUN
ajst-12267	2	55	of	of	ADP
ajst-12267	2	56	medical	medical	ADJ
ajst-12267	2	57	image	image	NOUN
ajst-12267	2	58	data	datum	NOUN
ajst-12267	2	59	is	be	AUX
ajst-12267	2	60	relatively	relatively	ADV
ajst-12267	2	61	small	small	ADJ
ajst-12267	2	62	,	,	PUNCT
ajst-12267	2	63	which	which	PRON
ajst-12267	2	64	further	far	ADV
ajst-12267	2	65	affects	affect	VERB
ajst-12267	2	66	the	the	DET
ajst-12267	2	67	effect	effect	NOUN
ajst-12267	2	68	of	of	ADP
ajst-12267	2	69	deep	deep	ADJ
ajst-12267	2	70	learning	learning	NOUN
ajst-12267	2	71	.	.	PUNCT
ajst-12267	3	1	using	use	VERB
ajst-12267	3	2	data	datum	NOUN
ajst-12267	3	3	enhancement	enhancement	NOUN
ajst-12267	3	4	technology	technology	NOUN
ajst-12267	3	5	to	to	PART
ajst-12267	3	6	expand	expand	VERB
ajst-12267	3	7	existing	exist	VERB
ajst-12267	3	8	data	datum	NOUN
ajst-12267	3	9	sets	set	NOUN
ajst-12267	3	10	can	can	AUX
ajst-12267	3	11	significantly	significantly	ADV
ajst-12267	3	12	alleviate	alleviate	VERB
ajst-12267	3	13	the	the	DET
ajst-12267	3	14	problem	problem	NOUN
ajst-12267	3	15	of	of	ADP
ajst-12267	3	16	insufficient	insufficient	ADJ
ajst-12267	3	17	data	datum	NOUN
ajst-12267	3	18	.	.	PUNCT
ajst-12267	4	1	deep	deep	ADJ
ajst-12267	4	2	convolutional	convolutional	ADJ
ajst-12267	4	3	generative	generative	ADJ
ajst-12267	4	4	adversarial	adversarial	ADJ
ajst-12267	4	5	network	network	NOUN
ajst-12267	4	6	(	(	PUNCT
ajst-12267	4	7	dcgan	dcgan	NOUN
ajst-12267	4	8	)	)	PUNCT
ajst-12267	4	9	technology	technology	NOUN
ajst-12267	4	10	has	have	AUX
ajst-12267	4	11	been	be	AUX
ajst-12267	4	12	widely	widely	ADV
ajst-12267	4	13	used	use	VERB
ajst-12267	4	14	in	in	ADP
ajst-12267	4	15	the	the	DET
ajst-12267	4	16	field	field	NOUN
ajst-12267	4	17	of	of	ADP
ajst-12267	4	18	medical	medical	ADJ
ajst-12267	4	19	image	image	NOUN
ajst-12267	4	20	data	datum	NOUN
ajst-12267	4	21	enhancement	enhancement	NOUN
ajst-12267	4	22	,	,	PUNCT
ajst-12267	4	23	but	but	CCONJ
ajst-12267	4	24	there	there	PRON
ajst-12267	4	25	are	be	VERB
ajst-12267	4	26	still	still	ADV
ajst-12267	4	27	some	some	DET
ajst-12267	4	28	problems	problem	NOUN
ajst-12267	4	29	such	such	ADJ
ajst-12267	4	30	as	as	ADP
ajst-12267	4	31	unstable	unstable	ADJ
ajst-12267	4	32	training	training	NOUN
ajst-12267	4	33	,	,	PUNCT
ajst-12267	4	34	difficulty	difficulty	NOUN
ajst-12267	4	35	in	in	ADP
ajst-12267	4	36	convergence	convergence	NOUN
ajst-12267	4	37	,	,	PUNCT
ajst-12267	4	38	easy	easy	ADJ
ajst-12267	4	39	to	to	PART
ajst-12267	4	40	produce	produce	VERB
ajst-12267	4	41	mode	mode	NOUN
ajst-12267	4	42	collapse	collapse	NOUN
ajst-12267	4	43	and	and	CCONJ
ajst-12267	4	44	insufficient	insufficient	ADJ
ajst-12267	4	45	quality	quality	NOUN
ajst-12267	4	46	of	of	ADP
ajst-12267	4	47	the	the	DET
ajst-12267	4	48	generated	generate	VERB
ajst-12267	4	49	images	image	NOUN
ajst-12267	4	50	.	.	PUNCT
ajst-12267	5	1	in	in	ADP
ajst-12267	5	2	this	this	DET
ajst-12267	5	3	paper	paper	NOUN
ajst-12267	5	4	,	,	PUNCT
ajst-12267	5	5	the	the	DET
ajst-12267	5	6	mode	mode	NOUN
ajst-12267	5	7	collapse	collapse	NOUN
ajst-12267	5	8	problem	problem	NOUN
ajst-12267	5	9	of	of	ADP
ajst-12267	5	10	deep	deep	ADJ
ajst-12267	5	11	convolutional	convolutional	ADJ
ajst-12267	5	12	generative	generative	ADJ
ajst-12267	5	13	adversarial	adversarial	ADJ
ajst-12267	5	14	network	network	NOUN
ajst-12267	5	15	(	(	PUNCT
ajst-12267	5	16	dcgan	dcgan	NOUN
ajst-12267	5	17	)	)	PUNCT
ajst-12267	5	18	is	be	AUX
ajst-12267	5	19	improved	improve	VERB
ajst-12267	5	20	by	by	ADP
ajst-12267	5	21	using	use	VERB
ajst-12267	5	22	the	the	DET
ajst-12267	5	23	fusion	fusion	NOUN
ajst-12267	5	24	attention	attention	NOUN
ajst-12267	5	25	mechanism	mechanism	NOUN
ajst-12267	5	26	and	and	CCONJ
ajst-12267	5	27	modifying	modify	VERB
ajst-12267	5	28	the	the	DET
ajst-12267	5	29	loss	loss	NOUN
ajst-12267	5	30	function	function	NOUN
ajst-12267	5	31	.	.	PUNCT
ajst-12267	6	1	experiments	experiment	NOUN
ajst-12267	6	2	show	show	VERB
ajst-12267	6	3	that	that	SCONJ
ajst-12267	6	4	the	the	DET
ajst-12267	6	5	method	method	NOUN
ajst-12267	6	6	proposed	propose	VERB
ajst-12267	6	7	in	in	ADP
ajst-12267	6	8	this	this	DET
ajst-12267	6	9	paper	paper	NOUN
ajst-12267	6	10	can	can	AUX
ajst-12267	6	11	effectively	effectively	ADV
ajst-12267	6	12	improve	improve	VERB
ajst-12267	6	13	the	the	DET
ajst-12267	6	14	enhancement	enhancement	NOUN
ajst-12267	6	15	effect	effect	NOUN
ajst-12267	6	16	of	of	ADP
ajst-12267	6	17	lung	lung	NOUN
ajst-12267	6	18	nodule	nodule	NOUN
ajst-12267	6	19	image	image	NOUN
ajst-12267	6	20	data	datum	NOUN
ajst-12267	6	21	.	.	PUNCT
ajst-12267	7	1	keywords	keyword	NOUN
ajst-12267	7	2	:	:	PUNCT
ajst-12267	7	3	data	datum	NOUN
ajst-12267	7	4	enhancement	enhancement	NOUN
ajst-12267	7	5	,	,	PUNCT
ajst-12267	7	6	pulmonary	pulmonary	ADJ
ajst-12267	7	7	nodules	nodule	NOUN
ajst-12267	7	8	,	,	PUNCT
ajst-12267	7	9	generation	generation	NOUN
ajst-12267	7	10	of	of	ADP
ajst-12267	7	11	adversarial	adversarial	ADJ
ajst-12267	7	12	networks	network	NOUN
ajst-12267	7	13	.	.	PUNCT
ajst-12267	8	1	1	1	X
ajst-12267	8	2	.	.	X
ajst-12267	8	3	introduction	introduction	NOUN
ajst-12267	8	4	at	at	ADP
ajst-12267	8	5	present	present	ADJ
ajst-12267	8	6	,	,	PUNCT
ajst-12267	8	7	the	the	DET
ajst-12267	8	8	interpretation	interpretation	NOUN
ajst-12267	8	9	of	of	ADP
ajst-12267	8	10	ct	ct	NUM
ajst-12267	8	11	images	image	NOUN
ajst-12267	8	12	of	of	ADP
ajst-12267	8	13	pulmonary	pulmonary	ADJ
ajst-12267	8	14	nodules	nodule	NOUN
ajst-12267	8	15	is	be	AUX
ajst-12267	8	16	mainly	mainly	ADV
ajst-12267	8	17	done	do	VERB
ajst-12267	8	18	by	by	ADP
ajst-12267	8	19	doctors	doctor	NOUN
ajst-12267	8	20	.	.	PUNCT
ajst-12267	9	1	however	however	ADV
ajst-12267	9	2	,	,	PUNCT
ajst-12267	9	3	due	due	ADP
ajst-12267	9	4	to	to	ADP
ajst-12267	9	5	the	the	DET
ajst-12267	9	6	different	different	ADJ
ajst-12267	9	7	personal	personal	ADJ
ajst-12267	9	8	experience	experience	NOUN
ajst-12267	9	9	of	of	ADP
ajst-12267	9	10	doctors	doctor	NOUN
ajst-12267	9	11	and	and	CCONJ
ajst-12267	9	12	the	the	DET
ajst-12267	9	13	fatigue	fatigue	NOUN
ajst-12267	9	14	of	of	ADP
ajst-12267	9	15	doctors	doctor	NOUN
ajst-12267	9	16	caused	cause	VERB
ajst-12267	9	17	by	by	ADP
ajst-12267	9	18	a	a	DET
ajst-12267	9	19	large	large	ADJ
ajst-12267	9	20	number	number	NOUN
ajst-12267	9	21	of	of	ADP
ajst-12267	9	22	film	film	NOUN
ajst-12267	9	23	reading	reading	NOUN
ajst-12267	9	24	,	,	PUNCT
ajst-12267	9	25	the	the	DET
ajst-12267	9	26	interpretation	interpretation	NOUN
ajst-12267	9	27	of	of	ADP
ajst-12267	9	28	ct	ct	NUM
ajst-12267	9	29	images	image	NOUN
ajst-12267	9	30	has	have	VERB
ajst-12267	9	31	some	some	DET
ajst-12267	9	32	problems	problem	NOUN
ajst-12267	9	33	such	such	ADJ
ajst-12267	9	34	as	as	ADP
ajst-12267	9	35	low	low	ADJ
ajst-12267	9	36	efficiency	efficiency	NOUN
ajst-12267	9	37	and	and	CCONJ
ajst-12267	9	38	uneven	uneven	ADJ
ajst-12267	9	39	accuracy	accuracy	NOUN
ajst-12267	9	40	level	level	NOUN
ajst-12267	9	41	.	.	PUNCT
ajst-12267	10	1	it	it	PRON
ajst-12267	10	2	is	be	AUX
ajst-12267	10	3	urgent	urgent	ADJ
ajst-12267	10	4	to	to	PART
ajst-12267	10	5	interpret	interpret	VERB
ajst-12267	10	6	ct	ct	NUM
ajst-12267	10	7	images	image	NOUN
ajst-12267	10	8	with	with	ADP
ajst-12267	10	9	computer	computer	NOUN
ajst-12267	10	10	-	-	PUNCT
ajst-12267	10	11	aided	aid	VERB
ajst-12267	10	12	diagnostic	diagnostic	ADJ
ajst-12267	10	13	cad	cad	NOUN
ajst-12267	10	14	.	.	PUNCT
ajst-12267	11	1	computer	computer	NOUN
ajst-12267	11	2	aided	aid	VERB
ajst-12267	11	3	diagnosis	diagnosis	NOUN
ajst-12267	11	4	(	(	PUNCT
ajst-12267	11	5	cad	cad	NOUN
ajst-12267	11	6	)	)	PUNCT
ajst-12267	11	7	is	be	AUX
ajst-12267	11	8	of	of	ADP
ajst-12267	11	9	great	great	ADJ
ajst-12267	11	10	help	help	NOUN
ajst-12267	11	11	to	to	ADP
ajst-12267	11	12	the	the	DET
ajst-12267	11	13	reading	reading	NOUN
ajst-12267	11	14	of	of	ADP
ajst-12267	11	15	lung	lung	NOUN
ajst-12267	11	16	ct	ct	NUM
ajst-12267	11	17	images	image	NOUN
ajst-12267	11	18	,	,	PUNCT
ajst-12267	11	19	which	which	PRON
ajst-12267	11	20	can	can	AUX
ajst-12267	11	21	relieve	relieve	VERB
ajst-12267	11	22	the	the	DET
ajst-12267	11	23	pressure	pressure	NOUN
ajst-12267	11	24	of	of	ADP
ajst-12267	11	25	doctors	doctor	NOUN
ajst-12267	11	26	to	to	PART
ajst-12267	11	27	read	read	VERB
ajst-12267	11	28	the	the	DET
ajst-12267	11	29	images	image	NOUN
ajst-12267	11	30	and	and	CCONJ
ajst-12267	11	31	guarantee	guarantee	VERB
ajst-12267	11	32	the	the	DET
ajst-12267	11	33	accuracy	accuracy	NOUN
ajst-12267	11	34	to	to	ADP
ajst-12267	11	35	some	some	DET
ajst-12267	11	36	extent	extent	NOUN
ajst-12267	11	37	.	.	PUNCT
ajst-12267	12	1	at	at	ADP
ajst-12267	12	2	present	present	ADJ
ajst-12267	12	3	,	,	PUNCT
ajst-12267	12	4	deep	deep	ADJ
ajst-12267	12	5	learning	learning	NOUN
ajst-12267	12	6	technology	technology	NOUN
ajst-12267	12	7	has	have	AUX
ajst-12267	12	8	been	be	AUX
ajst-12267	12	9	widely	widely	ADV
ajst-12267	12	10	used	use	VERB
ajst-12267	12	11	in	in	ADP
ajst-12267	12	12	pulmonary	pulmonary	ADJ
ajst-12267	12	13	cad	cad	NOUN
ajst-12267	12	14	systems	system	NOUN
ajst-12267	12	15	,	,	PUNCT
ajst-12267	12	16	but	but	CCONJ
ajst-12267	12	17	a	a	DET
ajst-12267	12	18	large	large	ADJ
ajst-12267	12	19	amount	amount	NOUN
ajst-12267	12	20	of	of	ADP
ajst-12267	12	21	labeled	label	VERB
ajst-12267	12	22	data	datum	NOUN
ajst-12267	12	23	is	be	AUX
ajst-12267	12	24	required	require	VERB
ajst-12267	12	25	[	[	X
ajst-12267	12	26	1	1	NUM
ajst-12267	12	27	]	]	PUNCT
ajst-12267	12	28	.	.	PUNCT
ajst-12267	13	1	due	due	ADP
ajst-12267	13	2	to	to	ADP
ajst-12267	13	3	the	the	DET
ajst-12267	13	4	strong	strong	ADJ
ajst-12267	13	5	privacy	privacy	NOUN
ajst-12267	13	6	of	of	ADP
ajst-12267	13	7	patients	patient	NOUN
ajst-12267	13	8	and	and	CCONJ
ajst-12267	13	9	the	the	DET
ajst-12267	13	10	lack	lack	NOUN
ajst-12267	13	11	of	of	ADP
ajst-12267	13	12	professional	professional	ADJ
ajst-12267	13	13	doctors	doctor	NOUN
ajst-12267	13	14	to	to	PART
ajst-12267	13	15	label	label	VERB
ajst-12267	13	16	the	the	DET
ajst-12267	13	17	data	datum	NOUN
ajst-12267	13	18	,	,	PUNCT
ajst-12267	13	19	there	there	PRON
ajst-12267	13	20	are	be	VERB
ajst-12267	13	21	few	few	ADJ
ajst-12267	13	22	datasets	dataset	NOUN
ajst-12267	13	23	available	available	ADJ
ajst-12267	13	24	for	for	ADP
ajst-12267	13	25	deep	deep	ADJ
ajst-12267	13	26	learning	learning	NOUN
ajst-12267	13	27	in	in	ADP
ajst-12267	13	28	medical	medical	ADJ
ajst-12267	13	29	images	image	NOUN
ajst-12267	13	30	.	.	PUNCT
ajst-12267	14	1	transfer	transfer	NOUN
ajst-12267	14	2	learning	learning	NOUN
ajst-12267	14	3	and	and	CCONJ
ajst-12267	14	4	data	datum	NOUN
ajst-12267	14	5	enhancement	enhancement	NOUN
ajst-12267	14	6	are	be	AUX
ajst-12267	14	7	two	two	NUM
ajst-12267	14	8	commonly	commonly	ADV
ajst-12267	14	9	used	use	VERB
ajst-12267	14	10	methods	method	NOUN
ajst-12267	14	11	to	to	PART
ajst-12267	14	12	solve	solve	VERB
ajst-12267	14	13	the	the	DET
ajst-12267	14	14	shortage	shortage	NOUN
ajst-12267	14	15	of	of	ADP
ajst-12267	14	16	training	training	NOUN
ajst-12267	14	17	data	datum	NOUN
ajst-12267	14	18	sets	set	NOUN
ajst-12267	14	19	[	[	X
ajst-12267	14	20	2	2	NUM
ajst-12267	14	21	]	]	PUNCT
ajst-12267	14	22	.	.	PUNCT
ajst-12267	15	1	generative	generative	ADJ
ajst-12267	15	2	adversarial	adversarial	ADJ
ajst-12267	15	3	network	network	NOUN
ajst-12267	15	4	(	(	PUNCT
ajst-12267	15	5	gan	gan	PROPN
ajst-12267	15	6	)	)	PUNCT
ajst-12267	15	7	has	have	VERB
ajst-12267	15	8	excellent	excellent	ADJ
ajst-12267	15	9	image	image	NOUN
ajst-12267	15	10	generation	generation	NOUN
ajst-12267	15	11	capability	capability	NOUN
ajst-12267	15	12	,	,	PUNCT
ajst-12267	15	13	so	so	ADV
ajst-12267	15	14	gan	gin	VERB
ajst-12267	15	15	as	as	ADP
ajst-12267	15	16	a	a	DET
ajst-12267	15	17	data	data	NOUN
ajst-12267	15	18	enhancement	enhancement	NOUN
ajst-12267	15	19	method	method	NOUN
ajst-12267	15	20	has	have	AUX
ajst-12267	15	21	been	be	AUX
ajst-12267	15	22	applied	apply	VERB
ajst-12267	15	23	in	in	ADP
ajst-12267	15	24	biomedical	biomedical	ADJ
ajst-12267	15	25	image	image	NOUN
ajst-12267	15	26	analysis	analysis	NOUN
ajst-12267	15	27	applications	application	NOUN
ajst-12267	15	28	[	[	X
ajst-12267	15	29	3	3	NUM
ajst-12267	15	30	]	]	PUNCT
ajst-12267	15	31	.	.	PUNCT
ajst-12267	16	1	however	however	ADV
ajst-12267	16	2	,	,	PUNCT
ajst-12267	16	3	gan	gan	PROPN
ajst-12267	16	4	image	image	NOUN
ajst-12267	16	5	data	datum	NOUN
ajst-12267	16	6	enhancement	enhancement	NOUN
ajst-12267	16	7	technology	technology	NOUN
ajst-12267	16	8	has	have	VERB
ajst-12267	16	9	problems	problem	NOUN
ajst-12267	16	10	such	such	ADJ
ajst-12267	16	11	as	as	ADP
ajst-12267	16	12	unstable	unstable	ADJ
ajst-12267	16	13	model	model	NOUN
ajst-12267	16	14	convergence	convergence	NOUN
ajst-12267	16	15	and	and	CCONJ
ajst-12267	16	16	low	low	ADJ
ajst-12267	16	17	quality	quality	NOUN
ajst-12267	16	18	of	of	ADP
ajst-12267	16	19	generated	generate	VERB
ajst-12267	16	20	images	image	NOUN
ajst-12267	16	21	[	[	X
ajst-12267	16	22	4	4	NUM
ajst-12267	16	23	]	]	PUNCT
ajst-12267	16	24	.	.	PUNCT
ajst-12267	17	1	in	in	ADP
ajst-12267	17	2	this	this	DET
ajst-12267	17	3	paper	paper	NOUN
ajst-12267	17	4	,	,	PUNCT
ajst-12267	17	5	the	the	DET
ajst-12267	17	6	improved	improved	ADJ
ajst-12267	17	7	gan	gan	PROPN
ajst-12267	17	8	model	model	NOUN
ajst-12267	17	9	is	be	AUX
ajst-12267	17	10	used	use	VERB
ajst-12267	17	11	to	to	PART
ajst-12267	17	12	generate	generate	VERB
ajst-12267	17	13	lung	lung	NOUN
ajst-12267	17	14	nodule	nodule	NOUN
ajst-12267	17	15	images	image	NOUN
ajst-12267	17	16	to	to	PART
ajst-12267	17	17	achieve	achieve	VERB
ajst-12267	17	18	the	the	DET
ajst-12267	17	19	purpose	purpose	NOUN
ajst-12267	17	20	of	of	ADP
ajst-12267	17	21	enhancing	enhance	VERB
ajst-12267	17	22	lung	lung	NOUN
ajst-12267	17	23	nodule	nodule	NOUN
ajst-12267	17	24	data	datum	NOUN
ajst-12267	17	25	.	.	PUNCT
ajst-12267	18	1	the	the	DET
ajst-12267	18	2	main	main	ADJ
ajst-12267	18	3	work	work	NOUN
ajst-12267	18	4	of	of	ADP
ajst-12267	18	5	this	this	DET
ajst-12267	18	6	paper	paper	NOUN
ajst-12267	18	7	is	be	AUX
ajst-12267	18	8	as	as	SCONJ
ajst-12267	18	9	follows	follow	VERB
ajst-12267	18	10	:	:	PUNCT
ajst-12267	18	11	(	(	PUNCT
ajst-12267	18	12	1	1	X
ajst-12267	18	13	)	)	PUNCT
ajst-12267	18	14	by	by	ADP
ajst-12267	18	15	improving	improve	VERB
ajst-12267	18	16	the	the	DET
ajst-12267	18	17	loss	loss	NOUN
ajst-12267	18	18	function	function	NOUN
ajst-12267	18	19	,	,	PUNCT
ajst-12267	18	20	using	use	VERB
ajst-12267	18	21	w	w	PROPN
ajst-12267	18	22	distance	distance	NOUN
ajst-12267	18	23	instead	instead	ADV
ajst-12267	18	24	of	of	ADP
ajst-12267	18	25	js	js	ADJ
ajst-12267	18	26	divergence	divergence	NOUN
ajst-12267	18	27	and	and	CCONJ
ajst-12267	18	28	adding	add	VERB
ajst-12267	18	29	regularization	regularization	NOUN
ajst-12267	18	30	terms	term	NOUN
ajst-12267	18	31	to	to	ADP
ajst-12267	18	32	the	the	DET
ajst-12267	18	33	loss	loss	NOUN
ajst-12267	18	34	function	function	NOUN
ajst-12267	18	35	,	,	PUNCT
ajst-12267	18	36	this	this	DET
ajst-12267	18	37	paper	paper	NOUN
ajst-12267	18	38	effectively	effectively	ADV
ajst-12267	18	39	improves	improve	VERB
ajst-12267	18	40	the	the	DET
ajst-12267	18	41	pattern	pattern	NOUN
ajst-12267	18	42	collapse	collapse	NOUN
ajst-12267	18	43	problem	problem	NOUN
ajst-12267	18	44	that	that	PRON
ajst-12267	18	45	is	be	AUX
ajst-12267	18	46	prone	prone	ADJ
ajst-12267	18	47	to	to	PART
ajst-12267	18	48	occur	occur	VERB
ajst-12267	18	49	in	in	ADP
ajst-12267	18	50	dcgan	dcgan	NOUN
ajst-12267	18	51	generated	generate	VERB
ajst-12267	18	52	images	image	NOUN
ajst-12267	18	53	.	.	PUNCT
ajst-12267	19	1	(	(	PUNCT
ajst-12267	19	2	2	2	X
ajst-12267	19	3	)	)	PUNCT
ajst-12267	19	4	add	add	VERB
ajst-12267	19	5	the	the	DET
ajst-12267	19	6	attention	attention	NOUN
ajst-12267	19	7	mechanism	mechanism	NOUN
ajst-12267	19	8	to	to	PART
ajst-12267	19	9	pay	pay	VERB
ajst-12267	19	10	attention	attention	NOUN
ajst-12267	19	11	to	to	ADP
ajst-12267	19	12	the	the	DET
ajst-12267	19	13	generation	generation	NOUN
ajst-12267	19	14	of	of	ADP
ajst-12267	19	15	small	small	ADJ
ajst-12267	19	16	nodules	nodule	NOUN
ajst-12267	19	17	to	to	PART
ajst-12267	19	18	further	far	ADV
ajst-12267	19	19	improve	improve	VERB
ajst-12267	19	20	the	the	DET
ajst-12267	19	21	diversity	diversity	NOUN
ajst-12267	19	22	of	of	ADP
ajst-12267	19	23	generated	generate	VERB
ajst-12267	19	24	images	image	NOUN
ajst-12267	19	25	.	.	PUNCT
ajst-12267	20	1	2	2	X
ajst-12267	20	2	.	.	NUM
ajst-12267	20	3	improved	improve	VERB
ajst-12267	20	4	generative	generative	ADJ
ajst-12267	20	5	adversarial	adversarial	ADJ
ajst-12267	20	6	network	network	NOUN
ajst-12267	20	7	2.1	2.1	NUM
ajst-12267	20	8	.	.	PUNCT
ajst-12267	21	1	generative	generative	ADJ
ajst-12267	21	2	adversarial	adversarial	ADJ
ajst-12267	21	3	network	network	NOUN
ajst-12267	21	4	(	(	PUNCT
ajst-12267	21	5	1	1	NUM
ajst-12267	21	6	)	)	PUNCT
ajst-12267	21	7	generate	generate	VERB
ajst-12267	21	8	adversarial	adversarial	ADJ
ajst-12267	21	9	network	network	NOUN
ajst-12267	21	10	introduction	introduction	NOUN
ajst-12267	21	11	generative	generative	ADJ
ajst-12267	21	12	adversarial	adversarial	ADJ
ajst-12267	21	13	network	network	NOUN
ajst-12267	21	14	(	(	PUNCT
ajst-12267	21	15	gan	gan	PROPN
ajst-12267	21	16	)	)	PUNCT
ajst-12267	22	1	[	[	X
ajst-12267	22	2	5	5	NUM
ajst-12267	22	3	]	]	PUNCT
ajst-12267	22	4	is	be	AUX
ajst-12267	22	5	a	a	DET
ajst-12267	22	6	probabilistic	probabilistic	ADJ
ajst-12267	22	7	generative	generative	ADJ
ajst-12267	22	8	model	model	NOUN
ajst-12267	22	9	proposed	propose	VERB
ajst-12267	22	10	by	by	ADP
ajst-12267	22	11	goodfellow	goodfellow	PROPN
ajst-12267	22	12	et	et	PROPN
ajst-12267	22	13	al	al	PROPN
ajst-12267	22	14	in	in	ADP
ajst-12267	22	15	2014	2014	NUM
ajst-12267	22	16	,	,	PUNCT
ajst-12267	22	17	inspired	inspire	VERB
ajst-12267	22	18	by	by	ADP
ajst-12267	22	19	minimax	minimax	NOUN
ajst-12267	22	20	two	two	NUM
ajst-12267	22	21	-	-	PUNCT
ajst-12267	22	22	player	player	NOUN
ajst-12267	22	23	games	game	NOUN
ajst-12267	22	24	.	.	PUNCT
ajst-12267	23	1	this	this	DET
ajst-12267	23	2	model	model	NOUN
ajst-12267	23	3	uses	use	VERB
ajst-12267	23	4	the	the	DET
ajst-12267	23	5	idea	idea	NOUN
ajst-12267	23	6	of	of	ADP
ajst-12267	23	7	zero	zero	NUM
ajst-12267	23	8	-	-	PUNCT
ajst-12267	23	9	sum	sum	NOUN
ajst-12267	23	10	game	game	NOUN
ajst-12267	23	11	in	in	ADP
ajst-12267	23	12	game	game	NOUN
ajst-12267	23	13	theory	theory	NOUN
ajst-12267	23	14	[	[	X
ajst-12267	23	15	6	6	NUM
ajst-12267	23	16	]	]	PUNCT
ajst-12267	23	17	.	.	PUNCT
ajst-12267	24	1	gan	gan	PROPN
ajst-12267	24	2	is	be	AUX
ajst-12267	24	3	divided	divide	VERB
ajst-12267	24	4	into	into	ADP
ajst-12267	24	5	two	two	NUM
ajst-12267	24	6	parts	part	NOUN
ajst-12267	24	7	,	,	PUNCT
ajst-12267	24	8	namely	namely	ADV
ajst-12267	24	9	generator	generator	VERB
ajst-12267	24	10	part	part	NOUN
ajst-12267	24	11	and	and	CCONJ
ajst-12267	24	12	discriminator	discriminator	ADJ
ajst-12267	24	13	part	part	NOUN
ajst-12267	24	14	.	.	PUNCT
ajst-12267	25	1	the	the	DET
ajst-12267	25	2	input	input	NOUN
ajst-12267	25	3	to	to	ADP
ajst-12267	25	4	the	the	DET
ajst-12267	25	5	generator	generator	NOUN
ajst-12267	25	6	is	be	AUX
ajst-12267	25	7	random	random	ADJ
ajst-12267	25	8	noise	noise	NOUN
ajst-12267	25	9	,	,	PUNCT
ajst-12267	25	10	and	and	CCONJ
ajst-12267	25	11	the	the	DET
ajst-12267	25	12	input	input	NOUN
ajst-12267	25	13	is	be	AUX
ajst-12267	25	14	mapped	map	VERB
ajst-12267	25	15	to	to	ADP
ajst-12267	25	16	a	a	DET
ajst-12267	25	17	hypothetical	hypothetical	ADJ
ajst-12267	25	18	data	datum	NOUN
ajst-12267	25	19	through	through	ADP
ajst-12267	25	20	the	the	DET
ajst-12267	25	21	generator	generator	NOUN
ajst-12267	25	22	to	to	PART
ajst-12267	25	23	generate	generate	VERB
ajst-12267	25	24	sample	sample	NOUN
ajst-12267	25	25	data	datum	NOUN
ajst-12267	25	26	,	,	PUNCT
ajst-12267	25	27	and	and	CCONJ
ajst-12267	25	28	if	if	SCONJ
ajst-12267	25	29	the	the	DET
ajst-12267	25	30	generator	generator	NOUN
ajst-12267	25	31	is	be	AUX
ajst-12267	25	32	good	good	ADJ
ajst-12267	25	33	enough	enough	ADV
ajst-12267	25	34	,	,	PUNCT
ajst-12267	25	35	all	all	DET
ajst-12267	25	36	the	the	DET
ajst-12267	25	37	generators	generator	NOUN
ajst-12267	25	38	will	will	AUX
ajst-12267	25	39	correspond	correspond	VERB
ajst-12267	25	40	to	to	ADP
ajst-12267	25	41	a	a	DET
ajst-12267	25	42	very	very	ADV
ajst-12267	25	43	real	real	ADJ
ajst-12267	25	44	data	datum	NOUN
ajst-12267	25	45	.	.	PUNCT
ajst-12267	26	1	the	the	DET
ajst-12267	26	2	function	function	NOUN
ajst-12267	26	3	of	of	ADP
ajst-12267	26	4	the	the	DET
ajst-12267	26	5	discriminator	discriminator	NOUN
ajst-12267	26	6	is	be	AUX
ajst-12267	26	7	to	to	PART
ajst-12267	26	8	determine	determine	VERB
ajst-12267	26	9	the	the	DET
ajst-12267	26	10	authenticity	authenticity	NOUN
ajst-12267	26	11	of	of	ADP
ajst-12267	26	12	the	the	DET
ajst-12267	26	13	generated	generate	VERB
ajst-12267	26	14	sample	sample	NOUN
ajst-12267	26	15	data	datum	NOUN
ajst-12267	26	16	relative	relative	ADJ
ajst-12267	26	17	to	to	ADP
ajst-12267	26	18	the	the	DET
ajst-12267	26	19	real	real	ADJ
ajst-12267	26	20	sample	sample	NOUN
ajst-12267	26	21	data	datum	NOUN
ajst-12267	26	22	.	.	PUNCT
ajst-12267	27	1	during	during	ADP
ajst-12267	27	2	the	the	DET
ajst-12267	27	3	training	training	NOUN
ajst-12267	27	4	process	process	NOUN
ajst-12267	27	5	,	,	PUNCT
ajst-12267	27	6	the	the	DET
ajst-12267	27	7	discriminator	discriminator	NOUN
ajst-12267	27	8	and	and	CCONJ
ajst-12267	27	9	generator	generator	NOUN
ajst-12267	27	10	take	take	VERB
ajst-12267	27	11	turns	turn	NOUN
ajst-12267	27	12	to	to	PART
ajst-12267	27	13	upgrade	upgrade	VERB
ajst-12267	27	14	,	,	PUNCT
ajst-12267	27	15	competing	compete	VERB
ajst-12267	27	16	with	with	ADP
ajst-12267	27	17	each	each	DET
ajst-12267	27	18	other	other	ADJ
ajst-12267	27	19	to	to	PART
ajst-12267	27	20	continuously	continuously	ADV
ajst-12267	27	21	optimize	optimize	VERB
ajst-12267	27	22	the	the	DET
ajst-12267	27	23	capabilities	capability	NOUN
ajst-12267	27	24	of	of	ADP
ajst-12267	27	25	the	the	DET
ajst-12267	27	26	generator	generator	NOUN
ajst-12267	27	27	.	.	PUNCT
ajst-12267	28	1	when	when	SCONJ
ajst-12267	28	2	the	the	DET
ajst-12267	28	3	discriminator	discriminator	NOUN
ajst-12267	28	4	can	can	AUX
ajst-12267	28	5	not	not	PART
ajst-12267	28	6	identify	identify	VERB
ajst-12267	28	7	the	the	DET
ajst-12267	28	8	true	true	ADJ
ajst-12267	28	9	data	datum	NOUN
ajst-12267	28	10	,	,	PUNCT
ajst-12267	28	11	it	it	PRON
ajst-12267	28	12	means	mean	VERB
ajst-12267	28	13	that	that	SCONJ
ajst-12267	28	14	the	the	DET
ajst-12267	28	15	generator	generator	NOUN
ajst-12267	28	16	has	have	AUX
ajst-12267	28	17	generated	generate	VERB
ajst-12267	28	18	data	datum	NOUN
ajst-12267	28	19	that	that	PRON
ajst-12267	28	20	is	be	AUX
ajst-12267	28	21	closer	close	ADJ
ajst-12267	28	22	to	to	ADP
ajst-12267	28	23	the	the	DET
ajst-12267	28	24	real	real	ADJ
ajst-12267	28	25	data	datum	NOUN
ajst-12267	28	26	.	.	PUNCT
ajst-12267	29	1	loss	loss	NOUN
ajst-12267	29	2	function	function	NOUN
ajst-12267	29	3	is	be	AUX
ajst-12267	29	4	the	the	DET
ajst-12267	29	5	core	core	NOUN
ajst-12267	29	6	of	of	ADP
ajst-12267	29	7	gan	gan	PROPN
ajst-12267	29	8	design	design	NOUN
ajst-12267	29	9	[	[	X
ajst-12267	29	10	7	7	NUM
ajst-12267	29	11	]	]	PUNCT
ajst-12267	29	12	,	,	PUNCT
ajst-12267	29	13	and	and	CCONJ
ajst-12267	29	14	its	its	PRON
ajst-12267	29	15	expression	expression	NOUN
ajst-12267	29	16	is	be	AUX
ajst-12267	29	17	as	as	SCONJ
ajst-12267	29	18	follows	follow	VERB
ajst-12267	29	19	:	:	PUNCT
ajst-12267	29	20	(	(	PUNCT
ajst-12267	29	21	,	,	PUNCT
ajst-12267	29	22	)	)	PUNCT
ajst-12267	29	23	(	(	PUNCT
ajst-12267	29	24	log	log	NOUN
ajst-12267	29	25	)	)	PUNCT
ajst-12267	30	1	[	[	X
ajst-12267	30	2	log(1	log(1	NOUN
ajst-12267	30	3	(	(	PUNCT
ajst-12267	30	4	(	(	PUNCT
ajst-12267	30	5	)	)	PUNCT
ajst-12267	30	6	)	)	PUNCT
ajst-12267	30	7	)	)	PUNCT
ajst-12267	30	8	]	]	X
ajst-12267	30	9	dl	dl	PROPN
ajst-12267	30	10	g	g	PROPN
ajst-12267	30	11	d	d	PROPN
ajst-12267	30	12	e	e	X
ajst-12267	30	13	d	d	X
ajst-12267	30	14	e	e	X
ajst-12267	30	15	d	d	X
ajst-12267	30	16	g	g	PROPN
ajst-12267	30	17	p	p	ADV
ajst-12267	30	18			ADV
ajst-12267	30	19			X
ajst-12267	30	20			X
ajst-12267	30	21	(	(	PUNCT
ajst-12267	30	22	1	1	NUM
ajst-12267	30	23	)	)	PUNCT
ajst-12267	30	24	where	where	SCONJ
ajst-12267	30	25	,	,	PUNCT
ajst-12267	30	26	g	g	PROPN
ajst-12267	30	27	represents	represent	VERB
ajst-12267	30	28	the	the	DET
ajst-12267	30	29	generator	generator	NOUN
ajst-12267	30	30	,	,	PUNCT
ajst-12267	30	31	d	d	PRON
ajst-12267	30	32	represents	represent	VERB
ajst-12267	30	33	the	the	DET
ajst-12267	30	34	discriminator	discriminator	NOUN
ajst-12267	30	35	,	,	PUNCT
ajst-12267	30	36	d	d	PRON
ajst-12267	30	37	represents	represent	VERB
ajst-12267	30	38	the	the	DET
ajst-12267	30	39	real	real	ADJ
ajst-12267	30	40	data	datum	NOUN
ajst-12267	30	41	,	,	PUNCT
ajst-12267	30	42	and	and	CCONJ
ajst-12267	30	43	dd	dd	NOUN
ajst-12267	30	44	represents	represent	VERB
ajst-12267	30	45	the	the	DET
ajst-12267	30	46	discriminant	discriminant	ADJ
ajst-12267	30	47	result	result	NOUN
ajst-12267	30	48	of	of	ADP
ajst-12267	30	49	the	the	DET
ajst-12267	30	50	discriminator	discriminator	NOUN
ajst-12267	30	51	on	on	ADP
ajst-12267	30	52	the	the	DET
ajst-12267	30	53	real	real	ADJ
ajst-12267	30	54	data	datum	NOUN
ajst-12267	30	55	.	.	PUNCT
ajst-12267	31	1	the	the	DET
ajst-12267	31	2	goal	goal	NOUN
ajst-12267	31	3	is	be	AUX
ajst-12267	31	4	to	to	PART
ajst-12267	31	5	generate	generate	VERB
ajst-12267	31	6	data	datum	NOUN
ajst-12267	31	7	that	that	PRON
ajst-12267	31	8	is	be	AUX
ajst-12267	31	9	consistent	consistent	ADJ
ajst-12267	31	10	with	with	ADP
ajst-12267	31	11	real	real	ADJ
ajst-12267	31	12	data	datum	NOUN
ajst-12267	31	13	.	.	PUNCT
ajst-12267	32	1	(	(	PUNCT
ajst-12267	32	2	2	2	X
ajst-12267	32	3	)	)	PUNCT
ajst-12267	32	4	disadvantages	disadvantage	NOUN
ajst-12267	32	5	of	of	ADP
ajst-12267	32	6	generating	generate	VERB
ajst-12267	32	7	adversarial	adversarial	ADJ
ajst-12267	32	8	networks	network	NOUN
ajst-12267	32	9	gan	gin	VERB
ajst-12267	32	10	is	be	AUX
ajst-12267	32	11	a	a	DET
ajst-12267	32	12	generative	generative	ADJ
ajst-12267	32	13	model	model	NOUN
ajst-12267	32	14	.	.	PUNCT
ajst-12267	33	1	compared	compare	VERB
ajst-12267	33	2	with	with	ADP
ajst-12267	33	3	other	other	ADJ
ajst-12267	33	4	generative	generative	ADJ
ajst-12267	33	5	models	model	NOUN
ajst-12267	33	6	,	,	PUNCT
ajst-12267	33	7	which	which	PRON
ajst-12267	33	8	only	only	ADV
ajst-12267	33	9	use	use	VERB
ajst-12267	33	10	backpropagation	backpropagation	NOUN
ajst-12267	33	11	instead	instead	ADV
ajst-12267	33	12	of	of	ADP
ajst-12267	33	13	complex	complex	ADJ
ajst-12267	33	14	markov	markov	NOUN
ajst-12267	33	15	chains	chain	NOUN
ajst-12267	33	16	,	,	PUNCT
ajst-12267	33	17	it	it	PRON
ajst-12267	33	18	can	can	AUX
ajst-12267	33	19	generate	generate	VERB
ajst-12267	33	20	clearer	clear	ADJ
ajst-12267	33	21	and	and	CCONJ
ajst-12267	33	22	more	more	ADV
ajst-12267	33	23	authentic	authentic	ADJ
ajst-12267	33	24	samples	sample	NOUN
ajst-12267	33	25	.	.	PUNCT
ajst-12267	34	1	moreover	moreover	ADV
ajst-12267	34	2	,	,	PUNCT
ajst-12267	34	3	it	it	PRON
ajst-12267	34	4	adopts	adopt	VERB
ajst-12267	34	5	unsupervised	unsupervised	ADJ
ajst-12267	34	6	learning	learning	NOUN
ajst-12267	34	7	training	training	NOUN
ajst-12267	34	8	,	,	PUNCT
ajst-12267	34	9	so	so	CCONJ
ajst-12267	34	10	it	it	PRON
ajst-12267	34	11	has	have	AUX
ajst-12267	34	12	been	be	AUX
ajst-12267	34	13	widely	widely	ADV
ajst-12267	34	14	used	use	VERB
ajst-12267	34	15	in	in	ADP
ajst-12267	34	16	medical	medical	ADJ
ajst-12267	34	17	image	image	NOUN
ajst-12267	34	18	enhancement	enhancement	NOUN
ajst-12267	34	19	.	.	PUNCT
ajst-12267	35	1	however	however	ADV
ajst-12267	35	2	,	,	PUNCT
ajst-12267	35	3	in	in	ADP
ajst-12267	35	4	terms	term	NOUN
ajst-12267	35	5	of	of	ADP
ajst-12267	35	6	medical	medical	ADJ
ajst-12267	35	7	image	image	NOUN
ajst-12267	35	8	enhancement	enhancement	NOUN
ajst-12267	35	9	,	,	PUNCT
ajst-12267	35	10	gan	gan	PROPN
ajst-12267	35	11	still	still	ADV
ajst-12267	35	12	has	have	VERB
ajst-12267	35	13	problems	problem	NOUN
ajst-12267	35	14	of	of	ADP
ajst-12267	35	15	unstable	unstable	ADJ
ajst-12267	35	16	training	training	NOUN
ajst-12267	35	17	and	and	CCONJ
ajst-12267	35	18	difficult	difficult	ADJ
ajst-12267	35	19	convergence	convergence	NOUN
ajst-12267	35	20	[	[	X
ajst-12267	35	21	8	8	NUM
ajst-12267	35	22	]	]	PUNCT
ajst-12267	35	23	.	.	PUNCT
ajst-12267	36	1	184	184	NUM
ajst-12267	36	2	2.2	2.2	NUM
ajst-12267	36	3	.	.	PUNCT
ajst-12267	37	1	improved	improve	VERB
ajst-12267	37	2	generative	generative	ADJ
ajst-12267	37	3	adversarial	adversarial	ADJ
ajst-12267	37	4	network	network	NOUN
ajst-12267	37	5	deep	deep	ADJ
ajst-12267	37	6	convolutional	convolutional	ADJ
ajst-12267	37	7	generative	generative	ADJ
ajst-12267	37	8	adversarial	adversarial	ADJ
ajst-12267	37	9	network	network	NOUN
ajst-12267	37	10	(	(	PUNCT
ajst-12267	37	11	dcgan	dcgan	NOUN
ajst-12267	37	12	)	)	PUNCT
ajst-12267	37	13	uses	use	VERB
ajst-12267	37	14	deep	deep	ADJ
ajst-12267	37	15	convolutional	convolutional	ADJ
ajst-12267	37	16	neural	neural	ADJ
ajst-12267	37	17	network	network	NOUN
ajst-12267	37	18	in	in	ADP
ajst-12267	37	19	both	both	CCONJ
ajst-12267	37	20	its	its	PRON
ajst-12267	37	21	generative	generative	ADJ
ajst-12267	37	22	model	model	NOUN
ajst-12267	37	23	and	and	CCONJ
ajst-12267	37	24	discriminant	discriminant	ADJ
ajst-12267	37	25	model	model	NOUN
ajst-12267	37	26	,	,	PUNCT
ajst-12267	37	27	and	and	CCONJ
ajst-12267	37	28	it	it	PRON
ajst-12267	37	29	is	be	AUX
ajst-12267	37	30	also	also	ADV
ajst-12267	37	31	the	the	DET
ajst-12267	37	32	first	first	ADJ
ajst-12267	37	33	network	network	NOUN
ajst-12267	37	34	that	that	PRON
ajst-12267	37	35	combines	combine	VERB
ajst-12267	37	36	gan	gan	ADJ
ajst-12267	37	37	and	and	CCONJ
ajst-12267	37	38	convolutional	convolutional	ADJ
ajst-12267	37	39	network	network	NOUN
ajst-12267	37	40	[	[	X
ajst-12267	37	41	2	2	NUM
ajst-12267	37	42	]	]	PUNCT
ajst-12267	37	43	.	.	PUNCT
ajst-12267	38	1	the	the	DET
ajst-12267	38	2	generator	generator	NOUN
ajst-12267	38	3	for	for	ADP
ajst-12267	38	4	dcgan	dcgan	NOUN
ajst-12267	38	5	is	be	AUX
ajst-12267	38	6	as	as	SCONJ
ajst-12267	38	7	follows	follow	VERB
ajst-12267	38	8	:	:	PUNCT
ajst-12267	38	9	figure	figure	NOUN
ajst-12267	38	10	1	1	NUM
ajst-12267	38	11	.	.	PUNCT
ajst-12267	38	12	dcgan	dcgan	PROPN
ajst-12267	38	13	model	model	PROPN
ajst-12267	38	14	generator	generator	PROPN
ajst-12267	38	15	dcgan	dcgan	NOUN
ajst-12267	38	16	has	have	AUX
ajst-12267	38	17	improved	improve	VERB
ajst-12267	38	18	the	the	DET
ajst-12267	38	19	pooling	pool	VERB
ajst-12267	38	20	layer	layer	NOUN
ajst-12267	38	21	,	,	PUNCT
ajst-12267	38	22	activation	activation	NOUN
ajst-12267	38	23	function	function	NOUN
ajst-12267	38	24	and	and	CCONJ
ajst-12267	38	25	other	other	ADJ
ajst-12267	38	26	aspects	aspect	NOUN
ajst-12267	38	27	to	to	PART
ajst-12267	38	28	solve	solve	VERB
ajst-12267	38	29	the	the	DET
ajst-12267	38	30	problem	problem	NOUN
ajst-12267	38	31	of	of	ADP
ajst-12267	38	32	gan	gan	ADJ
ajst-12267	38	33	training	training	NOUN
ajst-12267	38	34	instability	instability	NOUN
ajst-12267	38	35	,	,	PUNCT
ajst-12267	38	36	but	but	CCONJ
ajst-12267	38	37	there	there	PRON
ajst-12267	38	38	are	be	VERB
ajst-12267	38	39	still	still	ADV
ajst-12267	38	40	problems	problem	NOUN
ajst-12267	38	41	that	that	SCONJ
ajst-12267	38	42	the	the	DET
ajst-12267	38	43	generated	generate	VERB
ajst-12267	38	44	images	image	NOUN
ajst-12267	38	45	are	be	AUX
ajst-12267	38	46	not	not	PART
ajst-12267	38	47	diversified	diversify	VERB
ajst-12267	38	48	enough	enough	ADV
ajst-12267	38	49	,	,	PUNCT
ajst-12267	38	50	that	that	ADV
ajst-12267	38	51	is	is	ADV
ajst-12267	38	52	,	,	PUNCT
ajst-12267	38	53	mode	mode	NOUN
ajst-12267	38	54	collapse	collapse	NOUN
ajst-12267	38	55	,	,	PUNCT
ajst-12267	38	56	and	and	CCONJ
ajst-12267	38	57	small	small	ADJ
ajst-12267	38	58	nodules	nodule	NOUN
ajst-12267	38	59	are	be	AUX
ajst-12267	38	60	difficult	difficult	ADJ
ajst-12267	38	61	to	to	PART
ajst-12267	38	62	identify	identify	VERB
ajst-12267	38	63	.	.	PUNCT
ajst-12267	39	1	therefore	therefore	ADV
ajst-12267	39	2	,	,	PUNCT
ajst-12267	39	3	the	the	DET
ajst-12267	39	4	improved	improve	VERB
ajst-12267	39	5	generative	generative	ADJ
ajst-12267	39	6	adversarial	adversarial	ADJ
ajst-12267	39	7	network	network	NOUN
ajst-12267	39	8	proposed	propose	VERB
ajst-12267	39	9	in	in	ADP
ajst-12267	39	10	this	this	DET
ajst-12267	39	11	paper	paper	NOUN
ajst-12267	39	12	is	be	AUX
ajst-12267	39	13	mainly	mainly	ADV
ajst-12267	39	14	improved	improve	VERB
ajst-12267	39	15	from	from	ADP
ajst-12267	39	16	the	the	DET
ajst-12267	39	17	following	follow	VERB
ajst-12267	39	18	aspects	aspect	NOUN
ajst-12267	39	19	.	.	PUNCT
ajst-12267	40	1	2.2.1	2.2.1	NUM
ajst-12267	40	2	.	.	PUNCT
ajst-12267	40	3	loss	loss	NOUN
ajst-12267	40	4	function	function	NOUN
ajst-12267	40	5	improvement	improvement	NOUN
ajst-12267	40	6	like	like	ADP
ajst-12267	40	7	gan	gan	PROPN
ajst-12267	40	8	,	,	PUNCT
ajst-12267	40	9	dcgan	dcgan	NOUN
ajst-12267	40	10	is	be	AUX
ajst-12267	40	11	composed	compose	VERB
ajst-12267	40	12	of	of	ADP
ajst-12267	40	13	generator	generator	NOUN
ajst-12267	40	14	g	g	PROPN
ajst-12267	40	15	and	and	CCONJ
ajst-12267	40	16	discriminator	discriminator	PROPN
ajst-12267	40	17	d.	d.	PROPN
ajst-12267	40	18	the	the	DET
ajst-12267	40	19	loss	loss	NOUN
ajst-12267	40	20	function	function	NOUN
ajst-12267	40	21	of	of	ADP
ajst-12267	40	22	the	the	DET
ajst-12267	40	23	generator	generator	NOUN
ajst-12267	40	24	model	model	NOUN
ajst-12267	40	25	in	in	ADP
ajst-12267	40	26	dcgan	dcgan	NOUN
ajst-12267	40	27	is	be	AUX
ajst-12267	40	28	as	as	SCONJ
ajst-12267	40	29	follows	follow	VERB
ajst-12267	40	30	:	:	PUNCT
ajst-12267	40	31	(	(	PUNCT
ajst-12267	40	32	)	)	PUNCT
ajst-12267	40	33	~min	~min	X
ajst-12267	40	34	(	(	PUNCT
ajst-12267	40	35	,	,	PUNCT
ajst-12267	40	36	)	)	PUNCT
ajst-12267	41	1	[	[	X
ajst-12267	41	2	lg(1	lg(1	X
ajst-12267	41	3	(	(	PUNCT
ajst-12267	41	4	(	(	PUNCT
ajst-12267	41	5	)	)	PUNCT
ajst-12267	41	6	)	)	PUNCT
ajst-12267	41	7	)	)	PUNCT
ajst-12267	41	8	]	]	X
ajst-12267	41	9	g	g	PROPN
ajst-12267	41	10	zz	zz	PROPN
ajst-12267	41	11	p	p	NOUN
ajst-12267	41	12	g	g	PROPN
ajst-12267	41	13	v	v	NUM
ajst-12267	41	14	d	d	X
ajst-12267	41	15	g	g	NOUN
ajst-12267	41	16	e	e	PROPN
ajst-12267	41	17	d	d	X
ajst-12267	41	18	g	g	PROPN
ajst-12267	41	19	z	z	PROPN
ajst-12267	41	20			PROPN
ajst-12267	41	21			NUM
ajst-12267	41	22	(	(	PUNCT
ajst-12267	41	23	2	2	NUM
ajst-12267	41	24	)	)	PUNCT
ajst-12267	41	25	the	the	DET
ajst-12267	41	26	goal	goal	NOUN
ajst-12267	41	27	of	of	ADP
ajst-12267	41	28	the	the	DET
ajst-12267	41	29	generator	generator	NOUN
ajst-12267	41	30	is	be	AUX
ajst-12267	41	31	to	to	PART
ajst-12267	41	32	generate	generate	VERB
ajst-12267	41	33	data	datum	NOUN
ajst-12267	41	34	as	as	ADV
ajst-12267	41	35	close	close	ADV
ajst-12267	41	36	to	to	ADP
ajst-12267	41	37	the	the	DET
ajst-12267	41	38	real	real	ADJ
ajst-12267	41	39	data	datum	NOUN
ajst-12267	41	40	as	as	ADP
ajst-12267	41	41	possible	possible	ADJ
ajst-12267	41	42	,	,	PUNCT
ajst-12267	41	43	so	so	CCONJ
ajst-12267	41	44	the	the	DET
ajst-12267	41	45	goal	goal	NOUN
ajst-12267	41	46	of	of	ADP
ajst-12267	41	47	the	the	DET
ajst-12267	41	48	loss	loss	NOUN
ajst-12267	41	49	function	function	NOUN
ajst-12267	41	50	is	be	AUX
ajst-12267	41	51	to	to	PART
ajst-12267	41	52	get	get	VERB
ajst-12267	41	53	as	as	ADV
ajst-12267	41	54	close	close	ADJ
ajst-12267	41	55	to	to	PART
ajst-12267	41	56	1	1	NUM
ajst-12267	41	57	as	as	ADP
ajst-12267	41	58	possible	possible	ADJ
ajst-12267	41	59	.	.	PUNCT
ajst-12267	42	1	in	in	ADP
ajst-12267	42	2	the	the	DET
ajst-12267	42	3	discriminator	discriminator	NOUN
ajst-12267	42	4	model	model	NOUN
ajst-12267	42	5	,	,	PUNCT
ajst-12267	42	6	the	the	DET
ajst-12267	42	7	loss	loss	NOUN
ajst-12267	42	8	function	function	NOUN
ajst-12267	42	9	is	be	AUX
ajst-12267	42	10	as	as	SCONJ
ajst-12267	42	11	follows	follow	VERB
ajst-12267	42	12	:	:	PUNCT
ajst-12267	42	13	(	(	PUNCT
ajst-12267	42	14	)	)	PUNCT
ajst-12267	42	15	(	(	PUNCT
ajst-12267	42	16	)	)	PUNCT
ajst-12267	42	17	~	~	PUNCT
ajst-12267	42	18	~max	~max	NOUN
ajst-12267	42	19	(	(	PUNCT
ajst-12267	42	20	,	,	PUNCT
ajst-12267	42	21	)	)	PUNCT
ajst-12267	43	1	[	[	X
ajst-12267	43	2	lg	lg	NOUN
ajst-12267	43	3	(	(	PUNCT
ajst-12267	43	4	(	(	PUNCT
ajst-12267	43	5	)	)	PUNCT
ajst-12267	43	6	)	)	PUNCT
ajst-12267	43	7	]	]	PUNCT
ajst-12267	44	1	[	[	X
ajst-12267	44	2	lg(1	lg(1	X
ajst-12267	44	3	(	(	PUNCT
ajst-12267	44	4	(	(	PUNCT
ajst-12267	44	5	)	)	PUNCT
ajst-12267	44	6	)	)	PUNCT
ajst-12267	44	7	)	)	PUNCT
ajst-12267	44	8	]	]	PUNCT
ajst-12267	44	9	data	datum	NOUN
ajst-12267	44	10	x	x	PUNCT
ajst-12267	44	11	g	g	NOUN
ajst-12267	44	12	zx	zx	NUM
ajst-12267	44	13	p	p	NOUN
ajst-12267	44	14	z	z	PROPN
ajst-12267	44	15	p	p	NOUN
ajst-12267	44	16	d	d	X
ajst-12267	44	17	v	v	NOUN
ajst-12267	44	18	d	d	X
ajst-12267	44	19	g	g	NOUN
ajst-12267	44	20	e	e	PROPN
ajst-12267	44	21	d	d	X
ajst-12267	44	22	x	x	PUNCT
ajst-12267	44	23	e	e	NOUN
ajst-12267	44	24	d	d	X
ajst-12267	44	25	g	g	PROPN
ajst-12267	44	26	z	z	PROPN
ajst-12267	44	27			VERB
ajst-12267	44	28			X
ajst-12267	44	29	(	(	PUNCT
ajst-12267	44	30	3	3	NUM
ajst-12267	44	31	)	)	PUNCT
ajst-12267	44	32	in	in	ADP
ajst-12267	44	33	the	the	DET
ajst-12267	44	34	above	above	ADJ
ajst-12267	44	35	loss	loss	NOUN
ajst-12267	44	36	function	function	NOUN
ajst-12267	44	37	expression	expression	NOUN
ajst-12267	44	38	,	,	PUNCT
ajst-12267	44	39	the	the	DET
ajst-12267	44	40	former	former	ADJ
ajst-12267	44	41	expression	expression	NOUN
ajst-12267	44	42	represents	represent	VERB
ajst-12267	44	43	the	the	DET
ajst-12267	44	44	real	real	ADJ
ajst-12267	44	45	data	datum	NOUN
ajst-12267	44	46	set	set	VERB
ajst-12267	44	47	,	,	PUNCT
ajst-12267	44	48	so	so	CCONJ
ajst-12267	44	49	the	the	DET
ajst-12267	44	50	goal	goal	NOUN
ajst-12267	44	51	is	be	AUX
ajst-12267	44	52	close	close	ADJ
ajst-12267	44	53	to	to	ADP
ajst-12267	44	54	1	1	NUM
ajst-12267	44	55	;	;	PUNCT
ajst-12267	44	56	the	the	DET
ajst-12267	44	57	latter	latter	ADJ
ajst-12267	44	58	formula	formula	NOUN
ajst-12267	44	59	is	be	AUX
ajst-12267	44	60	the	the	DET
ajst-12267	44	61	generated	generate	VERB
ajst-12267	44	62	data	datum	NOUN
ajst-12267	44	63	set	set	VERB
ajst-12267	44	64	,	,	PUNCT
ajst-12267	44	65	and	and	CCONJ
ajst-12267	44	66	the	the	DET
ajst-12267	44	67	discriminator	discriminator	NOUN
ajst-12267	44	68	wants	want	VERB
ajst-12267	44	69	to	to	PART
ajst-12267	44	70	recognize	recognize	VERB
ajst-12267	44	71	it	it	PRON
ajst-12267	44	72	completely	completely	ADV
ajst-12267	44	73	,	,	PUNCT
ajst-12267	44	74	so	so	CCONJ
ajst-12267	44	75	the	the	DET
ajst-12267	44	76	goal	goal	NOUN
ajst-12267	44	77	of	of	ADP
ajst-12267	44	78	the	the	DET
ajst-12267	44	79	latter	latter	ADJ
ajst-12267	44	80	formula	formula	NOUN
ajst-12267	44	81	is	be	AUX
ajst-12267	44	82	close	close	ADJ
ajst-12267	44	83	to	to	ADP
ajst-12267	44	84	0	0	NUM
ajst-12267	44	85	,	,	PUNCT
ajst-12267	44	86	and	and	CCONJ
ajst-12267	44	87	the	the	DET
ajst-12267	44	88	final	final	ADJ
ajst-12267	44	89	goal	goal	NOUN
ajst-12267	44	90	of	of	ADP
ajst-12267	44	91	the	the	DET
ajst-12267	44	92	whole	whole	ADJ
ajst-12267	44	93	formula	formula	NOUN
ajst-12267	44	94	is	be	AUX
ajst-12267	44	95	close	close	ADJ
ajst-12267	44	96	to	to	ADP
ajst-12267	44	97	1	1	NUM
ajst-12267	44	98	.	.	PUNCT
ajst-12267	45	1	since	since	SCONJ
ajst-12267	45	2	dcgan	dcgan	NOUN
ajst-12267	45	3	still	still	ADV
ajst-12267	45	4	has	have	VERB
ajst-12267	45	5	the	the	DET
ajst-12267	45	6	problem	problem	NOUN
ajst-12267	45	7	of	of	ADP
ajst-12267	45	8	mode	mode	NOUN
ajst-12267	45	9	collapse	collapse	NOUN
ajst-12267	45	10	,	,	PUNCT
ajst-12267	45	11	this	this	DET
ajst-12267	45	12	paper	paper	NOUN
ajst-12267	45	13	will	will	AUX
ajst-12267	45	14	modify	modify	VERB
ajst-12267	45	15	the	the	DET
ajst-12267	45	16	model	model	NOUN
ajst-12267	45	17	from	from	ADP
ajst-12267	45	18	the	the	DET
ajst-12267	45	19	aspect	aspect	NOUN
ajst-12267	45	20	of	of	ADP
ajst-12267	45	21	loss	loss	NOUN
ajst-12267	45	22	function	function	NOUN
ajst-12267	45	23	.	.	PUNCT
ajst-12267	46	1	specific	specific	ADJ
ajst-12267	46	2	modifications	modification	NOUN
ajst-12267	46	3	are	be	AUX
ajst-12267	46	4	as	as	SCONJ
ajst-12267	46	5	follows	follow	VERB
ajst-12267	46	6	:	:	PUNCT
ajst-12267	46	7	(	(	PUNCT
ajst-12267	46	8	1	1	X
ajst-12267	46	9	)	)	PUNCT
ajst-12267	46	10	the	the	DET
ajst-12267	46	11	kl	kl	PROPN
ajst-12267	46	12	divergence	divergence	NOUN
ajst-12267	46	13	used	use	VERB
ajst-12267	46	14	in	in	ADP
ajst-12267	46	15	dcgan	dcgan	NOUN
ajst-12267	46	16	is	be	AUX
ajst-12267	46	17	asymmetrical	asymmetrical	ADJ
ajst-12267	46	18	,	,	PUNCT
ajst-12267	46	19	which	which	PRON
ajst-12267	46	20	means	mean	VERB
ajst-12267	46	21	that	that	SCONJ
ajst-12267	46	22	for	for	ADP
ajst-12267	46	23	the	the	DET
ajst-12267	46	24	same	same	ADJ
ajst-12267	46	25	distance	distance	NOUN
ajst-12267	46	26	,	,	PUNCT
ajst-12267	46	27	different	different	ADJ
ajst-12267	46	28	observation	observation	NOUN
ajst-12267	46	29	methods	method	NOUN
ajst-12267	46	30	will	will	AUX
ajst-12267	46	31	lead	lead	VERB
ajst-12267	46	32	to	to	ADP
ajst-12267	46	33	different	different	ADJ
ajst-12267	46	34	loss	loss	NOUN
ajst-12267	46	35	values	value	NOUN
ajst-12267	46	36	,	,	PUNCT
ajst-12267	46	37	resulting	result	VERB
ajst-12267	46	38	in	in	ADP
ajst-12267	46	39	the	the	DET
ajst-12267	46	40	overall	overall	ADJ
ajst-12267	46	41	loss	loss	NOUN
ajst-12267	46	42	decline	decline	NOUN
ajst-12267	46	43	direction	direction	NOUN
ajst-12267	46	44	tending	tend	VERB
ajst-12267	46	45	to	to	ADP
ajst-12267	46	46	a	a	DET
ajst-12267	46	47	specific	specific	ADJ
ajst-12267	46	48	direction	direction	NOUN
ajst-12267	46	49	,	,	PUNCT
ajst-12267	46	50	which	which	PRON
ajst-12267	46	51	is	be	AUX
ajst-12267	46	52	easy	easy	ADJ
ajst-12267	46	53	to	to	PART
ajst-12267	46	54	cause	cause	VERB
ajst-12267	46	55	the	the	DET
ajst-12267	46	56	problem	problem	NOUN
ajst-12267	46	57	of	of	ADP
ajst-12267	46	58	model	model	NOUN
ajst-12267	46	59	collapse	collapse	NOUN
ajst-12267	46	60	.	.	PUNCT
ajst-12267	47	1	in	in	ADP
ajst-12267	47	2	this	this	DET
ajst-12267	47	3	paper	paper	NOUN
ajst-12267	47	4	,	,	PUNCT
ajst-12267	47	5	w	w	NOUN
ajst-12267	47	6	distance	distance	NOUN
ajst-12267	47	7	is	be	AUX
ajst-12267	47	8	used	use	VERB
ajst-12267	47	9	to	to	PART
ajst-12267	47	10	replace	replace	VERB
ajst-12267	47	11	kl	kl	NOUN
ajst-12267	47	12	divergence	divergence	NOUN
ajst-12267	47	13	.	.	PUNCT
ajst-12267	48	1	the	the	DET
ajst-12267	48	2	w	w	PROPN
ajst-12267	48	3	distance	distance	NOUN
ajst-12267	48	4	is	be	AUX
ajst-12267	48	5	defined	define	VERB
ajst-12267	48	6	as	as	SCONJ
ajst-12267	48	7	follows	follow	VERB
ajst-12267	48	8	:	:	PUNCT
ajst-12267	48	9	(	(	PUNCT
ajst-12267	48	10	,	,	PUNCT
ajst-12267	48	11	)	)	PUNCT
ajst-12267	48	12	~	~	PUNCT
ajst-12267	48	13	(	(	PUNCT
ajst-12267	48	14	,	,	PUNCT
ajst-12267	48	15	)	)	PUNCT
ajst-12267	48	16	[	[	PUNCT
ajst-12267	48	17	,	,	PUNCT
ajst-12267	48	18	]	]	X
ajst-12267	48	19	[	[	PUNCT
ajst-12267	48	20	,	,	PUNCT
ajst-12267	48	21	]	]	X
ajst-12267	48	22	[	[	PUNCT
ajst-12267	48	23	,	,	PUNCT
ajst-12267	48	24	]	]	X
ajst-12267	48	25	inf	inf	PROPN
ajst-12267	48	26	(	(	PUNCT
ajst-12267	48	27	,	,	PUNCT
ajst-12267	48	28	)	)	PUNCT
ajst-12267	48	29	(	(	PUNCT
ajst-12267	48	30	,	,	PUNCT
ajst-12267	48	31	)	)	PUNCT
ajst-12267	48	32	inf	inf	NOUN
ajst-12267	48	33	[	[	PUNCT
ajst-12267	48	34	(	(	PUNCT
ajst-12267	48	35	,	,	PUNCT
ajst-12267	48	36	)	)	PUNCT
ajst-12267	48	37	]	]	X
ajst-12267	49	1	x	x	X
ajst-12267	49	2	y	y	NOUN
ajst-12267	49	3	r	r	NOUN
ajst-12267	49	4	x	x	PUNCT
ajst-12267	49	5	y	y	NOUN
ajst-12267	49	6	r	r	NOUN
ajst-12267	49	7	p	p	NOUN
ajst-12267	49	8	q	q	NOUN
ajst-12267	49	9	r	r	NOUN
ajst-12267	49	10	p	p	NOUN
ajst-12267	49	11	q	q	PROPN
ajst-12267	49	12	w	w	PROPN
ajst-12267	49	13	p	p	NOUN
ajst-12267	49	14	q	q	NOUN
ajst-12267	49	15	r	r	NOUN
ajst-12267	49	16	x	x	PUNCT
ajst-12267	49	17	y	y	NOUN
ajst-12267	49	18	d	d	NOUN
ajst-12267	49	19	x	x	X
ajst-12267	49	20	y	y	PROPN
ajst-12267	49	21	dxdy	dxdy	PROPN
ajst-12267	49	22	e	e	PROPN
ajst-12267	49	23	d	d	PROPN
ajst-12267	49	24	x	x	X
ajst-12267	49	25	y	y	PROPN
ajst-12267	49	26			PROPN
ajst-12267	49	27			NUM
ajst-12267	49	28			NUM
ajst-12267	49	29	∈	∈	NUM
ajst-12267	49	30	∈	∈	NOUN
ajst-12267	49	31	(	(	PUNCT
ajst-12267	49	32	4	4	NUM
ajst-12267	49	33	)	)	PUNCT
ajst-12267	49	34	(	(	PUNCT
ajst-12267	49	35	2	2	X
ajst-12267	49	36	)	)	PUNCT
ajst-12267	49	37	add	add	VERB
ajst-12267	49	38	the	the	DET
ajst-12267	49	39	variance	variance	NOUN
ajst-12267	49	40	regularization	regularization	NOUN
ajst-12267	49	41	term	term	NOUN
ajst-12267	49	42	to	to	ADP
ajst-12267	49	43	the	the	DET
ajst-12267	49	44	generator	generator	NOUN
ajst-12267	49	45	loss	loss	NOUN
ajst-12267	49	46	function	function	NOUN
ajst-12267	49	47	,	,	PUNCT
ajst-12267	49	48	and	and	CCONJ
ajst-12267	49	49	the	the	DET
ajst-12267	49	50	formula	formula	NOUN
ajst-12267	49	51	is	be	AUX
ajst-12267	49	52	as	as	SCONJ
ajst-12267	49	53	follows	follow	VERB
ajst-12267	49	54	:	:	PUNCT
ajst-12267	49	55	~	~	PUNCT
ajst-12267	49	56	(	(	PUNCT
ajst-12267	49	57	)	)	PUNCT
ajst-12267	50	1	[	[	X
ajst-12267	50	2	log(1	log(1	NOUN
ajst-12267	50	3	(	(	PUNCT
ajst-12267	50	4	(	(	PUNCT
ajst-12267	50	5	)	)	PUNCT
ajst-12267	50	6	)	)	PUNCT
ajst-12267	50	7	)	)	PUNCT
ajst-12267	50	8	]	]	PUNCT
ajst-12267	50	9	(	(	PUNCT
ajst-12267	50	10	(	(	PUNCT
ajst-12267	50	11	(	(	PUNCT
ajst-12267	50	12	)	)	PUNCT
ajst-12267	50	13	,	,	PUNCT
ajst-12267	50	14	(	(	PUNCT
ajst-12267	50	15	)	)	PUNCT
ajst-12267	50	16	)	)	PUNCT
ajst-12267	50	17	)	)	PUNCT
ajst-12267	51	1	x	x	X
ajst-12267	52	1	i	i	PRON
ajst-12267	52	2	j	j	PROPN
ajst-12267	52	3	loss	loss	NOUN
ajst-12267	52	4	z	z	PROPN
ajst-12267	52	5	p	p	NOUN
ajst-12267	52	6	zg	zg	PROPN
ajst-12267	52	7	e	e	PROPN
ajst-12267	52	8	d	d	X
ajst-12267	52	9	g	g	PROPN
ajst-12267	52	10	z	z	PROPN
ajst-12267	52	11	dis	dis	PROPN
ajst-12267	53	1	g	g	PROPN
ajst-12267	53	2	z	z	PROPN
ajst-12267	53	3	g	g	PROPN
ajst-12267	53	4	z	z	PROPN
ajst-12267	53	5			PROPN
ajst-12267	53	6			PUNCT
ajst-12267	53	7	(	(	PUNCT
ajst-12267	53	8	5	5	NUM
ajst-12267	53	9	)	)	PUNCT
ajst-12267	53	10	where	where	SCONJ
ajst-12267	53	11	gloss	gloss	NOUN
ajst-12267	53	12	represents	represent	VERB
ajst-12267	53	13	the	the	DET
ajst-12267	53	14	generator	generator	NOUN
ajst-12267	53	15	loss	loss	NOUN
ajst-12267	53	16	,	,	PUNCT
ajst-12267	53	17	g(z	g(z	PROPN
ajst-12267	53	18	)	)	PUNCT
ajst-12267	53	19	represents	represent	VERB
ajst-12267	53	20	the	the	DET
ajst-12267	53	21	generated	generate	VERB
ajst-12267	53	22	image	image	NOUN
ajst-12267	53	23	,	,	PUNCT
ajst-12267	53	24	d	d	X
ajst-12267	53	25	(	(	PUNCT
ajst-12267	53	26	g(z	g(z	PROPN
ajst-12267	53	27	)	)	PUNCT
ajst-12267	53	28	)	)	PUNCT
ajst-12267	53	29	represents	represent	VERB
ajst-12267	53	30	the	the	DET
ajst-12267	53	31	probability	probability	NOUN
ajst-12267	53	32	of	of	ADP
ajst-12267	53	33	the	the	DET
ajst-12267	53	34	discriminator	discriminator	NOUN
ajst-12267	53	35	on	on	ADP
ajst-12267	53	36	the	the	DET
ajst-12267	53	37	image	image	NOUN
ajst-12267	53	38	g(z	g(z	PROPN
ajst-12267	53	39	)	)	PUNCT
ajst-12267	53	40	,	,	PUNCT
ajst-12267	53	41	dis(g(z)i	dis(g(z)i	ADJ
ajst-12267	53	42	,	,	PUNCT
ajst-12267	53	43	g(z)j	g(z)j	PROPN
ajst-12267	53	44	)	)	PUNCT
ajst-12267	53	45	represents	represent	VERB
ajst-12267	53	46	the	the	DET
ajst-12267	53	47	distance	distance	NOUN
ajst-12267	53	48	between	between	ADP
ajst-12267	53	49	image	image	NOUN
ajst-12267	53	50	i	i	PRON
ajst-12267	53	51	and	and	CCONJ
ajst-12267	53	52	image	image	NOUN
ajst-12267	53	53	j	j	PROPN
ajst-12267	53	54	in	in	ADP
ajst-12267	53	55	the	the	DET
ajst-12267	53	56	generated	generate	VERB
ajst-12267	53	57	batch	batch	NOUN
ajst-12267	53	58	of	of	ADP
ajst-12267	53	59	images	image	NOUN
ajst-12267	53	60	,	,	PUNCT
ajst-12267	53	61	&	&	CCONJ
ajst-12267	53	62	is	be	AUX
ajst-12267	53	63	the	the	DET
ajst-12267	53	64	variance	variance	NOUN
ajst-12267	53	65	,	,	PUNCT
ajst-12267	53	66	and	and	CCONJ
ajst-12267	53	67	nimda	nimda	NOUN
ajst-12267	53	68	is	be	AUX
ajst-12267	53	69	the	the	DET
ajst-12267	53	70	regularization	regularization	NOUN
ajst-12267	53	71	coefficient	coefficient	NOUN
ajst-12267	53	72	.	.	PUNCT
ajst-12267	54	1	2.2.2	2.2.2	X
ajst-12267	54	2	.	.	PUNCT
ajst-12267	54	3	attentional	attentional	ADJ
ajst-12267	54	4	mechanism	mechanism	NOUN
ajst-12267	54	5	the	the	DET
ajst-12267	54	6	essence	essence	NOUN
ajst-12267	54	7	of	of	ADP
ajst-12267	54	8	the	the	DET
ajst-12267	54	9	attention	attention	NOUN
ajst-12267	54	10	mechanism	mechanism	NOUN
ajst-12267	54	11	is	be	AUX
ajst-12267	54	12	to	to	PART
ajst-12267	54	13	locate	locate	VERB
ajst-12267	54	14	the	the	DET
ajst-12267	54	15	interesting	interesting	ADJ
ajst-12267	54	16	information	information	NOUN
ajst-12267	54	17	and	and	CCONJ
ajst-12267	54	18	suppress	suppress	VERB
ajst-12267	54	19	the	the	DET
ajst-12267	54	20	useless	useless	ADJ
ajst-12267	54	21	information	information	NOUN
ajst-12267	54	22	.	.	PUNCT
ajst-12267	55	1	in	in	ADP
ajst-12267	55	2	lung	lung	NOUN
ajst-12267	55	3	nodule	nodule	NOUN
ajst-12267	55	4	images	image	NOUN
ajst-12267	55	5	,	,	PUNCT
ajst-12267	55	6	not	not	PART
ajst-12267	55	7	all	all	DET
ajst-12267	55	8	areas	area	NOUN
ajst-12267	55	9	in	in	ADP
ajst-12267	55	10	the	the	DET
ajst-12267	55	11	image	image	NOUN
ajst-12267	55	12	contribute	contribute	VERB
ajst-12267	55	13	equally	equally	ADV
ajst-12267	55	14	to	to	ADP
ajst-12267	55	15	the	the	DET
ajst-12267	55	16	generation	generation	NOUN
ajst-12267	55	17	of	of	ADP
ajst-12267	55	18	new	new	ADJ
ajst-12267	55	19	data	datum	NOUN
ajst-12267	55	20	.	.	PUNCT
ajst-12267	56	1	only	only	ADV
ajst-12267	56	2	the	the	DET
ajst-12267	56	3	small	small	ADJ
ajst-12267	56	4	proportion	proportion	NOUN
ajst-12267	56	5	of	of	ADP
ajst-12267	56	6	lung	lung	NOUN
ajst-12267	56	7	nodule	nodule	NOUN
ajst-12267	56	8	areas	area	NOUN
ajst-12267	56	9	should	should	AUX
ajst-12267	56	10	be	be	AUX
ajst-12267	56	11	concerned	concern	VERB
ajst-12267	56	12	,	,	PUNCT
ajst-12267	56	13	so	so	ADV
ajst-12267	56	14	important	important	ADJ
ajst-12267	56	15	parts	part	NOUN
ajst-12267	56	16	should	should	AUX
ajst-12267	56	17	be	be	AUX
ajst-12267	56	18	found	find	VERB
ajst-12267	56	19	for	for	ADP
ajst-12267	56	20	processing	processing	NOUN
ajst-12267	56	21	.	.	PUNCT
ajst-12267	57	1	therefore	therefore	ADV
ajst-12267	57	2	,	,	PUNCT
ajst-12267	57	3	this	this	DET
ajst-12267	57	4	paper	paper	NOUN
ajst-12267	57	5	adds	add	VERB
ajst-12267	57	6	the	the	DET
ajst-12267	57	7	spatial	spatial	ADJ
ajst-12267	57	8	attention	attention	NOUN
ajst-12267	57	9	mechanism	mechanism	NOUN
ajst-12267	57	10	to	to	PART
ajst-12267	57	11	make	make	VERB
ajst-12267	57	12	the	the	DET
ajst-12267	57	13	important	important	ADJ
ajst-12267	57	14	part	part	NOUN
ajst-12267	57	15	of	of	ADP
ajst-12267	57	16	the	the	DET
ajst-12267	57	17	generative	generative	ADJ
ajst-12267	57	18	network	network	NOUN
ajst-12267	57	19	more	more	ADV
ajst-12267	57	20	prominent	prominent	ADJ
ajst-12267	57	21	.	.	PUNCT
ajst-12267	58	1	stn[9	stn[9	NUM
ajst-12267	58	2	]	]	PUNCT
ajst-12267	58	3	(	(	PUNCT
ajst-12267	58	4	spatial	spatial	ADJ
ajst-12267	58	5	transformer	transformer	NOUN
ajst-12267	58	6	network	network	NOUN
ajst-12267	58	7	)	)	PUNCT
ajst-12267	58	8	is	be	AUX
ajst-12267	58	9	a	a	DET
ajst-12267	58	10	model	model	NOUN
ajst-12267	58	11	based	base	VERB
ajst-12267	58	12	on	on	ADP
ajst-12267	58	13	spatial	spatial	ADJ
ajst-12267	58	14	attention	attention	NOUN
ajst-12267	58	15	mechanism	mechanism	NOUN
ajst-12267	58	16	,	,	PUNCT
ajst-12267	58	17	and	and	CCONJ
ajst-12267	58	18	its	its	PRON
ajst-12267	58	19	structure	structure	NOUN
ajst-12267	58	20	is	be	AUX
ajst-12267	58	21	as	as	SCONJ
ajst-12267	58	22	follows	follow	VERB
ajst-12267	58	23	:	:	PUNCT
ajst-12267	58	24	figure	figure	NOUN
ajst-12267	58	25	2	2	NUM
ajst-12267	58	26	.	.	PUNCT
ajst-12267	59	1	stn	stn	PROPN
ajst-12267	59	2	model	model	PROPN
ajst-12267	59	3	structure	structure	PROPN
ajst-12267	59	4	diagram	diagram	PROPN
ajst-12267	59	5	185	185	NUM
ajst-12267	59	6	the	the	DET
ajst-12267	59	7	addition	addition	NOUN
ajst-12267	59	8	of	of	ADP
ajst-12267	59	9	attention	attention	NOUN
ajst-12267	59	10	mechanism	mechanism	NOUN
ajst-12267	59	11	in	in	ADP
ajst-12267	59	12	dcgan	dcgan	NOUN
ajst-12267	59	13	helps	help	VERB
ajst-12267	59	14	the	the	DET
ajst-12267	59	15	model	model	NOUN
ajst-12267	59	16	to	to	PART
ajst-12267	59	17	generate	generate	VERB
ajst-12267	59	18	micro	micro	NOUN
ajst-12267	59	19	-	-	NOUN
ajst-12267	59	20	nodules	nodule	NOUN
ajst-12267	59	21	.	.	PUNCT
ajst-12267	60	1	3	3	X
ajst-12267	60	2	.	.	X
ajst-12267	60	3	experiment	experiment	NOUN
ajst-12267	60	4	3.1	3.1	NUM
ajst-12267	60	5	.	.	PUNCT
ajst-12267	61	1	data	datum	NOUN
ajst-12267	61	2	set	set	VERB
ajst-12267	61	3	introduction	introduction	NOUN
ajst-12267	61	4	the	the	DET
ajst-12267	61	5	data	datum	NOUN
ajst-12267	61	6	set	set	VERB
ajst-12267	61	7	used	use	VERB
ajst-12267	61	8	in	in	ADP
ajst-12267	61	9	this	this	DET
ajst-12267	61	10	paper	paper	NOUN
ajst-12267	61	11	is	be	AUX
ajst-12267	61	12	the	the	DET
ajst-12267	61	13	luna16	luna16	ADJ
ajst-12267	61	14	dataset	dataset	NOUN
ajst-12267	61	15	,	,	PUNCT
ajst-12267	61	16	which	which	PRON
ajst-12267	61	17	contains	contain	VERB
ajst-12267	61	18	888	888	NUM
ajst-12267	61	19	low	low	ADV
ajst-12267	61	20	-	-	PUNCT
ajst-12267	61	21	dose	dose	VERB
ajst-12267	61	22	ct	ct	NUM
ajst-12267	61	23	images	image	NOUN
ajst-12267	61	24	.	.	PUNCT
ajst-12267	62	1	each	each	DET
ajst-12267	62	2	image	image	NOUN
ajst-12267	62	3	contains	contain	VERB
ajst-12267	62	4	a	a	DET
ajst-12267	62	5	series	series	NOUN
ajst-12267	62	6	of	of	ADP
ajst-12267	62	7	axial	axial	ADJ
ajst-12267	62	8	sections	section	NOUN
ajst-12267	62	9	of	of	ADP
ajst-12267	62	10	the	the	DET
ajst-12267	62	11	chest	chest	NOUN
ajst-12267	62	12	and	and	CCONJ
ajst-12267	62	13	is	be	AUX
ajst-12267	62	14	annotated	annotate	VERB
ajst-12267	62	15	by	by	ADP
ajst-12267	62	16	several	several	ADJ
ajst-12267	62	17	experienced	experienced	ADJ
ajst-12267	62	18	radiologists	radiologist	NOUN
ajst-12267	62	19	.	.	PUNCT
ajst-12267	63	1	this	this	DET
ajst-12267	63	2	dataset	dataset	NOUN
ajst-12267	63	3	records	record	VERB
ajst-12267	63	4	the	the	DET
ajst-12267	63	5	position	position	NOUN
ajst-12267	63	6	information	information	NOUN
ajst-12267	63	7	and	and	CCONJ
ajst-12267	63	8	diameter	diameter	NOUN
ajst-12267	63	9	size	size	NOUN
ajst-12267	63	10	of	of	ADP
ajst-12267	63	11	lung	lung	NOUN
ajst-12267	63	12	nodules	nodule	NOUN
ajst-12267	63	13	.	.	PUNCT
ajst-12267	64	1	3.2	3.2	NUM
ajst-12267	64	2	.	.	PUNCT
ajst-12267	64	3	experimental	experimental	ADJ
ajst-12267	64	4	environment	environment	NOUN
ajst-12267	64	5	and	and	CCONJ
ajst-12267	64	6	related	related	ADJ
ajst-12267	64	7	parameters	parameter	NOUN
ajst-12267	64	8	this	this	DET
ajst-12267	64	9	experiment	experiment	NOUN
ajst-12267	64	10	was	be	AUX
ajst-12267	64	11	conducted	conduct	VERB
ajst-12267	64	12	under	under	ADP
ajst-12267	64	13	windows10	windows10	NOUN
ajst-12267	64	14	platform	platform	NOUN
ajst-12267	64	15	,	,	PUNCT
ajst-12267	64	16	the	the	DET
ajst-12267	64	17	experimental	experimental	ADJ
ajst-12267	64	18	environment	environment	NOUN
ajst-12267	64	19	was	be	AUX
ajst-12267	64	20	pytorch1.9.0	pytorch1.9.0	ADJ
ajst-12267	65	1	+	+	ADJ
ajst-12267	65	2	cuda	cuda	NOUN
ajst-12267	65	3	12.1	12.1	NUM
ajst-12267	65	4	,	,	PUNCT
ajst-12267	65	5	the	the	DET
ajst-12267	65	6	graphics	graphic	NOUN
ajst-12267	65	7	unit	unit	NOUN
ajst-12267	65	8	(	(	PUNCT
ajst-12267	65	9	gpu	gpu	PROPN
ajst-12267	65	10	)	)	PUNCT
ajst-12267	65	11	used	use	VERB
ajst-12267	65	12	was	be	AUX
ajst-12267	65	13	nvidia	nvidia	PROPN
ajst-12267	65	14	geforce	geforce	NOUN
ajst-12267	65	15	mx450	mx450	PROPN
ajst-12267	65	16	,	,	PUNCT
ajst-12267	65	17	and	and	CCONJ
ajst-12267	65	18	the	the	DET
ajst-12267	65	19	memory	memory	NOUN
ajst-12267	65	20	size	size	NOUN
ajst-12267	65	21	was	be	AUX
ajst-12267	65	22	16	16	NUM
ajst-12267	65	23	gb	gb	NOUN
ajst-12267	65	24	.	.	PUNCT
ajst-12267	66	1	the	the	DET
ajst-12267	66	2	886	886	NUM
ajst-12267	66	3	images	image	NOUN
ajst-12267	66	4	in	in	ADP
ajst-12267	66	5	the	the	DET
ajst-12267	66	6	luna16	luna16	ADJ
ajst-12267	66	7	data	datum	NOUN
ajst-12267	66	8	set	set	VERB
ajst-12267	66	9	were	be	AUX
ajst-12267	66	10	randomly	randomly	ADV
ajst-12267	66	11	scattered	scatter	VERB
ajst-12267	66	12	,	,	PUNCT
ajst-12267	66	13	and	and	CCONJ
ajst-12267	66	14	the	the	DET
ajst-12267	66	15	training	training	NOUN
ajst-12267	66	16	set	set	NOUN
ajst-12267	66	17	and	and	CCONJ
ajst-12267	66	18	test	test	NOUN
ajst-12267	66	19	set	set	VERB
ajst-12267	66	20	were	be	AUX
ajst-12267	66	21	divided	divide	VERB
ajst-12267	66	22	according	accord	VERB
ajst-12267	66	23	to	to	ADP
ajst-12267	66	24	7:3	7:3	NUM
ajst-12267	66	25	,	,	PUNCT
ajst-12267	66	26	that	that	ADV
ajst-12267	66	27	is	is	ADV
ajst-12267	66	28	,	,	PUNCT
ajst-12267	66	29	620	620	NUM
ajst-12267	66	30	pictures	picture	NOUN
ajst-12267	66	31	were	be	AUX
ajst-12267	66	32	taken	take	VERB
ajst-12267	66	33	as	as	SCONJ
ajst-12267	66	34	the	the	DET
ajst-12267	66	35	training	training	NOUN
ajst-12267	66	36	set	set	NOUN
ajst-12267	66	37	and	and	CCONJ
ajst-12267	66	38	266	266	NUM
ajst-12267	66	39	pictures	picture	NOUN
ajst-12267	66	40	were	be	AUX
ajst-12267	66	41	taken	take	VERB
ajst-12267	66	42	as	as	ADP
ajst-12267	66	43	the	the	DET
ajst-12267	66	44	test	test	NOUN
ajst-12267	66	45	set	set	VERB
ajst-12267	66	46	.	.	PUNCT
ajst-12267	67	1	the	the	DET
ajst-12267	67	2	size	size	NOUN
ajst-12267	67	3	of	of	ADP
ajst-12267	67	4	the	the	DET
ajst-12267	67	5	input	input	NOUN
ajst-12267	67	6	picture	picture	NOUN
ajst-12267	67	7	was	be	AUX
ajst-12267	67	8	416	416	NUM
ajst-12267	67	9	*	*	SYM
ajst-12267	67	10	416	416	NUM
ajst-12267	67	11	,	,	PUNCT
ajst-12267	67	12	500	500	NUM
ajst-12267	67	13	rounds	round	NOUN
ajst-12267	67	14	of	of	ADP
ajst-12267	67	15	training	training	NOUN
ajst-12267	67	16	were	be	AUX
ajst-12267	67	17	conducted	conduct	VERB
ajst-12267	67	18	,	,	PUNCT
ajst-12267	67	19	and	and	CCONJ
ajst-12267	67	20	the	the	DET
ajst-12267	67	21	initial	initial	ADJ
ajst-12267	67	22	learning	learning	NOUN
ajst-12267	67	23	rate	rate	NOUN
ajst-12267	67	24	was	be	AUX
ajst-12267	67	25	10	10	NUM
ajst-12267	67	26	-	-	SYM
ajst-12267	67	27	3	3	NUM
ajst-12267	67	28	.	.	PUNCT
ajst-12267	68	1	the	the	DET
ajst-12267	68	2	evaluation	evaluation	NOUN
ajst-12267	68	3	indicators	indicator	NOUN
ajst-12267	68	4	used	use	VERB
ajst-12267	68	5	in	in	ADP
ajst-12267	68	6	this	this	DET
ajst-12267	68	7	paper	paper	NOUN
ajst-12267	68	8	are	be	AUX
ajst-12267	68	9	confidence	confidence	NOUN
ajst-12267	68	10	and	and	CCONJ
ajst-12267	68	11	the	the	DET
ajst-12267	68	12	initial	initial	ADJ
ajst-12267	68	13	distance	distance	NOUN
ajst-12267	68	14	(	(	PUNCT
ajst-12267	68	15	fid	fid	NOUN
ajst-12267	68	16	)	)	PUNCT
ajst-12267	68	17	between	between	ADP
ajst-12267	68	18	the	the	DET
ajst-12267	68	19	two	two	NUM
ajst-12267	68	20	enhancement	enhancement	NOUN
ajst-12267	68	21	frames	frame	NOUN
ajst-12267	68	22	to	to	PART
ajst-12267	68	23	generate	generate	VERB
ajst-12267	68	24	lung	lung	NOUN
ajst-12267	68	25	nodule	nodule	NOUN
ajst-12267	68	26	images	image	NOUN
ajst-12267	68	27	.	.	PUNCT
ajst-12267	69	1	the	the	PRON
ajst-12267	69	2	greater	great	ADJ
ajst-12267	69	3	the	the	DET
ajst-12267	69	4	confidence	confidence	NOUN
ajst-12267	69	5	,	,	PUNCT
ajst-12267	69	6	the	the	PRON
ajst-12267	69	7	greater	great	ADJ
ajst-12267	69	8	the	the	DET
ajst-12267	69	9	possibility	possibility	NOUN
ajst-12267	69	10	that	that	SCONJ
ajst-12267	69	11	roi	roi	NOUN
ajst-12267	69	12	is	be	AUX
ajst-12267	69	13	a	a	DET
ajst-12267	69	14	nodule	nodule	NOUN
ajst-12267	69	15	.	.	PUNCT
ajst-12267	70	1	fid	fid	NOUN
ajst-12267	70	2	represents	represent	VERB
ajst-12267	70	3	the	the	DET
ajst-12267	70	4	difference	difference	NOUN
ajst-12267	70	5	between	between	ADP
ajst-12267	70	6	the	the	DET
ajst-12267	70	7	distribution	distribution	NOUN
ajst-12267	70	8	of	of	ADP
ajst-12267	70	9	the	the	DET
ajst-12267	70	10	real	real	ADJ
ajst-12267	70	11	image	image	NOUN
ajst-12267	70	12	and	and	CCONJ
ajst-12267	70	13	the	the	DET
ajst-12267	70	14	generated	generate	VERB
ajst-12267	70	15	image	image	NOUN
ajst-12267	70	16	,	,	PUNCT
ajst-12267	70	17	and	and	CCONJ
ajst-12267	70	18	represents	represent	VERB
ajst-12267	70	19	the	the	DET
ajst-12267	70	20	distance	distance	NOUN
ajst-12267	70	21	of	of	ADP
ajst-12267	70	22	the	the	DET
ajst-12267	70	23	feature	feature	NOUN
ajst-12267	70	24	vector	vector	NOUN
ajst-12267	70	25	between	between	ADP
ajst-12267	70	26	the	the	DET
ajst-12267	70	27	generated	generate	VERB
ajst-12267	70	28	image	image	NOUN
ajst-12267	70	29	and	and	CCONJ
ajst-12267	70	30	the	the	DET
ajst-12267	70	31	real	real	ADJ
ajst-12267	70	32	image	image	NOUN
ajst-12267	70	33	,	,	PUNCT
ajst-12267	70	34	and	and	CCONJ
ajst-12267	70	35	the	the	DET
ajst-12267	70	36	smaller	small	ADJ
ajst-12267	70	37	the	the	DET
ajst-12267	70	38	distance	distance	NOUN
ajst-12267	70	39	,	,	PUNCT
ajst-12267	70	40	the	the	PRON
ajst-12267	70	41	closer	close	ADV
ajst-12267	70	42	the	the	DET
ajst-12267	70	43	generated	generate	VERB
ajst-12267	70	44	nodule	nodule	NOUN
ajst-12267	70	45	is	be	AUX
ajst-12267	70	46	to	to	ADP
ajst-12267	70	47	the	the	DET
ajst-12267	70	48	real	real	ADJ
ajst-12267	70	49	image	image	NOUN
ajst-12267	70	50	.	.	PUNCT
ajst-12267	71	1	the	the	DET
ajst-12267	71	2	calculation	calculation	NOUN
ajst-12267	71	3	formula	formula	NOUN
ajst-12267	71	4	for	for	ADP
ajst-12267	71	5	fid	fid	NOUN
ajst-12267	71	6	is	be	AUX
ajst-12267	71	7	as	as	SCONJ
ajst-12267	71	8	follows	follow	VERB
ajst-12267	71	9	:	:	PUNCT
ajst-12267	71	10	2	2	NUM
ajst-12267	71	11	0.5	0.5	NUM
ajst-12267	71	12	2	2	NUM
ajst-12267	71	13	(	(	PUNCT
ajst-12267	71	14	,	,	PUNCT
ajst-12267	71	15	)	)	PUNCT
ajst-12267	71	16	||	||	PUNCT
ajst-12267	72	1	||	||	NOUN
ajst-12267	72	2	(	(	PUNCT
ajst-12267	72	3	2	2	NUM
ajst-12267	72	4	)	)	PUNCT
ajst-12267	72	5	)	)	PUNCT
ajst-12267	73	1	gxfid	gxfid	VERB
ajst-12267	73	2	x	x	PUNCT
ajst-12267	73	3	g	g	NOUN
ajst-12267	73	4	tr	tr	NOUN
ajst-12267	73	5	x	x	X
ajst-12267	73	6	g	g	PROPN
ajst-12267	73	7	x	x	SYM
ajst-12267	73	8	g	g	PUNCT
ajst-12267	73	9			NOUN
ajst-12267	73	10			NOUN
ajst-12267	73	11			VERB
ajst-12267	73	12			PROPN
ajst-12267	73	13			PROPN
ajst-12267	73	14			X
ajst-12267	73	15			X
ajst-12267	73	16			X
ajst-12267	73	17	（	（	PUNCT
ajst-12267	73	18	(	(	PUNCT
ajst-12267	73	19	6	6	NUM
ajst-12267	73	20	)	)	PUNCT
ajst-12267	73	21	3.3	3.3	NUM
ajst-12267	73	22	.	.	PUNCT
ajst-12267	74	1	experimental	experimental	ADJ
ajst-12267	74	2	result	result	VERB
ajst-12267	74	3	the	the	DET
ajst-12267	74	4	original	original	ADJ
ajst-12267	74	5	training	training	NOUN
ajst-12267	74	6	data	datum	NOUN
ajst-12267	74	7	,	,	PUNCT
ajst-12267	74	8	the	the	DET
ajst-12267	74	9	image	image	NOUN
ajst-12267	74	10	generated	generate	VERB
ajst-12267	74	11	by	by	ADP
ajst-12267	74	12	unmodified	unmodified	ADJ
ajst-12267	74	13	dcgan	dcgan	NOUN
ajst-12267	74	14	and	and	CCONJ
ajst-12267	74	15	the	the	DET
ajst-12267	74	16	image	image	NOUN
ajst-12267	74	17	generated	generate	VERB
ajst-12267	74	18	after	after	ADP
ajst-12267	74	19	modifying	modify	VERB
ajst-12267	74	20	the	the	DET
ajst-12267	74	21	kl	kl	PROPN
ajst-12267	74	22	divergence	divergence	NOUN
ajst-12267	74	23	in	in	ADP
ajst-12267	74	24	dcgan	dcgan	NOUN
ajst-12267	74	25	to	to	PART
ajst-12267	74	26	w	w	NOUN
ajst-12267	74	27	distance	distance	NOUN
ajst-12267	74	28	are	be	AUX
ajst-12267	74	29	shown	show	VERB
ajst-12267	74	30	as	as	SCONJ
ajst-12267	74	31	follows	follow	VERB
ajst-12267	74	32	:	:	PUNCT
ajst-12267	74	33	figure	figure	NOUN
ajst-12267	74	34	3	3	NUM
ajst-12267	74	35	.	.	PUNCT
ajst-12267	74	36	improved	improve	VERB
ajst-12267	74	37	dcgan	dcgan	NOUN
ajst-12267	74	38	to	to	PART
ajst-12267	74	39	generate	generate	VERB
ajst-12267	74	40	images	image	NOUN
ajst-12267	74	41	figure	figure	NOUN
ajst-12267	74	42	4	4	NUM
ajst-12267	74	43	.	.	PUNCT
ajst-12267	75	1	dcgan	dcgan	NOUN
ajst-12267	75	2	generates	generate	VERB
ajst-12267	75	3	images	image	NOUN
ajst-12267	75	4	as	as	SCONJ
ajst-12267	75	5	can	can	AUX
ajst-12267	75	6	be	be	AUX
ajst-12267	75	7	seen	see	VERB
ajst-12267	75	8	from	from	ADP
ajst-12267	75	9	the	the	DET
ajst-12267	75	10	above	above	ADJ
ajst-12267	75	11	pictures	picture	NOUN
ajst-12267	75	12	,	,	PUNCT
ajst-12267	75	13	compared	compare	VERB
ajst-12267	75	14	with	with	ADP
ajst-12267	75	15	dcgan	dcgan	NOUN
ajst-12267	75	16	that	that	PRON
ajst-12267	75	17	has	have	AUX
ajst-12267	75	18	not	not	PART
ajst-12267	75	19	been	be	AUX
ajst-12267	75	20	improved	improve	VERB
ajst-12267	75	21	,	,	PUNCT
ajst-12267	75	22	the	the	DET
ajst-12267	75	23	improved	improved	ADJ
ajst-12267	75	24	model	model	NOUN
ajst-12267	75	25	generates	generate	VERB
ajst-12267	75	26	a	a	DET
ajst-12267	75	27	higher	high	ADJ
ajst-12267	75	28	diversity	diversity	NOUN
ajst-12267	75	29	of	of	ADP
ajst-12267	75	30	images	image	NOUN
ajst-12267	75	31	,	,	PUNCT
ajst-12267	75	32	and	and	CCONJ
ajst-12267	75	33	the	the	DET
ajst-12267	75	34	model	model	NOUN
ajst-12267	75	35	has	have	AUX
ajst-12267	75	36	improved	improve	VERB
ajst-12267	75	37	the	the	DET
ajst-12267	75	38	diversity	diversity	NOUN
ajst-12267	75	39	of	of	ADP
ajst-12267	75	40	generated	generate	VERB
ajst-12267	75	41	data	datum	NOUN
ajst-12267	75	42	to	to	ADP
ajst-12267	75	43	a	a	DET
ajst-12267	75	44	certain	certain	ADJ
ajst-12267	75	45	extent	extent	NOUN
ajst-12267	75	46	.	.	PUNCT
ajst-12267	76	1	the	the	DET
ajst-12267	76	2	fid	fid	NOUN
ajst-12267	76	3	data	data	NOUN
ajst-12267	76	4	results	result	NOUN
ajst-12267	76	5	are	be	AUX
ajst-12267	76	6	as	as	SCONJ
ajst-12267	76	7	follows	follow	VERB
ajst-12267	76	8	:	:	PUNCT
ajst-12267	76	9	table	table	NOUN
ajst-12267	76	10	1	1	NUM
ajst-12267	76	11	.	.	PUNCT
ajst-12267	76	12	comparison	comparison	NOUN
ajst-12267	76	13	of	of	ADP
ajst-12267	76	14	fid	fid	NOUN
ajst-12267	76	15	of	of	ADP
ajst-12267	76	16	different	different	ADJ
ajst-12267	76	17	frameworks	framework	NOUN
ajst-12267	76	18	frame	frame	VERB
ajst-12267	76	19	data	datum	NOUN
ajst-12267	76	20	set	set	VERB
ajst-12267	76	21	fid	fid	NOUN
ajst-12267	76	22	dcgan	dcgan	NOUN
ajst-12267	76	23	luna16	luna16	PROPN
ajst-12267	76	24	85.6	85.6	NUM
ajst-12267	76	25	improved	improve	VERB
ajst-12267	76	26	dcgan	dcgan	NOUN
ajst-12267	76	27	luna16	luna16	NOUN
ajst-12267	76	28	81.3	81.3	NUM
ajst-12267	76	29	from	from	ADP
ajst-12267	76	30	the	the	DET
ajst-12267	76	31	above	above	NOUN
ajst-12267	76	32	,	,	PUNCT
ajst-12267	76	33	it	it	PRON
ajst-12267	76	34	can	can	AUX
ajst-12267	76	35	be	be	AUX
ajst-12267	76	36	seen	see	VERB
ajst-12267	76	37	that	that	SCONJ
ajst-12267	76	38	the	the	DET
ajst-12267	76	39	improved	improved	ADJ
ajst-12267	76	40	dcgan	dcgan	NOUN
ajst-12267	76	41	can	can	AUX
ajst-12267	76	42	improve	improve	VERB
ajst-12267	76	43	the	the	DET
ajst-12267	76	44	diversity	diversity	NOUN
ajst-12267	76	45	of	of	ADP
ajst-12267	76	46	nodules	nodule	NOUN
ajst-12267	76	47	,	,	PUNCT
ajst-12267	76	48	and	and	CCONJ
ajst-12267	76	49	reduce	reduce	VERB
ajst-12267	76	50	the	the	DET
ajst-12267	76	51	fid	fid	NOUN
ajst-12267	76	52	value	value	NOUN
ajst-12267	76	53	from	from	ADP
ajst-12267	76	54	85.6	85.6	NUM
ajst-12267	76	55	to	to	ADP
ajst-12267	76	56	81.3	81.3	NUM
ajst-12267	76	57	,	,	PUNCT
ajst-12267	76	58	generating	generate	VERB
ajst-12267	76	59	more	more	ADV
ajst-12267	76	60	realistic	realistic	ADJ
ajst-12267	76	61	pulmonary	pulmonary	ADJ
ajst-12267	76	62	nodule	nodule	NOUN
ajst-12267	76	63	data	datum	NOUN
ajst-12267	76	64	.	.	PUNCT
ajst-12267	77	1	4	4	X
ajst-12267	77	2	.	.	X
ajst-12267	77	3	conclusion	conclusion	NOUN
ajst-12267	77	4	to	to	PART
ajst-12267	77	5	solve	solve	VERB
ajst-12267	77	6	the	the	DET
ajst-12267	77	7	problem	problem	NOUN
ajst-12267	77	8	of	of	ADP
ajst-12267	77	9	enhancement	enhancement	NOUN
ajst-12267	77	10	of	of	ADP
ajst-12267	77	11	lung	lung	NOUN
ajst-12267	77	12	nodule	nodule	NOUN
ajst-12267	77	13	data	datum	NOUN
ajst-12267	77	14	set	set	PROPN
ajst-12267	77	15	,	,	PUNCT
ajst-12267	77	16	the	the	DET
ajst-12267	77	17	dcgan	dcgan	NOUN
ajst-12267	77	18	model	model	NOUN
ajst-12267	77	19	is	be	AUX
ajst-12267	77	20	improved	improve	VERB
ajst-12267	77	21	in	in	ADP
ajst-12267	77	22	this	this	DET
ajst-12267	77	23	paper	paper	NOUN
ajst-12267	77	24	.	.	PUNCT
ajst-12267	78	1	by	by	ADP
ajst-12267	78	2	combining	combine	VERB
ajst-12267	78	3	the	the	DET
ajst-12267	78	4	loss	loss	NOUN
ajst-12267	78	5	function	function	NOUN
ajst-12267	78	6	and	and	CCONJ
ajst-12267	78	7	the	the	DET
ajst-12267	78	8	addition	addition	NOUN
ajst-12267	78	9	of	of	ADP
ajst-12267	78	10	attention	attention	NOUN
ajst-12267	78	11	mechanism	mechanism	NOUN
ajst-12267	78	12	,	,	PUNCT
ajst-12267	78	13	the	the	DET
ajst-12267	78	14	problems	problem	NOUN
ajst-12267	78	15	of	of	ADP
ajst-12267	78	16	pattern	pattern	NOUN
ajst-12267	78	17	collapse	collapse	NOUN
ajst-12267	78	18	in	in	ADP
ajst-12267	78	19	the	the	DET
ajst-12267	78	20	enhancement	enhancement	NOUN
ajst-12267	78	21	of	of	ADP
ajst-12267	78	22	pulmonary	pulmonary	ADJ
ajst-12267	78	23	nodule	nodule	NOUN
ajst-12267	78	24	data	datum	NOUN
ajst-12267	78	25	set	set	VERB
ajst-12267	78	26	are	be	AUX
ajst-12267	78	27	improved	improve	VERB
ajst-12267	78	28	.	.	PUNCT
ajst-12267	79	1	the	the	DET
ajst-12267	79	2	experimental	experimental	ADJ
ajst-12267	79	3	results	result	NOUN
ajst-12267	79	4	show	show	VERB
ajst-12267	79	5	that	that	SCONJ
ajst-12267	79	6	after	after	ADP
ajst-12267	79	7	modification	modification	NOUN
ajst-12267	79	8	,	,	PUNCT
ajst-12267	79	9	the	the	DET
ajst-12267	79	10	model	model	NOUN
ajst-12267	79	11	has	have	AUX
ajst-12267	79	12	been	be	AUX
ajst-12267	79	13	improved	improve	VERB
ajst-12267	79	14	in	in	ADP
ajst-12267	79	15	the	the	DET
ajst-12267	79	16	aspects	aspect	NOUN
ajst-12267	79	17	of	of	ADP
ajst-12267	79	18	picture	picture	NOUN
ajst-12267	79	19	diversity	diversity	NOUN
ajst-12267	79	20	,	,	PUNCT
ajst-12267	79	21	confidence	confidence	NOUN
ajst-12267	79	22	and	and	CCONJ
ajst-12267	79	23	fid	fid	NOUN
ajst-12267	79	24	.	.	PUNCT
ajst-12267	80	1	however	however	ADV
ajst-12267	80	2	,	,	PUNCT
ajst-12267	80	3	the	the	DET
ajst-12267	80	4	disadvantage	disadvantage	NOUN
ajst-12267	80	5	is	be	AUX
ajst-12267	80	6	that	that	SCONJ
ajst-12267	80	7	the	the	DET
ajst-12267	80	8	calculation	calculation	NOUN
ajst-12267	80	9	is	be	AUX
ajst-12267	80	10	relatively	relatively	ADV
ajst-12267	80	11	complicated	complicated	ADJ
ajst-12267	80	12	.	.	PUNCT
ajst-12267	81	1	in	in	ADP
ajst-12267	81	2	the	the	DET
ajst-12267	81	3	future	future	NOUN
ajst-12267	81	4	,	,	PUNCT
ajst-12267	81	5	a	a	DET
ajst-12267	81	6	relatively	relatively	ADV
ajst-12267	81	7	simple	simple	ADJ
ajst-12267	81	8	algorithm	algorithm	NOUN
ajst-12267	81	9	improvement	improvement	NOUN
ajst-12267	81	10	method	method	NOUN
ajst-12267	81	11	will	will	AUX
ajst-12267	81	12	be	be	AUX
ajst-12267	81	13	studied	study	VERB
ajst-12267	81	14	on	on	ADP
ajst-12267	81	15	the	the	DET
ajst-12267	81	16	basis	basis	NOUN
ajst-12267	81	17	of	of	ADP
ajst-12267	81	18	ensuring	ensure	VERB
ajst-12267	81	19	the	the	DET
ajst-12267	81	20	diversity	diversity	NOUN
ajst-12267	81	21	of	of	ADP
ajst-12267	81	22	the	the	DET
ajst-12267	81	23	generated	generate	VERB
ajst-12267	81	24	images	image	NOUN
ajst-12267	81	25	.	.	PUNCT
ajst-12267	82	1	references	reference	NOUN
ajst-12267	82	2	[	[	X
ajst-12267	82	3	1	1	NUM
ajst-12267	82	4	]	]	X
ajst-12267	82	5	li	li	PROPN
ajst-12267	82	6	yang	yang	PROPN
ajst-12267	82	7	,	,	PUNCT
ajst-12267	82	8	gao	gao	PROPN
ajst-12267	82	9	shiqi	shiqi	PROPN
ajst-12267	82	10	.	.	PUNCT
ajst-12267	83	1	lung	lung	NOUN
ajst-12267	83	2	nodule	nodule	NOUN
ajst-12267	83	3	detection	detection	NOUN
ajst-12267	83	4	system	system	NOUN
ajst-12267	83	5	based	base	VERB
ajst-12267	83	6	on	on	ADP
ajst-12267	83	7	data	datum	NOUN
ajst-12267	83	8	augmentation	augmentation	NOUN
ajst-12267	83	9	and	and	CCONJ
ajst-12267	83	10	attention	attention	NOUN
ajst-12267	83	11	mechanism[j	mechanism[j	PROPN
ajst-12267	83	12	]	]	PUNCT
ajst-12267	83	13	.	.	PUNCT
ajst-12267	84	1	journal	journal	PROPN
ajst-12267	84	2	of	of	ADP
ajst-12267	84	3	beijing	beijing	PROPN
ajst-12267	84	4	university	university	PROPN
ajst-12267	84	5	of	of	ADP
ajst-12267	84	6	posts	post	NOUN
ajst-12267	84	7	and	and	CCONJ
ajst-12267	84	8	telecommunications	telecommunication	NOUN
ajst-12267	84	9	,	,	PUNCT
ajst-12267	84	10	2022	2022	NUM
ajst-12267	84	11	,	,	PUNCT
ajst-12267	84	12	45(4	45(4	NUM
ajst-12267	84	13	):	):	PUNCT
ajst-12267	84	14	25	25	NUM
ajst-12267	84	15	-	-	SYM
ajst-12267	84	16	30	30	NUM
ajst-12267	84	17	.	.	PUNCT
ajst-12267	85	1	[	[	X
ajst-12267	85	2	2	2	X
ajst-12267	85	3	]	]	PUNCT
ajst-12267	85	4	qi	qi	PROPN
ajst-12267	85	5	yongjun	yongjun	PROPN
ajst-12267	85	6	,	,	PUNCT
ajst-12267	85	7	gu	gu	PROPN
ajst-12267	85	8	junhua	junhua	PROPN
ajst-12267	85	9	,	,	PUNCT
ajst-12267	85	10	li	li	PROPN
ajst-12267	86	1	weixun	weixun	PROPN
ajst-12267	86	2	et	et	PROPN
ajst-12267	86	3	al	al	PROPN
ajst-12267	86	4	.	.	PUNCT
ajst-12267	87	1	pulmonary	pulmonary	ADJ
ajst-12267	87	2	nodule	nodule	NOUN
ajst-12267	87	3	image	image	NOUN
ajst-12267	87	4	data	datum	NOUN
ajst-12267	87	5	enhancement	enhancement	NOUN
ajst-12267	87	6	method	method	NOUN
ajst-12267	87	7	based	base	VERB
ajst-12267	87	8	on	on	ADP
ajst-12267	87	9	generative	generative	ADJ
ajst-12267	87	10	adversarial	adversarial	ADJ
ajst-12267	87	11	network	network	NOUN
ajst-12267	87	12	[	[	PUNCT
ajst-12267	87	13	c]//	c]//	PROPN
ajst-12267	87	14	network	network	NOUN
ajst-12267	87	15	application	application	NOUN
ajst-12267	87	16	branch	branch	NOUN
ajst-12267	87	17	of	of	ADP
ajst-12267	87	18	china	china	PROPN
ajst-12267	87	19	computer	computer	PROPN
ajst-12267	87	20	users	users	PROPN
ajst-12267	87	21	association	association	PROPN
ajst-12267	87	22	.	.	PUNCT
ajst-12267	88	1	proceedings	proceeding	NOUN
ajst-12267	88	2	of	of	ADP
ajst-12267	88	3	the	the	DET
ajst-12267	88	4	23rd	23rd	ADJ
ajst-12267	88	5	annual	annual	ADJ
ajst-12267	88	6	conference	conference	NOUN
ajst-12267	88	7	on	on	ADP
ajst-12267	88	8	new	new	ADJ
ajst-12267	88	9	network	network	NOUN
ajst-12267	88	10	technologies	technology	NOUN
ajst-12267	88	11	and	and	CCONJ
ajst-12267	88	12	applications	application	NOUN
ajst-12267	88	13	,	,	PUNCT
ajst-12267	88	14	china	china	PROPN
ajst-12267	88	15	computer	computer	NOUN
ajst-12267	88	16	users	users	PROPN
ajst-12267	88	17	association	association	PROPN
ajst-12267	88	18	network	network	NOUN
ajst-12267	88	19	application	application	NOUN
ajst-12267	88	20	branch	branch	NOUN
ajst-12267	88	21	,	,	PUNCT
ajst-12267	88	22	2019	2019	NUM
ajst-12267	88	23	.	.	PUNCT
ajst-12267	89	1	editorial	editorial	PROPN
ajst-12267	89	2	board	board	NOUN
ajst-12267	89	3	of	of	ADP
ajst-12267	89	4	computer	computer	NOUN
ajst-12267	89	5	science	science	NOUN
ajst-12267	89	6	science	science	NOUN
ajst-12267	89	7	)	)	PUNCT
ajst-12267	89	8	,	,	PUNCT
ajst-12267	89	9	2019	2019	NUM
ajst-12267	89	10	-	-	SYM
ajst-12267	89	11	5	5	NUM
ajst-12267	89	12	.	.	PUNCT
ajst-12267	89	13	doi	doi	NOUN
ajst-12267	89	14	:	:	PUNCT
ajst-12267	89	15	10.26914	10.26914	NUM
ajst-12267	89	16	/	/	SYM
ajst-12267	89	17	arthur	arthur	PROPN
ajst-12267	89	18	c.	c.	PROPN
ajst-12267	89	19	nkihy	nkihy	PROPN
ajst-12267	89	20	.	.	PUNCT
ajst-12267	90	1	2019.004458	2019.004458	NOUN
ajst-12267	90	2	.	.	PUNCT
ajst-12267	91	1	[	[	X
ajst-12267	91	2	3	3	X
ajst-12267	91	3	]	]	X
ajst-12267	91	4	kazeminia	kazeminia	NOUN
ajst-12267	91	5	s	s	PROPN
ajst-12267	91	6	,	,	PUNCT
ajst-12267	91	7	baur	baur	PROPN
ajst-12267	91	8	c	c	PROPN
ajst-12267	91	9	,	,	PUNCT
ajst-12267	91	10	kuijper	kuijper	PROPN
ajst-12267	91	11	a	a	X
ajst-12267	91	12	,	,	PUNCT
ajst-12267	91	13	et	et	PROPN
ajst-12267	91	14	al	al	PROPN
ajst-12267	91	15	.	.	PROPN
ajst-12267	91	16	gans	gans	PROPN
ajst-12267	91	17	for	for	ADP
ajst-12267	91	18	medical	medical	ADJ
ajst-12267	91	19	image	image	NOUN
ajst-12267	91	20	analysis[j	analysis[j	PROPN
ajst-12267	91	21	]	]	PUNCT
ajst-12267	91	22	.	.	PUNCT
ajst-12267	92	1	artificial	artificial	ADJ
ajst-12267	92	2	intelligence	intelligence	NOUN
ajst-12267	92	3	in	in	ADP
ajst-12267	92	4	medicine	medicine	NOUN
ajst-12267	92	5	,	,	PUNCT
ajst-12267	92	6	2020	2020	NUM
ajst-12267	92	7	,	,	PUNCT
ajst-12267	92	8	109	109	NUM
ajst-12267	92	9	:	:	SYM
ajst-12267	92	10	101938	101938	NUM
ajst-12267	92	11	.	.	PUNCT
ajst-12267	93	1	[	[	X
ajst-12267	93	2	4	4	NUM
ajst-12267	93	3	]	]	X
ajst-12267	93	4	fan	fan	PROPN
ajst-12267	93	5	li	li	PROPN
ajst-12267	93	6	.	.	PROPN
ajst-12267	93	7	research	research	NOUN
ajst-12267	93	8	and	and	CCONJ
ajst-12267	93	9	application	application	NOUN
ajst-12267	93	10	of	of	ADP
ajst-12267	93	11	image	image	NOUN
ajst-12267	93	12	data	datum	NOUN
ajst-12267	93	13	enhancement	enhancement	NOUN
ajst-12267	93	14	technology	technology	NOUN
ajst-12267	93	15	based	base	VERB
ajst-12267	93	16	on	on	ADP
ajst-12267	93	17	generative	generative	ADJ
ajst-12267	93	18	adversarial	adversarial	ADJ
ajst-12267	93	19	network	network	NOUN
ajst-12267	94	1	[	[	X
ajst-12267	94	2	d	d	X
ajst-12267	94	3	]	]	X
ajst-12267	94	4	.	.	PUNCT
ajst-12267	95	1	186	186	NUM
ajst-12267	95	2	zhejiang	zhejiang	PROPN
ajst-12267	95	3	university	university	PROPN
ajst-12267	95	4	,	,	PUNCT
ajst-12267	95	5	2022	2022	NUM
ajst-12267	95	6	.	.	PUNCT
ajst-12267	96	1	doi	doi	NOUN
ajst-12267	96	2	:	:	PUNCT
ajst-12267	96	3	10.27461	10.27461	NUM
ajst-12267	96	4	/	/	SYM
ajst-12267	96	5	,	,	PUNCT
ajst-12267	97	1	dc	dc	PROPN
ajst-12267	97	2	nki	nki	PROPN
ajst-12267	97	3	.	.	PUNCT
ajst-12267	97	4	gzjdx	gzjdx	PROPN
ajst-12267	97	5	.	.	PUNCT
ajst-12267	98	1	2022.001690	2022.001690	VERB
ajst-12267	98	2	.	.	PUNCT
ajst-12267	99	1	[	[	X
ajst-12267	99	2	5	5	NUM
ajst-12267	99	3	]	]	X
ajst-12267	99	4	goodfellow	goodfellow	PROPN
ajst-12267	99	5	i	i	PROPN
ajst-12267	99	6	,	,	PUNCT
ajst-12267	99	7	pouget	pouget	NOUN
ajst-12267	99	8	-	-	PUNCT
ajst-12267	99	9	abadie	abadie	NOUN
ajst-12267	99	10	j	j	PROPN
ajst-12267	99	11	,	,	PUNCT
ajst-12267	99	12	mirza	mirza	PROPN
ajst-12267	99	13	m	m	PROPN
ajst-12267	99	14	,	,	PUNCT
ajst-12267	99	15	et	et	PROPN
ajst-12267	99	16	al	al	PROPN
ajst-12267	99	17	.	.	PROPN
ajst-12267	99	18	generative	generative	ADJ
ajst-12267	99	19	adversarial	adversarial	ADJ
ajst-12267	99	20	nets	net	NOUN
ajst-12267	100	1	[	[	X
ajst-12267	100	2	j	j	X
ajst-12267	100	3	]	]	X
ajst-12267	100	4	.	.	PUNCT
ajst-12267	101	1	advances	advance	NOUN
ajst-12267	101	2	in	in	ADP
ajst-12267	101	3	neural	neural	ADJ
ajst-12267	101	4	information	information	NOUN
ajst-12267	101	5	processing	processing	NOUN
ajst-12267	101	6	systems	system	NOUN
ajst-12267	101	7	,	,	PUNCT
ajst-12267	101	8	2014	2014	NUM
ajst-12267	101	9	,	,	PUNCT
ajst-12267	101	10	27	27	NUM
ajst-12267	102	1	[	[	SYM
ajst-12267	102	2	6	6	NUM
ajst-12267	102	3	]	]	PUNCT
ajst-12267	102	4	gong	gong	NOUN
ajst-12267	102	5	m	m	PROPN
ajst-12267	102	6	,	,	PUNCT
ajst-12267	102	7	niu	niu	PROPN
ajst-12267	102	8	x	x	PROPN
ajst-12267	102	9	,	,	PUNCT
ajst-12267	102	10	zhang	zhang	PROPN
ajst-12267	102	11	p	p	PROPN
ajst-12267	102	12	,	,	PUNCT
ajst-12267	102	13	et	et	PROPN
ajst-12267	102	14	al	al	PROPN
ajst-12267	102	15	.	.	PROPN
ajst-12267	102	16	generative	generative	ADJ
ajst-12267	102	17	adversarial	adversarial	ADJ
ajst-12267	102	18	networks	network	NOUN
ajst-12267	102	19	for	for	ADP
ajst-12267	102	20	change	change	NOUN
ajst-12267	102	21	detection	detection	NOUN
ajst-12267	102	22	in	in	ADP
ajst-12267	102	23	multispectral	multispectral	ADJ
ajst-12267	102	24	imagery	imagery	NOUN
ajst-12267	103	1	[	[	X
ajst-12267	103	2	j	j	X
ajst-12267	103	3	]	]	X
ajst-12267	103	4	.	.	PUNCT
ajst-12267	104	1	ieee	ieee	PROPN
ajst-12267	104	2	geoscience	geoscience	PROPN
ajst-12267	104	3	and	and	CCONJ
ajst-12267	104	4	remote	remote	ADJ
ajst-12267	104	5	sensing	sense	VERB
ajst-12267	104	6	letters	letter	NOUN
ajst-12267	104	7	,	,	PUNCT
ajst-12267	104	8	2017	2017	NUM
ajst-12267	104	9	,	,	PUNCT
ajst-12267	104	10	14	14	NUM
ajst-12267	104	11	(	(	PUNCT
ajst-12267	104	12	12	12	NUM
ajst-12267	104	13	)	)	PUNCT
ajst-12267	104	14	;	;	PUNCT
ajst-12267	104	15	2310	2310	NUM
ajst-12267	104	16	-	-	SYM
ajst-12267	104	17	2314	2314	NUM
ajst-12267	104	18	.	.	PUNCT
ajst-12267	105	1	[	[	X
ajst-12267	105	2	7	7	X
ajst-12267	105	3	]	]	X
ajst-12267	105	4	zhang	zhang	PROPN
ajst-12267	105	5	bin	bin	PROPN
ajst-12267	105	6	,	,	PUNCT
ajst-12267	105	7	et	et	PROPN
ajst-12267	105	8	al	al	PROPN
ajst-12267	105	9	.	.	PUNCT
ajst-12267	105	10	"	"	PUNCT
ajst-12267	106	1	overview	overview	NOUN
ajst-12267	106	2	of	of	ADP
ajst-12267	106	3	generative	generative	ADJ
ajst-12267	106	4	adversarial	adversarial	ADJ
ajst-12267	106	5	network	network	NOUN
ajst-12267	106	6	improvement	improvement	NOUN
ajst-12267	106	7	perspectives	perspective	NOUN
ajst-12267	106	8	and	and	CCONJ
ajst-12267	106	9	applications	application	NOUN
ajst-12267	106	10	.	.	PUNCT
ajst-12267	106	11	"	"	PUNCT
ajst-12267	107	1	computer	computer	NOUN
ajst-12267	107	2	application	application	NOUN
ajst-12267	107	3	research	research	NOUN
ajst-12267	107	4	...	...	PUNCT
ajst-12267	108	1	doi	doi	NOUN
ajst-12267	108	2	:	:	PUNCT
ajst-12267	108	3	10.19734	10.19734	NUM
ajst-12267	108	4	/	/	SYM
ajst-12267	108	5	j.i	j.i	X
ajst-12267	108	6	ssn	ssn	PROPN
ajst-12267	108	7	.	.	PUNCT
ajst-12267	108	8	10013695.2022.08.0410	10013695.2022.08.0410	NUM
ajst-12267	108	9	.	.	PUNCT
ajst-12267	109	1	[	[	X
ajst-12267	109	2	8	8	NUM
ajst-12267	109	3	]	]	X
ajst-12267	109	4	yang	yang	PROPN
ajst-12267	109	5	yu	yu	PROPN
ajst-12267	109	6	,	,	PUNCT
ajst-12267	109	7	based	base	VERB
ajst-12267	109	8	on	on	ADP
ajst-12267	109	9	generating	generate	VERB
ajst-12267	109	10	method	method	NOUN
ajst-12267	109	11	of	of	ADP
ajst-12267	109	12	image	image	NOUN
ajst-12267	109	13	data	datum	NOUN
ajst-12267	109	14	enhancement	enhancement	NOUN
ajst-12267	109	15	against	against	ADP
ajst-12267	109	16	network	network	NOUN
ajst-12267	109	17	research	research	NOUN
ajst-12267	109	18	[	[	X
ajst-12267	109	19	d	d	X
ajst-12267	109	20	]	]	X
ajst-12267	109	21	.	.	PUNCT
ajst-12267	110	1	strategic	strategic	ADJ
ajst-12267	110	2	support	support	NOUN
ajst-12267	110	3	troops	troop	NOUN
ajst-12267	110	4	information	information	NOUN
ajst-12267	110	5	engineering	engineering	PROPN
ajst-12267	110	6	university	university	NOUN
ajst-12267	110	7	,	,	PUNCT
ajst-12267	110	8	2022	2022	NUM
ajst-12267	110	9	.	.	PUNCT
ajst-12267	111	1	the	the	DET
ajst-12267	111	2	doi	doi	NOUN
ajst-12267	111	3	:	:	PUNCT
ajst-12267	111	4	10.27188	10.27188	NUM
ajst-12267	111	5	/	/	SYM
ajst-12267	111	6	,	,	PUNCT
ajst-12267	111	7	dc	dc	PROPN
ajst-12267	111	8	nki	nki	PROPN
ajst-12267	111	9	.	.	PUNCT
ajst-12267	111	10	gzjxu	gzjxu	PROPN
ajst-12267	111	11	.	.	PUNCT
ajst-12267	112	1	2022.000039	2022.000039	NUM
ajst-12267	112	2	.	.	PUNCT
ajst-12267	113	1	[	[	X
ajst-12267	113	2	9	9	NUM
ajst-12267	113	3	]	]	SYM
ajst-12267	113	4	jaderberg	jaderberg	PROPN
ajst-12267	113	5	,	,	PUNCT
ajst-12267	113	6	max	max	PROPN
ajst-12267	113	7	,	,	PUNCT
ajst-12267	113	8	karen	karen	PROPN
ajst-12267	113	9	simonyan	simonyan	PROPN
ajst-12267	113	10	,	,	PUNCT
ajst-12267	113	11	and	and	CCONJ
ajst-12267	113	12	andrew	andrew	PROPN
ajst-12267	113	13	zisserman	zisserman	PROPN
ajst-12267	113	14	.	.	PUNCT
ajst-12267	114	1	"	"	PUNCT
ajst-12267	114	2	spatial	spatial	ADJ
ajst-12267	114	3	transformer	transformer	NOUN
ajst-12267	114	4	networks	network	NOUN
ajst-12267	114	5	.	.	PUNCT
ajst-12267	114	6	"	"	PUNCT
ajst-12267	115	1	advances	advance	NOUN
ajst-12267	115	2	in	in	ADP
ajst-12267	115	3	neural	neural	ADJ
ajst-12267	115	4	information	information	NOUN
ajst-12267	115	5	processing	processing	NOUN
ajst-12267	115	6	systems	system	NOUN
ajst-12267	115	7	28	28	NUM
ajst-12267	115	8	(	(	PUNCT
ajst-12267	115	9	2015	2015	NUM
ajst-12267	115	10	)	)	PUNCT
ajst-12267	115	11	.	.	PUNCT