id	sid	tid	token	lemma	pos
cana-6113	1	1	communications	communication	NOUN
cana-6113	1	2	on	on	ADP
cana-6113	1	3	applied	apply	VERB
cana-6113	1	4	nonlinear	nonlinear	ADJ
cana-6113	1	5	analysis	analysis	NOUN
cana-6113	1	6	issn	issn	NOUN
cana-6113	1	7	:	:	PUNCT
cana-6113	1	8	1074	1074	NUM
cana-6113	1	9	-	-	PUNCT
cana-6113	1	10	133x	133x	NUM
cana-6113	1	11	vol	vol	VERB
cana-6113	1	12	32	32	NUM
cana-6113	1	13	no	no	NOUN
cana-6113	1	14	.	.	PUNCT
cana-6113	2	1	10s	10	NOUN
cana-6113	2	2	(	(	PUNCT
cana-6113	2	3	2025	2025	NUM
cana-6113	2	4	)	)	PUNCT
cana-6113	2	5	3403	3403	NOUN
cana-6113	2	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	2	7	deep	deep	ADJ
cana-6113	2	8	learning	learning	NOUN
cana-6113	2	9	-	-	PUNCT
cana-6113	2	10	powered	power	VERB
cana-6113	2	11	multi	multi	ADJ
cana-6113	2	12	-	-	ADJ
cana-6113	2	13	stage	stage	ADJ
cana-6113	2	14	image	image	NOUN
cana-6113	2	15	feature	feature	NOUN
cana-6113	2	16	processing	processing	NOUN
cana-6113	2	17	and	and	CCONJ
cana-6113	2	18	fusion	fusion	NOUN
cana-6113	2	19	approach	approach	NOUN
cana-6113	2	20	for	for	ADP
cana-6113	2	21	enhanced	enhanced	ADJ
cana-6113	2	22	visual	visual	ADJ
cana-6113	2	23	understanding	understanding	NOUN
cana-6113	2	24	cijin	cijin	PROPN
cana-6113	2	25	k	k	PROPN
cana-6113	2	26	paul1	paul1	PROPN
cana-6113	2	27	,	,	PUNCT
cana-6113	2	28	dr	dr	PROPN
cana-6113	2	29	.	.	PROPN
cana-6113	2	30	balaji	balaji	PROPN
cana-6113	2	31	venkateswaran2	venkateswaran2	PROPN
cana-6113	2	32	,	,	PUNCT
cana-6113	2	33	dr	dr	PROPN
cana-6113	2	34	.	.	PROPN
cana-6113	2	35	ajay	ajay	PROPN
cana-6113	2	36	sharma3	sharma3	PROPN
cana-6113	2	37	,	,	PUNCT
cana-6113	2	38	rohit	rohit	PROPN
cana-6113	2	39	gupta4	gupta4	PROPN
cana-6113	2	40	,	,	PUNCT
cana-6113	2	41	dr	dr	PROPN
cana-6113	2	42	.	.	PROPN
cana-6113	2	43	rakesh	rakesh	PROPN
cana-6113	2	44	sharma5	sharma5	PROPN
cana-6113	2	45	,	,	PUNCT
cana-6113	3	1	dr	dr	PROPN
cana-6113	3	2	.	.	PROPN
cana-6113	3	3	yashwant	yashwant	PROPN
cana-6113	3	4	singh	singh	PROPN
cana-6113	3	5	sangwan6	sangwan6	PROPN
cana-6113	3	6	1assistant	1assistant	NUM
cana-6113	3	7	professor	professor	NOUN
cana-6113	3	8	,	,	PUNCT
cana-6113	3	9	department	department	NOUN
cana-6113	3	10	of	of	ADP
cana-6113	3	11	computer	computer	NOUN
cana-6113	3	12	science	science	NOUN
cana-6113	3	13	,	,	PUNCT
cana-6113	3	14	union	union	PROPN
cana-6113	3	15	christian	christian	PROPN
cana-6113	3	16	college	college	PROPN
cana-6113	3	17	,	,	PUNCT
cana-6113	3	18	aluva	aluva	PROPN
cana-6113	3	19	(	(	PUNCT
cana-6113	3	20	kerala	kerala	PROPN
cana-6113	3	21	)	)	PUNCT
cana-6113	3	22	,	,	PUNCT
cana-6113	3	23	india	india	PROPN
cana-6113	3	24	.	.	PUNCT
cana-6113	3	25	email	email	NOUN
cana-6113	3	26	:	:	PUNCT
cana-6113	3	27	cijinkpaul@uccollege.edu.in	cijinkpaul@uccollege.edu.in	NOUN
cana-6113	3	28	2independent	2independent	NUM
cana-6113	3	29	researcher	researcher	NOUN
cana-6113	3	30	,	,	PUNCT
cana-6113	3	31	chennai	chennai	PROPN
cana-6113	3	32	(	(	PUNCT
cana-6113	3	33	tamil	tamil	PROPN
cana-6113	3	34	nadu	nadu	PROPN
cana-6113	3	35	)	)	PUNCT
cana-6113	3	36	,	,	PUNCT
cana-6113	3	37	india	india	PROPN
cana-6113	3	38	.	.	PUNCT
cana-6113	4	1	email	email	NOUN
cana-6113	4	2	:	:	PUNCT
cana-6113	4	3	balaji.venkateswaran@gmail.com	balaji.venkateswaran@gmail.com	X
cana-6113	4	4	3associate	3associate	NUM
cana-6113	4	5	professor	professor	NOUN
cana-6113	4	6	,	,	PUNCT
cana-6113	4	7	department	department	NOUN
cana-6113	4	8	of	of	ADP
cana-6113	4	9	computer	computer	NOUN
cana-6113	4	10	science	science	NOUN
cana-6113	4	11	,	,	PUNCT
cana-6113	4	12	gniot	gniot	PROPN
cana-6113	4	13	institute	institute	PROPN
cana-6113	4	14	of	of	ADP
cana-6113	4	15	professional	professional	PROPN
cana-6113	4	16	studies	study	NOUN
cana-6113	4	17	,	,	PUNCT
cana-6113	4	18	greater	great	ADJ
cana-6113	4	19	noida	noida	PROPN
cana-6113	4	20	(	(	PUNCT
cana-6113	4	21	u.p	u.p	PROPN
cana-6113	4	22	.	.	PROPN
cana-6113	4	23	)	)	PUNCT
cana-6113	4	24	,	,	PUNCT
cana-6113	4	25	india	india	PROPN
cana-6113	4	26	.	.	PUNCT
cana-6113	4	27	email	email	NOUN
cana-6113	4	28	:	:	PUNCT
cana-6113	4	29	ajay0202@gmail.com	ajay0202@gmail.com	X
cana-6113	5	1	4assistant	4assistant	NUM
cana-6113	5	2	professor	professor	NOUN
cana-6113	5	3	in	in	ADP
cana-6113	5	4	computer	computer	NOUN
cana-6113	5	5	applications	application	NOUN
cana-6113	5	6	,	,	PUNCT
cana-6113	5	7	psps	psps	NOUN
cana-6113	5	8	govt	govt	PROPN
cana-6113	5	9	college	college	PROPN
cana-6113	5	10	for	for	ADP
cana-6113	5	11	women	woman	NOUN
cana-6113	5	12	,	,	PUNCT
cana-6113	5	13	gandhi	gandhi	PROPN
cana-6113	5	14	nagar	nagar	PROPN
cana-6113	5	15	,	,	PUNCT
cana-6113	5	16	jammu	jammu	PROPN
cana-6113	5	17	(	(	PUNCT
cana-6113	5	18	j&k	j&k	PROPN
cana-6113	5	19	)	)	PUNCT
cana-6113	5	20	,	,	PUNCT
cana-6113	5	21	india	india	PROPN
cana-6113	5	22	,	,	PUNCT
cana-6113	5	23	email	email	NOUN
cana-6113	5	24	:	:	PUNCT
cana-6113	5	25	gupta00shagun@gmail.com	gupta00shagun@gmail.com	NUM
cana-6113	5	26	5assistant	5assistant	NUM
cana-6113	5	27	professor	professor	NOUN
cana-6113	5	28	of	of	ADP
cana-6113	5	29	computer	computer	NOUN
cana-6113	5	30	science	science	PROPN
cana-6113	5	31	,	,	PUNCT
cana-6113	5	32	crm	crm	PROPN
cana-6113	5	33	jat	jat	PROPN
cana-6113	5	34	college	college	PROPN
cana-6113	5	35	,	,	PUNCT
cana-6113	5	36	hisar	hisar	PROPN
cana-6113	5	37	(	(	PUNCT
cana-6113	5	38	haryana	haryana	PROPN
cana-6113	5	39	)	)	PUNCT
cana-6113	5	40	,	,	PUNCT
cana-6113	5	41	india	india	PROPN
cana-6113	5	42	email	email	NOUN
cana-6113	5	43	:	:	PUNCT
cana-6113	5	44	rakeshsharma3112@gmail.com	rakeshsharma3112@gmail.com	X
cana-6113	6	1	6assistant	6assistant	NUM
cana-6113	6	2	professor	professor	NOUN
cana-6113	6	3	of	of	ADP
cana-6113	6	4	computer	computer	NOUN
cana-6113	6	5	science	science	NOUN
cana-6113	6	6	,	,	PUNCT
cana-6113	6	7	g.g.j	g.g.j	NOUN
cana-6113	6	8	.	.	PUNCT
cana-6113	6	9	govt	govt	PROPN
cana-6113	6	10	.	.	PUNCT
cana-6113	7	1	college	college	NOUN
cana-6113	7	2	,	,	PUNCT
cana-6113	7	3	hisar	hisar	PROPN
cana-6113	7	4	(	(	PUNCT
cana-6113	7	5	haryana	haryana	PROPN
cana-6113	7	6	)	)	PUNCT
cana-6113	7	7	,	,	PUNCT
cana-6113	7	8	india	india	PROPN
cana-6113	7	9	email	email	NOUN
cana-6113	7	10	:	:	PUNCT
cana-6113	8	1	yssangwan@gmail.com	yssangwan@gmail.com	PROPN
cana-6113	8	2	article	article	PROPN
cana-6113	8	3	history	history	NOUN
cana-6113	8	4	:	:	PUNCT
cana-6113	8	5	received	receive	VERB
cana-6113	8	6	:	:	PUNCT
cana-6113	8	7	19	19	NUM
cana-6113	8	8	-	-	PUNCT
cana-6113	8	9	07	07	NUM
cana-6113	8	10	-	-	PUNCT
cana-6113	8	11	2025	2025	NUM
cana-6113	8	12	revised	revise	VERB
cana-6113	8	13	:	:	PUNCT
cana-6113	8	14	24	24	NUM
cana-6113	8	15	-	-	SYM
cana-6113	8	16	08	08	NUM
cana-6113	8	17	-	-	PUNCT
cana-6113	8	18	2025	2025	NUM
cana-6113	8	19	accepted	accept	VERB
cana-6113	8	20	:	:	PUNCT
cana-6113	8	21	21	21	NUM
cana-6113	8	22	-	-	SYM
cana-6113	8	23	09	09	NUM
cana-6113	8	24	-	-	PUNCT
cana-6113	8	25	2025	2025	NUM
cana-6113	8	26	abstract	abstract	NOUN
cana-6113	8	27	:	:	PUNCT
cana-6113	8	28	in	in	ADP
cana-6113	8	29	recent	recent	ADJ
cana-6113	8	30	years	year	NOUN
cana-6113	8	31	,	,	PUNCT
cana-6113	8	32	deep	deep	ADJ
cana-6113	8	33	learning	learning	NOUN
cana-6113	8	34	has	have	VERB
cana-6113	8	35	significantly	significantly	ADV
cana-6113	8	36	advanced	advance	VERB
cana-6113	8	37	medical	medical	ADJ
cana-6113	8	38	image	image	NOUN
cana-6113	8	39	analysis	analysis	NOUN
cana-6113	8	40	,	,	PUNCT
cana-6113	8	41	offering	offer	VERB
cana-6113	8	42	robust	robust	ADJ
cana-6113	8	43	capabilities	capability	NOUN
cana-6113	8	44	for	for	ADP
cana-6113	8	45	feature	feature	NOUN
cana-6113	8	46	extraction	extraction	NOUN
cana-6113	8	47	,	,	PUNCT
cana-6113	8	48	pattern	pattern	NOUN
cana-6113	8	49	recognition	recognition	NOUN
cana-6113	8	50	,	,	PUNCT
cana-6113	8	51	and	and	CCONJ
cana-6113	8	52	diagnostic	diagnostic	ADJ
cana-6113	8	53	prediction	prediction	NOUN
cana-6113	8	54	.	.	PUNCT
cana-6113	9	1	this	this	DET
cana-6113	9	2	paper	paper	NOUN
cana-6113	9	3	proposes	propose	VERB
cana-6113	9	4	a	a	DET
cana-6113	9	5	deep	deep	ADJ
cana-6113	9	6	learningbased	learningbase	VERB
cana-6113	9	7	multi	multi	ADJ
cana-6113	9	8	-	-	ADJ
cana-6113	9	9	stage	stage	ADJ
cana-6113	9	10	feature	feature	NOUN
cana-6113	9	11	processing	processing	NOUN
cana-6113	9	12	and	and	CCONJ
cana-6113	9	13	fusion	fusion	NOUN
cana-6113	9	14	framework	framework	NOUN
cana-6113	9	15	for	for	ADP
cana-6113	9	16	cervical	cervical	ADJ
cana-6113	9	17	lesion	lesion	NOUN
cana-6113	9	18	prediction	prediction	NOUN
cana-6113	9	19	,	,	PUNCT
cana-6113	9	20	which	which	DET
cana-6113	9	21	integrates	integrate	VERB
cana-6113	9	22	colposcopy	colposcopy	NOUN
cana-6113	9	23	images	image	NOUN
cana-6113	9	24	and	and	CCONJ
cana-6113	9	25	patient	patient	ADJ
cana-6113	9	26	clinical	clinical	ADJ
cana-6113	9	27	data	datum	NOUN
cana-6113	9	28	.	.	PUNCT
cana-6113	10	1	the	the	DET
cana-6113	10	2	framework	framework	NOUN
cana-6113	10	3	processes	process	NOUN
cana-6113	10	4	colposcopy	colposcopy	VERB
cana-6113	10	5	images	image	NOUN
cana-6113	10	6	through	through	ADP
cana-6113	10	7	hierarchical	hierarchical	ADJ
cana-6113	10	8	convolutional	convolutional	ADJ
cana-6113	10	9	neural	neural	ADJ
cana-6113	10	10	networks	network	NOUN
cana-6113	10	11	to	to	PART
cana-6113	10	12	extract	extract	VERB
cana-6113	10	13	lowand	lowand	ADJ
cana-6113	10	14	high	high	ADJ
cana-6113	10	15	-	-	PUNCT
cana-6113	10	16	level	level	NOUN
cana-6113	10	17	visual	visual	ADJ
cana-6113	10	18	features	feature	NOUN
cana-6113	10	19	,	,	PUNCT
cana-6113	10	20	while	while	SCONJ
cana-6113	10	21	clinical	clinical	ADJ
cana-6113	10	22	data	datum	NOUN
cana-6113	10	23	such	such	ADJ
cana-6113	10	24	as	as	ADP
cana-6113	10	25	age	age	NOUN
cana-6113	10	26	,	,	PUNCT
cana-6113	10	27	hpv	hpv	NOUN
cana-6113	10	28	status	status	NOUN
cana-6113	10	29	,	,	PUNCT
cana-6113	10	30	and	and	CCONJ
cana-6113	10	31	medical	medical	ADJ
cana-6113	10	32	history	history	NOUN
cana-6113	10	33	are	be	AUX
cana-6113	10	34	processed	process	VERB
cana-6113	10	35	using	use	VERB
cana-6113	10	36	machine	machine	NOUN
cana-6113	10	37	learning	learning	NOUN
cana-6113	10	38	algorithms	algorithm	NOUN
cana-6113	10	39	.	.	PUNCT
cana-6113	11	1	preprocessing	preprocesse	VERB
cana-6113	11	2	techniques	technique	NOUN
cana-6113	11	3	including	include	VERB
cana-6113	11	4	contrast	contrast	NOUN
cana-6113	11	5	enhancement	enhancement	NOUN
cana-6113	11	6	,	,	PUNCT
cana-6113	11	7	noise	noise	NOUN
cana-6113	11	8	reduction	reduction	NOUN
cana-6113	11	9	,	,	PUNCT
cana-6113	11	10	and	and	CCONJ
cana-6113	11	11	region	region	NOUN
cana-6113	11	12	-	-	PUNCT
cana-6113	11	13	of	of	ADP
cana-6113	11	14	-	-	PUNCT
cana-6113	11	15	interest	interest	NOUN
cana-6113	11	16	segmentation	segmentation	NOUN
cana-6113	11	17	are	be	AUX
cana-6113	11	18	applied	apply	VERB
cana-6113	11	19	to	to	ADP
cana-6113	11	20	images	image	NOUN
cana-6113	11	21	to	to	PART
cana-6113	11	22	improve	improve	VERB
cana-6113	11	23	feature	feature	NOUN
cana-6113	11	24	quality	quality	NOUN
cana-6113	11	25	.	.	PUNCT
cana-6113	12	1	the	the	DET
cana-6113	12	2	extracted	extract	VERB
cana-6113	12	3	features	feature	NOUN
cana-6113	12	4	from	from	ADP
cana-6113	12	5	both	both	DET
cana-6113	12	6	modalities	modality	NOUN
cana-6113	12	7	are	be	AUX
cana-6113	12	8	fused	fuse	VERB
cana-6113	12	9	into	into	ADP
cana-6113	12	10	a	a	DET
cana-6113	12	11	unified	unified	ADJ
cana-6113	12	12	representation	representation	NOUN
cana-6113	12	13	,	,	PUNCT
cana-6113	12	14	which	which	PRON
cana-6113	12	15	is	be	AUX
cana-6113	12	16	then	then	ADV
cana-6113	12	17	fed	feed	VERB
cana-6113	12	18	into	into	ADP
cana-6113	12	19	a	a	DET
cana-6113	12	20	classifier	classifier	NOUN
cana-6113	12	21	for	for	ADP
cana-6113	12	22	accurate	accurate	ADJ
cana-6113	12	23	prediction	prediction	NOUN
cana-6113	12	24	.	.	PUNCT
cana-6113	13	1	attention	attention	NOUN
cana-6113	13	2	mechanisms	mechanism	NOUN
cana-6113	13	3	are	be	AUX
cana-6113	13	4	employed	employ	VERB
cana-6113	13	5	to	to	PART
cana-6113	13	6	focus	focus	VERB
cana-6113	13	7	on	on	ADP
cana-6113	13	8	diagnostically	diagnostically	ADV
cana-6113	13	9	relevant	relevant	ADJ
cana-6113	13	10	regions	region	NOUN
cana-6113	13	11	,	,	PUNCT
cana-6113	13	12	enhancing	enhance	VERB
cana-6113	13	13	both	both	DET
cana-6113	13	14	model	model	NOUN
cana-6113	13	15	performance	performance	NOUN
cana-6113	13	16	and	and	CCONJ
cana-6113	13	17	interpretability	interpretability	NOUN
cana-6113	13	18	.	.	PUNCT
cana-6113	14	1	extensive	extensive	ADJ
cana-6113	14	2	experiments	experiment	NOUN
cana-6113	14	3	demonstrate	demonstrate	VERB
cana-6113	14	4	that	that	SCONJ
cana-6113	14	5	the	the	DET
cana-6113	14	6	proposed	propose	VERB
cana-6113	14	7	approach	approach	NOUN
cana-6113	14	8	outperforms	outperform	VERB
cana-6113	14	9	conventional	conventional	ADJ
cana-6113	14	10	deep	deep	ADJ
cana-6113	14	11	learning	learning	NOUN
cana-6113	14	12	models	model	NOUN
cana-6113	14	13	such	such	ADJ
cana-6113	14	14	as	as	ADP
cana-6113	14	15	alexnet	alexnet	NOUN
cana-6113	14	16	,	,	PUNCT
cana-6113	14	17	vgg16	vgg16	NOUN
cana-6113	14	18	,	,	PUNCT
cana-6113	14	19	and	and	CCONJ
cana-6113	14	20	inceptionv3	inceptionv3	NOUN
cana-6113	14	21	,	,	PUNCT
cana-6113	14	22	achieving	achieve	VERB
cana-6113	14	23	an	an	DET
cana-6113	14	24	accuracy	accuracy	NOUN
cana-6113	14	25	of	of	ADP
cana-6113	14	26	96.85	96.85	NUM
cana-6113	14	27	%	%	NOUN
cana-6113	14	28	,	,	PUNCT
cana-6113	14	29	with	with	ADP
cana-6113	14	30	corresponding	correspond	VERB
cana-6113	14	31	improvements	improvement	NOUN
cana-6113	14	32	in	in	ADP
cana-6113	14	33	precision	precision	NOUN
cana-6113	14	34	,	,	PUNCT
cana-6113	14	35	recall	recall	NOUN
cana-6113	14	36	,	,	PUNCT
cana-6113	14	37	and	and	CCONJ
cana-6113	14	38	f1	f1	NOUN
cana-6113	14	39	-	-	PUNCT
cana-6113	14	40	score	score	NOUN
cana-6113	14	41	.	.	PUNCT
cana-6113	15	1	the	the	DET
cana-6113	15	2	results	result	NOUN
cana-6113	15	3	highlight	highlight	VERB
cana-6113	15	4	the	the	DET
cana-6113	15	5	effectiveness	effectiveness	NOUN
cana-6113	15	6	of	of	ADP
cana-6113	15	7	combining	combine	VERB
cana-6113	15	8	multi	multi	ADJ
cana-6113	15	9	-	-	ADJ
cana-6113	15	10	modal	modal	ADJ
cana-6113	15	11	data	datum	NOUN
cana-6113	15	12	and	and	CCONJ
cana-6113	15	13	multi	multi	ADJ
cana-6113	15	14	-	-	ADJ
cana-6113	15	15	stage	stage	ADJ
cana-6113	15	16	feature	feature	NOUN
cana-6113	15	17	extraction	extraction	NOUN
cana-6113	15	18	in	in	ADP
cana-6113	15	19	medical	medical	ADJ
cana-6113	15	20	diagnostics	diagnostic	NOUN
cana-6113	15	21	.	.	PUNCT
cana-6113	16	1	the	the	DET
cana-6113	16	2	proposed	propose	VERB
cana-6113	16	3	framework	framework	NOUN
cana-6113	16	4	offers	offer	VERB
cana-6113	16	5	a	a	DET
cana-6113	16	6	reliable	reliable	ADJ
cana-6113	16	7	and	and	CCONJ
cana-6113	16	8	interpretable	interpretable	ADJ
cana-6113	16	9	solution	solution	NOUN
cana-6113	16	10	for	for	ADP
cana-6113	16	11	automated	automated	ADJ
cana-6113	16	12	cervical	cervical	ADJ
cana-6113	16	13	cancer	cancer	NOUN
cana-6113	16	14	screening	screening	NOUN
cana-6113	16	15	and	and	CCONJ
cana-6113	16	16	has	have	VERB
cana-6113	16	17	potential	potential	ADJ
cana-6113	16	18	applications	application	NOUN
cana-6113	16	19	in	in	ADP
cana-6113	16	20	other	other	ADJ
cana-6113	16	21	image	image	NOUN
cana-6113	16	22	-	-	PUNCT
cana-6113	16	23	based	base	VERB
cana-6113	16	24	clinical	clinical	ADJ
cana-6113	16	25	prediction	prediction	NOUN
cana-6113	16	26	tasks	task	NOUN
cana-6113	16	27	.	.	PUNCT
cana-6113	17	1	by	by	ADP
cana-6113	17	2	integrating	integrate	VERB
cana-6113	17	3	structured	structure	VERB
cana-6113	17	4	clinical	clinical	ADJ
cana-6113	17	5	information	information	NOUN
cana-6113	17	6	with	with	ADP
cana-6113	17	7	unstructured	unstructured	ADJ
cana-6113	17	8	image	image	NOUN
cana-6113	17	9	data	datum	NOUN
cana-6113	17	10	,	,	PUNCT
cana-6113	17	11	the	the	DET
cana-6113	17	12	mailto:balaji.venkateswaran@gmail.com	mailto:balaji.venkateswaran@gmail.com	PROPN
cana-6113	17	13	mailto:rakeshsharma3112@gmail.com	mailto:rakeshsharma3112@gmail.com	X
cana-6113	17	14	mailto:yssangwan@gmail.com	mailto:yssangwan@gmail.com	PROPN
cana-6113	17	15	communications	communication	NOUN
cana-6113	17	16	on	on	ADP
cana-6113	17	17	applied	apply	VERB
cana-6113	17	18	nonlinear	nonlinear	ADJ
cana-6113	17	19	analysis	analysis	NOUN
cana-6113	17	20	issn	issn	NOUN
cana-6113	17	21	:	:	PUNCT
cana-6113	17	22	1074	1074	NUM
cana-6113	17	23	-	-	PUNCT
cana-6113	17	24	133x	133x	NUM
cana-6113	17	25	vol	vol	VERB
cana-6113	17	26	32	32	NUM
cana-6113	17	27	no	no	NOUN
cana-6113	17	28	.	.	PUNCT
cana-6113	18	1	10s	10	NOUN
cana-6113	18	2	(	(	PUNCT
cana-6113	18	3	2025	2025	NUM
cana-6113	18	4	)	)	PUNCT
cana-6113	18	5	3404	3404	NUM
cana-6113	18	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	18	7	system	system	NOUN
cana-6113	18	8	provides	provide	VERB
cana-6113	18	9	a	a	DET
cana-6113	18	10	comprehensive	comprehensive	ADJ
cana-6113	18	11	and	and	CCONJ
cana-6113	18	12	robust	robust	ADJ
cana-6113	18	13	platform	platform	NOUN
cana-6113	18	14	for	for	ADP
cana-6113	18	15	early	early	ADJ
cana-6113	18	16	disease	disease	NOUN
cana-6113	18	17	detection	detection	NOUN
cana-6113	18	18	and	and	CCONJ
cana-6113	18	19	decision	decision	NOUN
cana-6113	18	20	support	support	NOUN
cana-6113	18	21	in	in	ADP
cana-6113	18	22	healthcare	healthcare	PROPN
cana-6113	18	23	.	.	PUNCT
cana-6113	19	1	keywords	keyword	NOUN
cana-6113	19	2	:	:	PUNCT
cana-6113	19	3	deep	deep	ADJ
cana-6113	19	4	learning	learning	NOUN
cana-6113	19	5	,	,	PUNCT
cana-6113	19	6	convolutional	convolutional	ADJ
cana-6113	19	7	neural	neural	ADJ
cana-6113	19	8	networks	network	NOUN
cana-6113	19	9	(	(	PUNCT
cana-6113	19	10	cnns	cnns	PROPN
cana-6113	19	11	)	)	PUNCT
cana-6113	19	12	,	,	PUNCT
cana-6113	19	13	feature	feature	NOUN
cana-6113	19	14	fusion	fusion	NOUN
cana-6113	19	15	,	,	PUNCT
cana-6113	19	16	visual	visual	ADJ
cana-6113	19	17	scene	scene	NOUN
cana-6113	19	18	understanding	understanding	NOUN
cana-6113	19	19	,	,	PUNCT
cana-6113	19	20	image	image	NOUN
cana-6113	19	21	classification	classification	NOUN
cana-6113	19	22	introduction	introduction	NOUN
cana-6113	19	23	in	in	ADP
cana-6113	19	24	recent	recent	ADJ
cana-6113	19	25	years	year	NOUN
cana-6113	19	26	,	,	PUNCT
cana-6113	19	27	deep	deep	ADJ
cana-6113	19	28	learning	learning	NOUN
cana-6113	19	29	has	have	AUX
cana-6113	19	30	emerged	emerge	VERB
cana-6113	19	31	as	as	ADP
cana-6113	19	32	a	a	DET
cana-6113	19	33	transformative	transformative	ADJ
cana-6113	19	34	technology	technology	NOUN
cana-6113	19	35	in	in	ADP
cana-6113	19	36	the	the	DET
cana-6113	19	37	field	field	NOUN
cana-6113	19	38	of	of	ADP
cana-6113	19	39	computer	computer	NOUN
cana-6113	19	40	vision	vision	NOUN
cana-6113	19	41	,	,	PUNCT
cana-6113	19	42	driving	drive	VERB
cana-6113	19	43	remarkable	remarkable	ADJ
cana-6113	19	44	progress	progress	NOUN
cana-6113	19	45	in	in	ADP
cana-6113	19	46	image	image	NOUN
cana-6113	19	47	recognition	recognition	NOUN
cana-6113	19	48	,	,	PUNCT
cana-6113	19	49	object	object	NOUN
cana-6113	19	50	detection	detection	NOUN
cana-6113	19	51	,	,	PUNCT
cana-6113	19	52	and	and	CCONJ
cana-6113	19	53	visual	visual	ADJ
cana-6113	19	54	scene	scene	NOUN
cana-6113	19	55	understanding	understand	VERB
cana-6113	19	56	.	.	PUNCT
cana-6113	20	1	the	the	DET
cana-6113	20	2	evolution	evolution	NOUN
cana-6113	20	3	from	from	ADP
cana-6113	20	4	traditional	traditional	ADJ
cana-6113	20	5	handcrafted	handcraft	VERB
cana-6113	20	6	feature	feature	NOUN
cana-6113	20	7	extraction	extraction	NOUN
cana-6113	20	8	methods	method	NOUN
cana-6113	20	9	to	to	ADP
cana-6113	20	10	data	data	NOUN
cana-6113	20	11	-	-	PUNCT
cana-6113	20	12	driven	drive	VERB
cana-6113	20	13	deep	deep	ADJ
cana-6113	20	14	neural	neural	ADJ
cana-6113	20	15	networks	network	NOUN
cana-6113	20	16	has	have	AUX
cana-6113	20	17	enabled	enable	VERB
cana-6113	20	18	models	model	NOUN
cana-6113	20	19	to	to	PART
cana-6113	20	20	learn	learn	VERB
cana-6113	20	21	complex	complex	ADJ
cana-6113	20	22	visual	visual	ADJ
cana-6113	20	23	representations	representation	NOUN
cana-6113	20	24	directly	directly	ADV
cana-6113	20	25	from	from	ADP
cana-6113	20	26	raw	raw	ADJ
cana-6113	20	27	image	image	NOUN
cana-6113	20	28	data	datum	NOUN
cana-6113	20	29	.	.	PUNCT
cana-6113	21	1	among	among	ADP
cana-6113	21	2	these	these	PRON
cana-6113	21	3	,	,	PUNCT
cana-6113	21	4	convolutional	convolutional	ADJ
cana-6113	21	5	neural	neural	ADJ
cana-6113	21	6	networks	network	NOUN
cana-6113	21	7	(	(	PUNCT
cana-6113	21	8	cnns	cnns	PROPN
cana-6113	21	9	)	)	PUNCT
cana-6113	22	1	[	[	X
cana-6113	22	2	11	11	NUM
cana-6113	22	3	-	-	SYM
cana-6113	22	4	12	12	NUM
cana-6113	22	5	]	]	PUNCT
cana-6113	22	6	have	have	AUX
cana-6113	22	7	played	play	VERB
cana-6113	22	8	a	a	DET
cana-6113	22	9	pivotal	pivotal	ADJ
cana-6113	22	10	role	role	NOUN
cana-6113	22	11	due	due	ADP
cana-6113	22	12	to	to	ADP
cana-6113	22	13	their	their	PRON
cana-6113	22	14	hierarchical	hierarchical	ADJ
cana-6113	22	15	learning	learning	NOUN
cana-6113	22	16	capability	capability	NOUN
cana-6113	22	17	,	,	PUNCT
cana-6113	22	18	where	where	SCONJ
cana-6113	22	19	successive	successive	ADJ
cana-6113	22	20	layers	layer	NOUN
cana-6113	22	21	capture	capture	VERB
cana-6113	22	22	increasingly	increasingly	ADV
cana-6113	22	23	abstract	abstract	ADJ
cana-6113	22	24	visual	visual	ADJ
cana-6113	22	25	features	feature	NOUN
cana-6113	22	26	from	from	ADP
cana-6113	22	27	edges	edge	NOUN
cana-6113	22	28	and	and	CCONJ
cana-6113	22	29	textures	texture	NOUN
cana-6113	22	30	in	in	ADP
cana-6113	22	31	lower	low	ADJ
cana-6113	22	32	layers	layer	NOUN
cana-6113	22	33	to	to	ADP
cana-6113	22	34	shapes	shape	NOUN
cana-6113	22	35	,	,	PUNCT
cana-6113	22	36	objects	object	NOUN
cana-6113	22	37	,	,	PUNCT
cana-6113	22	38	and	and	CCONJ
cana-6113	22	39	semantic	semantic	ADJ
cana-6113	22	40	patterns	pattern	NOUN
cana-6113	22	41	in	in	ADP
cana-6113	22	42	higher	high	ADJ
cana-6113	22	43	layers	layer	NOUN
cana-6113	22	44	.	.	PUNCT
cana-6113	23	1	this	this	DET
cana-6113	23	2	hierarchical	hierarchical	ADJ
cana-6113	23	3	feature	feature	NOUN
cana-6113	23	4	learning	learning	NOUN
cana-6113	23	5	has	have	AUX
cana-6113	23	6	made	make	VERB
cana-6113	23	7	cnns	cnn	NOUN
cana-6113	23	8	indispensable	indispensable	ADJ
cana-6113	23	9	in	in	ADP
cana-6113	23	10	applications	application	NOUN
cana-6113	23	11	such	such	ADJ
cana-6113	23	12	as	as	ADP
cana-6113	23	13	facial	facial	ADJ
cana-6113	23	14	recognition	recognition	NOUN
cana-6113	23	15	,	,	PUNCT
cana-6113	23	16	medical	medical	ADJ
cana-6113	23	17	imaging	imaging	NOUN
cana-6113	23	18	,	,	PUNCT
cana-6113	23	19	surveillance	surveillance	NOUN
cana-6113	23	20	analysis	analysis	NOUN
cana-6113	23	21	,	,	PUNCT
cana-6113	23	22	and	and	CCONJ
cana-6113	23	23	autonomous	autonomous	ADJ
cana-6113	23	24	navigation	navigation	NOUN
cana-6113	23	25	.	.	PUNCT
cana-6113	24	1	despite	despite	SCONJ
cana-6113	24	2	these	these	DET
cana-6113	24	3	advancements	advancement	NOUN
cana-6113	24	4	,	,	PUNCT
cana-6113	24	5	achieving	achieve	VERB
cana-6113	24	6	a	a	DET
cana-6113	24	7	comprehensive	comprehensive	ADJ
cana-6113	24	8	understanding	understanding	NOUN
cana-6113	24	9	of	of	ADP
cana-6113	24	10	visual	visual	ADJ
cana-6113	24	11	scenes	scene	NOUN
cana-6113	24	12	remains	remain	VERB
cana-6113	24	13	a	a	DET
cana-6113	24	14	significant	significant	ADJ
cana-6113	24	15	challenge	challenge	NOUN
cana-6113	24	16	.	.	PUNCT
cana-6113	25	1	traditional	traditional	ADJ
cana-6113	25	2	cnn	cnn	PROPN
cana-6113	25	3	architectures	architecture	NOUN
cana-6113	25	4	[	[	X
cana-6113	25	5	13	13	NUM
cana-6113	25	6	]	]	PUNCT
cana-6113	25	7	often	often	ADV
cana-6113	25	8	focus	focus	VERB
cana-6113	25	9	on	on	ADP
cana-6113	25	10	single	single	ADJ
cana-6113	25	11	-	-	PUNCT
cana-6113	25	12	level	level	NOUN
cana-6113	25	13	feature	feature	NOUN
cana-6113	25	14	extraction	extraction	NOUN
cana-6113	25	15	,	,	PUNCT
cana-6113	25	16	which	which	PRON
cana-6113	25	17	can	can	AUX
cana-6113	25	18	limit	limit	VERB
cana-6113	25	19	their	their	PRON
cana-6113	25	20	ability	ability	NOUN
cana-6113	25	21	to	to	PART
cana-6113	25	22	represent	represent	VERB
cana-6113	25	23	fine	fine	ADV
cana-6113	25	24	-	-	PUNCT
cana-6113	25	25	grained	grain	VERB
cana-6113	25	26	spatial	spatial	ADJ
cana-6113	25	27	and	and	CCONJ
cana-6113	25	28	contextual	contextual	ADJ
cana-6113	25	29	relationships	relationship	NOUN
cana-6113	25	30	present	present	ADJ
cana-6113	25	31	in	in	ADP
cana-6113	25	32	complex	complex	ADJ
cana-6113	25	33	images	image	NOUN
cana-6113	25	34	.	.	PUNCT
cana-6113	26	1	moreover	moreover	ADV
cana-6113	26	2	,	,	PUNCT
cana-6113	26	3	visual	visual	ADJ
cana-6113	26	4	understanding	understanding	NOUN
cana-6113	26	5	tasks	task	NOUN
cana-6113	26	6	frequently	frequently	ADV
cana-6113	26	7	demand	demand	VERB
cana-6113	26	8	the	the	DET
cana-6113	26	9	integration	integration	NOUN
cana-6113	26	10	of	of	ADP
cana-6113	26	11	features	feature	NOUN
cana-6113	26	12	at	at	ADP
cana-6113	26	13	multiple	multiple	ADJ
cana-6113	26	14	abstraction	abstraction	NOUN
cana-6113	26	15	levels	level	NOUN
cana-6113	26	16	combining	combine	VERB
cana-6113	26	17	both	both	CCONJ
cana-6113	26	18	global	global	ADJ
cana-6113	26	19	context	context	NOUN
cana-6113	26	20	and	and	CCONJ
cana-6113	26	21	localized	localized	ADJ
cana-6113	26	22	detail	detail	NOUN
cana-6113	26	23	to	to	PART
cana-6113	26	24	achieve	achieve	VERB
cana-6113	26	25	accurate	accurate	ADJ
cana-6113	26	26	and	and	CCONJ
cana-6113	26	27	interpretable	interpretable	ADJ
cana-6113	26	28	predictions	prediction	NOUN
cana-6113	26	29	.	.	PUNCT
cana-6113	27	1	in	in	ADP
cana-6113	27	2	this	this	DET
cana-6113	27	3	context	context	NOUN
cana-6113	27	4	,	,	PUNCT
cana-6113	27	5	multistage	multistage	NOUN
cana-6113	27	6	feature	feature	NOUN
cana-6113	27	7	extraction	extraction	NOUN
cana-6113	27	8	and	and	CCONJ
cana-6113	27	9	fusion	fusion	NOUN
cana-6113	27	10	techniques	technique	NOUN
cana-6113	27	11	have	have	AUX
cana-6113	27	12	gained	gain	VERB
cana-6113	27	13	growing	grow	VERB
cana-6113	27	14	attention	attention	NOUN
cana-6113	27	15	for	for	ADP
cana-6113	27	16	their	their	PRON
cana-6113	27	17	ability	ability	NOUN
cana-6113	27	18	to	to	PART
cana-6113	27	19	enhance	enhance	VERB
cana-6113	27	20	representation	representation	NOUN
cana-6113	27	21	learning	learning	NOUN
cana-6113	27	22	by	by	ADP
cana-6113	27	23	combining	combine	VERB
cana-6113	27	24	features	feature	NOUN
cana-6113	27	25	[	[	X
cana-6113	27	26	14	14	NUM
cana-6113	27	27	]	]	PUNCT
cana-6113	27	28	from	from	ADP
cana-6113	27	29	multiple	multiple	ADJ
cana-6113	27	30	layers	layer	NOUN
cana-6113	27	31	or	or	CCONJ
cana-6113	27	32	models	model	NOUN
cana-6113	27	33	,	,	PUNCT
cana-6113	27	34	thus	thus	ADV
cana-6113	27	35	providing	provide	VERB
cana-6113	27	36	richer	rich	ADJ
cana-6113	27	37	and	and	CCONJ
cana-6113	27	38	more	more	ADJ
cana-6113	27	39	discriminative	discriminative	ADJ
cana-6113	27	40	visual	visual	ADJ
cana-6113	27	41	information	information	NOUN
cana-6113	27	42	.	.	PUNCT
cana-6113	28	1	feature	feature	NOUN
cana-6113	28	2	fusion	fusion	NOUN
cana-6113	28	3	is	be	AUX
cana-6113	28	4	a	a	DET
cana-6113	28	5	critical	critical	ADJ
cana-6113	28	6	component	component	NOUN
cana-6113	28	7	in	in	ADP
cana-6113	28	8	modern	modern	ADJ
cana-6113	28	9	computer	computer	NOUN
cana-6113	28	10	vision	vision	NOUN
cana-6113	28	11	systems	system	NOUN
cana-6113	28	12	,	,	PUNCT
cana-6113	28	13	as	as	SCONJ
cana-6113	28	14	it	it	PRON
cana-6113	28	15	enables	enable	VERB
cana-6113	28	16	the	the	DET
cana-6113	28	17	aggregation	aggregation	NOUN
cana-6113	28	18	of	of	ADP
cana-6113	28	19	complementary	complementary	ADJ
cana-6113	28	20	cues	cue	NOUN
cana-6113	28	21	such	such	ADJ
cana-6113	28	22	as	as	ADP
cana-6113	28	23	color	color	NOUN
cana-6113	28	24	,	,	PUNCT
cana-6113	28	25	texture	texture	NOUN
cana-6113	28	26	,	,	PUNCT
cana-6113	28	27	shape	shape	NOUN
cana-6113	28	28	,	,	PUNCT
cana-6113	28	29	and	and	CCONJ
cana-6113	28	30	spatial	spatial	ADJ
cana-6113	28	31	relationships	relationship	NOUN
cana-6113	28	32	.	.	PUNCT
cana-6113	29	1	when	when	SCONJ
cana-6113	29	2	integrated	integrate	VERB
cana-6113	29	3	effectively	effectively	ADV
cana-6113	29	4	within	within	ADP
cana-6113	29	5	deep	deep	ADJ
cana-6113	29	6	neural	neural	ADJ
cana-6113	29	7	architectures	architecture	NOUN
cana-6113	29	8	,	,	PUNCT
cana-6113	29	9	fusion	fusion	NOUN
cana-6113	29	10	strategies	strategy	NOUN
cana-6113	29	11	can	can	AUX
cana-6113	29	12	improve	improve	VERB
cana-6113	29	13	the	the	DET
cana-6113	29	14	robustness	robustness	NOUN
cana-6113	29	15	and	and	CCONJ
cana-6113	29	16	generalization	generalization	NOUN
cana-6113	29	17	capability	capability	NOUN
cana-6113	29	18	of	of	ADP
cana-6113	29	19	models	model	NOUN
cana-6113	29	20	,	,	PUNCT
cana-6113	29	21	particularly	particularly	ADV
cana-6113	29	22	in	in	ADP
cana-6113	29	23	challenging	challenge	VERB
cana-6113	29	24	environments	environment	NOUN
cana-6113	29	25	with	with	ADP
cana-6113	29	26	noise	noise	NOUN
cana-6113	29	27	,	,	PUNCT
cana-6113	29	28	occlusion	occlusion	NOUN
cana-6113	29	29	,	,	PUNCT
cana-6113	29	30	or	or	CCONJ
cana-6113	29	31	illumination	illumination	NOUN
cana-6113	29	32	variations	variation	NOUN
cana-6113	29	33	[	[	X
cana-6113	29	34	15	15	NUM
cana-6113	29	35	]	]	PUNCT
cana-6113	29	36	.	.	PUNCT
cana-6113	30	1	however	however	ADV
cana-6113	30	2	,	,	PUNCT
cana-6113	30	3	existing	exist	VERB
cana-6113	30	4	fusion	fusion	NOUN
cana-6113	30	5	approaches	approach	NOUN
cana-6113	30	6	often	often	ADV
cana-6113	30	7	face	face	VERB
cana-6113	30	8	limitations	limitation	NOUN
cana-6113	30	9	in	in	ADP
cana-6113	30	10	handling	handle	VERB
cana-6113	30	11	diverse	diverse	ADJ
cana-6113	30	12	feature	feature	NOUN
cana-6113	30	13	scales	scale	NOUN
cana-6113	30	14	,	,	PUNCT
cana-6113	30	15	redundant	redundant	ADJ
cana-6113	30	16	information	information	NOUN
cana-6113	30	17	,	,	PUNCT
cana-6113	30	18	and	and	CCONJ
cana-6113	30	19	the	the	DET
cana-6113	30	20	computational	computational	ADJ
cana-6113	30	21	complexity	complexity	NOUN
cana-6113	30	22	involved	involve	VERB
cana-6113	30	23	in	in	ADP
cana-6113	30	24	multi	multi	ADJ
cana-6113	30	25	-	-	ADJ
cana-6113	30	26	level	level	ADJ
cana-6113	30	27	processing	processing	NOUN
cana-6113	30	28	.	.	PUNCT
cana-6113	31	1	therefore	therefore	ADV
cana-6113	31	2	,	,	PUNCT
cana-6113	31	3	there	there	PRON
cana-6113	31	4	is	be	VERB
cana-6113	31	5	a	a	DET
cana-6113	31	6	pressing	press	VERB
cana-6113	31	7	need	need	NOUN
cana-6113	31	8	for	for	ADP
cana-6113	31	9	a	a	DET
cana-6113	31	10	unified	unified	ADJ
cana-6113	31	11	framework	framework	NOUN
cana-6113	31	12	that	that	PRON
cana-6113	31	13	can	can	AUX
cana-6113	31	14	efficiently	efficiently	ADV
cana-6113	31	15	integrate	integrate	VERB
cana-6113	31	16	multi	multi	ADJ
cana-6113	31	17	-	-	ADJ
cana-6113	31	18	stage	stage	ADJ
cana-6113	31	19	feature	feature	NOUN
cana-6113	31	20	representations	representation	NOUN
cana-6113	31	21	while	while	SCONJ
cana-6113	31	22	maintaining	maintain	VERB
cana-6113	31	23	computational	computational	ADJ
cana-6113	31	24	efficiency	efficiency	NOUN
cana-6113	31	25	and	and	CCONJ
cana-6113	31	26	interpretability	interpretability	NOUN
cana-6113	31	27	.	.	PUNCT
cana-6113	32	1	this	this	DET
cana-6113	32	2	paper	paper	NOUN
cana-6113	32	3	proposes	propose	VERB
cana-6113	32	4	a	a	DET
cana-6113	32	5	deep	deep	ADJ
cana-6113	32	6	learning	learning	NOUN
cana-6113	32	7	-	-	PUNCT
cana-6113	32	8	powered	power	VERB
cana-6113	32	9	multi	multi	ADJ
cana-6113	32	10	-	-	ADJ
cana-6113	32	11	stage	stage	ADJ
cana-6113	32	12	image	image	NOUN
cana-6113	32	13	feature	feature	NOUN
cana-6113	32	14	processing	processing	NOUN
cana-6113	32	15	and	and	CCONJ
cana-6113	32	16	fusion	fusion	NOUN
cana-6113	32	17	framework	framework	NOUN
cana-6113	32	18	designed	design	VERB
cana-6113	32	19	to	to	PART
cana-6113	32	20	enhance	enhance	VERB
cana-6113	32	21	visual	visual	ADJ
cana-6113	32	22	understanding	understanding	NOUN
cana-6113	32	23	by	by	ADP
cana-6113	32	24	leveraging	leverage	VERB
cana-6113	32	25	hierarchical	hierarchical	ADJ
cana-6113	32	26	cnn	cnn	PROPN
cana-6113	32	27	architectures	architecture	NOUN
cana-6113	32	28	combined	combine	VERB
cana-6113	32	29	with	with	ADP
cana-6113	32	30	statistical	statistical	ADJ
cana-6113	32	31	and	and	CCONJ
cana-6113	32	32	contextual	contextual	ADJ
cana-6113	32	33	fusion	fusion	NOUN
cana-6113	32	34	techniques	technique	NOUN
cana-6113	32	35	[	[	X
cana-6113	32	36	16	16	NUM
cana-6113	32	37	-	-	SYM
cana-6113	32	38	17	17	NUM
cana-6113	32	39	]	]	PUNCT
cana-6113	32	40	.	.	PUNCT
cana-6113	33	1	the	the	DET
cana-6113	33	2	model	model	NOUN
cana-6113	33	3	incorporates	incorporate	VERB
cana-6113	33	4	preprocessing	preprocesse	VERB
cana-6113	33	5	modules	module	NOUN
cana-6113	33	6	for	for	ADP
cana-6113	33	7	normalization	normalization	NOUN
cana-6113	33	8	,	,	PUNCT
cana-6113	33	9	contrast	contrast	NOUN
cana-6113	33	10	enhancement	enhancement	NOUN
cana-6113	33	11	,	,	PUNCT
cana-6113	33	12	and	and	CCONJ
cana-6113	33	13	regionbased	regionbase	VERB
cana-6113	33	14	segmentation	segmentation	NOUN
cana-6113	33	15	to	to	PART
cana-6113	33	16	ensure	ensure	VERB
cana-6113	33	17	high	high	ADJ
cana-6113	33	18	-	-	PUNCT
cana-6113	33	19	quality	quality	NOUN
cana-6113	33	20	input	input	NOUN
cana-6113	33	21	features	feature	NOUN
cana-6113	33	22	.	.	PUNCT
cana-6113	34	1	additionally	additionally	ADV
cana-6113	34	2	,	,	PUNCT
cana-6113	34	3	adaptive	adaptive	ADJ
cana-6113	34	4	pooling	pooling	NOUN
cana-6113	34	5	and	and	CCONJ
cana-6113	34	6	attention	attention	NOUN
cana-6113	34	7	mechanisms	mechanism	NOUN
cana-6113	34	8	are	be	AUX
cana-6113	34	9	integrated	integrate	VERB
cana-6113	34	10	to	to	PART
cana-6113	34	11	focus	focus	VERB
cana-6113	34	12	on	on	ADP
cana-6113	34	13	the	the	DET
cana-6113	34	14	most	most	ADV
cana-6113	34	15	informative	informative	ADJ
cana-6113	34	16	regions	region	NOUN
cana-6113	34	17	of	of	ADP
cana-6113	34	18	an	an	DET
cana-6113	34	19	image	image	NOUN
cana-6113	34	20	,	,	PUNCT
cana-6113	34	21	improving	improve	VERB
cana-6113	34	22	both	both	DET
cana-6113	34	23	accuracy	accuracy	NOUN
cana-6113	34	24	and	and	CCONJ
cana-6113	34	25	interpretability	interpretability	NOUN
cana-6113	34	26	.	.	PUNCT
cana-6113	35	1	a	a	DET
cana-6113	35	2	weighted	weight	VERB
cana-6113	35	3	ensemble	ensemble	ADJ
cana-6113	35	4	of	of	ADP
cana-6113	35	5	cnn	cnn	PROPN
cana-6113	35	6	-	-	PUNCT
cana-6113	35	7	based	base	VERB
cana-6113	35	8	encoders	encoder	NOUN
cana-6113	35	9	and	and	CCONJ
cana-6113	35	10	regression	regression	NOUN
cana-6113	35	11	-	-	PUNCT
cana-6113	35	12	based	base	VERB
cana-6113	35	13	fusion	fusion	NOUN
cana-6113	35	14	layers	layer	NOUN
cana-6113	35	15	is	be	AUX
cana-6113	35	16	employed	employ	VERB
cana-6113	35	17	to	to	PART
cana-6113	35	18	merge	merge	VERB
cana-6113	35	19	multi	multi	ADJ
cana-6113	35	20	-	-	ADJ
cana-6113	35	21	level	level	ADJ
cana-6113	35	22	feature	feature	NOUN
cana-6113	35	23	maps	map	NOUN
cana-6113	35	24	,	,	PUNCT
cana-6113	35	25	effectively	effectively	ADV
cana-6113	35	26	combining	combine	VERB
cana-6113	35	27	semantic	semantic	ADJ
cana-6113	35	28	and	and	CCONJ
cana-6113	35	29	spatial	spatial	ADJ
cana-6113	35	30	cues	cue	NOUN
cana-6113	35	31	for	for	ADP
cana-6113	35	32	a	a	DET
cana-6113	35	33	deeper	deep	ADJ
cana-6113	35	34	and	and	CCONJ
cana-6113	35	35	more	more	ADV
cana-6113	35	36	coherent	coherent	ADJ
cana-6113	35	37	understanding	understanding	NOUN
cana-6113	35	38	of	of	ADP
cana-6113	35	39	visual	visual	ADJ
cana-6113	35	40	content	content	NOUN
cana-6113	35	41	.	.	PUNCT
cana-6113	36	1	communications	communication	NOUN
cana-6113	36	2	on	on	ADP
cana-6113	36	3	applied	apply	VERB
cana-6113	36	4	nonlinear	nonlinear	ADJ
cana-6113	36	5	analysis	analysis	NOUN
cana-6113	36	6	issn	issn	NOUN
cana-6113	36	7	:	:	PUNCT
cana-6113	36	8	1074	1074	NUM
cana-6113	36	9	-	-	PUNCT
cana-6113	36	10	133x	133x	NUM
cana-6113	36	11	vol	vol	VERB
cana-6113	36	12	32	32	NUM
cana-6113	36	13	no	no	NOUN
cana-6113	36	14	.	.	PUNCT
cana-6113	37	1	10s	10	NOUN
cana-6113	37	2	(	(	PUNCT
cana-6113	37	3	2025	2025	NUM
cana-6113	37	4	)	)	PUNCT
cana-6113	37	5	3405	3405	NUM
cana-6113	37	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	37	7	extensive	extensive	ADJ
cana-6113	37	8	experimentation	experimentation	NOUN
cana-6113	37	9	on	on	ADP
cana-6113	37	10	benchmark	benchmark	ADJ
cana-6113	37	11	datasets	dataset	NOUN
cana-6113	37	12	demonstrates	demonstrate	VERB
cana-6113	37	13	the	the	DET
cana-6113	37	14	superior	superior	ADJ
cana-6113	37	15	performance	performance	NOUN
cana-6113	37	16	of	of	ADP
cana-6113	37	17	the	the	DET
cana-6113	37	18	proposed	propose	VERB
cana-6113	37	19	framework	framework	NOUN
cana-6113	37	20	compared	compare	VERB
cana-6113	37	21	to	to	ADP
cana-6113	37	22	conventional	conventional	ADJ
cana-6113	37	23	deep	deep	ADJ
cana-6113	37	24	learning	learning	NOUN
cana-6113	37	25	models	model	NOUN
cana-6113	37	26	[	[	X
cana-6113	37	27	11	11	NUM
cana-6113	37	28	]	]	PUNCT
cana-6113	37	29	,	,	PUNCT
cana-6113	37	30	achieving	achieve	VERB
cana-6113	37	31	significant	significant	ADJ
cana-6113	37	32	improvements	improvement	NOUN
cana-6113	37	33	in	in	ADP
cana-6113	37	34	accuracy	accuracy	NOUN
cana-6113	37	35	,	,	PUNCT
cana-6113	37	36	precision	precision	NOUN
cana-6113	37	37	,	,	PUNCT
cana-6113	37	38	and	and	CCONJ
cana-6113	37	39	f1	f1	NOUN
cana-6113	37	40	-	-	PUNCT
cana-6113	37	41	score	score	NOUN
cana-6113	37	42	.	.	PUNCT
cana-6113	38	1	furthermore	furthermore	ADV
cana-6113	38	2	,	,	PUNCT
cana-6113	38	3	the	the	DET
cana-6113	38	4	model	model	NOUN
cana-6113	38	5	exhibits	exhibit	VERB
cana-6113	38	6	adaptability	adaptability	NOUN
cana-6113	38	7	across	across	ADP
cana-6113	38	8	multiple	multiple	ADJ
cana-6113	38	9	domains	domain	NOUN
cana-6113	38	10	,	,	PUNCT
cana-6113	38	11	making	make	VERB
cana-6113	38	12	it	it	PRON
cana-6113	38	13	suitable	suitable	ADJ
cana-6113	38	14	for	for	ADP
cana-6113	38	15	a	a	DET
cana-6113	38	16	wide	wide	ADJ
cana-6113	38	17	range	range	NOUN
cana-6113	38	18	of	of	ADP
cana-6113	38	19	applications	application	NOUN
cana-6113	38	20	,	,	PUNCT
cana-6113	38	21	including	include	VERB
cana-6113	38	22	medical	medical	ADJ
cana-6113	38	23	diagnostics	diagnostic	NOUN
cana-6113	38	24	,	,	PUNCT
cana-6113	38	25	remote	remote	ADJ
cana-6113	38	26	sensing	sensing	NOUN
cana-6113	38	27	,	,	PUNCT
cana-6113	38	28	and	and	CCONJ
cana-6113	38	29	autonomous	autonomous	ADJ
cana-6113	38	30	vision	vision	NOUN
cana-6113	38	31	systems	system	NOUN
cana-6113	38	32	.	.	PUNCT
cana-6113	39	1	1	1	X
cana-6113	39	2	.	.	X
cana-6113	39	3	review	review	NOUN
cana-6113	39	4	of	of	ADP
cana-6113	39	5	literature	literature	NOUN
cana-6113	39	6	recent	recent	ADJ
cana-6113	39	7	developments	development	NOUN
cana-6113	39	8	in	in	ADP
cana-6113	39	9	deep	deep	ADJ
cana-6113	39	10	learning	learning	NOUN
cana-6113	39	11	have	have	AUX
cana-6113	39	12	greatly	greatly	ADV
cana-6113	39	13	transformed	transform	VERB
cana-6113	39	14	the	the	DET
cana-6113	39	15	field	field	NOUN
cana-6113	39	16	of	of	ADP
cana-6113	39	17	computer	computer	NOUN
cana-6113	39	18	vision	vision	NOUN
cana-6113	39	19	by	by	ADP
cana-6113	39	20	enabling	enable	VERB
cana-6113	39	21	more	more	ADV
cana-6113	39	22	accurate	accurate	ADJ
cana-6113	39	23	and	and	CCONJ
cana-6113	39	24	detailed	detailed	ADJ
cana-6113	39	25	image	image	NOUN
cana-6113	39	26	understanding	understanding	NOUN
cana-6113	39	27	.	.	PUNCT
cana-6113	40	1	traditional	traditional	ADJ
cana-6113	40	2	single	single	ADJ
cana-6113	40	3	-	-	PUNCT
cana-6113	40	4	stage	stage	NOUN
cana-6113	40	5	convolutional	convolutional	ADJ
cana-6113	40	6	networks	network	NOUN
cana-6113	40	7	often	often	ADV
cana-6113	40	8	struggled	struggle	VERB
cana-6113	40	9	to	to	PART
cana-6113	40	10	capture	capture	VERB
cana-6113	40	11	the	the	DET
cana-6113	40	12	full	full	ADJ
cana-6113	40	13	range	range	NOUN
cana-6113	40	14	of	of	ADP
cana-6113	40	15	visual	visual	ADJ
cana-6113	40	16	information	information	NOUN
cana-6113	40	17	present	present	ADJ
cana-6113	40	18	in	in	ADP
cana-6113	40	19	complex	complex	ADJ
cana-6113	40	20	scenes	scene	NOUN
cana-6113	40	21	,	,	PUNCT
cana-6113	40	22	but	but	CCONJ
cana-6113	40	23	modern	modern	ADJ
cana-6113	40	24	architectures	architecture	NOUN
cana-6113	40	25	now	now	ADV
cana-6113	40	26	integrate	integrate	VERB
cana-6113	40	27	multi	multi	ADJ
cana-6113	40	28	-	-	ADJ
cana-6113	40	29	stage	stage	NOUN
cana-6113	40	30	processing	processing	NOUN
cana-6113	40	31	,	,	PUNCT
cana-6113	40	32	attention	attention	NOUN
cana-6113	40	33	mechanisms	mechanism	NOUN
cana-6113	40	34	,	,	PUNCT
cana-6113	40	35	and	and	CCONJ
cana-6113	40	36	feature	feature	NOUN
cana-6113	40	37	fusion	fusion	NOUN
cana-6113	40	38	strategies	strategy	NOUN
cana-6113	40	39	to	to	PART
cana-6113	40	40	overcome	overcome	VERB
cana-6113	40	41	these	these	DET
cana-6113	40	42	limitations	limitation	NOUN
cana-6113	40	43	.	.	PUNCT
cana-6113	41	1	by	by	ADP
cana-6113	41	2	combining	combine	VERB
cana-6113	41	3	lowlevel	lowlevel	VERB
cana-6113	41	4	texture	texture	NOUN
cana-6113	41	5	and	and	CCONJ
cana-6113	41	6	edge	edge	NOUN
cana-6113	41	7	features	feature	NOUN
cana-6113	41	8	with	with	ADP
cana-6113	41	9	high	high	ADJ
cana-6113	41	10	-	-	PUNCT
cana-6113	41	11	level	level	NOUN
cana-6113	41	12	semantic	semantic	ADJ
cana-6113	41	13	representations	representation	NOUN
cana-6113	41	14	,	,	PUNCT
cana-6113	41	15	these	these	DET
cana-6113	41	16	approaches	approach	NOUN
cana-6113	41	17	enhance	enhance	VERB
cana-6113	41	18	the	the	DET
cana-6113	41	19	model	model	NOUN
cana-6113	41	20	’s	’s	PART
cana-6113	41	21	ability	ability	NOUN
cana-6113	41	22	to	to	PART
cana-6113	41	23	recognize	recognize	VERB
cana-6113	41	24	patterns	pattern	NOUN
cana-6113	41	25	,	,	PUNCT
cana-6113	41	26	detect	detect	NOUN
cana-6113	41	27	anomalies	anomaly	NOUN
cana-6113	41	28	,	,	PUNCT
cana-6113	41	29	and	and	CCONJ
cana-6113	41	30	interpret	interpret	VERB
cana-6113	41	31	visual	visual	ADJ
cana-6113	41	32	contexts	context	NOUN
cana-6113	41	33	more	more	ADV
cana-6113	41	34	effectively	effectively	ADV
cana-6113	41	35	.	.	PUNCT
cana-6113	42	1	multi	multi	ADJ
cana-6113	42	2	-	-	ADJ
cana-6113	42	3	level	level	ADJ
cana-6113	42	4	feature	feature	NOUN
cana-6113	42	5	fusion	fusion	NOUN
cana-6113	42	6	techniques	technique	NOUN
cana-6113	42	7	,	,	PUNCT
cana-6113	42	8	in	in	ADP
cana-6113	42	9	particular	particular	ADJ
cana-6113	42	10	,	,	PUNCT
cana-6113	42	11	have	have	AUX
cana-6113	42	12	proven	prove	VERB
cana-6113	42	13	essential	essential	ADJ
cana-6113	42	14	for	for	ADP
cana-6113	42	15	unifying	unify	VERB
cana-6113	42	16	spatial	spatial	ADJ
cana-6113	42	17	,	,	PUNCT
cana-6113	42	18	contextual	contextual	ADJ
cana-6113	42	19	,	,	PUNCT
cana-6113	42	20	and	and	CCONJ
cana-6113	42	21	semantic	semantic	ADJ
cana-6113	42	22	cues	cue	NOUN
cana-6113	42	23	into	into	ADP
cana-6113	42	24	a	a	DET
cana-6113	42	25	coherent	coherent	ADJ
cana-6113	42	26	representation	representation	NOUN
cana-6113	42	27	,	,	PUNCT
cana-6113	42	28	improving	improve	VERB
cana-6113	42	29	both	both	DET
cana-6113	42	30	accuracy	accuracy	NOUN
cana-6113	42	31	and	and	CCONJ
cana-6113	42	32	generalization	generalization	NOUN
cana-6113	42	33	across	across	ADP
cana-6113	42	34	diverse	diverse	ADJ
cana-6113	42	35	datasets	dataset	NOUN
cana-6113	42	36	.	.	PUNCT
cana-6113	43	1	furthermore	furthermore	ADV
cana-6113	43	2	,	,	PUNCT
cana-6113	43	3	attentionbased	attentionbase	VERB
cana-6113	43	4	mechanisms	mechanism	NOUN
cana-6113	43	5	help	help	VERB
cana-6113	43	6	focus	focus	VERB
cana-6113	43	7	the	the	DET
cana-6113	43	8	model	model	NOUN
cana-6113	43	9	on	on	ADP
cana-6113	43	10	the	the	DET
cana-6113	43	11	most	most	ADV
cana-6113	43	12	discriminative	discriminative	ADJ
cana-6113	43	13	regions	region	NOUN
cana-6113	43	14	of	of	ADP
cana-6113	43	15	an	an	DET
cana-6113	43	16	image	image	NOUN
cana-6113	43	17	,	,	PUNCT
cana-6113	43	18	increasing	increase	VERB
cana-6113	43	19	interpretability	interpretability	NOUN
cana-6113	43	20	and	and	CCONJ
cana-6113	43	21	performance	performance	NOUN
cana-6113	43	22	in	in	ADP
cana-6113	43	23	tasks	task	NOUN
cana-6113	43	24	such	such	ADJ
cana-6113	43	25	as	as	ADP
cana-6113	43	26	classification	classification	NOUN
cana-6113	43	27	,	,	PUNCT
cana-6113	43	28	segmentation	segmentation	NOUN
cana-6113	43	29	,	,	PUNCT
cana-6113	43	30	and	and	CCONJ
cana-6113	43	31	recognition	recognition	NOUN
cana-6113	43	32	.	.	PUNCT
cana-6113	44	1	table	table	NOUN
cana-6113	44	2	1	1	NUM
cana-6113	44	3	:	:	PUNCT
cana-6113	44	4	review	review	NOUN
cana-6113	44	5	of	of	ADP
cana-6113	44	6	literature	literature	NOUN
cana-6113	44	7	for	for	ADP
cana-6113	44	8	deep	deep	ADJ
cana-6113	44	9	learning	learning	NOUN
cana-6113	44	10	-	-	PUNCT
cana-6113	44	11	powered	power	VERB
cana-6113	44	12	multi	multi	ADJ
cana-6113	44	13	-	-	ADJ
cana-6113	44	14	stage	stage	ADJ
cana-6113	44	15	image	image	NOUN
cana-6113	44	16	feature	feature	NOUN
cana-6113	44	17	processing	processing	NOUN
cana-6113	44	18	and	and	CCONJ
cana-6113	44	19	fusion	fusion	NOUN
cana-6113	44	20	approach	approach	NOUN
cana-6113	44	21	ref	ref	NOUN
cana-6113	44	22	.	.	PUNCT
cana-6113	45	1	algorithm	algorithm	NOUN
cana-6113	45	2	/	/	SYM
cana-6113	45	3	model	model	NOUN
cana-6113	45	4	used	use	VERB
cana-6113	45	5	dataset	dataset	NOUN
cana-6113	45	6	used	use	VERB
cana-6113	45	7	key	key	ADJ
cana-6113	45	8	findings	finding	NOUN
cana-6113	45	9	/	/	SYM
cana-6113	45	10	contributions	contribution	NOUN
cana-6113	46	1	[	[	X
cana-6113	46	2	1	1	NUM
cana-6113	46	3	]	]	SYM
cana-6113	46	4	two	two	NUM
cana-6113	46	5	-	-	PUNCT
cana-6113	46	6	stream	stream	NOUN
cana-6113	46	7	cnn	cnn	PROPN
cana-6113	46	8	ucf101	ucf101	PROPN
cana-6113	46	9	,	,	PUNCT
cana-6113	46	10	hmdb51	hmdb51	NOUN
cana-6113	46	11	combined	combine	VERB
cana-6113	46	12	spatial	spatial	ADJ
cana-6113	46	13	and	and	CCONJ
cana-6113	46	14	temporal	temporal	ADJ
cana-6113	46	15	streams	stream	NOUN
cana-6113	46	16	improved	improve	VERB
cana-6113	46	17	video	video	NOUN
cana-6113	46	18	action	action	NOUN
cana-6113	46	19	recognition	recognition	NOUN
cana-6113	46	20	accuracy	accuracy	NOUN
cana-6113	46	21	.	.	PUNCT
cana-6113	47	1	[	[	X
cana-6113	47	2	2	2	NUM
cana-6113	47	3	]	]	X
cana-6113	47	4	deep	deep	ADJ
cana-6113	47	5	multiple	multiple	ADJ
cana-6113	47	6	instance	instance	NOUN
cana-6113	47	7	learning	learn	VERB
cana-6113	47	8	(	(	PUNCT
cana-6113	47	9	mil	mil	NOUN
cana-6113	47	10	)	)	PUNCT
cana-6113	47	11	ucf	ucf	PROPN
cana-6113	47	12	-	-	PUNCT
cana-6113	47	13	crime	crime	NOUN
cana-6113	47	14	detected	detect	VERB
cana-6113	47	15	anomalies	anomaly	NOUN
cana-6113	47	16	in	in	ADP
cana-6113	47	17	untrimmed	untrimmed	ADJ
cana-6113	47	18	videos	video	NOUN
cana-6113	47	19	using	use	VERB
cana-6113	47	20	weakly	weakly	ADJ
cana-6113	47	21	labeled	label	VERB
cana-6113	47	22	data	datum	NOUN
cana-6113	47	23	with	with	ADP
cana-6113	47	24	high	high	ADJ
cana-6113	47	25	precision	precision	NOUN
cana-6113	47	26	.	.	PUNCT
cana-6113	48	1	[	[	X
cana-6113	48	2	3	3	NUM
cana-6113	48	3	]	]	X
cana-6113	48	4	temporal	temporal	ADJ
cana-6113	48	5	segment	segment	NOUN
cana-6113	48	6	network	network	NOUN
cana-6113	48	7	(	(	PUNCT
cana-6113	48	8	tsn	tsn	PROPN
cana-6113	48	9	)	)	PUNCT
cana-6113	48	10	kinetics	kinetic	NOUN
cana-6113	48	11	,	,	PUNCT
cana-6113	48	12	hmdb51	hmdb51	PROPN
cana-6113	48	13	modeled	model	VERB
cana-6113	48	14	long	long	ADJ
cana-6113	48	15	-	-	PUNCT
cana-6113	48	16	term	term	NOUN
cana-6113	48	17	temporal	temporal	ADJ
cana-6113	48	18	structure	structure	NOUN
cana-6113	48	19	and	and	CCONJ
cana-6113	48	20	improved	improve	VERB
cana-6113	48	21	human	human	ADJ
cana-6113	48	22	action	action	NOUN
cana-6113	48	23	recognition	recognition	NOUN
cana-6113	48	24	.	.	PUNCT
cana-6113	49	1	[	[	X
cana-6113	49	2	4	4	X
cana-6113	49	3	]	]	PUNCT
cana-6113	49	4	resnet	resnet	NOUN
cana-6113	49	5	(	(	PUNCT
cana-6113	49	6	residual	residual	ADJ
cana-6113	49	7	networks	network	NOUN
cana-6113	49	8	)	)	PUNCT
cana-6113	49	9	imagenet	imagenet	NOUN
cana-6113	49	10	introduced	introduce	VERB
cana-6113	49	11	residual	residual	ADJ
cana-6113	49	12	learning	learning	NOUN
cana-6113	49	13	to	to	PART
cana-6113	49	14	train	train	VERB
cana-6113	49	15	very	very	ADV
cana-6113	49	16	deep	deep	ADJ
cana-6113	49	17	cnns	cnn	NOUN
cana-6113	49	18	efficiently	efficiently	ADV
cana-6113	49	19	with	with	ADP
cana-6113	49	20	improved	improved	ADJ
cana-6113	49	21	accuracy	accuracy	NOUN
cana-6113	49	22	.	.	PUNCT
cana-6113	50	1	[	[	X
cana-6113	50	2	5	5	NUM
cana-6113	50	3	]	]	X
cana-6113	50	4	squeeze	squeeze	NOUN
cana-6113	50	5	-	-	PUNCT
cana-6113	50	6	and	and	CCONJ
cana-6113	50	7	-	-	PUNCT
cana-6113	50	8	excitation	excitation	NOUN
cana-6113	50	9	network	network	NOUN
cana-6113	50	10	(	(	PUNCT
cana-6113	50	11	senet	senet	NOUN
cana-6113	50	12	)	)	PUNCT
cana-6113	50	13	imagenet	imagenet	NOUN
cana-6113	50	14	used	use	VERB
cana-6113	50	15	channel	channel	NOUN
cana-6113	50	16	attention	attention	NOUN
cana-6113	50	17	to	to	PART
cana-6113	50	18	enhance	enhance	VERB
cana-6113	50	19	feature	feature	NOUN
cana-6113	50	20	importance	importance	NOUN
cana-6113	50	21	,	,	PUNCT
cana-6113	50	22	achieving	achieve	VERB
cana-6113	50	23	state	state	NOUN
cana-6113	50	24	-	-	PUNCT
cana-6113	50	25	ofthe	ofthe	NOUN
cana-6113	50	26	-	-	PUNCT
cana-6113	50	27	art	art	NOUN
cana-6113	50	28	accuracy	accuracy	NOUN
cana-6113	50	29	.	.	PUNCT
cana-6113	51	1	[	[	X
cana-6113	51	2	6	6	NUM
cana-6113	51	3	]	]	X
cana-6113	51	4	cbam	cbam	NOUN
cana-6113	51	5	(	(	PUNCT
cana-6113	51	6	convolutional	convolutional	ADJ
cana-6113	51	7	block	block	NOUN
cana-6113	51	8	attention	attention	NOUN
cana-6113	51	9	module	module	NOUN
cana-6113	51	10	)	)	PUNCT
cana-6113	51	11	imagenet	imagenet	NOUN
cana-6113	51	12	,	,	PUNCT
cana-6113	51	13	coco	coco	PROPN
cana-6113	51	14	combined	combine	VERB
cana-6113	51	15	spatial	spatial	PROPN
cana-6113	51	16	and	and	CCONJ
cana-6113	51	17	channel	channel	NOUN
cana-6113	51	18	attention	attention	NOUN
cana-6113	51	19	to	to	PART
cana-6113	51	20	boost	boost	VERB
cana-6113	51	21	model	model	NOUN
cana-6113	51	22	interpretability	interpretability	NOUN
cana-6113	51	23	and	and	CCONJ
cana-6113	51	24	accuracy	accuracy	NOUN
cana-6113	51	25	.	.	PUNCT
cana-6113	52	1	[	[	X
cana-6113	52	2	7	7	X
cana-6113	52	3	]	]	X
cana-6113	52	4	dilated	dilated	ADJ
cana-6113	52	5	convolutions	convolution	NOUN
cana-6113	52	6	pascal	pascal	PROPN
cana-6113	52	7	voc	voc	PROPN
cana-6113	52	8	captured	capture	VERB
cana-6113	52	9	multi	multi	ADJ
cana-6113	52	10	-	-	ADJ
cana-6113	52	11	scale	scale	ADJ
cana-6113	52	12	contextual	contextual	ADJ
cana-6113	52	13	features	feature	NOUN
cana-6113	52	14	without	without	ADP
cana-6113	52	15	loss	loss	NOUN
cana-6113	52	16	of	of	ADP
cana-6113	52	17	resolution	resolution	NOUN
cana-6113	52	18	for	for	ADP
cana-6113	52	19	segmentation	segmentation	NOUN
cana-6113	52	20	.	.	PUNCT
cana-6113	53	1	communications	communication	NOUN
cana-6113	53	2	on	on	ADP
cana-6113	53	3	applied	apply	VERB
cana-6113	53	4	nonlinear	nonlinear	ADJ
cana-6113	53	5	analysis	analysis	NOUN
cana-6113	53	6	issn	issn	NOUN
cana-6113	53	7	:	:	PUNCT
cana-6113	53	8	1074	1074	NUM
cana-6113	53	9	-	-	PUNCT
cana-6113	53	10	133x	133x	NUM
cana-6113	53	11	vol	vol	VERB
cana-6113	53	12	32	32	NUM
cana-6113	53	13	no	no	NOUN
cana-6113	53	14	.	.	PUNCT
cana-6113	54	1	10s	10	NOUN
cana-6113	54	2	(	(	PUNCT
cana-6113	54	3	2025	2025	NUM
cana-6113	54	4	)	)	PUNCT
cana-6113	54	5	3406	3406	NUM
cana-6113	54	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	55	1	[	[	X
cana-6113	55	2	8	8	X
cana-6113	55	3	]	]	X
cana-6113	55	4	deeplab	deeplab	NOUN
cana-6113	55	5	(	(	PUNCT
cana-6113	55	6	atrous	atrous	ADJ
cana-6113	55	7	convolution	convolution	NOUN
cana-6113	55	8	+	+	CCONJ
cana-6113	55	9	crf	crf	NOUN
cana-6113	55	10	)	)	PUNCT
cana-6113	55	11	pascal	pascal	PROPN
cana-6113	55	12	voc	voc	NOUN
cana-6113	55	13	,	,	PUNCT
cana-6113	55	14	cityscapes	cityscape	NOUN
cana-6113	55	15	improved	improve	VERB
cana-6113	55	16	semantic	semantic	ADJ
cana-6113	55	17	segmentation	segmentation	NOUN
cana-6113	55	18	using	use	VERB
cana-6113	55	19	multi	multi	ADJ
cana-6113	55	20	-	-	NOUN
cana-6113	55	21	scale	scale	ADJ
cana-6113	55	22	and	and	CCONJ
cana-6113	55	23	contextual	contextual	ADJ
cana-6113	55	24	feature	feature	NOUN
cana-6113	55	25	extraction	extraction	NOUN
cana-6113	55	26	.	.	PUNCT
cana-6113	56	1	[	[	X
cana-6113	56	2	9	9	NUM
cana-6113	56	3	]	]	PUNCT
cana-6113	56	4	efficientnet	efficientnet	NOUN
cana-6113	56	5	imagenet	imagenet	NOUN
cana-6113	56	6	achieved	achieve	VERB
cana-6113	56	7	high	high	ADJ
cana-6113	56	8	performance	performance	NOUN
cana-6113	56	9	using	use	VERB
cana-6113	56	10	compound	compound	NOUN
cana-6113	56	11	scaling	scaling	NOUN
cana-6113	56	12	with	with	ADP
cana-6113	56	13	reduced	reduced	ADJ
cana-6113	56	14	computational	computational	ADJ
cana-6113	56	15	cost	cost	NOUN
cana-6113	56	16	.	.	PUNCT
cana-6113	57	1	[	[	X
cana-6113	57	2	10	10	NUM
cana-6113	57	3	]	]	PUNCT
cana-6113	57	4	vision	vision	NOUN
cana-6113	57	5	transformer	transformer	NOUN
cana-6113	57	6	(	(	PUNCT
cana-6113	57	7	vit	vit	NOUN
cana-6113	57	8	)	)	PUNCT
cana-6113	57	9	imagenet	imagenet	NOUN
cana-6113	57	10	,	,	PUNCT
cana-6113	57	11	jft300	jft300	PROPN
cana-6113	57	12	m	m	PROPN
cana-6113	57	13	replaced	replace	VERB
cana-6113	57	14	cnns	cnn	NOUN
cana-6113	57	15	with	with	ADP
cana-6113	57	16	transformers	transformer	NOUN
cana-6113	57	17	,	,	PUNCT
cana-6113	57	18	achieving	achieve	VERB
cana-6113	57	19	superior	superior	ADJ
cana-6113	57	20	image	image	NOUN
cana-6113	57	21	classification	classification	NOUN
cana-6113	57	22	results	result	NOUN
cana-6113	57	23	.	.	PUNCT
cana-6113	58	1	2	2	X
cana-6113	58	2	.	.	X
cana-6113	58	3	image	image	NOUN
cana-6113	58	4	preprocessing	preprocessing	NOUN
cana-6113	58	5	and	and	CCONJ
cana-6113	58	6	feature	feature	NOUN
cana-6113	58	7	extraction	extraction	NOUN
cana-6113	58	8	the	the	DET
cana-6113	58	9	image	image	NOUN
cana-6113	58	10	preprocessing	preprocessing	NOUN
cana-6113	58	11	and	and	CCONJ
cana-6113	58	12	feature	feature	NOUN
cana-6113	58	13	extraction	extraction	NOUN
cana-6113	58	14	stages	stage	NOUN
cana-6113	58	15	play	play	VERB
cana-6113	58	16	a	a	DET
cana-6113	58	17	vital	vital	ADJ
cana-6113	58	18	role	role	NOUN
cana-6113	58	19	in	in	ADP
cana-6113	58	20	ensuring	ensure	VERB
cana-6113	58	21	the	the	DET
cana-6113	58	22	accuracy	accuracy	NOUN
cana-6113	58	23	and	and	CCONJ
cana-6113	58	24	reliability	reliability	NOUN
cana-6113	58	25	of	of	ADP
cana-6113	58	26	the	the	DET
cana-6113	58	27	proposed	propose	VERB
cana-6113	58	28	cervical	cervical	ADJ
cana-6113	58	29	lesion	lesion	NOUN
cana-6113	58	30	prediction	prediction	NOUN
cana-6113	58	31	model	model	NOUN
cana-6113	58	32	.	.	PUNCT
cana-6113	59	1	colposcopy	colposcopy	NOUN
cana-6113	59	2	images	image	NOUN
cana-6113	59	3	,	,	PUNCT
cana-6113	59	4	often	often	ADV
cana-6113	59	5	affected	affect	VERB
cana-6113	59	6	by	by	ADP
cana-6113	59	7	uneven	uneven	ADJ
cana-6113	59	8	illumination	illumination	NOUN
cana-6113	59	9	,	,	PUNCT
cana-6113	59	10	noise	noise	NOUN
cana-6113	59	11	,	,	PUNCT
cana-6113	59	12	and	and	CCONJ
cana-6113	59	13	contrast	contrast	NOUN
cana-6113	59	14	variations	variation	NOUN
cana-6113	59	15	,	,	PUNCT
cana-6113	59	16	require	require	VERB
cana-6113	59	17	careful	careful	ADJ
cana-6113	59	18	enhancement	enhancement	NOUN
cana-6113	59	19	to	to	PART
cana-6113	59	20	improve	improve	VERB
cana-6113	59	21	the	the	DET
cana-6113	59	22	visibility	visibility	NOUN
cana-6113	59	23	of	of	ADP
cana-6113	59	24	diagnostically	diagnostically	ADV
cana-6113	59	25	important	important	ADJ
cana-6113	59	26	regions	region	NOUN
cana-6113	59	27	before	before	SCONJ
cana-6113	59	28	they	they	PRON
cana-6113	59	29	are	be	AUX
cana-6113	59	30	used	use	VERB
cana-6113	59	31	for	for	ADP
cana-6113	59	32	feature	feature	NOUN
cana-6113	59	33	analysis	analysis	NOUN
cana-6113	59	34	.	.	PUNCT
cana-6113	60	1	the	the	DET
cana-6113	60	2	preprocessing	preprocessing	NOUN
cana-6113	60	3	pipeline	pipeline	NOUN
cana-6113	60	4	,	,	PUNCT
cana-6113	60	5	illustrated	illustrate	VERB
cana-6113	60	6	in	in	ADP
cana-6113	60	7	figure	figure	NOUN
cana-6113	60	8	3	3	NUM
cana-6113	60	9	,	,	PUNCT
cana-6113	60	10	ensures	ensure	VERB
cana-6113	60	11	that	that	SCONJ
cana-6113	60	12	the	the	DET
cana-6113	60	13	images	image	NOUN
cana-6113	60	14	are	be	AUX
cana-6113	60	15	standardized	standardize	VERB
cana-6113	60	16	,	,	PUNCT
cana-6113	60	17	high	high	ADJ
cana-6113	60	18	in	in	ADP
cana-6113	60	19	clarity	clarity	NOUN
cana-6113	60	20	,	,	PUNCT
cana-6113	60	21	and	and	CCONJ
cana-6113	60	22	suitable	suitable	ADJ
cana-6113	60	23	for	for	ADP
cana-6113	60	24	deep	deep	ADJ
cana-6113	60	25	feature	feature	NOUN
cana-6113	60	26	extraction	extraction	NOUN
cana-6113	60	27	and	and	CCONJ
cana-6113	60	28	classification	classification	NOUN
cana-6113	60	29	.	.	PUNCT
cana-6113	61	1	3.1	3.1	NUM
cana-6113	61	2	image	image	NOUN
cana-6113	61	3	pre	pre	ADJ
cana-6113	61	4	-	-	ADJ
cana-6113	61	5	processing	processing	ADJ
cana-6113	61	6	preprocessing	preprocessing	NOUN
cana-6113	61	7	aims	aim	NOUN
cana-6113	61	8	to	to	PART
cana-6113	61	9	enhance	enhance	VERB
cana-6113	61	10	the	the	DET
cana-6113	61	11	visual	visual	ADJ
cana-6113	61	12	quality	quality	NOUN
cana-6113	61	13	of	of	ADP
cana-6113	61	14	colposcopy	colposcopy	NOUN
cana-6113	61	15	images	image	NOUN
cana-6113	61	16	and	and	CCONJ
cana-6113	61	17	eliminate	eliminate	VERB
cana-6113	61	18	unwanted	unwanted	ADJ
cana-6113	61	19	artifacts	artifact	NOUN
cana-6113	61	20	that	that	PRON
cana-6113	61	21	may	may	AUX
cana-6113	61	22	interfere	interfere	VERB
cana-6113	61	23	with	with	ADP
cana-6113	61	24	accurate	accurate	ADJ
cana-6113	61	25	feature	feature	NOUN
cana-6113	61	26	extraction	extraction	NOUN
cana-6113	61	27	.	.	PUNCT
cana-6113	62	1	since	since	SCONJ
cana-6113	62	2	the	the	DET
cana-6113	62	3	diagnostic	diagnostic	ADJ
cana-6113	62	4	regions	region	NOUN
cana-6113	62	5	in	in	ADP
cana-6113	62	6	cervical	cervical	ADJ
cana-6113	62	7	images	image	NOUN
cana-6113	62	8	such	such	ADJ
cana-6113	62	9	as	as	ADP
cana-6113	62	10	acetowhite	acetowhite	ADJ
cana-6113	62	11	lesions	lesion	NOUN
cana-6113	62	12	,	,	PUNCT
cana-6113	62	13	mosaic	mosaic	ADJ
cana-6113	62	14	patterns	pattern	NOUN
cana-6113	62	15	,	,	PUNCT
cana-6113	62	16	and	and	CCONJ
cana-6113	62	17	vascular	vascular	ADJ
cana-6113	62	18	structures	structure	NOUN
cana-6113	62	19	are	be	AUX
cana-6113	62	20	often	often	ADV
cana-6113	62	21	subtle	subtle	ADJ
cana-6113	62	22	,	,	PUNCT
cana-6113	62	23	improving	improve	VERB
cana-6113	62	24	contrast	contrast	NOUN
cana-6113	62	25	and	and	CCONJ
cana-6113	62	26	sharpness	sharpness	NOUN
cana-6113	62	27	is	be	AUX
cana-6113	62	28	essential	essential	ADJ
cana-6113	62	29	for	for	ADP
cana-6113	62	30	effective	effective	ADJ
cana-6113	62	31	analysis	analysis	NOUN
cana-6113	62	32	.	.	PUNCT
cana-6113	63	1	•	•	NUM
cana-6113	63	2	image	image	NOUN
cana-6113	63	3	enhancement	enhancement	NOUN
cana-6113	63	4	:	:	PUNCT
cana-6113	63	5	image	image	NOUN
cana-6113	63	6	enhancement	enhancement	NOUN
cana-6113	63	7	techniques	technique	NOUN
cana-6113	63	8	such	such	ADJ
cana-6113	63	9	as	as	ADP
cana-6113	63	10	contrast	contrast	NOUN
cana-6113	63	11	adjustment	adjustment	NOUN
cana-6113	63	12	,	,	PUNCT
cana-6113	63	13	histogram	histogram	NOUN
cana-6113	63	14	equalization	equalization	NOUN
cana-6113	63	15	,	,	PUNCT
cana-6113	63	16	and	and	CCONJ
cana-6113	63	17	brightness	brightness	NOUN
cana-6113	63	18	normalization	normalization	NOUN
cana-6113	63	19	are	be	AUX
cana-6113	63	20	applied	apply	VERB
cana-6113	63	21	to	to	PART
cana-6113	63	22	emphasize	emphasize	VERB
cana-6113	63	23	lesion	lesion	NOUN
cana-6113	63	24	boundaries	boundary	NOUN
cana-6113	63	25	and	and	CCONJ
cana-6113	63	26	internal	internal	ADJ
cana-6113	63	27	textures	texture	NOUN
cana-6113	63	28	.	.	PUNCT
cana-6113	64	1	these	these	DET
cana-6113	64	2	methods	method	NOUN
cana-6113	64	3	improve	improve	VERB
cana-6113	64	4	the	the	DET
cana-6113	64	5	differentiation	differentiation	NOUN
cana-6113	64	6	between	between	ADP
cana-6113	64	7	normal	normal	ADJ
cana-6113	64	8	and	and	CCONJ
cana-6113	64	9	abnormal	abnormal	ADJ
cana-6113	64	10	tissue	tissue	NOUN
cana-6113	64	11	regions	region	NOUN
cana-6113	64	12	by	by	ADP
cana-6113	64	13	enhancing	enhance	VERB
cana-6113	64	14	edge	edge	NOUN
cana-6113	64	15	visibility	visibility	NOUN
cana-6113	64	16	and	and	CCONJ
cana-6113	64	17	color	color	NOUN
cana-6113	64	18	consistency	consistency	NOUN
cana-6113	64	19	.	.	PUNCT
cana-6113	65	1	histogram	histogram	PROPN
cana-6113	65	2	equalization	equalization	NOUN
cana-6113	65	3	,	,	PUNCT
cana-6113	65	4	in	in	ADP
cana-6113	65	5	particular	particular	ADJ
cana-6113	65	6	,	,	PUNCT
cana-6113	65	7	redistributes	redistribute	VERB
cana-6113	65	8	pixel	pixel	ADJ
cana-6113	65	9	intensity	intensity	NOUN
cana-6113	65	10	values	value	NOUN
cana-6113	65	11	,	,	PUNCT
cana-6113	65	12	providing	provide	VERB
cana-6113	65	13	a	a	DET
cana-6113	65	14	balanced	balanced	ADJ
cana-6113	65	15	contrast	contrast	NOUN
cana-6113	65	16	across	across	ADP
cana-6113	65	17	the	the	DET
cana-6113	65	18	entire	entire	ADJ
cana-6113	65	19	image	image	NOUN
cana-6113	65	20	.	.	PUNCT
cana-6113	66	1	this	this	DET
cana-6113	66	2	step	step	NOUN
cana-6113	66	3	ensures	ensure	VERB
cana-6113	66	4	that	that	SCONJ
cana-6113	66	5	faint	faint	ADJ
cana-6113	66	6	lesion	lesion	NOUN
cana-6113	66	7	features	feature	NOUN
cana-6113	66	8	become	become	VERB
cana-6113	66	9	more	more	ADV
cana-6113	66	10	discernible	discernible	ADJ
cana-6113	66	11	to	to	ADP
cana-6113	66	12	both	both	DET
cana-6113	66	13	human	human	ADJ
cana-6113	66	14	experts	expert	NOUN
cana-6113	66	15	and	and	CCONJ
cana-6113	66	16	the	the	DET
cana-6113	66	17	automated	automate	VERB
cana-6113	66	18	system	system	NOUN
cana-6113	66	19	.	.	PUNCT
cana-6113	67	1	•	•	NOUN
cana-6113	67	2	noise	noise	NOUN
cana-6113	67	3	reduction	reduction	NOUN
cana-6113	67	4	and	and	CCONJ
cana-6113	67	5	smoothing	smoothing	NOUN
cana-6113	67	6	:	:	PUNCT
cana-6113	67	7	colposcopy	colposcopy	NOUN
cana-6113	67	8	images	image	NOUN
cana-6113	67	9	may	may	AUX
cana-6113	67	10	contain	contain	VERB
cana-6113	67	11	random	random	ADJ
cana-6113	67	12	noise	noise	NOUN
cana-6113	67	13	due	due	ADP
cana-6113	67	14	to	to	ADP
cana-6113	67	15	lighting	lighting	NOUN
cana-6113	67	16	inconsistencies	inconsistency	NOUN
cana-6113	67	17	or	or	CCONJ
cana-6113	67	18	camera	camera	NOUN
cana-6113	67	19	artifacts	artifact	NOUN
cana-6113	67	20	.	.	PUNCT
cana-6113	68	1	to	to	PART
cana-6113	68	2	mitigate	mitigate	VERB
cana-6113	68	3	this	this	PRON
cana-6113	68	4	,	,	PUNCT
cana-6113	68	5	filters	filter	NOUN
cana-6113	68	6	such	such	ADJ
cana-6113	68	7	as	as	ADP
cana-6113	68	8	gaussian	gaussian	ADJ
cana-6113	68	9	smoothing	smoothing	NOUN
cana-6113	68	10	or	or	CCONJ
cana-6113	68	11	median	median	ADJ
cana-6113	68	12	filtering	filtering	NOUN
cana-6113	68	13	are	be	AUX
cana-6113	68	14	employed	employ	VERB
cana-6113	68	15	to	to	PART
cana-6113	68	16	suppress	suppress	VERB
cana-6113	68	17	noise	noise	NOUN
cana-6113	68	18	while	while	SCONJ
cana-6113	68	19	preserving	preserve	VERB
cana-6113	68	20	important	important	ADJ
cana-6113	68	21	structural	structural	ADJ
cana-6113	68	22	edges	edge	NOUN
cana-6113	68	23	.	.	PUNCT
cana-6113	69	1	this	this	DET
cana-6113	69	2	step	step	NOUN
cana-6113	69	3	results	result	VERB
cana-6113	69	4	in	in	ADP
cana-6113	69	5	cleaner	clean	ADJ
cana-6113	69	6	images	image	NOUN
cana-6113	69	7	that	that	PRON
cana-6113	69	8	facilitate	facilitate	VERB
cana-6113	69	9	more	more	ADV
cana-6113	69	10	accurate	accurate	ADJ
cana-6113	69	11	texture	texture	NOUN
cana-6113	69	12	and	and	CCONJ
cana-6113	69	13	edge	edge	NOUN
cana-6113	69	14	detection	detection	NOUN
cana-6113	69	15	.	.	PUNCT
cana-6113	70	1	•	•	NUM
cana-6113	70	2	region	region	NOUN
cana-6113	70	3	of	of	ADP
cana-6113	70	4	interest	interest	NOUN
cana-6113	70	5	(	(	PUNCT
cana-6113	70	6	roi	roi	NOUN
cana-6113	70	7	)	)	PUNCT
cana-6113	70	8	detection	detection	NOUN
cana-6113	70	9	:	:	PUNCT
cana-6113	70	10	in	in	ADP
cana-6113	70	11	many	many	ADJ
cana-6113	70	12	cases	case	NOUN
cana-6113	70	13	,	,	PUNCT
cana-6113	70	14	only	only	ADV
cana-6113	70	15	specific	specific	ADJ
cana-6113	70	16	portions	portion	NOUN
cana-6113	70	17	of	of	ADP
cana-6113	70	18	the	the	DET
cana-6113	70	19	cervix	cervix	NOUN
cana-6113	70	20	contain	contain	VERB
cana-6113	70	21	diagnostically	diagnostically	ADV
cana-6113	70	22	relevant	relevant	ADJ
cana-6113	70	23	information	information	NOUN
cana-6113	70	24	.	.	PUNCT
cana-6113	71	1	therefore	therefore	ADV
cana-6113	71	2	,	,	PUNCT
cana-6113	71	3	segmentation	segmentation	NOUN
cana-6113	71	4	or	or	CCONJ
cana-6113	71	5	roi	roi	NOUN
cana-6113	71	6	extraction	extraction	NOUN
cana-6113	71	7	methods	method	NOUN
cana-6113	71	8	are	be	AUX
cana-6113	71	9	applied	apply	VERB
cana-6113	71	10	to	to	PART
cana-6113	71	11	isolate	isolate	VERB
cana-6113	71	12	the	the	DET
cana-6113	71	13	cervical	cervical	ADJ
cana-6113	71	14	region	region	NOUN
cana-6113	71	15	and	and	CCONJ
cana-6113	71	16	exclude	exclude	VERB
cana-6113	71	17	irrelevant	irrelevant	ADJ
cana-6113	71	18	background	background	NOUN
cana-6113	71	19	areas	area	NOUN
cana-6113	71	20	.	.	PUNCT
cana-6113	72	1	this	this	DET
cana-6113	72	2	targeted	target	VERB
cana-6113	72	3	focus	focus	NOUN
cana-6113	72	4	improves	improve	VERB
cana-6113	72	5	computational	computational	ADJ
cana-6113	72	6	efficiency	efficiency	NOUN
cana-6113	72	7	and	and	CCONJ
cana-6113	72	8	ensures	ensure	VERB
cana-6113	72	9	that	that	SCONJ
cana-6113	72	10	the	the	DET
cana-6113	72	11	extracted	extract	VERB
cana-6113	72	12	features	feature	NOUN
cana-6113	72	13	pertain	pertain	VERB
cana-6113	72	14	directly	directly	ADV
cana-6113	72	15	to	to	ADP
cana-6113	72	16	clinically	clinically	ADV
cana-6113	72	17	significant	significant	ADJ
cana-6113	72	18	regions	region	NOUN
cana-6113	72	19	.	.	PUNCT
cana-6113	73	1	3.2	3.2	NUM
cana-6113	73	2	image	image	NOUN
cana-6113	73	3	feature	feature	NOUN
cana-6113	73	4	extraction	extraction	NOUN
cana-6113	73	5	after	after	ADP
cana-6113	73	6	preprocessing	preprocessing	NOUN
cana-6113	73	7	,	,	PUNCT
cana-6113	73	8	the	the	DET
cana-6113	73	9	system	system	NOUN
cana-6113	73	10	proceeds	proceed	VERB
cana-6113	73	11	to	to	PART
cana-6113	73	12	feature	feature	NOUN
cana-6113	73	13	extraction	extraction	NOUN
cana-6113	73	14	,	,	PUNCT
cana-6113	73	15	where	where	SCONJ
cana-6113	73	16	the	the	DET
cana-6113	73	17	goal	goal	NOUN
cana-6113	73	18	is	be	AUX
cana-6113	73	19	to	to	PART
cana-6113	73	20	identify	identify	VERB
cana-6113	73	21	meaningful	meaningful	ADJ
cana-6113	73	22	patterns	pattern	NOUN
cana-6113	73	23	and	and	CCONJ
cana-6113	73	24	numerical	numerical	ADJ
cana-6113	73	25	descriptors	descriptor	NOUN
cana-6113	73	26	that	that	PRON
cana-6113	73	27	capture	capture	VERB
cana-6113	73	28	the	the	DET
cana-6113	73	29	visual	visual	ADJ
cana-6113	73	30	characteristics	characteristic	NOUN
cana-6113	73	31	of	of	ADP
cana-6113	73	32	cervical	cervical	ADJ
cana-6113	73	33	communications	communication	NOUN
cana-6113	73	34	on	on	ADP
cana-6113	73	35	applied	apply	VERB
cana-6113	73	36	nonlinear	nonlinear	ADJ
cana-6113	73	37	analysis	analysis	NOUN
cana-6113	73	38	issn	issn	NOUN
cana-6113	73	39	:	:	PUNCT
cana-6113	73	40	1074	1074	NUM
cana-6113	73	41	-	-	PUNCT
cana-6113	73	42	133x	133x	NUM
cana-6113	73	43	vol	vol	VERB
cana-6113	73	44	32	32	NUM
cana-6113	73	45	no	no	NOUN
cana-6113	73	46	.	.	PUNCT
cana-6113	74	1	10s	10	NOUN
cana-6113	74	2	(	(	PUNCT
cana-6113	74	3	2025	2025	NUM
cana-6113	74	4	)	)	PUNCT
cana-6113	74	5	3407	3407	NUM
cana-6113	74	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	74	7	lesions	lesion	NOUN
cana-6113	74	8	.	.	PUNCT
cana-6113	75	1	these	these	DET
cana-6113	75	2	features	feature	NOUN
cana-6113	75	3	are	be	AUX
cana-6113	75	4	crucial	crucial	ADJ
cana-6113	75	5	for	for	ADP
cana-6113	75	6	distinguishing	distinguish	VERB
cana-6113	75	7	between	between	ADP
cana-6113	75	8	normal	normal	ADJ
cana-6113	75	9	,	,	PUNCT
cana-6113	75	10	precancerous	precancerous	ADJ
cana-6113	75	11	,	,	PUNCT
cana-6113	75	12	and	and	CCONJ
cana-6113	75	13	cancerous	cancerous	ADJ
cana-6113	75	14	tissues	tissue	NOUN
cana-6113	75	15	.	.	PUNCT
cana-6113	76	1	•	•	NUM
cana-6113	76	2	texture	texture	NOUN
cana-6113	76	3	features	feature	VERB
cana-6113	76	4	:	:	PUNCT
cana-6113	76	5	texture	texture	NOUN
cana-6113	76	6	plays	play	VERB
cana-6113	76	7	a	a	DET
cana-6113	76	8	fundamental	fundamental	ADJ
cana-6113	76	9	role	role	NOUN
cana-6113	76	10	in	in	ADP
cana-6113	76	11	describing	describe	VERB
cana-6113	76	12	the	the	DET
cana-6113	76	13	surface	surface	NOUN
cana-6113	76	14	irregularities	irregularity	NOUN
cana-6113	76	15	and	and	CCONJ
cana-6113	76	16	spatial	spatial	ADJ
cana-6113	76	17	arrangement	arrangement	NOUN
cana-6113	76	18	of	of	ADP
cana-6113	76	19	pixel	pixel	PROPN
cana-6113	76	20	intensities	intensity	NOUN
cana-6113	76	21	in	in	ADP
cana-6113	76	22	colposcopy	colposcopy	NOUN
cana-6113	76	23	images	image	NOUN
cana-6113	76	24	.	.	PUNCT
cana-6113	77	1	lesions	lesion	NOUN
cana-6113	77	2	often	often	ADV
cana-6113	77	3	exhibit	exhibit	VERB
cana-6113	77	4	distinct	distinct	ADJ
cana-6113	77	5	textural	textural	ADJ
cana-6113	77	6	properties	property	NOUN
cana-6113	77	7	,	,	PUNCT
cana-6113	77	8	such	such	ADJ
cana-6113	77	9	as	as	ADP
cana-6113	77	10	roughness	roughness	NOUN
cana-6113	77	11	or	or	CCONJ
cana-6113	77	12	homogeneity	homogeneity	NOUN
cana-6113	77	13	,	,	PUNCT
cana-6113	77	14	which	which	PRON
cana-6113	77	15	can	can	AUX
cana-6113	77	16	serve	serve	VERB
cana-6113	77	17	as	as	ADP
cana-6113	77	18	key	key	ADJ
cana-6113	77	19	indicators	indicator	NOUN
cana-6113	77	20	of	of	ADP
cana-6113	77	21	disease	disease	NOUN
cana-6113	77	22	progression	progression	NOUN
cana-6113	77	23	.	.	PUNCT
cana-6113	78	1	one	one	NUM
cana-6113	78	2	of	of	ADP
cana-6113	78	3	the	the	DET
cana-6113	78	4	most	most	ADV
cana-6113	78	5	widely	widely	ADV
cana-6113	78	6	used	use	VERB
cana-6113	78	7	techniques	technique	NOUN
cana-6113	78	8	for	for	ADP
cana-6113	78	9	texture	texture	ADJ
cana-6113	78	10	analysis	analysis	NOUN
cana-6113	78	11	is	be	AUX
cana-6113	78	12	the	the	DET
cana-6113	78	13	gray	gray	ADJ
cana-6113	78	14	-	-	PUNCT
cana-6113	78	15	level	level	NOUN
cana-6113	78	16	co	co	NOUN
cana-6113	78	17	-	-	NOUN
cana-6113	78	18	occurrence	occurrence	ADJ
cana-6113	78	19	matrix	matrix	NOUN
cana-6113	78	20	(	(	PUNCT
cana-6113	78	21	glcm	glcm	PROPN
cana-6113	78	22	)	)	PUNCT
cana-6113	78	23	.	.	PUNCT
cana-6113	79	1	glcm	glcm	PROPN
cana-6113	79	2	quantifies	quantify	VERB
cana-6113	79	3	the	the	DET
cana-6113	79	4	frequency	frequency	NOUN
cana-6113	79	5	at	at	ADP
cana-6113	79	6	which	which	PRON
cana-6113	79	7	pairs	pair	NOUN
cana-6113	79	8	of	of	ADP
cana-6113	79	9	pixels	pixel	NOUN
cana-6113	79	10	with	with	ADP
cana-6113	79	11	specific	specific	ADJ
cana-6113	79	12	gray	gray	ADJ
cana-6113	79	13	-	-	PUNCT
cana-6113	79	14	level	level	NOUN
cana-6113	79	15	values	value	NOUN
cana-6113	79	16	occur	occur	VERB
cana-6113	79	17	at	at	ADP
cana-6113	79	18	a	a	DET
cana-6113	79	19	defined	define	VERB
cana-6113	79	20	spatial	spatial	ADJ
cana-6113	79	21	relationship	relationship	NOUN
cana-6113	79	22	in	in	ADP
cana-6113	79	23	the	the	DET
cana-6113	79	24	image	image	NOUN
cana-6113	79	25	.	.	PUNCT
cana-6113	80	1	from	from	ADP
cana-6113	80	2	the	the	DET
cana-6113	80	3	glcm	glcm	NOUN
cana-6113	80	4	,	,	PUNCT
cana-6113	80	5	several	several	ADJ
cana-6113	80	6	statistical	statistical	ADJ
cana-6113	80	7	measures	measure	NOUN
cana-6113	80	8	such	such	ADJ
cana-6113	80	9	as	as	ADP
cana-6113	80	10	contrast	contrast	NOUN
cana-6113	80	11	,	,	PUNCT
cana-6113	80	12	correlation	correlation	NOUN
cana-6113	80	13	,	,	PUNCT
cana-6113	80	14	energy	energy	NOUN
cana-6113	80	15	,	,	PUNCT
cana-6113	80	16	and	and	CCONJ
cana-6113	80	17	homogeneity	homogeneity	NOUN
cana-6113	80	18	are	be	AUX
cana-6113	80	19	computed	compute	VERB
cana-6113	80	20	to	to	PART
cana-6113	80	21	characterize	characterize	VERB
cana-6113	80	22	texture	texture	NOUN
cana-6113	80	23	patterns	pattern	NOUN
cana-6113	80	24	.	.	PUNCT
cana-6113	81	1	o	o	NOUN
cana-6113	81	2	contrast	contrast	NOUN
cana-6113	81	3	measures	measure	NOUN
cana-6113	81	4	local	local	ADJ
cana-6113	81	5	intensity	intensity	NOUN
cana-6113	81	6	variations	variation	NOUN
cana-6113	81	7	,	,	PUNCT
cana-6113	81	8	highlighting	highlight	VERB
cana-6113	81	9	lesion	lesion	NOUN
cana-6113	81	10	boundaries	boundary	NOUN
cana-6113	81	11	.	.	PUNCT
cana-6113	82	1	o	o	NOUN
cana-6113	82	2	correlation	correlation	NOUN
cana-6113	82	3	identifies	identify	VERB
cana-6113	82	4	the	the	DET
cana-6113	82	5	degree	degree	NOUN
cana-6113	82	6	of	of	ADP
cana-6113	82	7	linear	linear	ADJ
cana-6113	82	8	dependency	dependency	NOUN
cana-6113	82	9	between	between	ADP
cana-6113	82	10	neighboring	neighboring	NOUN
cana-6113	82	11	pixels	pixel	NOUN
cana-6113	82	12	.	.	PUNCT
cana-6113	83	1	o	o	NOUN
cana-6113	83	2	energy	energy	NOUN
cana-6113	83	3	indicates	indicate	VERB
cana-6113	83	4	textural	textural	ADJ
cana-6113	83	5	uniformity	uniformity	NOUN
cana-6113	83	6	,	,	PUNCT
cana-6113	83	7	often	often	ADV
cana-6113	83	8	lower	low	ADJ
cana-6113	83	9	in	in	ADP
cana-6113	83	10	irregular	irregular	ADJ
cana-6113	83	11	or	or	CCONJ
cana-6113	83	12	abnormal	abnormal	ADJ
cana-6113	83	13	tissue	tissue	NOUN
cana-6113	83	14	regions	region	NOUN
cana-6113	83	15	.	.	PUNCT
cana-6113	84	1	o	o	NOUN
cana-6113	84	2	homogeneity	homogeneity	NOUN
cana-6113	84	3	assesses	assess	VERB
cana-6113	84	4	the	the	DET
cana-6113	84	5	smoothness	smoothness	NOUN
cana-6113	84	6	of	of	ADP
cana-6113	84	7	pixel	pixel	ADJ
cana-6113	84	8	transitions	transition	NOUN
cana-6113	84	9	,	,	PUNCT
cana-6113	84	10	distinguishing	distinguish	VERB
cana-6113	84	11	between	between	ADP
cana-6113	84	12	healthy	healthy	ADJ
cana-6113	84	13	and	and	CCONJ
cana-6113	84	14	rough	rough	ADJ
cana-6113	84	15	lesion	lesion	NOUN
cana-6113	84	16	surfaces	surface	NOUN
cana-6113	84	17	.	.	PUNCT
cana-6113	85	1	these	these	DET
cana-6113	85	2	extracted	extract	VERB
cana-6113	85	3	texture	texture	NOUN
cana-6113	85	4	features	feature	NOUN
cana-6113	85	5	provide	provide	VERB
cana-6113	85	6	quantitative	quantitative	ADJ
cana-6113	85	7	insights	insight	NOUN
cana-6113	85	8	into	into	ADP
cana-6113	85	9	the	the	DET
cana-6113	85	10	visual	visual	ADJ
cana-6113	85	11	patterns	pattern	NOUN
cana-6113	85	12	of	of	ADP
cana-6113	85	13	cervical	cervical	ADJ
cana-6113	85	14	tissue	tissue	NOUN
cana-6113	85	15	,	,	PUNCT
cana-6113	85	16	enabling	enable	VERB
cana-6113	85	17	the	the	DET
cana-6113	85	18	model	model	NOUN
cana-6113	85	19	to	to	PART
cana-6113	85	20	differentiate	differentiate	VERB
cana-6113	85	21	between	between	ADP
cana-6113	85	22	normal	normal	ADJ
cana-6113	85	23	and	and	CCONJ
cana-6113	85	24	abnormal	abnormal	ADJ
cana-6113	85	25	colposcopy	colposcopy	NOUN
cana-6113	85	26	images	image	NOUN
cana-6113	85	27	effectively	effectively	ADV
cana-6113	85	28	.	.	PUNCT
cana-6113	86	1	when	when	SCONJ
cana-6113	86	2	combined	combine	VERB
cana-6113	86	3	with	with	ADP
cana-6113	86	4	deep	deep	ADJ
cana-6113	86	5	learning	learning	NOUN
cana-6113	86	6	features	feature	NOUN
cana-6113	86	7	or	or	CCONJ
cana-6113	86	8	clinical	clinical	ADJ
cana-6113	86	9	data	datum	NOUN
cana-6113	86	10	in	in	ADP
cana-6113	86	11	later	later	ADJ
cana-6113	86	12	stages	stage	NOUN
cana-6113	86	13	,	,	PUNCT
cana-6113	86	14	they	they	PRON
cana-6113	86	15	contribute	contribute	VERB
cana-6113	86	16	significantly	significantly	ADV
cana-6113	86	17	to	to	ADP
cana-6113	86	18	the	the	DET
cana-6113	86	19	overall	overall	ADJ
cana-6113	86	20	diagnostic	diagnostic	ADJ
cana-6113	86	21	performance	performance	NOUN
cana-6113	86	22	of	of	ADP
cana-6113	86	23	the	the	DET
cana-6113	86	24	proposed	propose	VERB
cana-6113	86	25	system	system	NOUN
cana-6113	86	26	.	.	PUNCT
cana-6113	87	1	3	3	X
cana-6113	87	2	.	.	NUM
cana-6113	87	3	proposed	propose	VERB
cana-6113	87	4	system	system	NOUN
cana-6113	87	5	model	model	NOUN
cana-6113	87	6	the	the	DET
cana-6113	87	7	proposed	propose	VERB
cana-6113	87	8	system	system	NOUN
cana-6113	87	9	introduces	introduce	VERB
cana-6113	87	10	a	a	DET
cana-6113	87	11	multi	multi	ADJ
cana-6113	87	12	-	-	ADJ
cana-6113	87	13	modal	modal	ADJ
cana-6113	87	14	prediction	prediction	NOUN
cana-6113	87	15	framework	framework	NOUN
cana-6113	87	16	designed	design	VERB
cana-6113	87	17	to	to	PART
cana-6113	87	18	integrate	integrate	VERB
cana-6113	87	19	both	both	DET
cana-6113	87	20	clinical	clinical	ADJ
cana-6113	87	21	data	datum	NOUN
cana-6113	87	22	and	and	CCONJ
cana-6113	87	23	colposcopy	colposcopy	VERB
cana-6113	87	24	image	image	NOUN
cana-6113	87	25	features	feature	NOUN
cana-6113	87	26	for	for	ADP
cana-6113	87	27	improved	improved	ADJ
cana-6113	87	28	diagnostic	diagnostic	ADJ
cana-6113	87	29	accuracy	accuracy	NOUN
cana-6113	87	30	.	.	PUNCT
cana-6113	88	1	the	the	DET
cana-6113	88	2	model	model	NOUN
cana-6113	88	3	architecture	architecture	NOUN
cana-6113	88	4	follows	follow	VERB
cana-6113	88	5	a	a	DET
cana-6113	88	6	two	two	NUM
cana-6113	88	7	-	-	PUNCT
cana-6113	88	8	branch	branch	NOUN
cana-6113	88	9	structure	structure	NOUN
cana-6113	88	10	where	where	SCONJ
cana-6113	88	11	each	each	DET
cana-6113	88	12	data	data	NOUN
cana-6113	88	13	modality	modality	NOUN
cana-6113	88	14	—	—	PUNCT
cana-6113	88	15	tabular	tabular	VERB
cana-6113	88	16	clinical	clinical	ADJ
cana-6113	88	17	information	information	NOUN
cana-6113	88	18	and	and	CCONJ
cana-6113	88	19	visual	visual	ADJ
cana-6113	88	20	image	image	NOUN
cana-6113	88	21	data	datum	NOUN
cana-6113	88	22	is	be	AUX
cana-6113	88	23	processed	process	VERB
cana-6113	88	24	independently	independently	ADV
cana-6113	88	25	using	use	VERB
cana-6113	88	26	specialized	specialized	ADJ
cana-6113	88	27	techniques	technique	NOUN
cana-6113	88	28	before	before	ADP
cana-6113	88	29	being	be	AUX
cana-6113	88	30	fused	fuse	VERB
cana-6113	88	31	at	at	ADP
cana-6113	88	32	a	a	DET
cana-6113	88	33	later	later	ADJ
cana-6113	88	34	stage	stage	NOUN
cana-6113	88	35	for	for	ADP
cana-6113	88	36	final	final	ADJ
cana-6113	88	37	classification	classification	NOUN
cana-6113	88	38	.	.	PUNCT
cana-6113	89	1	this	this	DET
cana-6113	89	2	approach	approach	NOUN
cana-6113	89	3	ensures	ensure	VERB
cana-6113	89	4	that	that	SCONJ
cana-6113	89	5	both	both	DET
cana-6113	89	6	data	datum	NOUN
cana-6113	89	7	types	type	NOUN
cana-6113	89	8	are	be	AUX
cana-6113	89	9	utilized	utilize	VERB
cana-6113	89	10	effectively	effectively	ADV
cana-6113	89	11	,	,	PUNCT
cana-6113	89	12	capturing	capture	VERB
cana-6113	89	13	the	the	DET
cana-6113	89	14	unique	unique	ADJ
cana-6113	89	15	statistical	statistical	ADJ
cana-6113	89	16	and	and	CCONJ
cana-6113	89	17	visual	visual	ADJ
cana-6113	89	18	cues	cue	NOUN
cana-6113	89	19	relevant	relevant	ADJ
cana-6113	89	20	to	to	PART
cana-6113	89	21	disease	disease	NOUN
cana-6113	89	22	prediction	prediction	NOUN
cana-6113	89	23	and	and	CCONJ
cana-6113	89	24	diagnosis	diagnosis	NOUN
cana-6113	89	25	.	.	PUNCT
cana-6113	90	1	the	the	DET
cana-6113	90	2	overall	overall	ADJ
cana-6113	90	3	structure	structure	NOUN
cana-6113	90	4	of	of	ADP
cana-6113	90	5	the	the	DET
cana-6113	90	6	algorithm	algorithm	NOUN
cana-6113	90	7	is	be	AUX
cana-6113	90	8	depicted	depict	VERB
cana-6113	90	9	in	in	ADP
cana-6113	90	10	figure	figure	NOUN
cana-6113	90	11	1	1	NUM
cana-6113	90	12	.	.	PUNCT
cana-6113	90	13	figure	figure	NOUN
cana-6113	90	14	1	1	NUM
cana-6113	90	15	:	:	PUNCT
cana-6113	90	16	deep	deep	ADJ
cana-6113	90	17	learning	learning	NOUN
cana-6113	90	18	-	-	PUNCT
cana-6113	90	19	based	base	VERB
cana-6113	90	20	framework	framework	NOUN
cana-6113	90	21	for	for	ADP
cana-6113	90	22	image	image	NOUN
cana-6113	90	23	future	future	ADJ
cana-6113	90	24	extraction	extraction	NOUN
cana-6113	90	25	and	and	CCONJ
cana-6113	90	26	classification	classification	NOUN
cana-6113	90	27	communications	communication	NOUN
cana-6113	90	28	on	on	ADP
cana-6113	90	29	applied	apply	VERB
cana-6113	90	30	nonlinear	nonlinear	ADJ
cana-6113	90	31	analysis	analysis	NOUN
cana-6113	90	32	issn	issn	NOUN
cana-6113	90	33	:	:	PUNCT
cana-6113	90	34	1074	1074	NUM
cana-6113	90	35	-	-	PUNCT
cana-6113	90	36	133x	133x	NUM
cana-6113	90	37	vol	vol	VERB
cana-6113	90	38	32	32	NUM
cana-6113	90	39	no	no	NOUN
cana-6113	90	40	.	.	PUNCT
cana-6113	91	1	10s	10	NOUN
cana-6113	91	2	(	(	PUNCT
cana-6113	91	3	2025	2025	NUM
cana-6113	91	4	)	)	PUNCT
cana-6113	91	5	3408	3408	NUM
cana-6113	91	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	91	7	4.1	4.1	NUM
cana-6113	91	8	separate	separate	ADJ
cana-6113	91	9	processing	processing	NOUN
cana-6113	91	10	of	of	ADP
cana-6113	91	11	clinical	clinical	ADJ
cana-6113	91	12	data	datum	NOUN
cana-6113	91	13	and	and	CCONJ
cana-6113	91	14	colposcopy	colposcopy	VERB
cana-6113	91	15	images	image	NOUN
cana-6113	91	16	a	a	DET
cana-6113	91	17	)	)	PUNCT
cana-6113	91	18	clinical	clinical	ADJ
cana-6113	91	19	data	datum	NOUN
cana-6113	91	20	processing	processing	NOUN
cana-6113	91	21	:	:	PUNCT
cana-6113	91	22	the	the	DET
cana-6113	91	23	clinical	clinical	ADJ
cana-6113	91	24	dataset	dataset	NOUN
cana-6113	91	25	comprises	comprise	VERB
cana-6113	91	26	patient	patient	NOUN
cana-6113	91	27	-	-	PUNCT
cana-6113	91	28	specific	specific	ADJ
cana-6113	91	29	attributes	attribute	NOUN
cana-6113	91	30	such	such	ADJ
cana-6113	91	31	as	as	ADP
cana-6113	91	32	age	age	NOUN
cana-6113	91	33	,	,	PUNCT
cana-6113	91	34	hpv	hpv	NOUN
cana-6113	91	35	status	status	NOUN
cana-6113	91	36	,	,	PUNCT
cana-6113	91	37	smoking	smoking	NOUN
cana-6113	91	38	history	history	NOUN
cana-6113	91	39	,	,	PUNCT
cana-6113	91	40	reproductive	reproductive	ADJ
cana-6113	91	41	health	health	NOUN
cana-6113	91	42	indicators	indicator	NOUN
cana-6113	91	43	,	,	PUNCT
cana-6113	91	44	and	and	CCONJ
cana-6113	91	45	prior	prior	ADJ
cana-6113	91	46	infection	infection	NOUN
cana-6113	91	47	history	history	NOUN
cana-6113	91	48	.	.	PUNCT
cana-6113	92	1	since	since	SCONJ
cana-6113	92	2	this	this	DET
cana-6113	92	3	data	data	NOUN
cana-6113	92	4	is	be	AUX
cana-6113	92	5	primarily	primarily	ADV
cana-6113	92	6	tabular	tabular	ADJ
cana-6113	92	7	in	in	ADP
cana-6113	92	8	nature	nature	NOUN
cana-6113	92	9	,	,	PUNCT
cana-6113	92	10	it	it	PRON
cana-6113	92	11	is	be	AUX
cana-6113	92	12	first	first	ADV
cana-6113	92	13	preprocessed	preprocesse	VERB
cana-6113	92	14	to	to	PART
cana-6113	92	15	handle	handle	VERB
cana-6113	92	16	missing	miss	VERB
cana-6113	92	17	values	value	NOUN
cana-6113	92	18	,	,	PUNCT
cana-6113	92	19	normalize	normalize	VERB
cana-6113	92	20	numerical	numerical	ADJ
cana-6113	92	21	features	feature	NOUN
cana-6113	92	22	,	,	PUNCT
cana-6113	92	23	and	and	CCONJ
cana-6113	92	24	encode	encode	VERB
cana-6113	92	25	categorical	categorical	ADJ
cana-6113	92	26	variables	variable	NOUN
cana-6113	92	27	into	into	ADP
cana-6113	92	28	machine	machine	NOUN
cana-6113	92	29	-	-	PUNCT
cana-6113	92	30	readable	readable	ADJ
cana-6113	92	31	formats	format	NOUN
cana-6113	92	32	.	.	PUNCT
cana-6113	93	1	following	follow	VERB
cana-6113	93	2	preprocessing	preprocesse	VERB
cana-6113	93	3	,	,	PUNCT
cana-6113	93	4	traditional	traditional	ADJ
cana-6113	93	5	machine	machine	NOUN
cana-6113	93	6	learning	learn	VERB
cana-6113	93	7	algorithms	algorithm	NOUN
cana-6113	93	8	such	such	ADJ
cana-6113	93	9	as	as	ADP
cana-6113	93	10	random	random	ADJ
cana-6113	93	11	forest	forest	NOUN
cana-6113	93	12	(	(	PUNCT
cana-6113	93	13	rf	rf	NOUN
cana-6113	93	14	)	)	PUNCT
cana-6113	93	15	,	,	PUNCT
cana-6113	93	16	naïve	naïve	ADJ
cana-6113	93	17	bayes	bayes	NOUN
cana-6113	93	18	,	,	PUNCT
cana-6113	93	19	and	and	CCONJ
cana-6113	93	20	logistic	logistic	ADJ
cana-6113	93	21	regression	regression	NOUN
cana-6113	93	22	are	be	AUX
cana-6113	93	23	employed	employ	VERB
cana-6113	93	24	to	to	PART
cana-6113	93	25	learn	learn	VERB
cana-6113	93	26	from	from	ADP
cana-6113	93	27	the	the	DET
cana-6113	93	28	clinical	clinical	ADJ
cana-6113	93	29	data	datum	NOUN
cana-6113	93	30	.	.	PUNCT
cana-6113	94	1	these	these	DET
cana-6113	94	2	algorithms	algorithm	NOUN
cana-6113	94	3	are	be	AUX
cana-6113	94	4	effective	effective	ADJ
cana-6113	94	5	in	in	ADP
cana-6113	94	6	modeling	model	VERB
cana-6113	94	7	structured	structure	VERB
cana-6113	94	8	information	information	NOUN
cana-6113	94	9	and	and	CCONJ
cana-6113	94	10	identifying	identify	VERB
cana-6113	94	11	key	key	ADJ
cana-6113	94	12	clinical	clinical	ADJ
cana-6113	94	13	risk	risk	NOUN
cana-6113	94	14	factors	factor	NOUN
cana-6113	94	15	associated	associate	VERB
cana-6113	94	16	with	with	ADP
cana-6113	94	17	cervical	cervical	ADJ
cana-6113	94	18	abnormalities	abnormality	NOUN
cana-6113	94	19	.	.	PUNCT
cana-6113	95	1	each	each	DET
cana-6113	95	2	algorithm	algorithm	NOUN
cana-6113	95	3	independently	independently	ADV
cana-6113	95	4	extracts	extract	VERB
cana-6113	95	5	discriminative	discriminative	NOUN
cana-6113	95	6	features	feature	NOUN
cana-6113	95	7	and	and	CCONJ
cana-6113	95	8	computes	compute	VERB
cana-6113	95	9	probabilistic	probabilistic	ADJ
cana-6113	95	10	predictions	prediction	NOUN
cana-6113	95	11	,	,	PUNCT
cana-6113	95	12	forming	form	VERB
cana-6113	95	13	a	a	DET
cana-6113	95	14	robust	robust	ADJ
cana-6113	95	15	feature	feature	NOUN
cana-6113	95	16	representation	representation	NOUN
cana-6113	95	17	of	of	ADP
cana-6113	95	18	the	the	DET
cana-6113	95	19	clinical	clinical	ADJ
cana-6113	95	20	domain	domain	NOUN
cana-6113	95	21	.	.	PUNCT
cana-6113	96	1	b	b	X
cana-6113	96	2	)	)	PUNCT
cana-6113	96	3	colposcopy	colposcopy	NOUN
cana-6113	96	4	image	image	NOUN
cana-6113	96	5	processing	processing	NOUN
cana-6113	96	6	:	:	PUNCT
cana-6113	96	7	colposcopy	colposcopy	NOUN
cana-6113	96	8	images	image	NOUN
cana-6113	96	9	,	,	PUNCT
cana-6113	96	10	being	be	AUX
cana-6113	96	11	high	high	ADV
cana-6113	96	12	-	-	PUNCT
cana-6113	96	13	dimensional	dimensional	ADJ
cana-6113	96	14	and	and	CCONJ
cana-6113	96	15	rich	rich	ADJ
cana-6113	96	16	in	in	ADP
cana-6113	96	17	spatial	spatial	ADJ
cana-6113	96	18	detail	detail	NOUN
cana-6113	96	19	,	,	PUNCT
cana-6113	96	20	require	require	VERB
cana-6113	96	21	a	a	DET
cana-6113	96	22	deep	deep	ADJ
cana-6113	96	23	learning	learning	NOUN
cana-6113	96	24	-	-	PUNCT
cana-6113	96	25	based	base	VERB
cana-6113	96	26	approach	approach	NOUN
cana-6113	96	27	for	for	ADP
cana-6113	96	28	feature	feature	NOUN
cana-6113	96	29	extraction	extraction	NOUN
cana-6113	96	30	.	.	PUNCT
cana-6113	97	1	a	a	DET
cana-6113	97	2	convolutional	convolutional	ADJ
cana-6113	97	3	neural	neural	ADJ
cana-6113	97	4	network	network	NOUN
cana-6113	97	5	(	(	PUNCT
cana-6113	97	6	cnn	cnn	PROPN
cana-6113	97	7	)	)	PUNCT
cana-6113	97	8	architecture	architecture	NOUN
cana-6113	97	9	is	be	AUX
cana-6113	97	10	utilized	utilize	VERB
cana-6113	97	11	to	to	PART
cana-6113	97	12	hierarchically	hierarchically	ADV
cana-6113	97	13	extract	extract	VERB
cana-6113	97	14	features	feature	NOUN
cana-6113	97	15	at	at	ADP
cana-6113	97	16	multiple	multiple	ADJ
cana-6113	97	17	abstraction	abstraction	NOUN
cana-6113	97	18	levels	level	NOUN
cana-6113	97	19	.	.	PUNCT
cana-6113	98	1	in	in	ADP
cana-6113	98	2	the	the	DET
cana-6113	98	3	initial	initial	ADJ
cana-6113	98	4	layers	layer	NOUN
cana-6113	98	5	,	,	PUNCT
cana-6113	98	6	the	the	DET
cana-6113	98	7	cnn	cnn	PROPN
cana-6113	98	8	captures	capture	VERB
cana-6113	98	9	low	low	ADJ
cana-6113	98	10	-	-	PUNCT
cana-6113	98	11	level	level	NOUN
cana-6113	98	12	features	feature	NOUN
cana-6113	98	13	such	such	ADJ
cana-6113	98	14	as	as	ADP
cana-6113	98	15	edges	edge	NOUN
cana-6113	98	16	,	,	PUNCT
cana-6113	98	17	color	color	NOUN
cana-6113	98	18	gradients	gradient	NOUN
cana-6113	98	19	,	,	PUNCT
cana-6113	98	20	and	and	CCONJ
cana-6113	98	21	textures	texture	NOUN
cana-6113	98	22	,	,	PUNCT
cana-6113	98	23	while	while	SCONJ
cana-6113	98	24	deeper	deep	ADJ
cana-6113	98	25	layers	layer	NOUN
cana-6113	98	26	focus	focus	VERB
cana-6113	98	27	on	on	ADP
cana-6113	98	28	high	high	ADJ
cana-6113	98	29	-	-	PUNCT
cana-6113	98	30	level	level	NOUN
cana-6113	98	31	features	feature	NOUN
cana-6113	98	32	such	such	ADJ
cana-6113	98	33	as	as	ADP
cana-6113	98	34	lesion	lesion	NOUN
cana-6113	98	35	shape	shape	NOUN
cana-6113	98	36	,	,	PUNCT
cana-6113	98	37	vascular	vascular	ADJ
cana-6113	98	38	patterns	pattern	NOUN
cana-6113	98	39	,	,	PUNCT
cana-6113	98	40	and	and	CCONJ
cana-6113	98	41	surface	surface	NOUN
cana-6113	98	42	irregularities	irregularity	NOUN
cana-6113	98	43	.	.	PUNCT
cana-6113	99	1	image	image	NOUN
cana-6113	99	2	preprocessing	preprocesse	VERB
cana-6113	99	3	techniques	technique	NOUN
cana-6113	99	4	like	like	ADP
cana-6113	99	5	normalization	normalization	NOUN
cana-6113	99	6	,	,	PUNCT
cana-6113	99	7	contrast	contrast	NOUN
cana-6113	99	8	enhancement	enhancement	NOUN
cana-6113	99	9	,	,	PUNCT
cana-6113	99	10	and	and	CCONJ
cana-6113	99	11	region	region	NOUN
cana-6113	99	12	-	-	PUNCT
cana-6113	99	13	of	of	ADP
cana-6113	99	14	-	-	PUNCT
cana-6113	99	15	interest	interest	NOUN
cana-6113	99	16	(	(	PUNCT
cana-6113	99	17	roi	roi	NOUN
cana-6113	99	18	)	)	PUNCT
cana-6113	99	19	segmentation	segmentation	NOUN
cana-6113	99	20	are	be	AUX
cana-6113	99	21	applied	apply	VERB
cana-6113	99	22	beforehand	beforehand	ADV
cana-6113	99	23	to	to	PART
cana-6113	99	24	improve	improve	VERB
cana-6113	99	25	clarity	clarity	NOUN
cana-6113	99	26	and	and	CCONJ
cana-6113	99	27	highlight	highlight	VERB
cana-6113	99	28	diagnostically	diagnostically	ADV
cana-6113	99	29	relevant	relevant	ADJ
cana-6113	99	30	regions	region	NOUN
cana-6113	99	31	.	.	PUNCT
cana-6113	100	1	the	the	DET
cana-6113	100	2	resulting	result	VERB
cana-6113	100	3	feature	feature	NOUN
cana-6113	100	4	maps	map	NOUN
cana-6113	100	5	are	be	AUX
cana-6113	100	6	then	then	ADV
cana-6113	100	7	flattened	flatten	VERB
cana-6113	100	8	or	or	CCONJ
cana-6113	100	9	pooled	pool	VERB
cana-6113	100	10	into	into	ADP
cana-6113	100	11	compact	compact	ADJ
cana-6113	100	12	feature	feature	NOUN
cana-6113	100	13	vectors	vector	NOUN
cana-6113	100	14	,	,	PUNCT
cana-6113	100	15	representing	represent	VERB
cana-6113	100	16	the	the	DET
cana-6113	100	17	visual	visual	ADJ
cana-6113	100	18	information	information	NOUN
cana-6113	100	19	extracted	extract	VERB
cana-6113	100	20	from	from	ADP
cana-6113	100	21	colposcopic	colposcopic	NOUN
cana-6113	100	22	images	image	NOUN
cana-6113	100	23	.	.	PUNCT
cana-6113	101	1	4.2	4.2	NUM
cana-6113	101	2	.	.	PUNCT
cana-6113	101	3	fusion	fusion	NOUN
cana-6113	101	4	of	of	ADP
cana-6113	101	5	image	image	NOUN
cana-6113	101	6	and	and	CCONJ
cana-6113	101	7	clinical	clinical	ADJ
cana-6113	101	8	features	feature	NOUN
cana-6113	101	9	once	once	SCONJ
cana-6113	101	10	both	both	DET
cana-6113	101	11	data	datum	NOUN
cana-6113	101	12	streams	stream	NOUN
cana-6113	101	13	have	have	AUX
cana-6113	101	14	been	be	AUX
cana-6113	101	15	independently	independently	ADV
cana-6113	101	16	processed	process	VERB
cana-6113	101	17	,	,	PUNCT
cana-6113	101	18	their	their	PRON
cana-6113	101	19	extracted	extract	VERB
cana-6113	101	20	features	feature	NOUN
cana-6113	101	21	are	be	AUX
cana-6113	101	22	fused	fuse	VERB
cana-6113	101	23	to	to	PART
cana-6113	101	24	create	create	VERB
cana-6113	101	25	a	a	DET
cana-6113	101	26	unified	unified	ADJ
cana-6113	101	27	feature	feature	NOUN
cana-6113	101	28	representation	representation	NOUN
cana-6113	101	29	.	.	PUNCT
cana-6113	102	1	this	this	DET
cana-6113	102	2	feature	feature	NOUN
cana-6113	102	3	-	-	PUNCT
cana-6113	102	4	level	level	NOUN
cana-6113	102	5	fusion	fusion	NOUN
cana-6113	102	6	combines	combine	VERB
cana-6113	102	7	the	the	DET
cana-6113	102	8	complementary	complementary	ADJ
cana-6113	102	9	strengths	strength	NOUN
cana-6113	102	10	of	of	ADP
cana-6113	102	11	numerical	numerical	ADJ
cana-6113	102	12	clinical	clinical	ADJ
cana-6113	102	13	data	datum	NOUN
cana-6113	102	14	and	and	CCONJ
cana-6113	102	15	visual	visual	ADJ
cana-6113	102	16	image	image	NOUN
cana-6113	102	17	features	feature	NOUN
cana-6113	102	18	,	,	PUNCT
cana-6113	102	19	ensuring	ensure	VERB
cana-6113	102	20	that	that	SCONJ
cana-6113	102	21	both	both	CCONJ
cana-6113	102	22	physiological	physiological	ADJ
cana-6113	102	23	and	and	CCONJ
cana-6113	102	24	morphological	morphological	ADJ
cana-6113	102	25	aspects	aspect	NOUN
cana-6113	102	26	contribute	contribute	VERB
cana-6113	102	27	to	to	ADP
cana-6113	102	28	the	the	DET
cana-6113	102	29	diagnostic	diagnostic	ADJ
cana-6113	102	30	decision	decision	NOUN
cana-6113	102	31	.	.	PUNCT
cana-6113	103	1	the	the	DET
cana-6113	103	2	concatenation	concatenation	NOUN
cana-6113	103	3	of	of	ADP
cana-6113	103	4	the	the	DET
cana-6113	103	5	clinical	clinical	ADJ
cana-6113	103	6	feature	feature	NOUN
cana-6113	103	7	vector	vector	NOUN
cana-6113	103	8	and	and	CCONJ
cana-6113	103	9	image	image	NOUN
cana-6113	103	10	feature	feature	NOUN
cana-6113	103	11	vector	vector	NOUN
cana-6113	103	12	produces	produce	VERB
cana-6113	103	13	a	a	DET
cana-6113	103	14	single	single	ADJ
cana-6113	103	15	multidimensional	multidimensional	ADJ
cana-6113	103	16	vector	vector	NOUN
cana-6113	103	17	,	,	PUNCT
cana-6113	103	18	which	which	PRON
cana-6113	103	19	is	be	AUX
cana-6113	103	20	subsequently	subsequently	ADV
cana-6113	103	21	passed	pass	VERB
cana-6113	103	22	into	into	ADP
cana-6113	103	23	a	a	DET
cana-6113	103	24	fusion	fusion	NOUN
cana-6113	103	25	-	-	PUNCT
cana-6113	103	26	based	base	VERB
cana-6113	103	27	classifier	classifier	NOUN
cana-6113	103	28	typically	typically	ADV
cana-6113	103	29	a	a	DET
cana-6113	103	30	random	random	ADJ
cana-6113	103	31	forest	forest	NOUN
cana-6113	103	32	(	(	PUNCT
cana-6113	103	33	rf	rf	NOUN
cana-6113	103	34	)	)	PUNCT
cana-6113	103	35	model	model	NOUN
cana-6113	103	36	or	or	CCONJ
cana-6113	103	37	a	a	DET
cana-6113	103	38	fully	fully	ADV
cana-6113	103	39	connected	connected	ADJ
cana-6113	103	40	neural	neural	ADJ
cana-6113	103	41	network	network	NOUN
cana-6113	103	42	to	to	PART
cana-6113	103	43	generate	generate	VERB
cana-6113	103	44	the	the	DET
cana-6113	103	45	final	final	ADJ
cana-6113	103	46	prediction	prediction	NOUN
cana-6113	103	47	output	output	NOUN
cana-6113	103	48	.	.	PUNCT
cana-6113	104	1	this	this	DET
cana-6113	104	2	fusion	fusion	NOUN
cana-6113	104	3	strategy	strategy	NOUN
cana-6113	104	4	enhances	enhance	VERB
cana-6113	104	5	the	the	DET
cana-6113	104	6	predictive	predictive	ADJ
cana-6113	104	7	power	power	NOUN
cana-6113	104	8	of	of	ADP
cana-6113	104	9	the	the	DET
cana-6113	104	10	model	model	NOUN
cana-6113	104	11	by	by	ADP
cana-6113	104	12	integrating	integrate	VERB
cana-6113	104	13	heterogeneous	heterogeneous	ADJ
cana-6113	104	14	data	datum	NOUN
cana-6113	104	15	sources	source	NOUN
cana-6113	104	16	,	,	PUNCT
cana-6113	104	17	thereby	thereby	ADV
cana-6113	104	18	improving	improve	VERB
cana-6113	104	19	diagnostic	diagnostic	ADJ
cana-6113	104	20	accuracy	accuracy	NOUN
cana-6113	104	21	and	and	CCONJ
cana-6113	104	22	robustness	robustness	NOUN
cana-6113	104	23	.	.	PUNCT
cana-6113	105	1	by	by	ADP
cana-6113	105	2	jointly	jointly	ADV
cana-6113	105	3	leveraging	leverage	VERB
cana-6113	105	4	structured	structured	ADJ
cana-6113	105	5	clinical	clinical	ADJ
cana-6113	105	6	parameters	parameter	NOUN
cana-6113	105	7	and	and	CCONJ
cana-6113	105	8	unstructured	unstructured	ADJ
cana-6113	105	9	visual	visual	ADJ
cana-6113	105	10	cues	cue	NOUN
cana-6113	105	11	from	from	ADP
cana-6113	105	12	colposcopy	colposcopy	NOUN
cana-6113	105	13	images	image	NOUN
cana-6113	105	14	,	,	PUNCT
cana-6113	105	15	the	the	DET
cana-6113	105	16	proposed	propose	VERB
cana-6113	105	17	system	system	NOUN
cana-6113	105	18	achieves	achieve	VERB
cana-6113	105	19	a	a	DET
cana-6113	105	20	more	more	ADV
cana-6113	105	21	comprehensive	comprehensive	ADJ
cana-6113	105	22	understanding	understanding	NOUN
cana-6113	105	23	of	of	ADP
cana-6113	105	24	cervical	cervical	ADJ
cana-6113	105	25	health	health	NOUN
cana-6113	105	26	conditions	condition	NOUN
cana-6113	105	27	.	.	PUNCT
cana-6113	106	1	such	such	DET
cana-6113	106	2	a	a	DET
cana-6113	106	3	hybrid	hybrid	ADJ
cana-6113	106	4	architecture	architecture	NOUN
cana-6113	106	5	not	not	PART
cana-6113	106	6	only	only	ADV
cana-6113	106	7	aids	aid	NOUN
cana-6113	106	8	in	in	ADP
cana-6113	106	9	early	early	ADJ
cana-6113	106	10	and	and	CCONJ
cana-6113	106	11	reliable	reliable	ADJ
cana-6113	106	12	disease	disease	NOUN
cana-6113	106	13	detection	detection	NOUN
cana-6113	106	14	but	but	CCONJ
cana-6113	106	15	also	also	ADV
cana-6113	106	16	contributes	contribute	VERB
cana-6113	106	17	to	to	ADP
cana-6113	106	18	explainable	explainable	ADJ
cana-6113	106	19	and	and	CCONJ
cana-6113	106	20	data	datum	NOUN
cana-6113	106	21	-	-	PUNCT
cana-6113	106	22	driven	drive	VERB
cana-6113	106	23	clinical	clinical	ADJ
cana-6113	106	24	decision	decision	NOUN
cana-6113	106	25	-	-	PUNCT
cana-6113	106	26	making	making	NOUN
cana-6113	106	27	.	.	PUNCT
cana-6113	107	1	figure	figure	NOUN
cana-6113	107	2	2	2	NUM
cana-6113	107	3	.	.	PUNCT
cana-6113	108	1	representation	representation	NOUN
cana-6113	108	2	of	of	ADP
cana-6113	108	3	input	input	NOUN
cana-6113	108	4	image	image	NOUN
cana-6113	108	5	and	and	CCONJ
cana-6113	108	6	proposed	propose	VERB
cana-6113	108	7	enhanced	enhanced	ADJ
cana-6113	108	8	image	image	NOUN
cana-6113	108	9	communications	communication	NOUN
cana-6113	108	10	on	on	ADP
cana-6113	108	11	applied	apply	VERB
cana-6113	108	12	nonlinear	nonlinear	ADJ
cana-6113	108	13	analysis	analysis	NOUN
cana-6113	108	14	issn	issn	NOUN
cana-6113	108	15	:	:	PUNCT
cana-6113	108	16	1074	1074	NUM
cana-6113	108	17	-	-	PUNCT
cana-6113	108	18	133x	133x	NUM
cana-6113	108	19	vol	vol	VERB
cana-6113	108	20	32	32	NUM
cana-6113	108	21	no	no	NOUN
cana-6113	108	22	.	.	PUNCT
cana-6113	109	1	10s	10	NOUN
cana-6113	109	2	(	(	PUNCT
cana-6113	109	3	2025	2025	NUM
cana-6113	109	4	)	)	PUNCT
cana-6113	109	5	3409	3409	NUM
cana-6113	109	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	109	7	figure	figure	NOUN
cana-6113	109	8	2	2	NUM
cana-6113	109	9	illustrates	illustrate	VERB
cana-6113	109	10	the	the	DET
cana-6113	109	11	comparison	comparison	NOUN
cana-6113	109	12	between	between	ADP
cana-6113	109	13	the	the	DET
cana-6113	109	14	original	original	ADJ
cana-6113	109	15	input	input	NOUN
cana-6113	109	16	colposcopy	colposcopy	NOUN
cana-6113	109	17	image	image	NOUN
cana-6113	109	18	and	and	CCONJ
cana-6113	109	19	the	the	DET
cana-6113	109	20	proposed	propose	VERB
cana-6113	109	21	enhanced	enhance	VERB
cana-6113	109	22	image	image	NOUN
cana-6113	109	23	after	after	ADP
cana-6113	109	24	preprocessing	preprocesse	VERB
cana-6113	109	25	.	.	PUNCT
cana-6113	110	1	the	the	DET
cana-6113	110	2	enhancement	enhancement	NOUN
cana-6113	110	3	process	process	NOUN
cana-6113	110	4	includes	include	VERB
cana-6113	110	5	contrast	contrast	NOUN
cana-6113	110	6	adjustment	adjustment	NOUN
cana-6113	110	7	,	,	PUNCT
cana-6113	110	8	noise	noise	NOUN
cana-6113	110	9	reduction	reduction	NOUN
cana-6113	110	10	,	,	PUNCT
cana-6113	110	11	and	and	CCONJ
cana-6113	110	12	region	region	NOUN
cana-6113	110	13	-	-	PUNCT
cana-6113	110	14	of	of	ADP
cana-6113	110	15	-	-	PUNCT
cana-6113	110	16	interest	interest	NOUN
cana-6113	110	17	(	(	PUNCT
cana-6113	110	18	roi	roi	NOUN
cana-6113	110	19	)	)	PUNCT
cana-6113	110	20	highlighting	highlighting	NOUN
cana-6113	110	21	,	,	PUNCT
cana-6113	110	22	which	which	PRON
cana-6113	110	23	collectively	collectively	ADV
cana-6113	110	24	improve	improve	VERB
cana-6113	110	25	the	the	DET
cana-6113	110	26	visibility	visibility	NOUN
cana-6113	110	27	of	of	ADP
cana-6113	110	28	lesion	lesion	NOUN
cana-6113	110	29	boundaries	boundary	NOUN
cana-6113	110	30	and	and	CCONJ
cana-6113	110	31	fine	fine	ADJ
cana-6113	110	32	tissue	tissue	NOUN
cana-6113	110	33	structures	structure	NOUN
cana-6113	110	34	.	.	PUNCT
cana-6113	111	1	by	by	ADP
cana-6113	111	2	enhancing	enhance	VERB
cana-6113	111	3	these	these	DET
cana-6113	111	4	diagnostically	diagnostically	ADV
cana-6113	111	5	relevant	relevant	ADJ
cana-6113	111	6	features	feature	NOUN
cana-6113	111	7	,	,	PUNCT
cana-6113	111	8	the	the	DET
cana-6113	111	9	proposed	propose	VERB
cana-6113	111	10	preprocessing	preprocessing	NOUN
cana-6113	111	11	ensures	ensure	VERB
cana-6113	111	12	that	that	SCONJ
cana-6113	111	13	subsequent	subsequent	ADJ
cana-6113	111	14	feature	feature	NOUN
cana-6113	111	15	extraction	extraction	NOUN
cana-6113	111	16	and	and	CCONJ
cana-6113	111	17	classification	classification	NOUN
cana-6113	111	18	steps	step	NOUN
cana-6113	111	19	can	can	AUX
cana-6113	111	20	more	more	ADV
cana-6113	111	21	accurately	accurately	ADV
cana-6113	111	22	capture	capture	VERB
cana-6113	111	23	the	the	DET
cana-6113	111	24	subtle	subtle	ADJ
cana-6113	111	25	patterns	pattern	NOUN
cana-6113	111	26	indicative	indicative	ADJ
cana-6113	111	27	of	of	ADP
cana-6113	111	28	cervical	cervical	ADJ
cana-6113	111	29	abnormalities	abnormality	NOUN
cana-6113	111	30	,	,	PUNCT
cana-6113	111	31	ultimately	ultimately	ADV
cana-6113	111	32	improving	improve	VERB
cana-6113	111	33	the	the	DET
cana-6113	111	34	performance	performance	NOUN
cana-6113	111	35	of	of	ADP
cana-6113	111	36	the	the	DET
cana-6113	111	37	predictive	predictive	ADJ
cana-6113	111	38	model	model	NOUN
cana-6113	111	39	.	.	PUNCT
cana-6113	112	1	4	4	X
cana-6113	112	2	.	.	X
cana-6113	112	3	performace	performace	NOUN
cana-6113	112	4	evaluation	evaluation	NOUN
cana-6113	112	5	precision	precision	NOUN
cana-6113	112	6	,	,	PUNCT
cana-6113	112	7	recall	recall	NOUN
cana-6113	112	8	,	,	PUNCT
cana-6113	112	9	accuracy	accuracy	NOUN
cana-6113	112	10	,	,	PUNCT
cana-6113	112	11	and	and	CCONJ
cana-6113	112	12	f1	f1	NOUN
cana-6113	112	13	score	score	NOUN
cana-6113	112	14	are	be	AUX
cana-6113	112	15	widely	widely	ADV
cana-6113	112	16	used	use	VERB
cana-6113	112	17	evaluation	evaluation	NOUN
cana-6113	112	18	metrics	metric	NOUN
cana-6113	112	19	in	in	ADP
cana-6113	112	20	classification	classification	NOUN
cana-6113	112	21	tasks	task	NOUN
cana-6113	112	22	.	.	PUNCT
cana-6113	113	1	each	each	DET
cana-6113	113	2	metric	metric	NOUN
cana-6113	113	3	provides	provide	VERB
cana-6113	113	4	a	a	DET
cana-6113	113	5	different	different	ADJ
cana-6113	113	6	aspect	aspect	NOUN
cana-6113	113	7	of	of	ADP
cana-6113	113	8	model	model	NOUN
cana-6113	113	9	performance	performance	NOUN
cana-6113	113	10	.	.	PUNCT
cana-6113	114	1	these	these	DET
cana-6113	114	2	metrics	metric	NOUN
cana-6113	114	3	are	be	AUX
cana-6113	114	4	valuable	valuable	ADJ
cana-6113	114	5	in	in	ADP
cana-6113	114	6	evaluating	evaluate	VERB
cana-6113	114	7	the	the	DET
cana-6113	114	8	performance	performance	NOUN
cana-6113	114	9	of	of	ADP
cana-6113	114	10	a	a	DET
cana-6113	114	11	classification	classification	NOUN
cana-6113	114	12	model	model	NOUN
cana-6113	114	13	and	and	CCONJ
cana-6113	114	14	can	can	AUX
cana-6113	114	15	provide	provide	VERB
cana-6113	114	16	insights	insight	NOUN
cana-6113	114	17	into	into	ADP
cana-6113	114	18	its	its	PRON
cana-6113	114	19	effectiveness	effectiveness	NOUN
cana-6113	114	20	in	in	ADP
cana-6113	114	21	correctly	correctly	ADV
cana-6113	114	22	predicting	predict	VERB
cana-6113	114	23	positive	positive	ADJ
cana-6113	114	24	and	and	CCONJ
cana-6113	114	25	negative	negative	ADJ
cana-6113	114	26	instances	instance	NOUN
cana-6113	114	27	[	[	X
cana-6113	114	28	12	12	NUM
cana-6113	114	29	-	-	SYM
cana-6113	114	30	13	13	NUM
cana-6113	114	31	]	]	PUNCT
cana-6113	114	32	as	as	SCONJ
cana-6113	114	33	depicted	depict	VERB
cana-6113	114	34	in	in	ADP
cana-6113	114	35	table	table	NOUN
cana-6113	115	1	2	2	NUM
cana-6113	115	2	.	.	NOUN
cana-6113	115	3	•	•	NUM
cana-6113	115	4	accuracy	accuracy	NOUN
cana-6113	115	5	measures	measure	VERB
cana-6113	115	6	the	the	DET
cana-6113	115	7	overall	overall	ADJ
cana-6113	115	8	correctness	correctness	NOUN
cana-6113	115	9	of	of	ADP
cana-6113	115	10	the	the	DET
cana-6113	115	11	model	model	NOUN
cana-6113	115	12	by	by	ADP
cana-6113	115	13	calculating	calculate	VERB
cana-6113	115	14	the	the	DET
cana-6113	115	15	ratio	ratio	NOUN
cana-6113	115	16	of	of	ADP
cana-6113	115	17	correctly	correctly	ADV
cana-6113	115	18	predicted	predict	VERB
cana-6113	115	19	samples	sample	NOUN
cana-6113	115	20	(	(	PUNCT
cana-6113	115	21	both	both	CCONJ
cana-6113	115	22	positive	positive	ADJ
cana-6113	115	23	and	and	CCONJ
cana-6113	115	24	negative	negative	ADJ
cana-6113	115	25	)	)	PUNCT
cana-6113	115	26	to	to	ADP
cana-6113	115	27	the	the	DET
cana-6113	115	28	total	total	ADJ
cana-6113	115	29	number	number	NOUN
cana-6113	115	30	of	of	ADP
cana-6113	115	31	samples	sample	NOUN
cana-6113	115	32	.	.	PUNCT
cana-6113	116	1	•	•	NUM
cana-6113	116	2	precision	precision	NOUN
cana-6113	116	3	quantifies	quantify	VERB
cana-6113	116	4	the	the	DET
cana-6113	116	5	proportion	proportion	NOUN
cana-6113	116	6	of	of	ADP
cana-6113	116	7	positive	positive	ADJ
cana-6113	116	8	predictions	prediction	NOUN
cana-6113	116	9	that	that	PRON
cana-6113	116	10	are	be	AUX
cana-6113	116	11	actually	actually	ADV
cana-6113	116	12	correct	correct	ADJ
cana-6113	116	13	,	,	PUNCT
cana-6113	116	14	highlighting	highlight	VERB
cana-6113	116	15	how	how	SCONJ
cana-6113	116	16	reliable	reliable	ADJ
cana-6113	116	17	the	the	DET
cana-6113	116	18	model	model	NOUN
cana-6113	116	19	’s	’s	PART
cana-6113	116	20	positive	positive	ADJ
cana-6113	116	21	predictions	prediction	NOUN
cana-6113	116	22	are	be	AUX
cana-6113	116	23	.	.	PUNCT
cana-6113	117	1	•	•	NUM
cana-6113	117	2	f1	f1	NOUN
cana-6113	117	3	-	-	PUNCT
cana-6113	117	4	score	score	NOUN
cana-6113	117	5	provides	provide	VERB
cana-6113	117	6	a	a	DET
cana-6113	117	7	harmonic	harmonic	ADJ
cana-6113	117	8	mean	mean	NOUN
cana-6113	117	9	of	of	ADP
cana-6113	117	10	precision	precision	NOUN
cana-6113	117	11	and	and	CCONJ
cana-6113	117	12	recall	recall	NOUN
cana-6113	117	13	,	,	PUNCT
cana-6113	117	14	offering	offer	VERB
cana-6113	117	15	a	a	DET
cana-6113	117	16	single	single	ADJ
cana-6113	117	17	measure	measure	NOUN
cana-6113	117	18	that	that	PRON
cana-6113	117	19	balances	balance	VERB
cana-6113	117	20	both	both	DET
cana-6113	117	21	false	false	ADJ
cana-6113	117	22	positives	positive	NOUN
cana-6113	117	23	and	and	CCONJ
cana-6113	117	24	false	false	ADJ
cana-6113	117	25	negatives	negative	NOUN
cana-6113	117	26	.	.	PUNCT
cana-6113	118	1	these	these	DET
cana-6113	118	2	metrics	metric	NOUN
cana-6113	118	3	provide	provide	VERB
cana-6113	118	4	a	a	DET
cana-6113	118	5	comprehensive	comprehensive	ADJ
cana-6113	118	6	understanding	understanding	NOUN
cana-6113	118	7	of	of	ADP
cana-6113	118	8	the	the	DET
cana-6113	118	9	model	model	NOUN
cana-6113	118	10	’s	’s	PART
cana-6113	118	11	predictive	predictive	ADJ
cana-6113	118	12	capability	capability	NOUN
cana-6113	118	13	and	and	CCONJ
cana-6113	118	14	its	its	PRON
cana-6113	118	15	ability	ability	NOUN
cana-6113	118	16	to	to	PART
cana-6113	118	17	correctly	correctly	ADV
cana-6113	118	18	identify	identify	VERB
cana-6113	118	19	cervical	cervical	ADJ
cana-6113	118	20	abnormalities	abnormality	NOUN
cana-6113	118	21	.	.	PUNCT
cana-6113	119	1	specifically	specifically	ADV
cana-6113	119	2	,	,	PUNCT
cana-6113	119	3	the	the	DET
cana-6113	119	4	metrics	metric	NOUN
cana-6113	119	5	calculated	calculate	VERB
cana-6113	119	6	include	include	VERB
cana-6113	119	7	accuracy	accuracy	NOUN
cana-6113	119	8	,	,	PUNCT
cana-6113	119	9	specificity	specificity	NOUN
cana-6113	119	10	,	,	PUNCT
cana-6113	119	11	sensitivity	sensitivity	NOUN
cana-6113	119	12	,	,	PUNCT
cana-6113	119	13	precision	precision	NOUN
cana-6113	119	14	,	,	PUNCT
cana-6113	119	15	recall	recall	NOUN
cana-6113	119	16	,	,	PUNCT
cana-6113	119	17	and	and	CCONJ
cana-6113	119	18	f1	f1	NOUN
cana-6113	119	19	-	-	PUNCT
cana-6113	119	20	score	score	NOUN
cana-6113	119	21	.	.	PUNCT
cana-6113	120	1	each	each	PRON
cana-6113	120	2	of	of	ADP
cana-6113	120	3	these	these	DET
cana-6113	120	4	metrics	metric	NOUN
cana-6113	120	5	quantifies	quantifie	NOUN
cana-6113	120	6	a	a	DET
cana-6113	120	7	different	different	ADJ
cana-6113	120	8	aspect	aspect	NOUN
cana-6113	120	9	of	of	ADP
cana-6113	120	10	the	the	DET
cana-6113	120	11	model	model	NOUN
cana-6113	120	12	’s	’s	PART
cana-6113	120	13	performance	performance	NOUN
cana-6113	120	14	and	and	CCONJ
cana-6113	120	15	is	be	AUX
cana-6113	120	16	defined	define	VERB
cana-6113	120	17	based	base	VERB
cana-6113	120	18	on	on	ADP
cana-6113	120	19	four	four	NUM
cana-6113	120	20	key	key	ADJ
cana-6113	120	21	components	component	NOUN
cana-6113	120	22	of	of	ADP
cana-6113	120	23	the	the	DET
cana-6113	120	24	confusion	confusion	NOUN
cana-6113	120	25	matrix	matrix	NOUN
cana-6113	120	26	:	:	PUNCT
cana-6113	120	27	true	true	ADJ
cana-6113	120	28	positive	positive	ADJ
cana-6113	120	29	(	(	PUNCT
cana-6113	120	30	tp	tp	NOUN
cana-6113	120	31	)	)	PUNCT
cana-6113	120	32	,	,	PUNCT
cana-6113	120	33	true	true	ADJ
cana-6113	120	34	negative	negative	ADJ
cana-6113	120	35	(	(	PUNCT
cana-6113	120	36	tn	tn	NOUN
cana-6113	120	37	)	)	PUNCT
cana-6113	120	38	,	,	PUNCT
cana-6113	120	39	false	false	ADJ
cana-6113	120	40	positive	positive	ADJ
cana-6113	120	41	(	(	PUNCT
cana-6113	120	42	fp	fp	NOUN
cana-6113	120	43	)	)	PUNCT
cana-6113	120	44	,	,	PUNCT
cana-6113	120	45	and	and	CCONJ
cana-6113	120	46	false	false	ADJ
cana-6113	120	47	negative	negative	ADJ
cana-6113	120	48	(	(	PUNCT
cana-6113	120	49	fn	fn	NOUN
cana-6113	120	50	)	)	PUNCT
cana-6113	120	51	.	.	PUNCT
cana-6113	121	1	table	table	NOUN
cana-6113	121	2	2	2	NUM
cana-6113	121	3	.	.	PUNCT
cana-6113	121	4	performance	performance	NOUN
cana-6113	121	5	evaluation	evaluation	NOUN
cana-6113	121	6	metrics	metric	NOUN
cana-6113	121	7	metric	metric	ADJ
cana-6113	121	8	definition	definition	NOUN
cana-6113	121	9	formulas	formula	NOUN
cana-6113	121	10	precision	precision	VERB
cana-6113	121	11	positive	positive	ADJ
cana-6113	121	12	predictive	predictive	ADJ
cana-6113	121	13	value	value	NOUN
cana-6113	121	14	precision	precision	NOUN
cana-6113	121	15	=	=	PUNCT
cana-6113	121	16	tp	tp	NOUN
cana-6113	121	17	/	/	PUNCT
cana-6113	121	18	(	(	PUNCT
cana-6113	121	19	tp	tp	AUX
cana-6113	121	20	+	+	CCONJ
cana-6113	121	21	fp	fp	X
cana-6113	121	22	)	)	PUNCT
cana-6113	121	23	recall	recall	VERB
cana-6113	121	24	true	true	ADJ
cana-6113	121	25	positive	positive	ADJ
cana-6113	121	26	rate	rate	NOUN
cana-6113	121	27	recall	recall	NOUN
cana-6113	121	28	=	=	VERB
cana-6113	121	29	tp	tp	NOUN
cana-6113	121	30	/	/	PUNCT
cana-6113	121	31	(	(	PUNCT
cana-6113	121	32	tp	tp	ADP
cana-6113	121	33	+	+	CCONJ
cana-6113	121	34	fn	fn	NOUN
cana-6113	121	35	)	)	PUNCT
cana-6113	121	36	accuracy	accuracy	NOUN
cana-6113	121	37	overall	overall	ADJ
cana-6113	121	38	accuracy	accuracy	NOUN
cana-6113	121	39	accuracy	accuracy	NOUN
cana-6113	122	1	=	=	SYM
cana-6113	122	2	(	(	PUNCT
cana-6113	122	3	tp	tp	X
cana-6113	122	4	+	+	CCONJ
cana-6113	122	5	tn	tn	NOUN
cana-6113	122	6	)	)	PUNCT
cana-6113	122	7	/	/	PUNCT
cana-6113	122	8	(	(	PUNCT
cana-6113	122	9	tp	tp	ADP
cana-6113	122	10	+	+	CCONJ
cana-6113	122	11	tn	tn	NOUN
cana-6113	122	12	+	+	CCONJ
cana-6113	122	13	fp	fp	PROPN
cana-6113	122	14	+	+	NUM
cana-6113	122	15	fn	fn	NOUN
cana-6113	122	16	)	)	PUNCT
cana-6113	122	17	f1	f1	NOUN
cana-6113	122	18	score	score	NOUN
cana-6113	122	19	harmonic	harmonic	NOUN
cana-6113	122	20	mean	mean	NOUN
cana-6113	122	21	of	of	ADP
cana-6113	122	22	precision	precision	NOUN
cana-6113	122	23	and	and	CCONJ
cana-6113	122	24	recall	recall	NOUN
cana-6113	122	25	f1	f1	NOUN
cana-6113	122	26	score	score	NOUN
cana-6113	122	27	=	=	SYM
cana-6113	122	28	2	2	NUM
cana-6113	122	29	*	*	PUNCT
cana-6113	122	30	(	(	PUNCT
cana-6113	122	31	precision	precision	NOUN
cana-6113	122	32	*	*	PUNCT
cana-6113	122	33	recall	recall	PROPN
cana-6113	122	34	)	)	PUNCT
cana-6113	122	35	/	/	PUNCT
cana-6113	122	36	(	(	PUNCT
cana-6113	122	37	precision	precision	NOUN
cana-6113	122	38	+	+	CCONJ
cana-6113	122	39	recall	recall	NOUN
cana-6113	122	40	)	)	PUNCT
cana-6113	122	41	5	5	NUM
cana-6113	122	42	.	.	X
cana-6113	122	43	result	result	NOUN
cana-6113	122	44	and	and	CCONJ
cana-6113	122	45	analysis	analysis	NOUN
cana-6113	122	46	the	the	DET
cana-6113	122	47	performance	performance	NOUN
cana-6113	122	48	evaluation	evaluation	NOUN
cana-6113	122	49	table	table	NOUN
cana-6113	122	50	3	3	NUM
cana-6113	122	51	and	and	CCONJ
cana-6113	122	52	figure	figure	VERB
cana-6113	122	53	4	4	NUM
cana-6113	122	54	presents	present	VERB
cana-6113	122	55	a	a	DET
cana-6113	122	56	comparative	comparative	ADJ
cana-6113	122	57	analysis	analysis	NOUN
cana-6113	122	58	of	of	ADP
cana-6113	122	59	four	four	NUM
cana-6113	122	60	models	model	NOUN
cana-6113	122	61	—	—	PUNCT
cana-6113	122	62	alexnet	alexnet	ADJ
cana-6113	122	63	,	,	PUNCT
cana-6113	122	64	vgg16	vgg16	PROPN
cana-6113	122	65	,	,	PUNCT
cana-6113	122	66	inceptionv3	inceptionv3	NOUN
cana-6113	122	67	,	,	PUNCT
cana-6113	122	68	and	and	CCONJ
cana-6113	122	69	the	the	DET
cana-6113	122	70	proposed	propose	VERB
cana-6113	122	71	deep	deep	ADJ
cana-6113	122	72	learning	learning	NOUN
cana-6113	122	73	framework	framework	NOUN
cana-6113	122	74	based	base	VERB
cana-6113	122	75	on	on	ADP
cana-6113	122	76	key	key	ADJ
cana-6113	122	77	metrics	metric	NOUN
cana-6113	122	78	including	include	VERB
cana-6113	122	79	accuracy	accuracy	NOUN
cana-6113	122	80	,	,	PUNCT
cana-6113	122	81	precision	precision	NOUN
cana-6113	122	82	,	,	PUNCT
cana-6113	122	83	recall	recall	NOUN
cana-6113	122	84	,	,	PUNCT
cana-6113	122	85	and	and	CCONJ
cana-6113	122	86	f1	f1	NOUN
cana-6113	122	87	-	-	PUNCT
cana-6113	122	88	score	score	NOUN
cana-6113	122	89	.	.	PUNCT
cana-6113	123	1	among	among	ADP
cana-6113	123	2	the	the	DET
cana-6113	123	3	pre	pre	ADJ
cana-6113	123	4	-	-	ADJ
cana-6113	123	5	existing	existing	ADJ
cana-6113	123	6	architectures	architecture	NOUN
cana-6113	123	7	,	,	PUNCT
cana-6113	123	8	alexnet	alexnet	PROPN
cana-6113	123	9	achieved	achieve	VERB
cana-6113	123	10	an	an	DET
cana-6113	123	11	accuracy	accuracy	NOUN
cana-6113	123	12	of	of	ADP
cana-6113	123	13	88.45	88.45	NUM
cana-6113	123	14	%	%	NOUN
cana-6113	123	15	,	,	PUNCT
cana-6113	123	16	with	with	ADP
cana-6113	123	17	slightly	slightly	ADV
cana-6113	123	18	lower	low	ADJ
cana-6113	123	19	precision	precision	NOUN
cana-6113	123	20	(	(	PUNCT
cana-6113	123	21	87.90	87.90	NUM
cana-6113	123	22	%	%	NOUN
cana-6113	123	23	)	)	PUNCT
cana-6113	123	24	communications	communication	NOUN
cana-6113	123	25	on	on	ADP
cana-6113	123	26	applied	apply	VERB
cana-6113	123	27	nonlinear	nonlinear	ADJ
cana-6113	123	28	analysis	analysis	NOUN
cana-6113	123	29	issn	issn	NOUN
cana-6113	123	30	:	:	PUNCT
cana-6113	123	31	1074	1074	NUM
cana-6113	123	32	-	-	PUNCT
cana-6113	123	33	133x	133x	NUM
cana-6113	123	34	vol	vol	VERB
cana-6113	123	35	32	32	NUM
cana-6113	123	36	no	no	NOUN
cana-6113	123	37	.	.	PUNCT
cana-6113	124	1	10s	10	NOUN
cana-6113	124	2	(	(	PUNCT
cana-6113	124	3	2025	2025	NUM
cana-6113	124	4	)	)	PUNCT
cana-6113	124	5	3410	3410	NUM
cana-6113	124	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	124	7	and	and	CCONJ
cana-6113	124	8	recall	recall	PROPN
cana-6113	124	9	(	(	PUNCT
cana-6113	124	10	86.75	86.75	NUM
cana-6113	124	11	%	%	NOUN
cana-6113	124	12	)	)	PUNCT
cana-6113	124	13	,	,	PUNCT
cana-6113	124	14	resulting	result	VERB
cana-6113	124	15	in	in	ADP
cana-6113	124	16	an	an	DET
cana-6113	124	17	f1	f1	NOUN
cana-6113	124	18	-	-	PUNCT
cana-6113	124	19	score	score	NOUN
cana-6113	124	20	of	of	ADP
cana-6113	124	21	87.32	87.32	NUM
cana-6113	124	22	%	%	NOUN
cana-6113	124	23	.	.	PUNCT
cana-6113	125	1	this	this	PRON
cana-6113	125	2	indicates	indicate	VERB
cana-6113	125	3	that	that	SCONJ
cana-6113	125	4	while	while	SCONJ
cana-6113	125	5	alexnet	alexnet	NOUN
cana-6113	125	6	is	be	AUX
cana-6113	125	7	capable	capable	ADJ
cana-6113	125	8	of	of	ADP
cana-6113	125	9	extracting	extract	VERB
cana-6113	125	10	relevant	relevant	ADJ
cana-6113	125	11	features	feature	NOUN
cana-6113	125	12	from	from	ADP
cana-6113	125	13	colposcopy	colposcopy	NOUN
cana-6113	125	14	images	image	NOUN
cana-6113	125	15	,	,	PUNCT
cana-6113	125	16	its	its	PRON
cana-6113	125	17	performance	performance	NOUN
cana-6113	125	18	is	be	AUX
cana-6113	125	19	limited	limit	VERB
cana-6113	125	20	by	by	ADP
cana-6113	125	21	its	its	PRON
cana-6113	125	22	relatively	relatively	ADV
cana-6113	125	23	shallow	shallow	ADJ
cana-6113	125	24	architecture	architecture	NOUN
cana-6113	125	25	.	.	PUNCT
cana-6113	126	1	vgg16	vgg16	NOUN
cana-6113	126	2	showed	show	VERB
cana-6113	126	3	improved	improved	ADJ
cana-6113	126	4	results	result	NOUN
cana-6113	126	5	,	,	PUNCT
cana-6113	126	6	achieving	achieve	VERB
cana-6113	126	7	91.20	91.20	NUM
cana-6113	126	8	%	%	NOUN
cana-6113	126	9	accuracy	accuracy	NOUN
cana-6113	126	10	,	,	PUNCT
cana-6113	126	11	90.85	90.85	NUM
cana-6113	126	12	%	%	NOUN
cana-6113	126	13	precision	precision	NOUN
cana-6113	126	14	,	,	PUNCT
cana-6113	126	15	and	and	CCONJ
cana-6113	126	16	90.10	90.10	NUM
cana-6113	126	17	%	%	NOUN
cana-6113	126	18	recall	recall	NOUN
cana-6113	126	19	,	,	PUNCT
cana-6113	126	20	with	with	ADP
cana-6113	126	21	an	an	DET
cana-6113	126	22	f1	f1	NOUN
cana-6113	126	23	-	-	PUNCT
cana-6113	126	24	score	score	NOUN
cana-6113	126	25	of	of	ADP
cana-6113	126	26	90.47	90.47	NUM
cana-6113	126	27	%	%	NOUN
cana-6113	126	28	.	.	PUNCT
cana-6113	127	1	this	this	DET
cana-6113	127	2	improvement	improvement	NOUN
cana-6113	127	3	is	be	AUX
cana-6113	127	4	attributed	attribute	VERB
cana-6113	127	5	to	to	ADP
cana-6113	127	6	vgg16	vgg16	PROPN
cana-6113	127	7	’s	’s	PART
cana-6113	127	8	deeper	deep	ADJ
cana-6113	127	9	convolutional	convolutional	ADJ
cana-6113	127	10	layers	layer	NOUN
cana-6113	127	11	and	and	CCONJ
cana-6113	127	12	more	more	ADJ
cana-6113	127	13	refined	refined	ADJ
cana-6113	127	14	feature	feature	NOUN
cana-6113	127	15	extraction	extraction	NOUN
cana-6113	127	16	capabilities	capability	NOUN
cana-6113	127	17	.	.	PUNCT
cana-6113	128	1	inceptionv3	inceptionv3	NOUN
cana-6113	128	2	further	far	ADV
cana-6113	128	3	enhanced	enhance	VERB
cana-6113	128	4	performance	performance	NOUN
cana-6113	128	5	with	with	ADP
cana-6113	128	6	an	an	DET
cana-6113	128	7	accuracy	accuracy	NOUN
cana-6113	128	8	of	of	ADP
cana-6113	128	9	93.50	93.50	NUM
cana-6113	128	10	%	%	NOUN
cana-6113	128	11	,	,	PUNCT
cana-6113	128	12	precision	precision	NOUN
cana-6113	128	13	of	of	ADP
cana-6113	128	14	93.10	93.10	NUM
cana-6113	128	15	%	%	NOUN
cana-6113	128	16	,	,	PUNCT
cana-6113	128	17	and	and	CCONJ
cana-6113	128	18	recall	recall	NOUN
cana-6113	128	19	of	of	ADP
cana-6113	128	20	92.75	92.75	NUM
cana-6113	128	21	%	%	NOUN
cana-6113	128	22	,	,	PUNCT
cana-6113	128	23	resulting	result	VERB
cana-6113	128	24	in	in	ADP
cana-6113	128	25	an	an	DET
cana-6113	128	26	f1	f1	NOUN
cana-6113	128	27	-	-	PUNCT
cana-6113	128	28	score	score	NOUN
cana-6113	128	29	of	of	ADP
cana-6113	128	30	92.92	92.92	NUM
cana-6113	128	31	%	%	NOUN
cana-6113	128	32	.	.	PUNCT
cana-6113	129	1	its	its	PRON
cana-6113	129	2	multi	multi	ADJ
cana-6113	129	3	-	-	ADJ
cana-6113	129	4	scale	scale	ADJ
cana-6113	129	5	convolutional	convolutional	ADJ
cana-6113	129	6	modules	module	NOUN
cana-6113	129	7	and	and	CCONJ
cana-6113	129	8	efficient	efficient	ADJ
cana-6113	129	9	architecture	architecture	NOUN
cana-6113	129	10	allow	allow	VERB
cana-6113	129	11	it	it	PRON
cana-6113	129	12	to	to	PART
cana-6113	129	13	capture	capture	VERB
cana-6113	129	14	complex	complex	ADJ
cana-6113	129	15	spatial	spatial	ADJ
cana-6113	129	16	patterns	pattern	NOUN
cana-6113	129	17	more	more	ADV
cana-6113	129	18	effectively	effectively	ADV
cana-6113	129	19	,	,	PUNCT
cana-6113	129	20	demonstrating	demonstrate	VERB
cana-6113	129	21	the	the	DET
cana-6113	129	22	advantage	advantage	NOUN
cana-6113	129	23	of	of	ADP
cana-6113	129	24	advanced	advanced	ADJ
cana-6113	129	25	deep	deep	ADJ
cana-6113	129	26	learning	learning	NOUN
cana-6113	129	27	models	model	NOUN
cana-6113	129	28	over	over	ADP
cana-6113	129	29	simpler	simple	ADJ
cana-6113	129	30	cnn	cnn	PROPN
cana-6113	129	31	architectures	architecture	NOUN
cana-6113	129	32	.	.	PUNCT
cana-6113	130	1	table	table	NOUN
cana-6113	130	2	3	3	NUM
cana-6113	130	3	:	:	PUNCT
cana-6113	130	4	performance	performance	NOUN
cana-6113	130	5	evaluation	evaluation	NOUN
cana-6113	130	6	of	of	ADP
cana-6113	130	7	alexnet	alexnet	NOUN
cana-6113	130	8	,	,	PUNCT
cana-6113	130	9	vgg16	vgg16	PROPN
cana-6113	130	10	,	,	PUNCT
cana-6113	130	11	inceptionv3	inceptionv3	NOUN
cana-6113	130	12	,	,	PUNCT
cana-6113	130	13	and	and	CCONJ
cana-6113	130	14	the	the	DET
cana-6113	130	15	proposed	propose	VERB
cana-6113	130	16	model	model	NOUN
cana-6113	130	17	using	use	VERB
cana-6113	130	18	accuracy	accuracy	NOUN
cana-6113	130	19	,	,	PUNCT
cana-6113	130	20	precision	precision	NOUN
cana-6113	130	21	,	,	PUNCT
cana-6113	130	22	recall	recall	NOUN
cana-6113	130	23	,	,	PUNCT
cana-6113	130	24	and	and	CCONJ
cana-6113	130	25	f1	f1	ADJ
cana-6113	130	26	-	-	PUNCT
cana-6113	130	27	score	score	NOUN
cana-6113	130	28	metrics	metric	NOUN
cana-6113	130	29	.	.	PUNCT
cana-6113	131	1	model	model	NOUN
cana-6113	131	2	accuracy	accuracy	NOUN
cana-6113	131	3	(	(	PUNCT
cana-6113	131	4	%	%	INTJ
cana-6113	131	5	)	)	PUNCT
cana-6113	131	6	precision	precision	NOUN
cana-6113	131	7	(	(	PUNCT
cana-6113	131	8	%	%	INTJ
cana-6113	131	9	)	)	PUNCT
cana-6113	131	10	recall	recall	NOUN
cana-6113	131	11	(	(	PUNCT
cana-6113	131	12	%	%	NOUN
cana-6113	131	13	)	)	PUNCT
cana-6113	131	14	f1	f1	NOUN
cana-6113	131	15	-	-	PUNCT
cana-6113	131	16	score	score	NOUN
cana-6113	131	17	(	(	PUNCT
cana-6113	131	18	%	%	INTJ
cana-6113	131	19	)	)	PUNCT
cana-6113	131	20	alexnet	alexnet	NOUN
cana-6113	132	1	[	[	X
cana-6113	132	2	11	11	NUM
cana-6113	132	3	]	]	PUNCT
cana-6113	132	4	88.45	88.45	NUM
cana-6113	132	5	87.90	87.90	NUM
cana-6113	132	6	86.75	86.75	NUM
cana-6113	132	7	87.32	87.32	NUM
cana-6113	132	8	vgg16	vgg16	PROPN
cana-6113	132	9	[	[	X
cana-6113	132	10	12	12	NUM
cana-6113	132	11	]	]	PUNCT
cana-6113	132	12	91.20	91.20	NUM
cana-6113	132	13	90.85	90.85	NUM
cana-6113	132	14	90.10	90.10	NUM
cana-6113	132	15	90.47	90.47	NUM
cana-6113	132	16	inceptionv3	inceptionv3	NOUN
cana-6113	133	1	[	[	X
cana-6113	133	2	13	13	NUM
cana-6113	133	3	]	]	PUNCT
cana-6113	133	4	93.50	93.50	NUM
cana-6113	133	5	93.10	93.10	NUM
cana-6113	133	6	92.75	92.75	NUM
cana-6113	133	7	92.92	92.92	NUM
cana-6113	133	8	proposed	propose	VERB
cana-6113	133	9	model	model	NOUN
cana-6113	133	10	96.85	96.85	NUM
cana-6113	133	11	96.50	96.50	NUM
cana-6113	133	12	96.20	96.20	NUM
cana-6113	133	13	96.35	96.35	NUM
cana-6113	133	14	figure	figure	NOUN
cana-6113	133	15	3	3	NUM
cana-6113	133	16	:	:	PUNCT
cana-6113	133	17	performance	performance	NOUN
cana-6113	133	18	comparison	comparison	NOUN
cana-6113	133	19	of	of	ADP
cana-6113	133	20	alexnet	alexnet	NOUN
cana-6113	133	21	,	,	PUNCT
cana-6113	133	22	vgg16	vgg16	PROPN
cana-6113	133	23	,	,	PUNCT
cana-6113	133	24	inceptionv3	inceptionv3	NOUN
cana-6113	133	25	,	,	PUNCT
cana-6113	133	26	and	and	CCONJ
cana-6113	133	27	the	the	DET
cana-6113	133	28	proposed	propose	VERB
cana-6113	133	29	model	model	NOUN
cana-6113	133	30	the	the	DET
cana-6113	133	31	proposed	propose	VERB
cana-6113	133	32	model	model	NOUN
cana-6113	133	33	,	,	PUNCT
cana-6113	133	34	however	however	ADV
cana-6113	133	35	,	,	PUNCT
cana-6113	133	36	outperformed	outperform	VERB
cana-6113	133	37	all	all	DET
cana-6113	133	38	baseline	baseline	PROPN
cana-6113	133	39	architectures	architecture	NOUN
cana-6113	133	40	,	,	PUNCT
cana-6113	133	41	achieving	achieve	VERB
cana-6113	133	42	96.85	96.85	NUM
cana-6113	133	43	%	%	NOUN
cana-6113	133	44	accuracy	accuracy	NOUN
cana-6113	133	45	,	,	PUNCT
cana-6113	133	46	96.50	96.50	NUM
cana-6113	133	47	%	%	NOUN
cana-6113	133	48	precision	precision	NOUN
cana-6113	133	49	,	,	PUNCT
cana-6113	133	50	96.20	96.20	NUM
cana-6113	133	51	%	%	NOUN
cana-6113	133	52	recall	recall	NOUN
cana-6113	133	53	,	,	PUNCT
cana-6113	133	54	and	and	CCONJ
cana-6113	133	55	an	an	DET
cana-6113	133	56	f1	f1	NOUN
cana-6113	133	57	-	-	PUNCT
cana-6113	133	58	score	score	NOUN
cana-6113	133	59	of	of	ADP
cana-6113	133	60	96.35	96.35	NUM
cana-6113	133	61	%	%	NOUN
cana-6113	133	62	.	.	PUNCT
cana-6113	134	1	the	the	DET
cana-6113	134	2	superior	superior	ADJ
cana-6113	134	3	performance	performance	NOUN
cana-6113	134	4	can	can	AUX
cana-6113	134	5	be	be	AUX
cana-6113	134	6	attributed	attribute	VERB
cana-6113	134	7	to	to	ADP
cana-6113	134	8	the	the	DET
cana-6113	134	9	multi	multi	NOUN
cana-6113	134	10	-	-	ADJ
cana-6113	134	11	modal	modal	ADJ
cana-6113	134	12	,	,	PUNCT
cana-6113	134	13	multi	multi	ADJ
cana-6113	134	14	-	-	ADJ
cana-6113	134	15	stage	stage	ADJ
cana-6113	134	16	feature	feature	NOUN
cana-6113	134	17	extraction	extraction	NOUN
cana-6113	134	18	and	and	CCONJ
cana-6113	134	19	fusion	fusion	NOUN
cana-6113	134	20	framework	framework	NOUN
cana-6113	134	21	,	,	PUNCT
cana-6113	134	22	which	which	PRON
cana-6113	134	23	integrates	integrate	VERB
cana-6113	134	24	both	both	CCONJ
cana-6113	134	25	clinical	clinical	ADJ
cana-6113	134	26	data	datum	NOUN
cana-6113	134	27	and	and	CCONJ
cana-6113	134	28	colposcopy	colposcopy	VERB
cana-6113	134	29	image	image	NOUN
cana-6113	134	30	features	feature	NOUN
cana-6113	134	31	.	.	PUNCT
cana-6113	135	1	by	by	ADP
cana-6113	135	2	combining	combine	VERB
cana-6113	135	3	structured	structure	VERB
cana-6113	135	4	and	and	CCONJ
cana-6113	135	5	unstructured	unstructured	ADJ
cana-6113	135	6	data	datum	NOUN
cana-6113	135	7	,	,	PUNCT
cana-6113	135	8	and	and	CCONJ
cana-6113	135	9	employing	employ	VERB
cana-6113	135	10	attention	attention	NOUN
cana-6113	135	11	mechanisms	mechanism	NOUN
cana-6113	135	12	and	and	CCONJ
cana-6113	135	13	hierarchical	hierarchical	ADJ
cana-6113	135	14	cnn	cnn	PROPN
cana-6113	135	15	layers	layer	NOUN
cana-6113	135	16	,	,	PUNCT
cana-6113	135	17	the	the	DET
cana-6113	135	18	proposed	propose	VERB
cana-6113	135	19	model	model	NOUN
cana-6113	135	20	captures	capture	VERB
cana-6113	135	21	subtle	subtle	ADJ
cana-6113	135	22	patterns	pattern	NOUN
cana-6113	135	23	in	in	ADP
cana-6113	135	24	the	the	DET
cana-6113	135	25	lesion	lesion	NOUN
cana-6113	135	26	areas	area	NOUN
cana-6113	135	27	while	while	SCONJ
cana-6113	135	28	also	also	ADV
cana-6113	135	29	leveraging	leverage	VERB
cana-6113	135	30	patient	patient	NOUN
cana-6113	135	31	-	-	PUNCT
cana-6113	135	32	specific	specific	ADJ
cana-6113	135	33	clinical	clinical	ADJ
cana-6113	135	34	information	information	NOUN
cana-6113	135	35	.	.	PUNCT
cana-6113	136	1	this	this	PRON
cana-6113	136	2	results	result	VERB
cana-6113	136	3	in	in	ADP
cana-6113	136	4	a	a	DET
cana-6113	136	5	more	more	ADV
cana-6113	136	6	robust	robust	ADJ
cana-6113	136	7	and	and	CCONJ
cana-6113	136	8	reliable	reliable	ADJ
cana-6113	136	9	prediction	prediction	NOUN
cana-6113	136	10	,	,	PUNCT
cana-6113	136	11	reducing	reduce	VERB
cana-6113	136	12	both	both	DET
cana-6113	136	13	false	false	ADJ
cana-6113	136	14	positives	positive	NOUN
cana-6113	136	15	and	and	CCONJ
cana-6113	136	16	false	false	ADJ
cana-6113	136	17	negatives	negative	NOUN
cana-6113	136	18	,	,	PUNCT
cana-6113	136	19	which	which	PRON
cana-6113	136	20	is	be	AUX
cana-6113	136	21	particularly	particularly	ADV
cana-6113	136	22	critical	critical	ADJ
cana-6113	136	23	in	in	ADP
cana-6113	136	24	medical	medical	ADJ
cana-6113	136	25	diagnostics	diagnostic	NOUN
cana-6113	136	26	.	.	PUNCT
cana-6113	137	1	80	80	NUM
cana-6113	137	2	82	82	NUM
cana-6113	137	3	84	84	NUM
cana-6113	137	4	86	86	NUM
cana-6113	137	5	88	88	NUM
cana-6113	137	6	90	90	NUM
cana-6113	137	7	92	92	NUM
cana-6113	137	8	94	94	NUM
cana-6113	137	9	96	96	NUM
cana-6113	137	10	98	98	NUM
cana-6113	137	11	accuracy	accuracy	NOUN
cana-6113	137	12	(	(	PUNCT
cana-6113	137	13	%	%	INTJ
cana-6113	137	14	)	)	PUNCT
cana-6113	137	15	precision	precision	NOUN
cana-6113	137	16	(	(	PUNCT
cana-6113	137	17	%	%	INTJ
cana-6113	137	18	)	)	PUNCT
cana-6113	137	19	recall	recall	NOUN
cana-6113	137	20	(	(	PUNCT
cana-6113	137	21	%	%	NOUN
cana-6113	137	22	)	)	PUNCT
cana-6113	137	23	f1	f1	NOUN
cana-6113	137	24	-	-	PUNCT
cana-6113	137	25	score	score	NOUN
cana-6113	137	26	(	(	PUNCT
cana-6113	137	27	%	%	INTJ
cana-6113	137	28	)	)	PUNCT
cana-6113	137	29	performance	performance	NOUN
cana-6113	137	30	evaluation	evaluation	NOUN
cana-6113	137	31	alexnet	alexnet	NOUN
cana-6113	137	32	[	[	X
cana-6113	137	33	11	11	NUM
cana-6113	137	34	]	]	X
cana-6113	137	35	vgg16	vgg16	NOUN
cana-6113	137	36	[	[	X
cana-6113	137	37	12	12	NUM
cana-6113	137	38	]	]	PUNCT
cana-6113	137	39	inceptionv3	inceptionv3	NOUN
cana-6113	138	1	[	[	X
cana-6113	138	2	13	13	NUM
cana-6113	138	3	]	]	PUNCT
cana-6113	138	4	proposed	propose	VERB
cana-6113	138	5	model	model	NOUN
cana-6113	138	6	communications	communication	NOUN
cana-6113	138	7	on	on	ADP
cana-6113	138	8	applied	apply	VERB
cana-6113	138	9	nonlinear	nonlinear	ADJ
cana-6113	138	10	analysis	analysis	NOUN
cana-6113	138	11	issn	issn	NOUN
cana-6113	138	12	:	:	PUNCT
cana-6113	138	13	1074	1074	NUM
cana-6113	138	14	-	-	PUNCT
cana-6113	138	15	133x	133x	NUM
cana-6113	138	16	vol	vol	VERB
cana-6113	138	17	32	32	NUM
cana-6113	138	18	no	no	NOUN
cana-6113	138	19	.	.	PUNCT
cana-6113	139	1	10s	10	NOUN
cana-6113	139	2	(	(	PUNCT
cana-6113	139	3	2025	2025	NUM
cana-6113	139	4	)	)	PUNCT
cana-6113	139	5	3411	3411	NUM
cana-6113	139	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	139	7	6	6	NUM
cana-6113	139	8	.	.	PUNCT
cana-6113	139	9	conclusion	conclusion	NOUN
cana-6113	139	10	this	this	DET
cana-6113	139	11	paper	paper	NOUN
cana-6113	139	12	presented	present	VERB
cana-6113	139	13	a	a	DET
cana-6113	139	14	deep	deep	ADJ
cana-6113	139	15	learning	learning	NOUN
cana-6113	139	16	-	-	PUNCT
cana-6113	139	17	based	base	VERB
cana-6113	139	18	multi	multi	ADJ
cana-6113	139	19	-	-	ADJ
cana-6113	139	20	stage	stage	ADJ
cana-6113	139	21	feature	feature	NOUN
cana-6113	139	22	processing	processing	NOUN
cana-6113	139	23	and	and	CCONJ
cana-6113	139	24	fusion	fusion	NOUN
cana-6113	139	25	framework	framework	NOUN
cana-6113	139	26	for	for	ADP
cana-6113	139	27	enhanced	enhanced	ADJ
cana-6113	139	28	visual	visual	ADJ
cana-6113	139	29	understanding	understanding	NOUN
cana-6113	139	30	,	,	PUNCT
cana-6113	139	31	focusing	focus	VERB
cana-6113	139	32	on	on	ADP
cana-6113	139	33	cervical	cervical	ADJ
cana-6113	139	34	lesion	lesion	NOUN
cana-6113	139	35	prediction	prediction	NOUN
cana-6113	139	36	using	use	VERB
cana-6113	139	37	both	both	CCONJ
cana-6113	139	38	clinical	clinical	ADJ
cana-6113	139	39	data	datum	NOUN
cana-6113	139	40	and	and	CCONJ
cana-6113	139	41	colposcopy	colposcopy	NOUN
cana-6113	139	42	images	image	NOUN
cana-6113	139	43	.	.	PUNCT
cana-6113	140	1	the	the	DET
cana-6113	140	2	proposed	propose	VERB
cana-6113	140	3	approach	approach	NOUN
cana-6113	140	4	combines	combine	VERB
cana-6113	140	5	hierarchical	hierarchical	ADJ
cana-6113	140	6	convolutional	convolutional	ADJ
cana-6113	140	7	neural	neural	ADJ
cana-6113	140	8	networks	network	NOUN
cana-6113	140	9	for	for	ADP
cana-6113	140	10	extracting	extract	VERB
cana-6113	140	11	image	image	NOUN
cana-6113	140	12	features	feature	NOUN
cana-6113	140	13	with	with	ADP
cana-6113	140	14	machine	machine	NOUN
cana-6113	140	15	learning	learning	NOUN
cana-6113	140	16	models	model	NOUN
cana-6113	140	17	for	for	ADP
cana-6113	140	18	clinical	clinical	ADJ
cana-6113	140	19	data	datum	NOUN
cana-6113	140	20	,	,	PUNCT
cana-6113	140	21	followed	follow	VERB
cana-6113	140	22	by	by	ADP
cana-6113	140	23	feature	feature	NOUN
cana-6113	140	24	-	-	PUNCT
cana-6113	140	25	level	level	NOUN
cana-6113	140	26	fusion	fusion	NOUN
cana-6113	140	27	to	to	PART
cana-6113	140	28	generate	generate	VERB
cana-6113	140	29	a	a	DET
cana-6113	140	30	unified	unified	ADJ
cana-6113	140	31	representation	representation	NOUN
cana-6113	140	32	.	.	PUNCT
cana-6113	141	1	performance	performance	NOUN
cana-6113	141	2	evaluation	evaluation	NOUN
cana-6113	141	3	demonstrated	demonstrate	VERB
cana-6113	141	4	that	that	SCONJ
cana-6113	141	5	the	the	DET
cana-6113	141	6	proposed	propose	VERB
cana-6113	141	7	model	model	NOUN
cana-6113	141	8	achieved	achieve	VERB
cana-6113	141	9	an	an	DET
cana-6113	141	10	accuracy	accuracy	NOUN
cana-6113	141	11	of	of	ADP
cana-6113	141	12	96.85	96.85	NUM
cana-6113	141	13	%	%	NOUN
cana-6113	141	14	,	,	PUNCT
cana-6113	141	15	outperforming	outperform	VERB
cana-6113	141	16	conventional	conventional	ADJ
cana-6113	141	17	architectures	architecture	NOUN
cana-6113	141	18	such	such	ADJ
cana-6113	141	19	as	as	ADP
cana-6113	141	20	alexnet	alexnet	NOUN
cana-6113	141	21	(	(	PUNCT
cana-6113	141	22	88.45	88.45	NUM
cana-6113	141	23	%	%	NOUN
cana-6113	141	24	)	)	PUNCT
cana-6113	141	25	,	,	PUNCT
cana-6113	141	26	vgg16	vgg16	NOUN
cana-6113	141	27	(	(	PUNCT
cana-6113	141	28	91.20	91.20	NUM
cana-6113	141	29	%	%	NOUN
cana-6113	141	30	)	)	PUNCT
cana-6113	141	31	,	,	PUNCT
cana-6113	141	32	and	and	CCONJ
cana-6113	141	33	inceptionv3	inceptionv3	NOUN
cana-6113	141	34	(	(	PUNCT
cana-6113	141	35	93.50	93.50	NUM
cana-6113	141	36	%	%	NOUN
cana-6113	141	37	)	)	PUNCT
cana-6113	141	38	.	.	PUNCT
cana-6113	142	1	the	the	DET
cana-6113	142	2	results	result	NOUN
cana-6113	142	3	also	also	ADV
cana-6113	142	4	show	show	VERB
cana-6113	142	5	improvements	improvement	NOUN
cana-6113	142	6	in	in	ADP
cana-6113	142	7	precision	precision	NOUN
cana-6113	142	8	,	,	PUNCT
cana-6113	142	9	recall	recall	NOUN
cana-6113	142	10	,	,	PUNCT
cana-6113	142	11	and	and	CCONJ
cana-6113	142	12	f1	f1	NOUN
cana-6113	142	13	-	-	PUNCT
cana-6113	142	14	score	score	NOUN
cana-6113	142	15	,	,	PUNCT
cana-6113	142	16	indicating	indicate	VERB
cana-6113	142	17	that	that	SCONJ
cana-6113	142	18	the	the	DET
cana-6113	142	19	model	model	NOUN
cana-6113	142	20	can	can	AUX
cana-6113	142	21	reliably	reliably	ADV
cana-6113	142	22	detect	detect	VERB
cana-6113	142	23	positive	positive	ADJ
cana-6113	142	24	and	and	CCONJ
cana-6113	142	25	negative	negative	ADJ
cana-6113	142	26	cases	case	NOUN
cana-6113	142	27	while	while	SCONJ
cana-6113	142	28	minimizing	minimize	VERB
cana-6113	142	29	false	false	ADJ
cana-6113	142	30	predictions	prediction	NOUN
cana-6113	142	31	.	.	PUNCT
cana-6113	143	1	this	this	DET
cana-6113	143	2	highlights	highlight	VERB
cana-6113	143	3	the	the	DET
cana-6113	143	4	effectiveness	effectiveness	NOUN
cana-6113	143	5	of	of	ADP
cana-6113	143	6	integrating	integrate	VERB
cana-6113	143	7	multi	multi	ADJ
cana-6113	143	8	-	-	ADJ
cana-6113	143	9	modal	modal	ADJ
cana-6113	143	10	data	datum	NOUN
cana-6113	143	11	and	and	CCONJ
cana-6113	143	12	advanced	advanced	ADJ
cana-6113	143	13	feature	feature	NOUN
cana-6113	143	14	extraction	extraction	NOUN
cana-6113	143	15	strategies	strategy	NOUN
cana-6113	143	16	in	in	ADP
cana-6113	143	17	medical	medical	ADJ
cana-6113	143	18	image	image	NOUN
cana-6113	143	19	analysis	analysis	NOUN
cana-6113	143	20	.	.	PUNCT
cana-6113	144	1	the	the	DET
cana-6113	144	2	study	study	NOUN
cana-6113	144	3	emphasizes	emphasize	VERB
cana-6113	144	4	the	the	DET
cana-6113	144	5	potential	potential	NOUN
cana-6113	144	6	of	of	ADP
cana-6113	144	7	multi	multi	ADJ
cana-6113	144	8	-	-	NOUN
cana-6113	144	9	stage	stage	NOUN
cana-6113	144	10	,	,	PUNCT
cana-6113	144	11	attention	attention	NOUN
cana-6113	144	12	-	-	PUNCT
cana-6113	144	13	driven	drive	VERB
cana-6113	144	14	deep	deep	ADJ
cana-6113	144	15	learning	learning	NOUN
cana-6113	144	16	frameworks	framework	NOUN
cana-6113	144	17	in	in	ADP
cana-6113	144	18	clinical	clinical	ADJ
cana-6113	144	19	decision	decision	NOUN
cana-6113	144	20	support	support	NOUN
cana-6113	144	21	systems	system	NOUN
cana-6113	144	22	.	.	PUNCT
cana-6113	145	1	by	by	ADP
cana-6113	145	2	leveraging	leverage	VERB
cana-6113	145	3	both	both	DET
cana-6113	145	4	image	image	NOUN
cana-6113	145	5	-	-	PUNCT
cana-6113	145	6	based	base	VERB
cana-6113	145	7	and	and	CCONJ
cana-6113	145	8	patient	patient	NOUN
cana-6113	145	9	-	-	PUNCT
cana-6113	145	10	specific	specific	ADJ
cana-6113	145	11	data	datum	NOUN
cana-6113	145	12	,	,	PUNCT
cana-6113	145	13	the	the	DET
cana-6113	145	14	proposed	propose	VERB
cana-6113	145	15	model	model	NOUN
cana-6113	145	16	ensures	ensure	VERB
cana-6113	145	17	more	more	ADV
cana-6113	145	18	reliable	reliable	ADJ
cana-6113	145	19	and	and	CCONJ
cana-6113	145	20	interpretable	interpretable	ADJ
cana-6113	145	21	predictions	prediction	NOUN
cana-6113	145	22	,	,	PUNCT
cana-6113	145	23	reducing	reduce	VERB
cana-6113	145	24	diagnostic	diagnostic	ADJ
cana-6113	145	25	errors	error	NOUN
cana-6113	145	26	in	in	ADP
cana-6113	145	27	critical	critical	ADJ
cana-6113	145	28	medical	medical	ADJ
cana-6113	145	29	applications	application	NOUN
cana-6113	145	30	.	.	PUNCT
cana-6113	146	1	references	reference	NOUN
cana-6113	146	2	[	[	X
cana-6113	146	3	1	1	NUM
cana-6113	146	4	]	]	X
cana-6113	146	5	simonyan	simonyan	ADJ
cana-6113	146	6	,	,	PUNCT
cana-6113	146	7	k.	k.	PROPN
cana-6113	146	8	,	,	PUNCT
cana-6113	146	9	&	&	CCONJ
cana-6113	146	10	zisserman	zisserman	PROPN
cana-6113	146	11	,	,	PUNCT
cana-6113	146	12	a.	a.	NOUN
cana-6113	146	13	(	(	PUNCT
cana-6113	146	14	2014	2014	NUM
cana-6113	146	15	)	)	PUNCT
cana-6113	146	16	.	.	PUNCT
cana-6113	147	1	two	two	NUM
cana-6113	147	2	-	-	PUNCT
cana-6113	147	3	stream	stream	NOUN
cana-6113	147	4	convolutional	convolutional	ADJ
cana-6113	147	5	networks	network	NOUN
cana-6113	147	6	for	for	ADP
cana-6113	147	7	action	action	NOUN
cana-6113	147	8	recognition	recognition	NOUN
cana-6113	147	9	in	in	ADP
cana-6113	147	10	videos	video	NOUN
cana-6113	147	11	.	.	PUNCT
cana-6113	148	1	advances	advance	NOUN
cana-6113	148	2	in	in	ADP
cana-6113	148	3	neural	neural	ADJ
cana-6113	148	4	information	information	NOUN
cana-6113	148	5	processing	processing	NOUN
cana-6113	148	6	systems	system	NOUN
cana-6113	148	7	(	(	PUNCT
cana-6113	148	8	neurips	neurip	NOUN
cana-6113	148	9	)	)	PUNCT
cana-6113	148	10	,	,	PUNCT
cana-6113	148	11	27	27	NUM
cana-6113	148	12	,	,	PUNCT
cana-6113	148	13	568–576	568–576	NUM
cana-6113	148	14	.	.	PUNCT
cana-6113	149	1	[	[	X
cana-6113	149	2	2	2	NUM
cana-6113	149	3	]	]	X
cana-6113	149	4	sultani	sultani	NOUN
cana-6113	149	5	,	,	PUNCT
cana-6113	149	6	w.	w.	PROPN
cana-6113	149	7	,	,	PUNCT
cana-6113	149	8	chen	chen	PROPN
cana-6113	149	9	,	,	PUNCT
cana-6113	149	10	c.	c.	PROPN
cana-6113	149	11	,	,	PUNCT
cana-6113	149	12	&	&	CCONJ
cana-6113	149	13	shah	shah	PROPN
cana-6113	149	14	,	,	PUNCT
cana-6113	149	15	m.	m.	NOUN
cana-6113	149	16	(	(	PUNCT
cana-6113	149	17	2018	2018	NUM
cana-6113	149	18	)	)	PUNCT
cana-6113	149	19	.	.	PUNCT
cana-6113	150	1	real	real	ADJ
cana-6113	150	2	-	-	PUNCT
cana-6113	150	3	world	world	NOUN
cana-6113	150	4	anomaly	anomaly	NOUN
cana-6113	150	5	detection	detection	NOUN
cana-6113	150	6	in	in	ADP
cana-6113	150	7	surveillance	surveillance	NOUN
cana-6113	150	8	videos	video	NOUN
cana-6113	150	9	.	.	PUNCT
cana-6113	151	1	proceedings	proceeding	NOUN
cana-6113	151	2	of	of	ADP
cana-6113	151	3	the	the	DET
cana-6113	151	4	ieee	ieee	NOUN
cana-6113	151	5	conference	conference	NOUN
cana-6113	151	6	on	on	ADP
cana-6113	151	7	computer	computer	NOUN
cana-6113	151	8	vision	vision	NOUN
cana-6113	151	9	and	and	CCONJ
cana-6113	151	10	pattern	pattern	NOUN
cana-6113	151	11	recognition	recognition	NOUN
cana-6113	151	12	(	(	PUNCT
cana-6113	151	13	cvpr	cvpr	NOUN
cana-6113	151	14	)	)	PUNCT
cana-6113	151	15	,	,	PUNCT
cana-6113	151	16	6479–6488	6479–6488	NUM
cana-6113	151	17	.	.	PUNCT
cana-6113	152	1	[	[	X
cana-6113	152	2	3	3	NUM
cana-6113	152	3	]	]	X
cana-6113	152	4	wang	wang	PROPN
cana-6113	152	5	,	,	PUNCT
cana-6113	152	6	l.	l.	PROPN
cana-6113	152	7	,	,	PUNCT
cana-6113	152	8	xiong	xiong	PROPN
cana-6113	152	9	,	,	PUNCT
cana-6113	152	10	y.	y.	PROPN
cana-6113	152	11	,	,	PUNCT
cana-6113	152	12	wang	wang	PROPN
cana-6113	152	13	,	,	PUNCT
cana-6113	152	14	z.	z.	PROPN
cana-6113	152	15	,	,	PUNCT
cana-6113	152	16	qiao	qiao	PROPN
cana-6113	152	17	,	,	PUNCT
cana-6113	152	18	y.	y.	PROPN
cana-6113	152	19	,	,	PUNCT
cana-6113	152	20	lin	lin	PROPN
cana-6113	152	21	,	,	PUNCT
cana-6113	152	22	d.	d.	PROPN
cana-6113	152	23	,	,	PUNCT
cana-6113	152	24	tang	tang	PROPN
cana-6113	152	25	,	,	PUNCT
cana-6113	152	26	x.	x.	PROPN
cana-6113	152	27	,	,	PUNCT
cana-6113	152	28	&	&	CCONJ
cana-6113	152	29	van	van	PROPN
cana-6113	152	30	gool	gool	PROPN
cana-6113	152	31	,	,	PUNCT
cana-6113	152	32	l.	l.	PROPN
cana-6113	152	33	(	(	PUNCT
cana-6113	152	34	2016	2016	NUM
cana-6113	152	35	)	)	PUNCT
cana-6113	152	36	.	.	PUNCT
cana-6113	153	1	temporal	temporal	ADJ
cana-6113	153	2	segment	segment	NOUN
cana-6113	153	3	networks	network	NOUN
cana-6113	153	4	for	for	ADP
cana-6113	153	5	action	action	NOUN
cana-6113	153	6	recognition	recognition	NOUN
cana-6113	153	7	in	in	ADP
cana-6113	153	8	videos	video	NOUN
cana-6113	153	9	.	.	PUNCT
cana-6113	154	1	ieee	ieee	NOUN
cana-6113	154	2	transactions	transaction	NOUN
cana-6113	154	3	on	on	ADP
cana-6113	154	4	pattern	pattern	NOUN
cana-6113	154	5	analysis	analysis	NOUN
cana-6113	154	6	and	and	CCONJ
cana-6113	154	7	machine	machine	NOUN
cana-6113	154	8	intelligence	intelligence	NOUN
cana-6113	154	9	(	(	PUNCT
cana-6113	154	10	tpami	tpami	NOUN
cana-6113	154	11	)	)	PUNCT
cana-6113	154	12	,	,	PUNCT
cana-6113	154	13	41(11	41(11	NUM
cana-6113	154	14	)	)	PUNCT
cana-6113	154	15	,	,	PUNCT
cana-6113	154	16	2740–2755	2740–2755	NUM
cana-6113	154	17	.	.	PUNCT
cana-6113	155	1	[	[	X
cana-6113	155	2	4	4	X
cana-6113	155	3	]	]	PUNCT
cana-6113	155	4	he	he	PRON
cana-6113	155	5	,	,	PUNCT
cana-6113	155	6	k.	k.	PROPN
cana-6113	155	7	,	,	PUNCT
cana-6113	155	8	zhang	zhang	PROPN
cana-6113	155	9	,	,	PUNCT
cana-6113	155	10	x.	x.	PROPN
cana-6113	155	11	,	,	PUNCT
cana-6113	155	12	ren	ren	PROPN
cana-6113	155	13	,	,	PUNCT
cana-6113	155	14	s.	s.	PROPN
cana-6113	155	15	,	,	PUNCT
cana-6113	155	16	&	&	CCONJ
cana-6113	155	17	sun	sun	PROPN
cana-6113	155	18	,	,	PUNCT
cana-6113	155	19	j.	j.	PROPN
cana-6113	155	20	(	(	PUNCT
cana-6113	155	21	2016	2016	NUM
cana-6113	155	22	)	)	PUNCT
cana-6113	155	23	.	.	PUNCT
cana-6113	156	1	deep	deep	ADJ
cana-6113	156	2	residual	residual	ADJ
cana-6113	156	3	learning	learning	NOUN
cana-6113	156	4	for	for	ADP
cana-6113	156	5	image	image	NOUN
cana-6113	156	6	recognition	recognition	NOUN
cana-6113	156	7	.	.	PUNCT
cana-6113	157	1	proceedings	proceeding	NOUN
cana-6113	157	2	of	of	ADP
cana-6113	157	3	the	the	DET
cana-6113	157	4	ieee	ieee	NOUN
cana-6113	157	5	conference	conference	NOUN
cana-6113	157	6	on	on	ADP
cana-6113	157	7	computer	computer	NOUN
cana-6113	157	8	vision	vision	NOUN
cana-6113	157	9	and	and	CCONJ
cana-6113	157	10	pattern	pattern	NOUN
cana-6113	157	11	recognition	recognition	NOUN
cana-6113	157	12	(	(	PUNCT
cana-6113	157	13	cvpr	cvpr	NOUN
cana-6113	157	14	)	)	PUNCT
cana-6113	157	15	,	,	PUNCT
cana-6113	157	16	770–778	770–778	NUM
cana-6113	157	17	.	.	PUNCT
cana-6113	158	1	[	[	X
cana-6113	158	2	5	5	NUM
cana-6113	158	3	]	]	SYM
cana-6113	158	4	hu	hu	PROPN
cana-6113	158	5	,	,	PUNCT
cana-6113	158	6	j.	j.	PROPN
cana-6113	158	7	,	,	PUNCT
cana-6113	158	8	shen	shen	PROPN
cana-6113	158	9	,	,	PUNCT
cana-6113	158	10	l.	l.	PROPN
cana-6113	158	11	,	,	PUNCT
cana-6113	158	12	&	&	CCONJ
cana-6113	158	13	sun	sun	PROPN
cana-6113	158	14	,	,	PUNCT
cana-6113	158	15	g.	g.	PROPN
cana-6113	158	16	(	(	PUNCT
cana-6113	158	17	2018	2018	NUM
cana-6113	158	18	)	)	PUNCT
cana-6113	158	19	.	.	PUNCT
cana-6113	159	1	squeeze	squeeze	NOUN
cana-6113	159	2	-	-	PUNCT
cana-6113	159	3	and	and	CCONJ
cana-6113	159	4	-	-	PUNCT
cana-6113	159	5	excitation	excitation	NOUN
cana-6113	159	6	networks	network	NOUN
cana-6113	159	7	.	.	PUNCT
cana-6113	160	1	proceedings	proceeding	NOUN
cana-6113	160	2	of	of	ADP
cana-6113	160	3	the	the	DET
cana-6113	160	4	ieee	ieee	NOUN
cana-6113	160	5	conference	conference	NOUN
cana-6113	160	6	on	on	ADP
cana-6113	160	7	computer	computer	NOUN
cana-6113	160	8	vision	vision	NOUN
cana-6113	160	9	and	and	CCONJ
cana-6113	160	10	pattern	pattern	NOUN
cana-6113	160	11	recognition	recognition	NOUN
cana-6113	160	12	(	(	PUNCT
cana-6113	160	13	cvpr	cvpr	NOUN
cana-6113	160	14	)	)	PUNCT
cana-6113	160	15	,	,	PUNCT
cana-6113	160	16	7132	7132	NUM
cana-6113	160	17	–	–	PUNCT
cana-6113	160	18	7141	7141	NUM
cana-6113	160	19	.	.	PUNCT
cana-6113	161	1	[	[	X
cana-6113	161	2	6	6	NUM
cana-6113	161	3	]	]	X
cana-6113	161	4	woo	woo	PROPN
cana-6113	161	5	,	,	PUNCT
cana-6113	161	6	s.	s.	PROPN
cana-6113	161	7	,	,	PUNCT
cana-6113	161	8	park	park	PROPN
cana-6113	161	9	,	,	PUNCT
cana-6113	161	10	j.	j.	PROPN
cana-6113	161	11	,	,	PUNCT
cana-6113	161	12	lee	lee	PROPN
cana-6113	161	13	,	,	PUNCT
cana-6113	161	14	j.-y	j.-y	PROPN
cana-6113	161	15	.	.	PROPN
cana-6113	161	16	,	,	PUNCT
cana-6113	161	17	&	&	CCONJ
cana-6113	161	18	kweon	kweon	PROPN
cana-6113	161	19	,	,	PUNCT
cana-6113	161	20	i.	i.	PROPN
cana-6113	161	21	s.	s.	PROPN
cana-6113	161	22	(	(	PUNCT
cana-6113	161	23	2018	2018	NUM
cana-6113	161	24	)	)	PUNCT
cana-6113	161	25	.	.	PUNCT
cana-6113	162	1	cbam	cbam	NOUN
cana-6113	162	2	:	:	PUNCT
cana-6113	162	3	convolutional	convolutional	ADJ
cana-6113	162	4	block	block	NOUN
cana-6113	162	5	attention	attention	NOUN
cana-6113	162	6	module	module	NOUN
cana-6113	162	7	.	.	PUNCT
cana-6113	163	1	proceedings	proceeding	NOUN
cana-6113	163	2	of	of	ADP
cana-6113	163	3	the	the	DET
cana-6113	163	4	european	european	PROPN
cana-6113	163	5	conference	conference	PROPN
cana-6113	163	6	on	on	ADP
cana-6113	163	7	computer	computer	NOUN
cana-6113	163	8	vision	vision	NOUN
cana-6113	163	9	(	(	PUNCT
cana-6113	163	10	eccv	eccv	ADV
cana-6113	163	11	)	)	PUNCT
cana-6113	163	12	,	,	PUNCT
cana-6113	163	13	3–19	3–19	NOUN
cana-6113	163	14	.	.	PUNCT
cana-6113	164	1	[	[	X
cana-6113	164	2	7	7	NUM
cana-6113	164	3	]	]	X
cana-6113	164	4	zhang	zhang	PROPN
cana-6113	164	5	,	,	PUNCT
cana-6113	164	6	f.	f.	PROPN
cana-6113	164	7	,	,	PUNCT
cana-6113	164	8	zhou	zhou	PROPN
cana-6113	164	9	,	,	PUNCT
cana-6113	164	10	d.	d.	PROPN
cana-6113	164	11	,	,	PUNCT
cana-6113	164	12	lin	lin	PROPN
cana-6113	164	13	,	,	PUNCT
cana-6113	164	14	y.	y.	PROPN
cana-6113	164	15	,	,	PUNCT
cana-6113	164	16	&	&	CCONJ
cana-6113	164	17	zhang	zhang	PROPN
cana-6113	164	18	,	,	PUNCT
cana-6113	164	19	s.	s.	PROPN
cana-6113	164	20	(	(	PUNCT
cana-6113	164	21	2019	2019	NUM
cana-6113	164	22	)	)	PUNCT
cana-6113	164	23	.	.	PUNCT
cana-6113	165	1	multi	multi	ADJ
cana-6113	165	2	-	-	ADJ
cana-6113	165	3	scale	scale	ADJ
cana-6113	165	4	context	context	NOUN
cana-6113	165	5	aggregation	aggregation	NOUN
cana-6113	165	6	by	by	ADP
cana-6113	165	7	dilated	dilated	ADJ
cana-6113	165	8	convolutions	convolution	NOUN
cana-6113	165	9	.	.	PUNCT
cana-6113	166	1	ieee	ieee	NOUN
cana-6113	166	2	transactions	transaction	NOUN
cana-6113	166	3	on	on	ADP
cana-6113	166	4	image	image	NOUN
cana-6113	166	5	processing	processing	NOUN
cana-6113	166	6	,	,	PUNCT
cana-6113	166	7	28(5	28(5	NUM
cana-6113	166	8	)	)	PUNCT
cana-6113	166	9	,	,	PUNCT
cana-6113	166	10	2169–2180	2169–2180	NUM
cana-6113	166	11	.	.	PUNCT
cana-6113	167	1	[	[	X
cana-6113	167	2	8	8	NUM
cana-6113	167	3	]	]	X
cana-6113	167	4	chen	chen	PROPN
cana-6113	167	5	,	,	PUNCT
cana-6113	167	6	l.-c	l.-c	PROPN
cana-6113	167	7	.	.	PUNCT
cana-6113	167	8	,	,	PUNCT
cana-6113	167	9	papandreou	papandreou	PROPN
cana-6113	167	10	,	,	PUNCT
cana-6113	167	11	g.	g.	PROPN
cana-6113	167	12	,	,	PUNCT
cana-6113	167	13	kokkinos	kokkinos	PROPN
cana-6113	167	14	,	,	PUNCT
cana-6113	167	15	i.	i.	PROPN
cana-6113	167	16	,	,	PUNCT
cana-6113	167	17	murphy	murphy	PROPN
cana-6113	167	18	,	,	PUNCT
cana-6113	167	19	k.	k.	PROPN
cana-6113	167	20	,	,	PUNCT
cana-6113	167	21	&	&	CCONJ
cana-6113	167	22	yuille	yuille	PROPN
cana-6113	167	23	,	,	PUNCT
cana-6113	167	24	a.	a.	PROPN
cana-6113	167	25	l.	l.	PROPN
cana-6113	167	26	(	(	PUNCT
cana-6113	167	27	2017	2017	NUM
cana-6113	167	28	)	)	PUNCT
cana-6113	167	29	.	.	PUNCT
cana-6113	168	1	deeplab	deeplab	NOUN
cana-6113	168	2	:	:	PUNCT
cana-6113	168	3	semantic	semantic	ADJ
cana-6113	168	4	image	image	NOUN
cana-6113	168	5	segmentation	segmentation	NOUN
cana-6113	168	6	with	with	ADP
cana-6113	168	7	atrous	atrous	ADJ
cana-6113	168	8	convolution	convolution	NOUN
cana-6113	168	9	and	and	CCONJ
cana-6113	168	10	fully	fully	ADV
cana-6113	168	11	connected	connect	VERB
cana-6113	168	12	crfs	crf	NOUN
cana-6113	168	13	.	.	PUNCT
cana-6113	169	1	ieee	ieee	NOUN
cana-6113	169	2	transactions	transaction	NOUN
cana-6113	169	3	on	on	ADP
cana-6113	169	4	pattern	pattern	NOUN
cana-6113	169	5	analysis	analysis	NOUN
cana-6113	169	6	and	and	CCONJ
cana-6113	169	7	machine	machine	NOUN
cana-6113	169	8	intelligence	intelligence	NOUN
cana-6113	169	9	(	(	PUNCT
cana-6113	169	10	tpami	tpami	NOUN
cana-6113	169	11	)	)	PUNCT
cana-6113	169	12	,	,	PUNCT
cana-6113	169	13	40(4	40(4	NUM
cana-6113	169	14	)	)	PUNCT
cana-6113	169	15	,	,	PUNCT
cana-6113	169	16	834–848	834–848	NUM
cana-6113	169	17	.	.	PUNCT
cana-6113	170	1	[	[	X
cana-6113	170	2	9	9	NUM
cana-6113	170	3	]	]	X
cana-6113	170	4	tan	tan	PROPN
cana-6113	170	5	,	,	PUNCT
cana-6113	170	6	m.	m.	NOUN
cana-6113	170	7	,	,	PUNCT
cana-6113	170	8	&	&	CCONJ
cana-6113	170	9	le	le	PROPN
cana-6113	170	10	,	,	PUNCT
cana-6113	170	11	q.	q.	PROPN
cana-6113	170	12	(	(	PUNCT
cana-6113	170	13	2019	2019	NUM
cana-6113	170	14	)	)	PUNCT
cana-6113	170	15	.	.	PUNCT
cana-6113	171	1	efficientnet	efficientnet	PROPN
cana-6113	171	2	:	:	PUNCT
cana-6113	171	3	rethinking	rethink	VERB
cana-6113	171	4	model	model	NOUN
cana-6113	171	5	scaling	scale	VERB
cana-6113	171	6	for	for	ADP
cana-6113	171	7	convolutional	convolutional	ADJ
cana-6113	171	8	neural	neural	ADJ
cana-6113	171	9	networks	network	NOUN
cana-6113	171	10	.	.	PUNCT
cana-6113	172	1	proceedings	proceeding	NOUN
cana-6113	172	2	of	of	ADP
cana-6113	172	3	the	the	DET
cana-6113	172	4	international	international	ADJ
cana-6113	172	5	conference	conference	NOUN
cana-6113	172	6	on	on	ADP
cana-6113	172	7	machine	machine	NOUN
cana-6113	172	8	learning	learning	NOUN
cana-6113	172	9	(	(	PUNCT
cana-6113	172	10	icml	icml	PROPN
cana-6113	172	11	)	)	PUNCT
cana-6113	172	12	,	,	PUNCT
cana-6113	172	13	6105–6114	6105–6114	NUM
cana-6113	172	14	.	.	PUNCT
cana-6113	173	1	communications	communication	NOUN
cana-6113	173	2	on	on	ADP
cana-6113	173	3	applied	apply	VERB
cana-6113	173	4	nonlinear	nonlinear	ADJ
cana-6113	173	5	analysis	analysis	NOUN
cana-6113	173	6	issn	issn	NOUN
cana-6113	173	7	:	:	PUNCT
cana-6113	173	8	1074	1074	NUM
cana-6113	173	9	-	-	PUNCT
cana-6113	173	10	133x	133x	NUM
cana-6113	173	11	vol	vol	VERB
cana-6113	173	12	32	32	NUM
cana-6113	173	13	no	no	NOUN
cana-6113	173	14	.	.	PUNCT
cana-6113	174	1	10s	10	NOUN
cana-6113	174	2	(	(	PUNCT
cana-6113	174	3	2025	2025	NUM
cana-6113	174	4	)	)	PUNCT
cana-6113	174	5	3412	3412	NUM
cana-6113	174	6	https://internationalpubls.com	https://internationalpubls.com	X
cana-6113	175	1	[	[	X
cana-6113	175	2	10	10	NUM
cana-6113	175	3	]	]	X
cana-6113	175	4	dosovitskiy	dosovitskiy	NOUN
cana-6113	175	5	,	,	PUNCT
cana-6113	175	6	a.	a.	PROPN
cana-6113	175	7	,	,	PUNCT
cana-6113	175	8	beyer	beyer	PROPN
cana-6113	175	9	,	,	PUNCT
cana-6113	175	10	l.	l.	PROPN
cana-6113	175	11	,	,	PUNCT
cana-6113	175	12	kolesnikov	kolesnikov	PROPN
cana-6113	175	13	,	,	PUNCT
cana-6113	175	14	a.	a.	NOUN
cana-6113	175	15	,	,	PUNCT
cana-6113	175	16	weissenborn	weissenborn	ADJ
cana-6113	175	17	,	,	PUNCT
cana-6113	175	18	d.	d.	PROPN
cana-6113	175	19	,	,	PUNCT
cana-6113	175	20	zhai	zhai	PROPN
cana-6113	175	21	,	,	PUNCT
cana-6113	175	22	x.	x.	PROPN
cana-6113	175	23	,	,	PUNCT
cana-6113	175	24	unterthiner	unterthiner	NOUN
cana-6113	175	25	,	,	PUNCT
cana-6113	175	26	t.	t.	PROPN
cana-6113	175	27	,	,	PUNCT
cana-6113	175	28	...	...	PUNCT
cana-6113	175	29	&	&	CCONJ
cana-6113	175	30	houlsby	houlsby	ADJ
cana-6113	175	31	,	,	PUNCT
cana-6113	175	32	n.	n.	NOUN
cana-6113	175	33	(	(	PUNCT
cana-6113	175	34	2020	2020	NUM
cana-6113	175	35	)	)	PUNCT
cana-6113	175	36	.	.	PUNCT
cana-6113	176	1	an	an	DET
cana-6113	176	2	image	image	NOUN
cana-6113	176	3	is	be	AUX
cana-6113	176	4	worth	worth	ADJ
cana-6113	176	5	16x16	16x16	NUM
cana-6113	176	6	words	word	NOUN
cana-6113	176	7	:	:	PUNCT
cana-6113	176	8	transformers	transformer	NOUN
cana-6113	176	9	for	for	ADP
cana-6113	176	10	image	image	NOUN
cana-6113	176	11	recognition	recognition	NOUN
cana-6113	176	12	at	at	ADP
cana-6113	176	13	scale	scale	NOUN
cana-6113	176	14	.	.	PUNCT
cana-6113	177	1	international	international	ADJ
cana-6113	177	2	conference	conference	NOUN
cana-6113	177	3	on	on	ADP
cana-6113	177	4	learning	learn	VERB
cana-6113	177	5	representations	representation	NOUN
cana-6113	177	6	(	(	PUNCT
cana-6113	177	7	iclr	iclr	NOUN
cana-6113	177	8	)	)	PUNCT
cana-6113	177	9	.	.	PUNCT
cana-6113	178	1	[	[	X
cana-6113	178	2	11	11	NUM
cana-6113	178	3	]	]	PUNCT
cana-6113	178	4	a.	a.	NOUN
cana-6113	178	5	krizhevsky	krizhevsky	PROPN
cana-6113	178	6	,	,	PUNCT
cana-6113	178	7	i.	i.	PROPN
cana-6113	178	8	sutskever	sutskever	PROPN
cana-6113	178	9	,	,	PUNCT
cana-6113	178	10	and	and	CCONJ
cana-6113	178	11	g.	g.	PROPN
cana-6113	178	12	e.	e.	PROPN
cana-6113	178	13	hinton	hinton	PROPN
cana-6113	178	14	,	,	PUNCT
cana-6113	178	15	"	"	PUNCT
cana-6113	178	16	imagenet	imagenet	ADJ
cana-6113	178	17	classification	classification	NOUN
cana-6113	178	18	with	with	ADP
cana-6113	178	19	deep	deep	ADJ
cana-6113	178	20	convolutional	convolutional	ADJ
cana-6113	178	21	neural	neural	ADJ
cana-6113	178	22	networks	network	NOUN
cana-6113	178	23	,	,	PUNCT
cana-6113	178	24	"	"	PUNCT
cana-6113	178	25	advances	advance	NOUN
cana-6113	178	26	in	in	ADP
cana-6113	178	27	neural	neural	ADJ
cana-6113	178	28	information	information	NOUN
cana-6113	178	29	processing	processing	NOUN
cana-6113	178	30	systems	system	NOUN
cana-6113	178	31	(	(	PUNCT
cana-6113	178	32	neurips	neurip	NOUN
cana-6113	178	33	)	)	PUNCT
cana-6113	178	34	,	,	PUNCT
cana-6113	178	35	vol	vol	NOUN
cana-6113	178	36	.	.	PROPN
cana-6113	178	37	25	25	NUM
cana-6113	178	38	,	,	PUNCT
cana-6113	178	39	pp	pp	ADJ
cana-6113	178	40	.	.	PUNCT
cana-6113	179	1	1097–1105	1097–1105	NUM
cana-6113	179	2	,	,	PUNCT
cana-6113	179	3	2012	2012	NUM
cana-6113	179	4	.	.	PUNCT
cana-6113	180	1	[	[	X
cana-6113	180	2	12	12	NUM
cana-6113	180	3	]	]	PUNCT
cana-6113	180	4	k.	k.	PROPN
cana-6113	180	5	simonyan	simonyan	PROPN
cana-6113	180	6	and	and	CCONJ
cana-6113	180	7	a.	a.	NOUN
cana-6113	180	8	zisserman	zisserman	PROPN
cana-6113	180	9	,	,	PUNCT
cana-6113	180	10	"	"	PUNCT
cana-6113	180	11	very	very	ADV
cana-6113	180	12	deep	deep	ADJ
cana-6113	180	13	convolutional	convolutional	ADJ
cana-6113	180	14	networks	network	NOUN
cana-6113	180	15	for	for	ADP
cana-6113	180	16	large	large	ADJ
cana-6113	180	17	-	-	PUNCT
cana-6113	180	18	scale	scale	NOUN
cana-6113	180	19	image	image	NOUN
cana-6113	180	20	recognition	recognition	NOUN
cana-6113	180	21	,	,	PUNCT
cana-6113	180	22	"	"	PUNCT
cana-6113	180	23	arxiv	arxiv	PROPN
cana-6113	180	24	preprint	preprint	NOUN
cana-6113	180	25	arxiv:1409.1556	arxiv:1409.1556	NOUN
cana-6113	180	26	,	,	PUNCT
cana-6113	180	27	2014	2014	NUM
cana-6113	180	28	.	.	PUNCT
cana-6113	181	1	[	[	X
cana-6113	181	2	13	13	NUM
cana-6113	181	3	]	]	X
cana-6113	181	4	c.	c.	NOUN
cana-6113	181	5	szegedy	szegedy	PROPN
cana-6113	181	6	,	,	PUNCT
cana-6113	181	7	v.	v.	PROPN
cana-6113	181	8	vanhoucke	vanhoucke	PROPN
cana-6113	181	9	,	,	PUNCT
cana-6113	181	10	s.	s.	PROPN
cana-6113	181	11	ioffe	ioffe	PROPN
cana-6113	181	12	,	,	PUNCT
cana-6113	181	13	j.	j.	PROPN
cana-6113	181	14	shlens	shlens	PROPN
cana-6113	181	15	,	,	PUNCT
cana-6113	181	16	and	and	CCONJ
cana-6113	181	17	z.	z.	PROPN
cana-6113	181	18	wojna	wojna	PROPN
cana-6113	181	19	,	,	PUNCT
cana-6113	181	20	"	"	PUNCT
cana-6113	181	21	rethinking	rethink	VERB
cana-6113	181	22	the	the	DET
cana-6113	181	23	inception	inception	ADJ
cana-6113	181	24	architecture	architecture	NOUN
cana-6113	181	25	for	for	ADP
cana-6113	181	26	computer	computer	NOUN
cana-6113	181	27	vision	vision	NOUN
cana-6113	181	28	,	,	PUNCT
cana-6113	181	29	"	"	PUNCT
cana-6113	181	30	in	in	ADP
cana-6113	181	31	proc	proc	NOUN
cana-6113	181	32	.	.	PUNCT
cana-6113	182	1	ieee	ieee	NOUN
cana-6113	182	2	conf	conf	NOUN
cana-6113	182	3	.	.	PUNCT
cana-6113	183	1	comput	comput	NOUN
cana-6113	183	2	.	.	PUNCT
cana-6113	184	1	vis	vis	X
cana-6113	184	2	.	.	X
cana-6113	184	3	pattern	pattern	NOUN
cana-6113	184	4	recognit	recognit	VERB
cana-6113	184	5	.	.	PUNCT
cana-6113	185	1	(	(	PUNCT
cana-6113	185	2	cvpr	cvpr	NOUN
cana-6113	185	3	)	)	PUNCT
cana-6113	185	4	,	,	PUNCT
cana-6113	185	5	2016	2016	NUM
cana-6113	185	6	,	,	PUNCT
cana-6113	185	7	pp	pp	ADJ
cana-6113	185	8	.	.	PUNCT
cana-6113	186	1	2818–2826	2818–2826	NUM
cana-6113	186	2	.	.	PUNCT
cana-6113	187	1	[	[	X
cana-6113	187	2	14	14	NUM
cana-6113	187	3	]	]	PUNCT
cana-6113	187	4	z.	z.	PROPN
cana-6113	187	5	alyafeai	alyafeai	PROPN
cana-6113	187	6	and	and	CCONJ
cana-6113	187	7	l.	l.	PROPN
cana-6113	187	8	ghouti	ghouti	PROPN
cana-6113	187	9	,	,	PUNCT
cana-6113	187	10	"	"	PUNCT
cana-6113	187	11	a	a	DET
cana-6113	187	12	fully	fully	ADV
cana-6113	187	13	-	-	PUNCT
cana-6113	187	14	automated	automate	VERB
cana-6113	187	15	deep	deep	ADJ
cana-6113	187	16	learning	learning	NOUN
cana-6113	187	17	pipeline	pipeline	NOUN
cana-6113	187	18	for	for	ADP
cana-6113	187	19	cervical	cervical	ADJ
cana-6113	187	20	cancer	cancer	NOUN
cana-6113	187	21	classification	classification	NOUN
cana-6113	187	22	,	,	PUNCT
cana-6113	187	23	"	"	PUNCT
cana-6113	187	24	expert	expert	NOUN
cana-6113	187	25	systems	system	NOUN
cana-6113	187	26	with	with	ADP
cana-6113	187	27	applications	application	NOUN
cana-6113	187	28	,	,	PUNCT
cana-6113	187	29	vol	vol	NOUN
cana-6113	187	30	.	.	PROPN
cana-6113	187	31	141	141	NUM
cana-6113	187	32	,	,	PUNCT
cana-6113	187	33	p.	p.	NOUN
cana-6113	187	34	112951	112951	NUM
cana-6113	187	35	,	,	PUNCT
cana-6113	187	36	2020	2020	NUM
cana-6113	187	37	.	.	PUNCT
cana-6113	188	1	[	[	X
cana-6113	188	2	15	15	NUM
cana-6113	188	3	]	]	X
cana-6113	188	4	k.	k.	PROPN
cana-6113	188	5	matsuo	matsuo	PROPN
cana-6113	188	6	,	,	PUNCT
cana-6113	188	7	s.	s.	PROPN
cana-6113	188	8	purushotham	purushotham	PROPN
cana-6113	188	9	,	,	PUNCT
cana-6113	188	10	b.	b.	PROPN
cana-6113	188	11	jiang	jiang	PROPN
cana-6113	188	12	,	,	PUNCT
cana-6113	188	13	r.	r.	PROPN
cana-6113	188	14	s.	s.	PROPN
cana-6113	188	15	mandelbaum	mandelbaum	PROPN
cana-6113	188	16	,	,	PUNCT
cana-6113	188	17	t.	t.	PROPN
cana-6113	188	18	takiuchi	takiuchi	PROPN
cana-6113	188	19	,	,	PUNCT
cana-6113	188	20	y.	y.	PROPN
cana-6113	188	21	liu	liu	PROPN
cana-6113	188	22	,	,	PUNCT
cana-6113	188	23	and	and	CCONJ
cana-6113	188	24	l.	l.	PROPN
cana-6113	188	25	d.	d.	PROPN
cana-6113	188	26	roman	roman	PROPN
cana-6113	188	27	,	,	PUNCT
cana-6113	188	28	"	"	PUNCT
cana-6113	188	29	survival	survival	NOUN
cana-6113	188	30	outcome	outcome	NOUN
cana-6113	188	31	prediction	prediction	NOUN
cana-6113	188	32	in	in	ADP
cana-6113	188	33	cervical	cervical	ADJ
cana-6113	188	34	cancer	cancer	NOUN
cana-6113	188	35	:	:	PUNCT
cana-6113	188	36	cox	cox	PROPN
cana-6113	188	37	models	model	NOUN
cana-6113	188	38	vs	vs	ADP
cana-6113	188	39	deeplearning	deeplearning	NOUN
cana-6113	188	40	model	model	NOUN
cana-6113	188	41	,	,	PUNCT
cana-6113	188	42	"	"	PUNCT
cana-6113	188	43	american	american	ADJ
cana-6113	188	44	journal	journal	NOUN
cana-6113	188	45	of	of	ADP
cana-6113	188	46	obstetrics	obstetric	NOUN
cana-6113	188	47	and	and	CCONJ
cana-6113	188	48	gynecology	gynecology	NOUN
cana-6113	188	49	,	,	PUNCT
cana-6113	188	50	vol	vol	NOUN
cana-6113	188	51	.	.	NOUN
cana-6113	188	52	220	220	NUM
cana-6113	188	53	,	,	PUNCT
cana-6113	188	54	no	no	INTJ
cana-6113	188	55	.	.	NOUN
cana-6113	188	56	4	4	NUM
cana-6113	188	57	,	,	PUNCT
cana-6113	188	58	pp	pp	ADJ
cana-6113	188	59	.	.	PUNCT
cana-6113	189	1	381	381	NUM
cana-6113	189	2	-	-	PUNCT
cana-6113	189	3	e1	e1	NOUN
cana-6113	189	4	,	,	PUNCT
cana-6113	189	5	2019	2019	NUM
cana-6113	189	6	.	.	PUNCT
cana-6113	190	1	[	[	X
cana-6113	190	2	16	16	X
cana-6113	190	3	]	]	X
cana-6113	190	4	v.	v.	CCONJ
cana-6113	190	5	chandran	chandran	PROPN
cana-6113	190	6	,	,	PUNCT
cana-6113	190	7	m.	m.	PROPN
cana-6113	190	8	g.	g.	PROPN
cana-6113	190	9	sumithra	sumithra	PROPN
cana-6113	190	10	,	,	PUNCT
cana-6113	190	11	a.	a.	PROPN
cana-6113	190	12	karthick	karthick	PROPN
cana-6113	190	13	,	,	PUNCT
cana-6113	190	14	t.	t.	PROPN
cana-6113	190	15	george	george	PROPN
cana-6113	190	16	,	,	PUNCT
cana-6113	190	17	m.	m.	NOUN
cana-6113	190	18	deivakani	deivakani	PROPN
cana-6113	190	19	,	,	PUNCT
cana-6113	190	20	b.	b.	PROPN
cana-6113	190	21	elakkiya	elakkiya	PROPN
cana-6113	190	22	,	,	PUNCT
cana-6113	190	23	...	...	PUNCT
cana-6113	190	24	and	and	CCONJ
cana-6113	190	25	s.	s.	PROPN
cana-6113	190	26	manoharan	manoharan	PROPN
cana-6113	190	27	,	,	PUNCT
cana-6113	190	28	"	"	PUNCT
cana-6113	190	29	diagnosis	diagnosis	NOUN
cana-6113	190	30	of	of	ADP
cana-6113	190	31	cervical	cervical	ADJ
cana-6113	190	32	cancer	cancer	NOUN
cana-6113	190	33	based	base	VERB
cana-6113	190	34	on	on	ADP
cana-6113	190	35	ensemble	ensemble	ADJ
cana-6113	190	36	deep	deep	ADJ
cana-6113	190	37	learning	learning	NOUN
cana-6113	190	38	network	network	NOUN
cana-6113	190	39	using	use	VERB
cana-6113	190	40	colposcopy	colposcopy	NOUN
cana-6113	190	41	images	image	NOUN
cana-6113	190	42	,	,	PUNCT
cana-6113	190	43	"	"	PUNCT
cana-6113	190	44	biomed	biome	VERB
cana-6113	190	45	research	research	NOUN
cana-6113	190	46	international	international	NOUN
cana-6113	190	47	,	,	PUNCT
cana-6113	190	48	2021	2021	NUM
cana-6113	190	49	.	.	PUNCT
cana-6113	191	1	[	[	X
cana-6113	191	2	17	17	NUM
cana-6113	191	3	]	]	PUNCT
cana-6113	191	4	x.	x.	PROPN
cana-6113	191	5	jiang	jiang	PROPN
cana-6113	191	6	,	,	PUNCT
cana-6113	191	7	j.	j.	PROPN
cana-6113	191	8	li	li	PROPN
cana-6113	191	9	,	,	PUNCT
cana-6113	191	10	y.	y.	PROPN
cana-6113	191	11	kan	kan	PROPN
cana-6113	191	12	,	,	PUNCT
cana-6113	191	13	t.	t.	PROPN
cana-6113	191	14	yu	yu	PROPN
cana-6113	191	15	,	,	PUNCT
cana-6113	191	16	s.	s.	PROPN
cana-6113	191	17	chang	chang	PROPN
cana-6113	191	18	,	,	PUNCT
cana-6113	191	19	x.	x.	PROPN
cana-6113	191	20	sha	sha	PROPN
cana-6113	191	21	,	,	PUNCT
cana-6113	191	22	...	...	PUNCT
cana-6113	191	23	and	and	CCONJ
cana-6113	191	24	s.	s.	PROPN
cana-6113	191	25	wang	wang	PROPN
cana-6113	191	26	,	,	PUNCT
cana-6113	191	27	"	"	PUNCT
cana-6113	191	28	mri	mri	NOUN
cana-6113	191	29	-	-	PUNCT
cana-6113	191	30	based	base	VERB
cana-6113	191	31	radiomics	radiomic	NOUN
cana-6113	191	32	approach	approach	VERB
cana-6113	191	33	with	with	ADP
cana-6113	191	34	deep	deep	ADJ
cana-6113	191	35	learning	learning	NOUN
cana-6113	191	36	for	for	ADP
cana-6113	191	37	prediction	prediction	NOUN
cana-6113	191	38	of	of	ADP
cana-6113	191	39	vessel	vessel	NOUN
cana-6113	191	40	invasion	invasion	NOUN
cana-6113	191	41	in	in	ADP
cana-6113	191	42	early	early	ADJ
cana-6113	191	43	-	-	PUNCT
cana-6113	191	44	stage	stage	NOUN
cana-6113	191	45	cervical	cervical	ADJ
cana-6113	191	46	cancer	cancer	NOUN
cana-6113	191	47	,	,	PUNCT
cana-6113	191	48	"	"	PUNCT
cana-6113	191	49	ieee	ieee	NOUN
cana-6113	191	50	/	/	SYM
cana-6113	191	51	acm	acm	NOUN
cana-6113	191	52	transactions	transaction	NOUN
cana-6113	191	53	on	on	ADP
cana-6113	191	54	computational	computational	ADJ
cana-6113	191	55	biology	biology	NOUN
cana-6113	191	56	and	and	CCONJ
cana-6113	191	57	bioinformatics	bioinformatics	NOUN
cana-6113	191	58	,	,	PUNCT
cana-6113	191	59	vol	vol	NOUN
cana-6113	191	60	.	.	PROPN
cana-6113	191	61	18	18	NUM
cana-6113	191	62	,	,	PUNCT
cana-6113	191	63	no	no	INTJ
cana-6113	191	64	.	.	NOUN
cana-6113	191	65	3	3	NUM
cana-6113	191	66	,	,	PUNCT
cana-6113	191	67	pp	pp	ADJ
cana-6113	191	68	.	.	PUNCT
cana-6113	192	1	995–1002	995–1002	NUM
cana-6113	192	2	,	,	PUNCT
cana-6113	192	3	2020	2020	NUM
cana-6113	192	4	.	.	PUNCT