id	sid	tid	token	lemma	pos
fcis-18739	1	1	frontiers	frontier	NOUN
fcis-18739	1	2	in	in	ADP
fcis-18739	1	3	computing	computing	NOUN
fcis-18739	1	4	and	and	CCONJ
fcis-18739	1	5	intelligent	intelligent	ADJ
fcis-18739	1	6	systems	system	NOUN
fcis-18739	1	7	issn	issn	VERB
fcis-18739	1	8	:	:	PUNCT
fcis-18739	1	9	2832	2832	NUM
fcis-18739	1	10	-	-	SYM
fcis-18739	1	11	6024	6024	NUM
fcis-18739	1	12	|	|	NOUN
fcis-18739	1	13	vol	vol	NOUN
fcis-18739	1	14	.	.	PROPN
fcis-18739	2	1	7	7	NUM
fcis-18739	2	2	,	,	PUNCT
fcis-18739	2	3	no	no	INTJ
fcis-18739	2	4	.	.	NOUN
fcis-18739	2	5	3	3	NUM
fcis-18739	2	6	,	,	PUNCT
fcis-18739	2	7	2024	2024	NUM
fcis-18739	2	8	17	17	NUM
fcis-18739	2	9	application	application	NOUN
fcis-18739	2	10	of	of	ADP
fcis-18739	2	11	image	image	NOUN
fcis-18739	2	12	segmentation	segmentation	NOUN
fcis-18739	2	13	algorithms	algorithm	NOUN
fcis-18739	2	14	in	in	ADP
fcis-18739	2	15	computer	computer	NOUN
fcis-18739	2	16	vision	vision	NOUN
fcis-18739	3	1	yuhang	yuhang	PROPN
fcis-18739	3	2	xu	xu	PROPN
fcis-18739	4	1	*	*	PUNCT
fcis-18739	4	2	school	school	NOUN
fcis-18739	4	3	of	of	ADP
fcis-18739	4	4	computer	computer	NOUN
fcis-18739	4	5	and	and	CCONJ
fcis-18739	4	6	software	software	NOUN
fcis-18739	4	7	,	,	PUNCT
fcis-18739	4	8	university	university	NOUN
fcis-18739	4	9	of	of	ADP
fcis-18739	4	10	science	science	NOUN
fcis-18739	4	11	and	and	CCONJ
fcis-18739	4	12	technology	technology	NOUN
fcis-18739	4	13	liaoning	liaoning	NOUN
fcis-18739	4	14	,	,	PUNCT
fcis-18739	4	15	anshan	anshan	PROPN
fcis-18739	4	16	liaoning	liaoning	NOUN
fcis-18739	4	17	,	,	PUNCT
fcis-18739	4	18	114000	114000	NUM
fcis-18739	4	19	,	,	PUNCT
fcis-18739	4	20	china	china	PROPN
fcis-18739	4	21	*	*	PUNCT
fcis-18739	4	22	corresponding	correspond	VERB
fcis-18739	4	23	author	author	NOUN
fcis-18739	4	24	email	email	NOUN
fcis-18739	4	25	:	:	PUNCT
fcis-18739	5	1	xuy21955@gmail.com	xuy21955@gmail.com	X
fcis-18739	5	2	abstract	abstract	PROPN
fcis-18739	5	3	:	:	PUNCT
fcis-18739	5	4	in	in	ADP
fcis-18739	5	5	the	the	DET
fcis-18739	5	6	field	field	NOUN
fcis-18739	5	7	of	of	ADP
fcis-18739	5	8	computer	computer	NOUN
fcis-18739	5	9	vision	vision	NOUN
fcis-18739	5	10	(	(	PUNCT
fcis-18739	5	11	cv	cv	PROPN
fcis-18739	5	12	)	)	PUNCT
fcis-18739	5	13	,	,	PUNCT
fcis-18739	5	14	image	image	NOUN
fcis-18739	5	15	segmentation	segmentation	NOUN
fcis-18739	5	16	technology	technology	NOUN
fcis-18739	5	17	,	,	PUNCT
fcis-18739	5	18	as	as	ADP
fcis-18739	5	19	a	a	DET
fcis-18739	5	20	fundamental	fundamental	ADJ
fcis-18739	5	21	part	part	NOUN
fcis-18739	5	22	,	,	PUNCT
fcis-18739	5	23	has	have	VERB
fcis-18739	5	24	a	a	DET
fcis-18739	5	25	crucial	crucial	ADJ
fcis-18739	5	26	impact	impact	NOUN
fcis-18739	5	27	on	on	ADP
fcis-18739	5	28	the	the	DET
fcis-18739	5	29	accuracy	accuracy	NOUN
fcis-18739	5	30	of	of	ADP
fcis-18739	5	31	subsequent	subsequent	ADJ
fcis-18739	5	32	image	image	NOUN
fcis-18739	5	33	processing	processing	NOUN
fcis-18739	5	34	tasks	task	NOUN
fcis-18739	5	35	.	.	PUNCT
fcis-18739	6	1	image	image	NOUN
fcis-18739	6	2	segmentation	segmentation	NOUN
fcis-18739	6	3	is	be	AUX
fcis-18739	6	4	not	not	PART
fcis-18739	6	5	only	only	ADV
fcis-18739	6	6	a	a	DET
fcis-18739	6	7	crucial	crucial	ADJ
fcis-18739	6	8	transitional	transitional	ADJ
fcis-18739	6	9	step	step	NOUN
fcis-18739	6	10	from	from	ADP
fcis-18739	6	11	image	image	NOUN
fcis-18739	6	12	processing	processing	NOUN
fcis-18739	6	13	to	to	PART
fcis-18739	6	14	image	image	VERB
fcis-18739	6	15	analysis	analysis	NOUN
fcis-18739	6	16	,	,	PUNCT
fcis-18739	6	17	but	but	CCONJ
fcis-18739	6	18	also	also	ADV
fcis-18739	6	19	a	a	DET
fcis-18739	6	20	hot	hot	ADJ
fcis-18739	6	21	and	and	CCONJ
fcis-18739	6	22	difficult	difficult	ADJ
fcis-18739	6	23	research	research	NOUN
fcis-18739	6	24	topic	topic	NOUN
fcis-18739	6	25	in	in	ADP
fcis-18739	6	26	the	the	DET
fcis-18739	6	27	field	field	NOUN
fcis-18739	6	28	of	of	ADP
fcis-18739	6	29	cv	cv	PROPN
fcis-18739	6	30	.	.	PUNCT
fcis-18739	7	1	although	although	SCONJ
fcis-18739	7	2	significant	significant	ADJ
fcis-18739	7	3	progress	progress	NOUN
fcis-18739	7	4	has	have	AUX
fcis-18739	7	5	been	be	AUX
fcis-18739	7	6	made	make	VERB
fcis-18739	7	7	in	in	ADP
fcis-18739	7	8	the	the	DET
fcis-18739	7	9	research	research	NOUN
fcis-18739	7	10	of	of	ADP
fcis-18739	7	11	image	image	NOUN
fcis-18739	7	12	segmentation	segmentation	NOUN
fcis-18739	7	13	algorithms	algorithm	NOUN
fcis-18739	7	14	,	,	PUNCT
fcis-18739	7	15	existing	exist	VERB
fcis-18739	7	16	segmentation	segmentation	NOUN
fcis-18739	7	17	algorithms	algorithm	NOUN
fcis-18739	7	18	may	may	AUX
fcis-18739	7	19	still	still	ADV
fcis-18739	7	20	face	face	VERB
fcis-18739	7	21	challenges	challenge	NOUN
fcis-18739	7	22	in	in	ADP
fcis-18739	7	23	certain	certain	ADJ
fcis-18739	7	24	specific	specific	ADJ
fcis-18739	7	25	scenarios	scenario	NOUN
fcis-18739	7	26	due	due	ADP
fcis-18739	7	27	to	to	ADP
fcis-18739	7	28	the	the	DET
fcis-18739	7	29	complexity	complexity	NOUN
fcis-18739	7	30	and	and	CCONJ
fcis-18739	7	31	diversity	diversity	NOUN
fcis-18739	7	32	of	of	ADP
fcis-18739	7	33	images	image	NOUN
fcis-18739	7	34	,	,	PUNCT
fcis-18739	7	35	making	make	VERB
fcis-18739	7	36	it	it	PRON
fcis-18739	7	37	difficult	difficult	ADJ
fcis-18739	7	38	to	to	PART
fcis-18739	7	39	achieve	achieve	VERB
fcis-18739	7	40	ideal	ideal	ADJ
fcis-18739	7	41	segmentation	segmentation	NOUN
fcis-18739	7	42	results	result	NOUN
fcis-18739	7	43	.	.	PUNCT
fcis-18739	8	1	in	in	ADP
fcis-18739	8	2	recent	recent	ADJ
fcis-18739	8	3	years	year	NOUN
fcis-18739	8	4	,	,	PUNCT
fcis-18739	8	5	the	the	DET
fcis-18739	8	6	rapid	rapid	ADJ
fcis-18739	8	7	development	development	NOUN
fcis-18739	8	8	of	of	ADP
fcis-18739	8	9	deep	deep	ADJ
fcis-18739	8	10	learning	learning	NOUN
fcis-18739	8	11	(	(	PUNCT
fcis-18739	8	12	dl	dl	NOUN
fcis-18739	8	13	)	)	PUNCT
fcis-18739	8	14	technology	technology	NOUN
fcis-18739	8	15	has	have	AUX
fcis-18739	8	16	brought	bring	VERB
fcis-18739	8	17	new	new	ADJ
fcis-18739	8	18	breakthroughs	breakthrough	NOUN
fcis-18739	8	19	to	to	ADP
fcis-18739	8	20	the	the	DET
fcis-18739	8	21	field	field	NOUN
fcis-18739	8	22	of	of	ADP
fcis-18739	8	23	image	image	NOUN
fcis-18739	8	24	segmentation	segmentation	NOUN
fcis-18739	8	25	.	.	PUNCT
fcis-18739	9	1	dl	dl	PROPN
fcis-18739	9	2	models	model	NOUN
fcis-18739	9	3	,	,	PUNCT
fcis-18739	9	4	especially	especially	ADV
fcis-18739	9	5	convolutional	convolutional	ADJ
fcis-18739	9	6	neural	neural	ADJ
fcis-18739	9	7	networks	network	NOUN
fcis-18739	9	8	(	(	PUNCT
fcis-18739	9	9	cnns	cnns	PROPN
fcis-18739	9	10	)	)	PUNCT
fcis-18739	9	11	,	,	PUNCT
fcis-18739	9	12	can	can	AUX
fcis-18739	9	13	capture	capture	VERB
fcis-18739	9	14	semantic	semantic	ADJ
fcis-18739	9	15	information	information	NOUN
fcis-18739	9	16	of	of	ADP
fcis-18739	9	17	images	image	NOUN
fcis-18739	9	18	more	more	ADV
fcis-18739	9	19	accurately	accurately	ADV
fcis-18739	9	20	by	by	ADP
fcis-18739	9	21	automatically	automatically	ADV
fcis-18739	9	22	learning	learn	VERB
fcis-18739	9	23	feature	feature	NOUN
fcis-18739	9	24	representations	representation	NOUN
fcis-18739	9	25	in	in	ADP
fcis-18739	9	26	images	image	NOUN
fcis-18739	9	27	,	,	PUNCT
fcis-18739	9	28	thereby	thereby	ADV
fcis-18739	9	29	achieving	achieve	VERB
fcis-18739	9	30	more	more	ADV
fcis-18739	9	31	precise	precise	ADJ
fcis-18739	9	32	image	image	NOUN
fcis-18739	9	33	segmentation	segmentation	NOUN
fcis-18739	9	34	.	.	PUNCT
fcis-18739	10	1	this	this	DET
fcis-18739	10	2	article	article	NOUN
fcis-18739	10	3	delves	delve	VERB
fcis-18739	10	4	into	into	ADP
fcis-18739	10	5	the	the	DET
fcis-18739	10	6	research	research	NOUN
fcis-18739	10	7	and	and	CCONJ
fcis-18739	10	8	application	application	NOUN
fcis-18739	10	9	of	of	ADP
fcis-18739	10	10	image	image	NOUN
fcis-18739	10	11	segmentation	segmentation	NOUN
fcis-18739	10	12	algorithms	algorithm	NOUN
fcis-18739	10	13	in	in	ADP
fcis-18739	10	14	cv	cv	PROPN
fcis-18739	10	15	,	,	PUNCT
fcis-18739	10	16	with	with	ADP
fcis-18739	10	17	a	a	DET
fcis-18739	10	18	focus	focus	NOUN
fcis-18739	10	19	on	on	ADP
fcis-18739	10	20	the	the	DET
fcis-18739	10	21	application	application	NOUN
fcis-18739	10	22	of	of	ADP
fcis-18739	10	23	dl	dl	PROPN
fcis-18739	10	24	in	in	ADP
fcis-18739	10	25	the	the	DET
fcis-18739	10	26	field	field	NOUN
fcis-18739	10	27	of	of	ADP
fcis-18739	10	28	image	image	NOUN
fcis-18739	10	29	segmentation	segmentation	NOUN
fcis-18739	10	30	.	.	PUNCT
fcis-18739	11	1	with	with	ADP
fcis-18739	11	2	the	the	DET
fcis-18739	11	3	continuous	continuous	ADJ
fcis-18739	11	4	development	development	NOUN
fcis-18739	11	5	of	of	ADP
fcis-18739	11	6	advanced	advanced	ADJ
fcis-18739	11	7	technologies	technology	NOUN
fcis-18739	11	8	such	such	ADJ
fcis-18739	11	9	as	as	ADP
fcis-18739	11	10	dl	dl	PROPN
fcis-18739	11	11	,	,	PUNCT
fcis-18739	11	12	it	it	PRON
fcis-18739	11	13	is	be	AUX
fcis-18739	11	14	believed	believe	VERB
fcis-18739	11	15	that	that	SCONJ
fcis-18739	11	16	image	image	NOUN
fcis-18739	11	17	segmentation	segmentation	NOUN
fcis-18739	11	18	technology	technology	NOUN
fcis-18739	11	19	will	will	AUX
fcis-18739	11	20	play	play	VERB
fcis-18739	11	21	a	a	DET
fcis-18739	11	22	greater	great	ADJ
fcis-18739	11	23	role	role	NOUN
fcis-18739	11	24	in	in	ADP
fcis-18739	11	25	more	more	ADJ
fcis-18739	11	26	fields	field	NOUN
fcis-18739	11	27	in	in	ADP
fcis-18739	11	28	the	the	DET
fcis-18739	11	29	future	future	NOUN
fcis-18739	11	30	.	.	PUNCT
fcis-18739	12	1	keywords	keyword	NOUN
fcis-18739	12	2	:	:	PUNCT
fcis-18739	12	3	computer	computer	NOUN
fcis-18739	12	4	vision	vision	NOUN
fcis-18739	12	5	;	;	PUNCT
fcis-18739	12	6	image	image	NOUN
fcis-18739	12	7	segmentation	segmentation	NOUN
fcis-18739	12	8	algorithms	algorithm	NOUN
fcis-18739	12	9	;	;	PUNCT
fcis-18739	12	10	research	research	NOUN
fcis-18739	12	11	and	and	CCONJ
fcis-18739	12	12	application	application	NOUN
fcis-18739	12	13	.	.	PUNCT
fcis-18739	13	1	1	1	X
fcis-18739	13	2	.	.	X
fcis-18739	13	3	introduction	introduction	NOUN
fcis-18739	13	4	image	image	NOUN
fcis-18739	13	5	segmentation	segmentation	NOUN
fcis-18739	13	6	is	be	AUX
fcis-18739	13	7	one	one	NUM
fcis-18739	13	8	of	of	ADP
fcis-18739	13	9	the	the	DET
fcis-18739	13	10	fundamental	fundamental	ADJ
fcis-18739	13	11	technologies	technology	NOUN
fcis-18739	13	12	in	in	ADP
fcis-18739	13	13	the	the	DET
fcis-18739	13	14	field	field	NOUN
fcis-18739	13	15	of	of	ADP
fcis-18739	13	16	cv	cv	PROPN
fcis-18739	13	17	,	,	PUNCT
fcis-18739	13	18	and	and	CCONJ
fcis-18739	13	19	it	it	PRON
fcis-18739	13	20	is	be	AUX
fcis-18739	13	21	also	also	ADV
fcis-18739	13	22	the	the	DET
fcis-18739	13	23	most	most	ADV
fcis-18739	13	24	basic	basic	ADJ
fcis-18739	13	25	method	method	NOUN
fcis-18739	13	26	for	for	ADP
fcis-18739	13	27	early	early	ADJ
fcis-18739	13	28	processing	processing	NOUN
fcis-18739	13	29	of	of	ADP
fcis-18739	13	30	image	image	NOUN
fcis-18739	13	31	data	datum	NOUN
fcis-18739	13	32	[	[	X
fcis-18739	13	33	1	1	NUM
fcis-18739	13	34	]	]	PUNCT
fcis-18739	13	35	.	.	PUNCT
fcis-18739	14	1	in	in	ADP
fcis-18739	14	2	cv	cv	PROPN
fcis-18739	14	3	tasks	task	NOUN
fcis-18739	14	4	,	,	PUNCT
fcis-18739	14	5	image	image	NOUN
fcis-18739	14	6	segmentation	segmentation	NOUN
fcis-18739	14	7	plays	play	VERB
fcis-18739	14	8	a	a	DET
fcis-18739	14	9	crucial	crucial	ADJ
fcis-18739	14	10	role	role	NOUN
fcis-18739	14	11	in	in	ADP
fcis-18739	14	12	dividing	divide	VERB
fcis-18739	14	13	images	image	NOUN
fcis-18739	14	14	into	into	ADP
fcis-18739	14	15	multiple	multiple	ADJ
fcis-18739	14	16	regions	region	NOUN
fcis-18739	14	17	with	with	ADP
fcis-18739	14	18	similar	similar	ADJ
fcis-18739	14	19	properties	property	NOUN
fcis-18739	14	20	,	,	PUNCT
fcis-18739	14	21	which	which	PRON
fcis-18739	14	22	facilitates	facilitate	VERB
fcis-18739	14	23	subsequent	subsequent	ADJ
fcis-18739	14	24	tasks	task	NOUN
fcis-18739	14	25	such	such	ADJ
fcis-18739	14	26	as	as	ADP
fcis-18739	14	27	image	image	NOUN
fcis-18739	14	28	analysis	analysis	NOUN
fcis-18739	14	29	,	,	PUNCT
fcis-18739	14	30	object	object	NOUN
fcis-18739	14	31	detection	detection	NOUN
fcis-18739	14	32	,	,	PUNCT
fcis-18739	14	33	and	and	CCONJ
fcis-18739	14	34	scene	scene	VERB
fcis-18739	14	35	understanding	understand	VERB
fcis-18739	14	36	[	[	X
fcis-18739	14	37	2	2	NUM
fcis-18739	14	38	]	]	PUNCT
fcis-18739	14	39	.	.	PUNCT
fcis-18739	15	1	high	high	ADJ
fcis-18739	15	2	quality	quality	NOUN
fcis-18739	15	3	image	image	NOUN
fcis-18739	15	4	segmentation	segmentation	NOUN
fcis-18739	15	5	results	result	NOUN
fcis-18739	15	6	can	can	AUX
fcis-18739	15	7	provide	provide	VERB
fcis-18739	15	8	more	more	ADV
fcis-18739	15	9	accurate	accurate	ADJ
fcis-18739	15	10	information	information	NOUN
fcis-18739	15	11	for	for	ADP
fcis-18739	15	12	other	other	ADJ
fcis-18739	15	13	cv	cv	PROPN
fcis-18739	15	14	tasks	task	NOUN
fcis-18739	15	15	,	,	PUNCT
fcis-18739	15	16	thereby	thereby	ADV
fcis-18739	15	17	improving	improve	VERB
fcis-18739	15	18	the	the	DET
fcis-18739	15	19	performance	performance	NOUN
fcis-18739	15	20	of	of	ADP
fcis-18739	15	21	the	the	DET
fcis-18739	15	22	entire	entire	ADJ
fcis-18739	15	23	system	system	NOUN
fcis-18739	15	24	[	[	X
fcis-18739	15	25	3	3	NUM
fcis-18739	15	26	]	]	PUNCT
fcis-18739	15	27	.	.	PUNCT
fcis-18739	16	1	therefore	therefore	ADV
fcis-18739	16	2	,	,	PUNCT
fcis-18739	16	3	studying	study	VERB
fcis-18739	16	4	efficient	efficient	ADJ
fcis-18739	16	5	and	and	CCONJ
fcis-18739	16	6	robust	robust	ADJ
fcis-18739	16	7	image	image	NOUN
fcis-18739	16	8	segmentation	segmentation	NOUN
fcis-18739	16	9	methods	method	NOUN
fcis-18739	16	10	has	have	VERB
fcis-18739	16	11	significant	significant	ADJ
fcis-18739	16	12	practical	practical	ADJ
fcis-18739	16	13	significance	significance	NOUN
fcis-18739	16	14	and	and	CCONJ
fcis-18739	16	15	application	application	NOUN
fcis-18739	16	16	value	value	NOUN
fcis-18739	16	17	.	.	PUNCT
fcis-18739	17	1	with	with	ADP
fcis-18739	17	2	the	the	DET
fcis-18739	17	3	continuous	continuous	ADJ
fcis-18739	17	4	development	development	NOUN
fcis-18739	17	5	of	of	ADP
fcis-18739	17	6	cv	cv	PROPN
fcis-18739	17	7	technology	technology	NOUN
fcis-18739	17	8	,	,	PUNCT
fcis-18739	17	9	image	image	NOUN
fcis-18739	17	10	segmentation	segmentation	NOUN
fcis-18739	17	11	algorithms	algorithm	NOUN
fcis-18739	17	12	have	have	AUX
fcis-18739	17	13	also	also	ADV
fcis-18739	17	14	undergone	undergo	VERB
fcis-18739	17	15	an	an	DET
fcis-18739	17	16	evolutionary	evolutionary	ADJ
fcis-18739	17	17	process	process	NOUN
fcis-18739	17	18	from	from	ADP
fcis-18739	17	19	simple	simple	ADJ
fcis-18739	17	20	to	to	ADP
fcis-18739	17	21	complex	complex	ADJ
fcis-18739	17	22	,	,	PUNCT
fcis-18739	17	23	from	from	ADP
fcis-18739	17	24	single	single	ADJ
fcis-18739	17	25	to	to	PART
fcis-18739	17	26	diverse	diverse	VERB
fcis-18739	17	27	[	[	X
fcis-18739	17	28	4	4	NUM
fcis-18739	17	29	]	]	PUNCT
fcis-18739	17	30	.	.	PUNCT
fcis-18739	18	1	traditional	traditional	ADJ
fcis-18739	18	2	image	image	NOUN
fcis-18739	18	3	segmentation	segmentation	NOUN
fcis-18739	18	4	methods	method	NOUN
fcis-18739	18	5	mainly	mainly	ADV
fcis-18739	18	6	rely	rely	VERB
fcis-18739	18	7	on	on	ADP
fcis-18739	18	8	low	low	ADJ
fcis-18739	18	9	-	-	PUNCT
fcis-18739	18	10	level	level	NOUN
fcis-18739	18	11	features	feature	NOUN
fcis-18739	18	12	such	such	ADJ
fcis-18739	18	13	as	as	ADP
fcis-18739	18	14	intensity	intensity	NOUN
fcis-18739	18	15	,	,	PUNCT
fcis-18739	18	16	texture	texture	NOUN
fcis-18739	18	17	,	,	PUNCT
fcis-18739	18	18	and	and	CCONJ
fcis-18739	18	19	color	color	NOUN
fcis-18739	18	20	of	of	ADP
fcis-18739	18	21	the	the	DET
fcis-18739	18	22	image	image	NOUN
fcis-18739	18	23	for	for	ADP
fcis-18739	18	24	region	region	NOUN
fcis-18739	18	25	division	division	NOUN
fcis-18739	18	26	[	[	X
fcis-18739	18	27	5	5	NUM
fcis-18739	18	28	]	]	PUNCT
fcis-18739	18	29	.	.	PUNCT
fcis-18739	19	1	among	among	ADP
fcis-18739	19	2	them	they	PRON
fcis-18739	19	3	,	,	PUNCT
fcis-18739	19	4	threshold	threshold	NOUN
fcis-18739	19	5	based	base	VERB
fcis-18739	19	6	algorithms	algorithm	NOUN
fcis-18739	19	7	divide	divide	VERB
fcis-18739	19	8	the	the	DET
fcis-18739	19	9	pixel	pixel	PROPN
fcis-18739	19	10	values	value	NOUN
fcis-18739	19	11	of	of	ADP
fcis-18739	19	12	an	an	DET
fcis-18739	19	13	image	image	NOUN
fcis-18739	19	14	into	into	ADP
fcis-18739	19	15	different	different	ADJ
fcis-18739	19	16	categories	category	NOUN
fcis-18739	19	17	by	by	ADP
fcis-18739	19	18	setting	set	VERB
fcis-18739	19	19	one	one	NUM
fcis-18739	19	20	or	or	CCONJ
fcis-18739	19	21	more	more	ADJ
fcis-18739	19	22	thresholds	threshold	NOUN
fcis-18739	19	23	,	,	PUNCT
fcis-18739	19	24	thereby	thereby	ADV
fcis-18739	19	25	achieving	achieve	VERB
fcis-18739	19	26	image	image	NOUN
fcis-18739	19	27	segmentation	segmentation	NOUN
fcis-18739	19	28	[	[	X
fcis-18739	19	29	6	6	NUM
fcis-18739	19	30	]	]	PUNCT
fcis-18739	19	31	.	.	PUNCT
fcis-18739	20	1	this	this	DET
fcis-18739	20	2	method	method	NOUN
fcis-18739	20	3	is	be	AUX
fcis-18739	20	4	simple	simple	ADJ
fcis-18739	20	5	and	and	CCONJ
fcis-18739	20	6	easy	easy	ADJ
fcis-18739	20	7	to	to	PART
fcis-18739	20	8	implement	implement	VERB
fcis-18739	20	9	,	,	PUNCT
fcis-18739	20	10	but	but	CCONJ
fcis-18739	20	11	its	its	PRON
fcis-18739	20	12	segmentation	segmentation	NOUN
fcis-18739	20	13	effect	effect	NOUN
fcis-18739	20	14	is	be	AUX
fcis-18739	20	15	often	often	ADV
fcis-18739	20	16	unsatisfactory	unsatisfactory	ADJ
fcis-18739	20	17	for	for	ADP
fcis-18739	20	18	complex	complex	ADJ
fcis-18739	20	19	scenes	scene	NOUN
fcis-18739	20	20	and	and	CCONJ
fcis-18739	20	21	images	image	NOUN
fcis-18739	20	22	under	under	ADP
fcis-18739	20	23	variable	variable	ADJ
fcis-18739	20	24	lighting	lighting	NOUN
fcis-18739	20	25	conditions	condition	NOUN
fcis-18739	20	26	.	.	PUNCT
fcis-18739	21	1	edge	edge	NOUN
fcis-18739	21	2	based	base	VERB
fcis-18739	21	3	algorithms	algorithm	NOUN
fcis-18739	21	4	extract	extract	VERB
fcis-18739	21	5	edge	edge	NOUN
fcis-18739	21	6	information	information	NOUN
fcis-18739	21	7	by	by	ADP
fcis-18739	21	8	detecting	detect	VERB
fcis-18739	21	9	areas	area	NOUN
fcis-18739	21	10	in	in	ADP
fcis-18739	21	11	the	the	DET
fcis-18739	21	12	image	image	NOUN
fcis-18739	21	13	where	where	SCONJ
fcis-18739	21	14	pixel	pixel	PROPN
fcis-18739	21	15	values	value	NOUN
fcis-18739	21	16	change	change	VERB
fcis-18739	21	17	dramatically	dramatically	ADV
fcis-18739	21	18	,	,	PUNCT
fcis-18739	21	19	thereby	thereby	ADV
fcis-18739	21	20	achieving	achieve	VERB
fcis-18739	21	21	image	image	NOUN
fcis-18739	21	22	segmentation	segmentation	NOUN
fcis-18739	21	23	.	.	PUNCT
fcis-18739	22	1	this	this	DET
fcis-18739	22	2	method	method	NOUN
fcis-18739	22	3	has	have	VERB
fcis-18739	22	4	good	good	ADJ
fcis-18739	22	5	performance	performance	NOUN
fcis-18739	22	6	for	for	ADP
fcis-18739	22	7	images	image	NOUN
fcis-18739	22	8	with	with	ADP
fcis-18739	22	9	obvious	obvious	ADJ
fcis-18739	22	10	edges	edge	NOUN
fcis-18739	22	11	,	,	PUNCT
fcis-18739	22	12	but	but	CCONJ
fcis-18739	22	13	its	its	PRON
fcis-18739	22	14	performance	performance	NOUN
fcis-18739	22	15	will	will	AUX
fcis-18739	22	16	be	be	AUX
fcis-18739	22	17	greatly	greatly	ADV
fcis-18739	22	18	affected	affect	VERB
fcis-18739	22	19	when	when	SCONJ
fcis-18739	22	20	facing	face	VERB
fcis-18739	22	21	edge	edge	NOUN
fcis-18739	22	22	blur	blur	NOUN
fcis-18739	22	23	or	or	CCONJ
fcis-18739	22	24	noise	noise	NOUN
fcis-18739	22	25	interference	interference	NOUN
fcis-18739	22	26	.	.	PUNCT
fcis-18739	23	1	in	in	ADP
fcis-18739	23	2	addition	addition	NOUN
fcis-18739	23	3	to	to	ADP
fcis-18739	23	4	traditional	traditional	ADJ
fcis-18739	23	5	image	image	NOUN
fcis-18739	23	6	segmentation	segmentation	NOUN
fcis-18739	23	7	methods	method	NOUN
fcis-18739	23	8	,	,	PUNCT
fcis-18739	23	9	there	there	PRON
fcis-18739	23	10	are	be	VERB
fcis-18739	23	11	also	also	ADV
fcis-18739	23	12	some	some	DET
fcis-18739	23	13	clustering	clustering	ADJ
fcis-18739	23	14	based	base	VERB
fcis-18739	23	15	algorithms	algorithm	NOUN
fcis-18739	23	16	,	,	PUNCT
fcis-18739	23	17	such	such	ADJ
fcis-18739	23	18	as	as	ADP
fcis-18739	23	19	k	k	NOUN
fcis-18739	23	20	-	-	PUNCT
fcis-18739	23	21	means	means	NOUN
fcis-18739	23	22	clustering	clustering	NOUN
fcis-18739	23	23	,	,	PUNCT
fcis-18739	23	24	fuzzy	fuzzy	ADJ
fcis-18739	23	25	c	c	NOUN
fcis-18739	23	26	-	-	PUNCT
fcis-18739	23	27	means	means	NOUN
fcis-18739	23	28	clustering	clustering	NOUN
fcis-18739	23	29	,	,	PUNCT
fcis-18739	23	30	etc	etc	X
fcis-18739	23	31	.	.	X
fcis-18739	24	1	these	these	DET
fcis-18739	24	2	algorithms	algorithm	NOUN
fcis-18739	24	3	divide	divide	VERB
fcis-18739	24	4	similar	similar	ADJ
fcis-18739	24	5	pixels	pixel	NOUN
fcis-18739	24	6	into	into	ADP
fcis-18739	24	7	the	the	DET
fcis-18739	24	8	same	same	ADJ
fcis-18739	24	9	category	category	NOUN
fcis-18739	24	10	by	by	ADP
fcis-18739	24	11	calculating	calculate	VERB
fcis-18739	24	12	the	the	DET
fcis-18739	24	13	similarity	similarity	NOUN
fcis-18739	24	14	between	between	ADP
fcis-18739	24	15	pixels	pixel	NOUN
fcis-18739	24	16	,	,	PUNCT
fcis-18739	24	17	thereby	thereby	ADV
fcis-18739	24	18	achieving	achieve	VERB
fcis-18739	24	19	image	image	NOUN
fcis-18739	24	20	segmentation	segmentation	NOUN
fcis-18739	24	21	.	.	PUNCT
fcis-18739	25	1	however	however	ADV
fcis-18739	25	2	,	,	PUNCT
fcis-18739	25	3	these	these	DET
fcis-18739	25	4	methods	method	NOUN
fcis-18739	25	5	may	may	AUX
fcis-18739	25	6	face	face	VERB
fcis-18739	25	7	problems	problem	NOUN
fcis-18739	25	8	such	such	ADJ
fcis-18739	25	9	as	as	ADP
fcis-18739	25	10	high	high	ADJ
fcis-18739	25	11	computational	computational	ADJ
fcis-18739	25	12	complexity	complexity	NOUN
fcis-18739	25	13	and	and	CCONJ
fcis-18739	25	14	low	low	ADJ
fcis-18739	25	15	segmentation	segmentation	NOUN
fcis-18739	25	16	accuracy	accuracy	NOUN
fcis-18739	25	17	when	when	SCONJ
fcis-18739	25	18	dealing	deal	VERB
fcis-18739	25	19	with	with	ADP
fcis-18739	25	20	large	large	ADJ
fcis-18739	25	21	-	-	PUNCT
fcis-18739	25	22	scale	scale	NOUN
fcis-18739	25	23	images	image	NOUN
fcis-18739	25	24	or	or	CCONJ
fcis-18739	25	25	complex	complex	ADJ
fcis-18739	25	26	scenes	scene	NOUN
fcis-18739	25	27	[	[	X
fcis-18739	25	28	7	7	NUM
fcis-18739	25	29	]	]	PUNCT
fcis-18739	25	30	.	.	PUNCT
fcis-18739	26	1	in	in	ADP
fcis-18739	26	2	recent	recent	ADJ
fcis-18739	26	3	years	year	NOUN
fcis-18739	26	4	,	,	PUNCT
fcis-18739	26	5	with	with	ADP
fcis-18739	26	6	the	the	DET
fcis-18739	26	7	rise	rise	NOUN
fcis-18739	26	8	of	of	ADP
fcis-18739	26	9	dl	dl	PROPN
fcis-18739	26	10	technology	technology	NOUN
fcis-18739	26	11	,	,	PUNCT
fcis-18739	26	12	image	image	NOUN
fcis-18739	26	13	segmentation	segmentation	NOUN
fcis-18739	26	14	algorithms	algorithm	NOUN
fcis-18739	26	15	based	base	VERB
fcis-18739	26	16	on	on	ADP
fcis-18739	26	17	dl	dl	PROPN
fcis-18739	26	18	have	have	AUX
fcis-18739	26	19	gradually	gradually	ADV
fcis-18739	26	20	become	become	VERB
fcis-18739	26	21	a	a	DET
fcis-18739	26	22	research	research	NOUN
fcis-18739	26	23	hotspot	hotspot	NOUN
fcis-18739	26	24	.	.	PUNCT
fcis-18739	27	1	the	the	DET
fcis-18739	27	2	dl	dl	PROPN
fcis-18739	27	3	model	model	NOUN
fcis-18739	27	4	can	can	AUX
fcis-18739	27	5	capture	capture	VERB
fcis-18739	27	6	the	the	DET
fcis-18739	27	7	semantic	semantic	ADJ
fcis-18739	27	8	information	information	NOUN
fcis-18739	27	9	of	of	ADP
fcis-18739	27	10	images	image	NOUN
fcis-18739	27	11	more	more	ADV
fcis-18739	27	12	accurately	accurately	ADV
fcis-18739	27	13	by	by	ADP
fcis-18739	27	14	automatically	automatically	ADV
fcis-18739	27	15	learning	learn	VERB
fcis-18739	27	16	high	high	ADJ
fcis-18739	27	17	-	-	PUNCT
fcis-18739	27	18	level	level	NOUN
fcis-18739	27	19	feature	feature	NOUN
fcis-18739	27	20	representations	representation	NOUN
fcis-18739	27	21	in	in	ADP
fcis-18739	27	22	images	image	NOUN
fcis-18739	27	23	,	,	PUNCT
fcis-18739	27	24	thereby	thereby	ADV
fcis-18739	27	25	achieving	achieve	VERB
fcis-18739	27	26	more	more	ADV
fcis-18739	27	27	precise	precise	ADJ
fcis-18739	27	28	image	image	NOUN
fcis-18739	27	29	segmentation	segmentation	NOUN
fcis-18739	27	30	.	.	PUNCT
fcis-18739	28	1	compared	compare	VERB
fcis-18739	28	2	with	with	ADP
fcis-18739	28	3	traditional	traditional	ADJ
fcis-18739	28	4	image	image	NOUN
fcis-18739	28	5	segmentation	segmentation	NOUN
fcis-18739	28	6	methods	method	NOUN
fcis-18739	28	7	,	,	PUNCT
fcis-18739	28	8	dl	dl	PROPN
fcis-18739	28	9	based	base	VERB
fcis-18739	28	10	image	image	NOUN
fcis-18739	28	11	segmentation	segmentation	NOUN
fcis-18739	28	12	algorithms	algorithm	NOUN
fcis-18739	28	13	have	have	VERB
fcis-18739	28	14	higher	high	ADJ
fcis-18739	28	15	segmentation	segmentation	NOUN
fcis-18739	28	16	accuracy	accuracy	NOUN
fcis-18739	28	17	and	and	CCONJ
fcis-18739	28	18	stronger	strong	ADJ
fcis-18739	28	19	robustness	robustness	NOUN
fcis-18739	28	20	.	.	PUNCT
fcis-18739	29	1	at	at	ADP
fcis-18739	29	2	present	present	NOUN
fcis-18739	29	3	,	,	PUNCT
fcis-18739	29	4	dl	dl	PROPN
fcis-18739	29	5	based	base	VERB
fcis-18739	29	6	image	image	NOUN
fcis-18739	29	7	segmentation	segmentation	NOUN
fcis-18739	29	8	methods	method	NOUN
fcis-18739	29	9	are	be	AUX
fcis-18739	29	10	mainly	mainly	ADV
fcis-18739	29	11	divided	divide	VERB
fcis-18739	29	12	into	into	ADP
fcis-18739	29	13	three	three	NUM
fcis-18739	29	14	categories	category	NOUN
fcis-18739	29	15	:	:	PUNCT
fcis-18739	29	16	image	image	NOUN
fcis-18739	29	17	semantic	semantic	ADJ
fcis-18739	29	18	segmentation	segmentation	NOUN
fcis-18739	29	19	,	,	PUNCT
fcis-18739	29	20	instance	instance	NOUN
fcis-18739	29	21	segmentation	segmentation	NOUN
fcis-18739	29	22	,	,	PUNCT
fcis-18739	29	23	and	and	CCONJ
fcis-18739	29	24	panoramic	panoramic	ADJ
fcis-18739	29	25	segmentation	segmentation	NOUN
fcis-18739	29	26	.	.	PUNCT
fcis-18739	30	1	image	image	NOUN
fcis-18739	30	2	semantic	semantic	ADJ
fcis-18739	30	3	segmentation	segmentation	NOUN
fcis-18739	30	4	is	be	AUX
fcis-18739	30	5	one	one	NUM
fcis-18739	30	6	of	of	ADP
fcis-18739	30	7	the	the	DET
fcis-18739	30	8	important	important	ADJ
fcis-18739	30	9	applications	application	NOUN
fcis-18739	30	10	of	of	ADP
fcis-18739	30	11	dl	dl	PROPN
fcis-18739	30	12	in	in	ADP
fcis-18739	30	13	the	the	DET
fcis-18739	30	14	field	field	NOUN
fcis-18739	30	15	of	of	ADP
fcis-18739	30	16	image	image	NOUN
fcis-18739	30	17	segmentation	segmentation	NOUN
fcis-18739	30	18	.	.	PUNCT
fcis-18739	31	1	its	its	PRON
fcis-18739	31	2	goal	goal	NOUN
fcis-18739	31	3	is	be	AUX
fcis-18739	31	4	to	to	PART
fcis-18739	31	5	assign	assign	VERB
fcis-18739	31	6	a	a	DET
fcis-18739	31	7	predefined	predefine	VERB
fcis-18739	31	8	category	category	NOUN
fcis-18739	31	9	label	label	NOUN
fcis-18739	31	10	to	to	ADP
fcis-18739	31	11	each	each	DET
fcis-18739	31	12	pixel	pixel	NOUN
fcis-18739	31	13	in	in	ADP
fcis-18739	31	14	the	the	DET
fcis-18739	31	15	image	image	NOUN
fcis-18739	31	16	,	,	PUNCT
fcis-18739	31	17	thereby	thereby	ADV
fcis-18739	31	18	achieving	achieve	VERB
fcis-18739	31	19	pixel	pixel	ADJ
fcis-18739	31	20	level	level	NOUN
fcis-18739	31	21	image	image	NOUN
fcis-18739	31	22	understanding	understanding	NOUN
fcis-18739	31	23	.	.	PUNCT
fcis-18739	32	1	with	with	ADP
fcis-18739	32	2	the	the	DET
fcis-18739	32	3	development	development	NOUN
fcis-18739	32	4	of	of	ADP
fcis-18739	32	5	cnn	cnn	PROPN
fcis-18739	32	6	,	,	PUNCT
fcis-18739	32	7	the	the	DET
fcis-18739	32	8	performance	performance	NOUN
fcis-18739	32	9	of	of	ADP
fcis-18739	32	10	image	image	NOUN
fcis-18739	32	11	semantic	semantic	ADJ
fcis-18739	32	12	segmentation	segmentation	NOUN
fcis-18739	32	13	has	have	AUX
fcis-18739	32	14	been	be	AUX
fcis-18739	32	15	significantly	significantly	ADV
fcis-18739	32	16	improved	improve	VERB
fcis-18739	32	17	.	.	PUNCT
fcis-18739	33	1	many	many	ADJ
fcis-18739	33	2	cnn	cnn	PROPN
fcis-18739	33	3	based	base	VERB
fcis-18739	33	4	semantic	semantic	ADJ
fcis-18739	33	5	segmentation	segmentation	NOUN
fcis-18739	33	6	models	model	NOUN
fcis-18739	33	7	have	have	AUX
fcis-18739	33	8	been	be	AUX
fcis-18739	33	9	proposed	propose	VERB
fcis-18739	33	10	,	,	PUNCT
fcis-18739	33	11	which	which	PRON
fcis-18739	33	12	further	far	ADV
fcis-18739	33	13	improve	improve	VERB
fcis-18739	33	14	the	the	DET
fcis-18739	33	15	accuracy	accuracy	NOUN
fcis-18739	33	16	and	and	CCONJ
fcis-18739	33	17	efficiency	efficiency	NOUN
fcis-18739	33	18	of	of	ADP
fcis-18739	33	19	image	image	NOUN
fcis-18739	33	20	semantic	semantic	ADJ
fcis-18739	33	21	segmentation	segmentation	NOUN
fcis-18739	33	22	by	by	ADP
fcis-18739	33	23	introducing	introduce	VERB
fcis-18739	33	24	different	different	ADJ
fcis-18739	33	25	network	network	NOUN
fcis-18739	33	26	structures	structure	NOUN
fcis-18739	33	27	and	and	CCONJ
fcis-18739	33	28	optimization	optimization	NOUN
fcis-18739	33	29	strategies	strategy	NOUN
fcis-18739	33	30	.	.	PUNCT
fcis-18739	34	1	2	2	X
fcis-18739	34	2	.	.	X
fcis-18739	34	3	the	the	DET
fcis-18739	34	4	application	application	NOUN
fcis-18739	34	5	of	of	ADP
fcis-18739	34	6	dl	dl	PROPN
fcis-18739	34	7	in	in	ADP
fcis-18739	34	8	image	image	NOUN
fcis-18739	34	9	segmentation	segmentation	NOUN
fcis-18739	34	10	as	as	ADP
fcis-18739	34	11	a	a	DET
fcis-18739	34	12	fundamental	fundamental	ADJ
fcis-18739	34	13	task	task	NOUN
fcis-18739	34	14	in	in	ADP
fcis-18739	34	15	the	the	DET
fcis-18739	34	16	field	field	NOUN
fcis-18739	34	17	of	of	ADP
fcis-18739	34	18	cv	cv	PROPN
fcis-18739	34	19	,	,	PUNCT
fcis-18739	34	20	image	image	NOUN
fcis-18739	34	21	segmentation	segmentation	NOUN
fcis-18739	34	22	aims	aim	VERB
fcis-18739	34	23	to	to	PART
fcis-18739	34	24	divide	divide	VERB
fcis-18739	34	25	images	image	NOUN
fcis-18739	34	26	into	into	ADP
fcis-18739	34	27	multiple	multiple	ADJ
fcis-18739	34	28	regions	region	NOUN
fcis-18739	34	29	with	with	ADP
fcis-18739	34	30	similar	similar	ADJ
fcis-18739	34	31	properties	property	NOUN
fcis-18739	34	32	,	,	PUNCT
fcis-18739	34	33	providing	provide	VERB
fcis-18739	34	34	basic	basic	ADJ
fcis-18739	34	35	data	datum	NOUN
fcis-18739	34	36	for	for	ADP
fcis-18739	34	37	subsequent	subsequent	ADJ
fcis-18739	34	38	tasks	task	NOUN
fcis-18739	34	39	such	such	ADJ
fcis-18739	34	40	as	as	ADP
fcis-18739	34	41	object	object	NOUN
fcis-18739	34	42	detection	detection	NOUN
fcis-18739	34	43	and	and	CCONJ
fcis-18739	34	44	scene	scene	NOUN
fcis-18739	34	45	understanding	understand	VERB
fcis-18739	34	46	[	[	X
fcis-18739	34	47	8	8	NUM
fcis-18739	34	48	]	]	PUNCT
fcis-18739	34	49	.	.	PUNCT
fcis-18739	35	1	however	however	ADV
fcis-18739	35	2	,	,	PUNCT
fcis-18739	35	3	due	due	ADP
fcis-18739	35	4	to	to	ADP
fcis-18739	35	5	the	the	DET
fcis-18739	35	6	complexity	complexity	NOUN
fcis-18739	35	7	and	and	CCONJ
fcis-18739	35	8	correlation	correlation	NOUN
fcis-18739	35	9	of	of	ADP
fcis-18739	35	10	images	image	NOUN
fcis-18739	35	11	themselves	themselves	PRON
fcis-18739	35	12	,	,	PUNCT
fcis-18739	35	13	traditional	traditional	ADJ
fcis-18739	35	14	image	image	NOUN
fcis-18739	35	15	segmentation	segmentation	NOUN
fcis-18739	35	16	methods	method	NOUN
fcis-18739	35	17	often	often	ADV
fcis-18739	35	18	have	have	VERB
fcis-18739	35	19	poor	poor	ADJ
fcis-18739	35	20	performance	performance	NOUN
fcis-18739	35	21	in	in	ADP
fcis-18739	35	22	dealing	deal	VERB
fcis-18739	35	23	with	with	ADP
fcis-18739	35	24	ambiguity	ambiguity	NOUN
fcis-18739	35	25	and	and	CCONJ
fcis-18739	35	26	uncertainty	uncertainty	NOUN
fcis-18739	35	27	[	[	X
fcis-18739	35	28	9	9	NUM
fcis-18739	35	29	]	]	PUNCT
fcis-18739	35	30	.	.	PUNCT
fcis-18739	36	1	in	in	ADP
fcis-18739	36	2	recent	recent	ADJ
fcis-18739	36	3	years	year	NOUN
fcis-18739	36	4	,	,	PUNCT
fcis-18739	36	5	with	with	ADP
fcis-18739	36	6	the	the	DET
fcis-18739	36	7	rapid	rapid	ADJ
fcis-18739	36	8	development	development	NOUN
fcis-18739	36	9	18	18	NUM
fcis-18739	36	10	of	of	ADP
fcis-18739	36	11	dl	dl	PROPN
fcis-18739	36	12	technology	technology	NOUN
fcis-18739	36	13	,	,	PUNCT
fcis-18739	36	14	especially	especially	ADV
fcis-18739	36	15	the	the	DET
fcis-18739	36	16	widespread	widespread	ADJ
fcis-18739	36	17	application	application	NOUN
fcis-18739	36	18	of	of	ADP
fcis-18739	36	19	cnn	cnn	PROPN
fcis-18739	36	20	,	,	PUNCT
fcis-18739	36	21	significant	significant	ADJ
fcis-18739	36	22	progress	progress	NOUN
fcis-18739	36	23	has	have	AUX
fcis-18739	36	24	been	be	AUX
fcis-18739	36	25	made	make	VERB
fcis-18739	36	26	in	in	ADP
fcis-18739	36	27	the	the	DET
fcis-18739	36	28	application	application	NOUN
fcis-18739	36	29	of	of	ADP
fcis-18739	36	30	dl	dl	PROPN
fcis-18739	36	31	in	in	ADP
fcis-18739	36	32	image	image	NOUN
fcis-18739	36	33	segmentation	segmentation	NOUN
fcis-18739	36	34	[	[	X
fcis-18739	36	35	10	10	NUM
fcis-18739	36	36	]	]	PUNCT
fcis-18739	36	37	.	.	PUNCT
fcis-18739	37	1	traditional	traditional	ADJ
fcis-18739	37	2	image	image	NOUN
fcis-18739	37	3	segmentation	segmentation	NOUN
fcis-18739	37	4	methods	method	NOUN
fcis-18739	37	5	,	,	PUNCT
fcis-18739	37	6	such	such	ADJ
fcis-18739	37	7	as	as	ADP
fcis-18739	37	8	threshold	threshold	NOUN
fcis-18739	37	9	based	base	VERB
fcis-18739	37	10	segmentation	segmentation	NOUN
fcis-18739	37	11	and	and	CCONJ
fcis-18739	37	12	edge	edge	NOUN
fcis-18739	37	13	based	base	VERB
fcis-18739	37	14	segmentation	segmentation	NOUN
fcis-18739	37	15	,	,	PUNCT
fcis-18739	37	16	mainly	mainly	ADV
fcis-18739	37	17	rely	rely	VERB
fcis-18739	37	18	on	on	ADP
fcis-18739	37	19	the	the	DET
fcis-18739	37	20	underlying	underlie	VERB
fcis-18739	37	21	features	feature	NOUN
fcis-18739	37	22	of	of	ADP
fcis-18739	37	23	the	the	DET
fcis-18739	37	24	image	image	NOUN
fcis-18739	37	25	for	for	ADP
fcis-18739	37	26	segmentation	segmentation	NOUN
fcis-18739	37	27	.	.	PUNCT
fcis-18739	38	1	however	however	ADV
fcis-18739	38	2	,	,	PUNCT
fcis-18739	38	3	these	these	DET
fcis-18739	38	4	methods	method	NOUN
fcis-18739	38	5	often	often	ADV
fcis-18739	38	6	struggle	struggle	VERB
fcis-18739	38	7	to	to	PART
fcis-18739	38	8	achieve	achieve	VERB
fcis-18739	38	9	ideal	ideal	ADJ
fcis-18739	38	10	segmentation	segmentation	NOUN
fcis-18739	38	11	results	result	NOUN
fcis-18739	38	12	when	when	SCONJ
fcis-18739	38	13	facing	face	VERB
fcis-18739	38	14	complex	complex	ADJ
fcis-18739	38	15	image	image	NOUN
fcis-18739	38	16	scenes	scene	NOUN
fcis-18739	38	17	and	and	CCONJ
fcis-18739	38	18	variable	variable	ADJ
fcis-18739	38	19	lighting	lighting	NOUN
fcis-18739	38	20	conditions	condition	NOUN
fcis-18739	38	21	.	.	PUNCT
fcis-18739	39	1	in	in	ADP
fcis-18739	39	2	addition	addition	NOUN
fcis-18739	39	3	,	,	PUNCT
fcis-18739	39	4	traditional	traditional	ADJ
fcis-18739	39	5	methods	method	NOUN
fcis-18739	39	6	often	often	ADV
fcis-18739	39	7	require	require	VERB
fcis-18739	39	8	preprocessing	preprocessing	NOUN
fcis-18739	39	9	and	and	CCONJ
fcis-18739	39	10	post	post	ADJ
fcis-18739	39	11	-	-	ADJ
fcis-18739	39	12	processing	processing	NOUN
fcis-18739	39	13	of	of	ADP
fcis-18739	39	14	images	image	NOUN
fcis-18739	39	15	,	,	PUNCT
fcis-18739	39	16	which	which	PRON
fcis-18739	39	17	is	be	AUX
fcis-18739	39	18	cumbersome	cumbersome	ADJ
fcis-18739	39	19	and	and	CCONJ
fcis-18739	39	20	has	have	VERB
fcis-18739	39	21	limited	limited	ADJ
fcis-18739	39	22	effectiveness	effectiveness	NOUN
fcis-18739	39	23	.	.	PUNCT
fcis-18739	40	1	dl	dl	INTJ
fcis-18739	40	2	,	,	PUNCT
fcis-18739	40	3	especially	especially	ADV
fcis-18739	40	4	cnn	cnn	PROPN
fcis-18739	40	5	,	,	PUNCT
fcis-18739	40	6	can	can	AUX
fcis-18739	40	7	more	more	ADV
fcis-18739	40	8	accurately	accurately	ADV
fcis-18739	40	9	capture	capture	VERB
fcis-18739	40	10	the	the	DET
fcis-18739	40	11	semantic	semantic	ADJ
fcis-18739	40	12	information	information	NOUN
fcis-18739	40	13	of	of	ADP
fcis-18739	40	14	images	image	NOUN
fcis-18739	40	15	by	by	ADP
fcis-18739	40	16	automatically	automatically	ADV
fcis-18739	40	17	learning	learn	VERB
fcis-18739	40	18	deep	deep	ADJ
fcis-18739	40	19	level	level	NOUN
fcis-18739	40	20	feature	feature	NOUN
fcis-18739	40	21	representations	representation	NOUN
fcis-18739	40	22	.	.	PUNCT
fcis-18739	41	1	in	in	ADP
fcis-18739	41	2	image	image	NOUN
fcis-18739	41	3	segmentation	segmentation	NOUN
fcis-18739	41	4	tasks	task	NOUN
fcis-18739	41	5	,	,	PUNCT
fcis-18739	41	6	dl	dl	PROPN
fcis-18739	41	7	models	model	NOUN
fcis-18739	41	8	can	can	AUX
fcis-18739	41	9	learn	learn	VERB
fcis-18739	41	10	the	the	DET
fcis-18739	41	11	spatial	spatial	ADJ
fcis-18739	41	12	relationships	relationship	NOUN
fcis-18739	41	13	and	and	CCONJ
fcis-18739	41	14	contextual	contextual	ADJ
fcis-18739	41	15	information	information	NOUN
fcis-18739	41	16	between	between	ADP
fcis-18739	41	17	pixels	pixel	NOUN
fcis-18739	41	18	,	,	PUNCT
fcis-18739	41	19	thereby	thereby	ADV
fcis-18739	41	20	achieving	achieve	VERB
fcis-18739	41	21	fine	fine	ADJ
fcis-18739	41	22	segmentation	segmentation	NOUN
fcis-18739	41	23	of	of	ADP
fcis-18739	41	24	images	image	NOUN
fcis-18739	41	25	.	.	PUNCT
fcis-18739	42	1	the	the	DET
fcis-18739	42	2	application	application	NOUN
fcis-18739	42	3	of	of	ADP
fcis-18739	42	4	cnn	cnn	PROPN
fcis-18739	42	5	in	in	ADP
fcis-18739	42	6	image	image	NOUN
fcis-18739	42	7	segmentation	segmentation	NOUN
fcis-18739	42	8	is	be	AUX
fcis-18739	42	9	mainly	mainly	ADV
fcis-18739	42	10	reflected	reflect	VERB
fcis-18739	42	11	in	in	ADP
fcis-18739	42	12	two	two	NUM
fcis-18739	42	13	aspects	aspect	NOUN
fcis-18739	42	14	:	:	PUNCT
fcis-18739	42	15	firstly	firstly	ADV
fcis-18739	42	16	,	,	PUNCT
fcis-18739	42	17	by	by	ADP
fcis-18739	42	18	constructing	construct	VERB
fcis-18739	42	19	an	an	DET
fcis-18739	42	20	end	end	NOUN
fcis-18739	42	21	-	-	PUNCT
fcis-18739	42	22	to	to	ADP
fcis-18739	42	23	-	-	PUNCT
fcis-18739	42	24	end	end	NOUN
fcis-18739	42	25	network	network	NOUN
fcis-18739	42	26	structure	structure	NOUN
fcis-18739	42	27	,	,	PUNCT
fcis-18739	42	28	the	the	DET
fcis-18739	42	29	segmentation	segmentation	NOUN
fcis-18739	42	30	results	result	NOUN
fcis-18739	42	31	of	of	ADP
fcis-18739	42	32	the	the	DET
fcis-18739	42	33	image	image	NOUN
fcis-18739	42	34	are	be	AUX
fcis-18739	42	35	directly	directly	ADV
fcis-18739	42	36	output	output	ADJ
fcis-18739	42	37	;	;	PUNCT
fcis-18739	42	38	the	the	DET
fcis-18739	42	39	second	second	NOUN
fcis-18739	42	40	is	be	AUX
fcis-18739	42	41	to	to	PART
fcis-18739	42	42	utilize	utilize	VERB
fcis-18739	42	43	the	the	DET
fcis-18739	42	44	feature	feature	NOUN
fcis-18739	42	45	learning	learn	VERB
fcis-18739	42	46	ability	ability	NOUN
fcis-18739	42	47	of	of	ADP
fcis-18739	42	48	dl	dl	PROPN
fcis-18739	42	49	to	to	PART
fcis-18739	42	50	extract	extract	VERB
fcis-18739	42	51	deep	deep	ADJ
fcis-18739	42	52	level	level	NOUN
fcis-18739	42	53	features	feature	NOUN
fcis-18739	42	54	of	of	ADP
fcis-18739	42	55	the	the	DET
fcis-18739	42	56	image	image	NOUN
fcis-18739	42	57	,	,	PUNCT
fcis-18739	42	58	providing	provide	VERB
fcis-18739	42	59	strong	strong	ADJ
fcis-18739	42	60	support	support	NOUN
fcis-18739	42	61	for	for	ADP
fcis-18739	42	62	subsequent	subsequent	ADJ
fcis-18739	42	63	segmentation	segmentation	NOUN
fcis-18739	42	64	tasks	task	NOUN
fcis-18739	42	65	.	.	PUNCT
fcis-18739	43	1	its	its	PRON
fcis-18739	43	2	structure	structure	NOUN
fcis-18739	43	3	is	be	AUX
fcis-18739	43	4	shown	show	VERB
fcis-18739	43	5	in	in	ADP
fcis-18739	43	6	figure	figure	NOUN
fcis-18739	43	7	1	1	NUM
fcis-18739	43	8	.	.	PUNCT
fcis-18739	44	1	in	in	ADP
fcis-18739	44	2	terms	term	NOUN
fcis-18739	44	3	of	of	ADP
fcis-18739	44	4	end	end	NOUN
fcis-18739	44	5	-	-	PUNCT
fcis-18739	44	6	to	to	ADP
fcis-18739	44	7	-	-	PUNCT
fcis-18739	44	8	end	end	NOUN
fcis-18739	44	9	network	network	NOUN
fcis-18739	44	10	structure	structure	NOUN
fcis-18739	44	11	,	,	PUNCT
fcis-18739	44	12	some	some	DET
fcis-18739	44	13	classic	classic	ADJ
fcis-18739	44	14	models	model	NOUN
fcis-18739	44	15	such	such	ADJ
fcis-18739	44	16	as	as	ADP
fcis-18739	44	17	fully	fully	ADV
fcis-18739	44	18	convolutional	convolutional	ADJ
fcis-18739	44	19	networks	network	NOUN
fcis-18739	44	20	(	(	PUNCT
fcis-18739	44	21	fcn	fcn	NOUN
fcis-18739	44	22	)	)	PUNCT
fcis-18739	44	23	and	and	CCONJ
fcis-18739	44	24	u	u	NOUN
fcis-18739	44	25	-	-	NOUN
fcis-18739	44	26	net	net	ADJ
fcis-18739	44	27	achieve	achieve	VERB
fcis-18739	44	28	direct	direct	ADJ
fcis-18739	44	29	mapping	mapping	NOUN
fcis-18739	44	30	from	from	ADP
fcis-18739	44	31	feature	feature	NOUN
fcis-18739	44	32	maps	map	NOUN
fcis-18739	44	33	to	to	ADP
fcis-18739	44	34	pixel	pixel	PROPN
fcis-18739	44	35	level	level	NOUN
fcis-18739	44	36	segmentation	segmentation	NOUN
fcis-18739	44	37	results	result	NOUN
fcis-18739	44	38	by	by	ADP
fcis-18739	44	39	introducing	introduce	VERB
fcis-18739	44	40	deconvolution	deconvolution	NOUN
fcis-18739	44	41	or	or	CCONJ
fcis-18739	44	42	upsampling	upsampling	NOUN
fcis-18739	44	43	operations	operation	NOUN
fcis-18739	44	44	.	.	PUNCT
fcis-18739	45	1	these	these	DET
fcis-18739	45	2	models	model	NOUN
fcis-18739	45	3	gradually	gradually	ADV
fcis-18739	45	4	integrate	integrate	VERB
fcis-18739	45	5	multi	multi	ADJ
fcis-18739	45	6	-	-	ADJ
fcis-18739	45	7	scale	scale	ADJ
fcis-18739	45	8	features	feature	NOUN
fcis-18739	45	9	of	of	ADP
fcis-18739	45	10	the	the	DET
fcis-18739	45	11	image	image	NOUN
fcis-18739	45	12	through	through	ADP
fcis-18739	45	13	multiple	multiple	ADJ
fcis-18739	45	14	downsampling	downsample	VERB
fcis-18739	45	15	and	and	CCONJ
fcis-18739	45	16	upsampling	upsample	VERB
fcis-18739	45	17	operations	operation	NOUN
fcis-18739	45	18	,	,	PUNCT
fcis-18739	45	19	thereby	thereby	ADV
fcis-18739	45	20	improving	improve	VERB
fcis-18739	45	21	the	the	DET
fcis-18739	45	22	accuracy	accuracy	NOUN
fcis-18739	45	23	and	and	CCONJ
fcis-18739	45	24	robustness	robustness	NOUN
fcis-18739	45	25	of	of	ADP
fcis-18739	45	26	segmentation	segmentation	NOUN
fcis-18739	45	27	.	.	PUNCT
fcis-18739	46	1	in	in	ADP
fcis-18739	46	2	terms	term	NOUN
fcis-18739	46	3	of	of	ADP
fcis-18739	46	4	feature	feature	NOUN
fcis-18739	46	5	learning	learning	NOUN
fcis-18739	46	6	,	,	PUNCT
fcis-18739	46	7	dl	dl	PROPN
fcis-18739	46	8	models	model	NOUN
fcis-18739	46	9	learn	learn	VERB
fcis-18739	46	10	deep	deep	ADJ
fcis-18739	46	11	level	level	NOUN
fcis-18739	46	12	feature	feature	NOUN
fcis-18739	46	13	representations	representation	NOUN
fcis-18739	46	14	of	of	ADP
fcis-18739	46	15	images	image	NOUN
fcis-18739	46	16	through	through	ADP
fcis-18739	46	17	a	a	DET
fcis-18739	46	18	large	large	ADJ
fcis-18739	46	19	amount	amount	NOUN
fcis-18739	46	20	of	of	ADP
fcis-18739	46	21	training	training	NOUN
fcis-18739	46	22	data	datum	NOUN
fcis-18739	46	23	.	.	PUNCT
fcis-18739	47	1	these	these	PRON
fcis-18739	47	2	features	feature	VERB
fcis-18739	47	3	not	not	PART
fcis-18739	47	4	only	only	ADV
fcis-18739	47	5	contain	contain	AUX
fcis-18739	47	6	lowlevel	lowlevel	VERB
fcis-18739	47	7	information	information	NOUN
fcis-18739	47	8	such	such	ADJ
fcis-18739	47	9	as	as	ADP
fcis-18739	47	10	color	color	NOUN
fcis-18739	47	11	and	and	CCONJ
fcis-18739	47	12	texture	texture	NOUN
fcis-18739	47	13	of	of	ADP
fcis-18739	47	14	the	the	DET
fcis-18739	47	15	image	image	NOUN
fcis-18739	47	16	,	,	PUNCT
fcis-18739	47	17	but	but	CCONJ
fcis-18739	47	18	also	also	ADV
fcis-18739	47	19	high	high	ADJ
fcis-18739	47	20	-	-	PUNCT
fcis-18739	47	21	level	level	NOUN
fcis-18739	47	22	information	information	NOUN
fcis-18739	47	23	such	such	ADJ
fcis-18739	47	24	as	as	ADP
fcis-18739	47	25	the	the	DET
fcis-18739	47	26	shape	shape	NOUN
fcis-18739	47	27	and	and	CCONJ
fcis-18739	47	28	structure	structure	NOUN
fcis-18739	47	29	of	of	ADP
fcis-18739	47	30	the	the	DET
fcis-18739	47	31	object	object	NOUN
fcis-18739	47	32	.	.	PUNCT
fcis-18739	48	1	by	by	ADP
fcis-18739	48	2	utilizing	utilize	VERB
fcis-18739	48	3	these	these	DET
fcis-18739	48	4	features	feature	NOUN
fcis-18739	48	5	,	,	PUNCT
fcis-18739	48	6	dl	dl	PROPN
fcis-18739	48	7	models	model	NOUN
fcis-18739	48	8	can	can	AUX
fcis-18739	48	9	better	well	ADV
fcis-18739	48	10	understand	understand	VERB
fcis-18739	48	11	the	the	DET
fcis-18739	48	12	semantic	semantic	ADJ
fcis-18739	48	13	content	content	NOUN
fcis-18739	48	14	of	of	ADP
fcis-18739	48	15	images	image	NOUN
fcis-18739	48	16	,	,	PUNCT
fcis-18739	48	17	thereby	thereby	ADV
fcis-18739	48	18	achieving	achieve	VERB
fcis-18739	48	19	accurate	accurate	ADJ
fcis-18739	48	20	segmentation	segmentation	NOUN
fcis-18739	48	21	of	of	ADP
fcis-18739	48	22	images	image	NOUN
fcis-18739	48	23	.	.	PUNCT
fcis-18739	49	1	figure	figure	NOUN
fcis-18739	49	2	1	1	NUM
fcis-18739	49	3	.	.	PUNCT
fcis-18739	49	4	cnn	cnn	PROPN
fcis-18739	49	5	structure	structure	NOUN
fcis-18739	49	6	in	in	ADP
fcis-18739	49	7	addition	addition	NOUN
fcis-18739	49	8	,	,	PUNCT
fcis-18739	49	9	dl	dl	PROPN
fcis-18739	49	10	also	also	ADV
fcis-18739	49	11	combines	combine	VERB
fcis-18739	49	12	some	some	DET
fcis-18739	49	13	advanced	advanced	ADJ
fcis-18739	49	14	optimization	optimization	NOUN
fcis-18739	49	15	algorithms	algorithm	NOUN
fcis-18739	49	16	and	and	CCONJ
fcis-18739	49	17	loss	loss	NOUN
fcis-18739	49	18	functions	function	NOUN
fcis-18739	49	19	to	to	PART
fcis-18739	49	20	further	far	ADV
fcis-18739	49	21	improve	improve	VERB
fcis-18739	49	22	the	the	DET
fcis-18739	49	23	performance	performance	NOUN
fcis-18739	49	24	of	of	ADP
fcis-18739	49	25	image	image	NOUN
fcis-18739	49	26	segmentation	segmentation	NOUN
fcis-18739	49	27	.	.	PUNCT
fcis-18739	50	1	for	for	ADP
fcis-18739	50	2	example	example	NOUN
fcis-18739	50	3	,	,	PUNCT
fcis-18739	50	4	some	some	DET
fcis-18739	50	5	models	model	NOUN
fcis-18739	50	6	adopt	adopt	VERB
fcis-18739	50	7	a	a	DET
fcis-18739	50	8	multi	multi	ADJ
fcis-18739	50	9	task	task	NOUN
fcis-18739	50	10	learning	learn	VERB
fcis-18739	50	11	approach	approach	NOUN
fcis-18739	50	12	that	that	PRON
fcis-18739	50	13	simultaneously	simultaneously	ADV
fcis-18739	50	14	learns	learn	VERB
fcis-18739	50	15	image	image	NOUN
fcis-18739	50	16	segmentation	segmentation	NOUN
fcis-18739	50	17	and	and	CCONJ
fcis-18739	50	18	classification	classification	NOUN
fcis-18739	50	19	tasks	task	NOUN
fcis-18739	50	20	,	,	PUNCT
fcis-18739	50	21	improving	improve	VERB
fcis-18739	50	22	segmentation	segmentation	NOUN
fcis-18739	50	23	accuracy	accuracy	NOUN
fcis-18739	50	24	by	by	ADP
fcis-18739	50	25	sharing	share	VERB
fcis-18739	50	26	underlying	underlie	VERB
fcis-18739	50	27	features	feature	NOUN
fcis-18739	50	28	.	.	PUNCT
fcis-18739	51	1	some	some	DET
fcis-18739	51	2	models	model	NOUN
fcis-18739	51	3	also	also	ADV
fcis-18739	51	4	introduce	introduce	VERB
fcis-18739	51	5	attention	attention	NOUN
fcis-18739	51	6	mechanisms	mechanism	NOUN
fcis-18739	51	7	,	,	PUNCT
fcis-18739	51	8	allowing	allow	VERB
fcis-18739	51	9	them	they	PRON
fcis-18739	51	10	to	to	PART
fcis-18739	51	11	focus	focus	VERB
fcis-18739	51	12	on	on	ADP
fcis-18739	51	13	key	key	ADJ
fcis-18739	51	14	regions	region	NOUN
fcis-18739	51	15	in	in	ADP
fcis-18739	51	16	the	the	DET
fcis-18739	51	17	image	image	NOUN
fcis-18739	51	18	,	,	PUNCT
fcis-18739	51	19	further	far	ADV
fcis-18739	51	20	improving	improve	VERB
fcis-18739	51	21	segmentation	segmentation	NOUN
fcis-18739	51	22	accuracy	accuracy	NOUN
fcis-18739	51	23	.	.	PUNCT
fcis-18739	52	1	the	the	DET
fcis-18739	52	2	application	application	NOUN
fcis-18739	52	3	of	of	ADP
fcis-18739	52	4	dl	dl	PROPN
fcis-18739	52	5	in	in	ADP
fcis-18739	52	6	image	image	NOUN
fcis-18739	52	7	segmentation	segmentation	NOUN
fcis-18739	52	8	not	not	PART
fcis-18739	52	9	only	only	ADV
fcis-18739	52	10	improves	improve	VERB
fcis-18739	52	11	the	the	DET
fcis-18739	52	12	accuracy	accuracy	NOUN
fcis-18739	52	13	and	and	CCONJ
fcis-18739	52	14	efficiency	efficiency	NOUN
fcis-18739	52	15	of	of	ADP
fcis-18739	52	16	segmentation	segmentation	NOUN
fcis-18739	52	17	,	,	PUNCT
fcis-18739	52	18	but	but	CCONJ
fcis-18739	52	19	also	also	ADV
fcis-18739	52	20	provides	provide	VERB
fcis-18739	52	21	more	more	ADV
fcis-18739	52	22	accurate	accurate	ADJ
fcis-18739	52	23	data	data	NOUN
fcis-18739	52	24	support	support	NOUN
fcis-18739	52	25	for	for	ADP
fcis-18739	52	26	subsequent	subsequent	ADJ
fcis-18739	52	27	image	image	NOUN
fcis-18739	52	28	processing	processing	NOUN
fcis-18739	52	29	and	and	CCONJ
fcis-18739	52	30	analysis	analysis	NOUN
fcis-18739	52	31	tasks	task	NOUN
fcis-18739	52	32	.	.	PUNCT
fcis-18739	53	1	for	for	ADP
fcis-18739	53	2	example	example	NOUN
fcis-18739	53	3	,	,	PUNCT
fcis-18739	53	4	in	in	ADP
fcis-18739	53	5	the	the	DET
fcis-18739	53	6	field	field	NOUN
fcis-18739	53	7	of	of	ADP
fcis-18739	53	8	medical	medical	ADJ
fcis-18739	53	9	image	image	NOUN
fcis-18739	53	10	analysis	analysis	NOUN
fcis-18739	53	11	,	,	PUNCT
fcis-18739	53	12	dl	dl	PROPN
fcis-18739	53	13	models	model	NOUN
fcis-18739	53	14	can	can	AUX
fcis-18739	53	15	achieve	achieve	VERB
fcis-18739	53	16	automatic	automatic	ADJ
fcis-18739	53	17	segmentation	segmentation	NOUN
fcis-18739	53	18	and	and	CCONJ
fcis-18739	53	19	recognition	recognition	NOUN
fcis-18739	53	20	of	of	ADP
fcis-18739	53	21	lesion	lesion	NOUN
fcis-18739	53	22	areas	area	NOUN
fcis-18739	53	23	,	,	PUNCT
fcis-18739	53	24	providing	provide	VERB
fcis-18739	53	25	doctors	doctor	NOUN
fcis-18739	53	26	with	with	ADP
fcis-18739	53	27	auxiliary	auxiliary	ADJ
fcis-18739	53	28	diagnostic	diagnostic	ADJ
fcis-18739	53	29	basis	basis	NOUN
fcis-18739	53	30	;	;	PUNCT
fcis-18739	53	31	in	in	ADP
fcis-18739	53	32	the	the	DET
fcis-18739	53	33	field	field	NOUN
fcis-18739	53	34	of	of	ADP
fcis-18739	53	35	autonomous	autonomous	ADJ
fcis-18739	53	36	driving	driving	NOUN
fcis-18739	53	37	,	,	PUNCT
fcis-18739	53	38	dl	dl	PROPN
fcis-18739	53	39	models	model	NOUN
fcis-18739	53	40	can	can	AUX
fcis-18739	53	41	achieve	achieve	VERB
fcis-18739	53	42	precise	precise	ADJ
fcis-18739	53	43	segmentation	segmentation	NOUN
fcis-18739	53	44	of	of	ADP
fcis-18739	53	45	targets	target	NOUN
fcis-18739	53	46	such	such	ADJ
fcis-18739	53	47	as	as	ADP
fcis-18739	53	48	roads	road	NOUN
fcis-18739	53	49	,	,	PUNCT
fcis-18739	53	50	vehicles	vehicle	NOUN
fcis-18739	53	51	,	,	PUNCT
fcis-18739	53	52	and	and	CCONJ
fcis-18739	53	53	pedestrians	pedestrian	NOUN
fcis-18739	53	54	,	,	PUNCT
fcis-18739	53	55	providing	provide	VERB
fcis-18739	53	56	assurance	assurance	NOUN
fcis-18739	53	57	for	for	ADP
fcis-18739	53	58	the	the	DET
fcis-18739	53	59	safe	safe	ADJ
fcis-18739	53	60	driving	driving	NOUN
fcis-18739	53	61	of	of	ADP
fcis-18739	53	62	vehicles	vehicle	NOUN
fcis-18739	53	63	.	.	PUNCT
fcis-18739	54	1	3	3	X
fcis-18739	54	2	.	.	X
fcis-18739	54	3	algorithm	algorithm	PROPN
fcis-18739	54	4	principle	principle	NOUN
fcis-18739	54	5	in	in	ADP
fcis-18739	54	6	the	the	DET
fcis-18739	54	7	network	network	NOUN
fcis-18739	54	8	,	,	PUNCT
fcis-18739	54	9	convolutional	convolutional	ADJ
fcis-18739	54	10	layers	layer	NOUN
fcis-18739	54	11	are	be	AUX
fcis-18739	54	12	the	the	DET
fcis-18739	54	13	core	core	ADJ
fcis-18739	54	14	components	component	NOUN
fcis-18739	54	15	mainly	mainly	ADV
fcis-18739	54	16	used	use	VERB
fcis-18739	54	17	for	for	ADP
fcis-18739	54	18	feature	feature	NOUN
fcis-18739	54	19	extraction	extraction	NOUN
fcis-18739	54	20	of	of	ADP
fcis-18739	54	21	images	image	NOUN
fcis-18739	54	22	.	.	PUNCT
fcis-18739	55	1	through	through	ADP
fcis-18739	55	2	convolutional	convolutional	ADJ
fcis-18739	55	3	operations	operation	NOUN
fcis-18739	55	4	,	,	PUNCT
fcis-18739	55	5	convolutional	convolutional	ADJ
fcis-18739	55	6	layers	layer	NOUN
fcis-18739	55	7	can	can	AUX
fcis-18739	55	8	learn	learn	VERB
fcis-18739	55	9	local	local	ADJ
fcis-18739	55	10	spatial	spatial	ADJ
fcis-18739	55	11	patterns	pattern	NOUN
fcis-18739	55	12	and	and	CCONJ
fcis-18739	55	13	texture	texture	ADJ
fcis-18739	55	14	information	information	NOUN
fcis-18739	55	15	from	from	ADP
fcis-18739	55	16	the	the	DET
fcis-18739	55	17	original	original	ADJ
fcis-18739	55	18	image	image	NOUN
fcis-18739	55	19	,	,	PUNCT
fcis-18739	55	20	and	and	CCONJ
fcis-18739	55	21	output	output	NOUN
fcis-18739	55	22	corresponding	correspond	VERB
fcis-18739	55	23	feature	feature	NOUN
fcis-18739	55	24	maps	map	NOUN
fcis-18739	55	25	.	.	PUNCT
fcis-18739	56	1	however	however	ADV
fcis-18739	56	2	,	,	PUNCT
fcis-18739	56	3	traditional	traditional	ADJ
fcis-18739	56	4	convolution	convolution	NOUN
fcis-18739	56	5	operations	operation	NOUN
fcis-18739	56	6	often	often	ADV
fcis-18739	56	7	result	result	VERB
fcis-18739	56	8	in	in	ADP
fcis-18739	56	9	a	a	DET
fcis-18739	56	10	loss	loss	NOUN
fcis-18739	56	11	of	of	ADP
fcis-18739	56	12	spatial	spatial	ADJ
fcis-18739	56	13	resolution	resolution	NOUN
fcis-18739	56	14	while	while	SCONJ
fcis-18739	56	15	increasing	increase	VERB
fcis-18739	56	16	the	the	DET
fcis-18739	56	17	receptive	receptive	ADJ
fcis-18739	56	18	field	field	NOUN
fcis-18739	56	19	,	,	PUNCT
fcis-18739	56	20	which	which	PRON
fcis-18739	56	21	may	may	AUX
fcis-18739	56	22	affect	affect	VERB
fcis-18739	56	23	the	the	DET
fcis-18739	56	24	accuracy	accuracy	NOUN
fcis-18739	56	25	of	of	ADP
fcis-18739	56	26	tasks	task	NOUN
fcis-18739	56	27	such	such	ADJ
fcis-18739	56	28	as	as	ADP
fcis-18739	56	29	image	image	NOUN
fcis-18739	56	30	segmentation	segmentation	NOUN
fcis-18739	56	31	.	.	PUNCT
fcis-18739	57	1	to	to	PART
fcis-18739	57	2	solve	solve	VERB
fcis-18739	57	3	this	this	DET
fcis-18739	57	4	problem	problem	NOUN
fcis-18739	57	5	,	,	PUNCT
fcis-18739	57	6	dilated	dilated	ADJ
fcis-18739	57	7	convolution	convolution	NOUN
fcis-18739	57	8	is	be	AUX
fcis-18739	57	9	introduced	introduce	VERB
fcis-18739	57	10	into	into	ADP
fcis-18739	57	11	the	the	DET
fcis-18739	57	12	network	network	NOUN
fcis-18739	57	13	.	.	PUNCT
fcis-18739	58	1	outf	outf	PROPN
fcis-18739	58	2	represents	represent	VERB
fcis-18739	58	3	the	the	DET
fcis-18739	58	4	output	output	NOUN
fcis-18739	58	5	image	image	NOUN
fcis-18739	58	6	size	size	NOUN
fcis-18739	58	7	after	after	ADP
fcis-18739	58	8	passing	pass	VERB
fcis-18739	58	9	through	through	ADP
fcis-18739	58	10	the	the	DET
fcis-18739	58	11	convolutional	convolutional	ADJ
fcis-18739	58	12	layer	layer	NOUN
fcis-18739	58	13	,	,	PUNCT
fcis-18739	58	14	inf	inf	PROPN
fcis-18739	58	15	represents	represent	VERB
fcis-18739	58	16	the	the	DET
fcis-18739	58	17	input	input	NOUN
fcis-18739	58	18	image	image	NOUN
fcis-18739	58	19	size	size	NOUN
fcis-18739	58	20	before	before	ADP
fcis-18739	58	21	passing	pass	VERB
fcis-18739	58	22	through	through	ADP
fcis-18739	58	23	the	the	DET
fcis-18739	58	24	convolutional	convolutional	ADJ
fcis-18739	58	25	layer	layer	NOUN
fcis-18739	58	26	,	,	PUNCT
fcis-18739	58	27	p	p	NOUN
fcis-18739	58	28	represents	represent	VERB
fcis-18739	58	29	padding	padding	NOUN
fcis-18739	58	30	,	,	PUNCT
fcis-18739	58	31	k	k	PROPN
fcis-18739	58	32	represents	represent	VERB
fcis-18739	58	33	the	the	DET
fcis-18739	58	34	convolutional	convolutional	ADJ
fcis-18739	58	35	kernel	kernel	NOUN
fcis-18739	58	36	size	size	NOUN
fcis-18739	58	37	,	,	PUNCT
fcis-18739	58	38	and	and	CCONJ
fcis-18739	58	39	s	s	VERB
fcis-18739	58	40	represents	represent	VERB
fcis-18739	58	41	the	the	DET
fcis-18739	58	42	convolutional	convolutional	ADJ
fcis-18739	58	43	kernel	kernel	NOUN
fcis-18739	58	44	step	step	NOUN
fcis-18739	58	45	size	size	NOUN
fcis-18739	58	46	.	.	PUNCT
fcis-18739	59	1	1	1	NUM
fcis-18739	59	2	2	2	NUM
fcis-18739	59	3			NOUN
fcis-18739	59	4			NOUN
fcis-18739	59	5			NOUN
fcis-18739	59	6			PROPN
fcis-18739	59	7			PROPN
fcis-18739	59	8			PROPN
fcis-18739	59	9	s	s	PART
fcis-18739	59	10	kpf	kpf	NOUN
fcis-18739	59	11	f	f	PROPN
fcis-18739	59	12	in	in	ADP
fcis-18739	59	13	out	out	ADV
fcis-18739	59	14	(	(	PUNCT
fcis-18739	59	15	1	1	NUM
fcis-18739	59	16	)	)	PUNCT
fcis-18739	59	17	in	in	ADP
fcis-18739	59	18	cnn	cnn	PROPN
fcis-18739	59	19	,	,	PUNCT
fcis-18739	59	20	the	the	DET
fcis-18739	59	21	output	output	NOUN
fcis-18739	59	22	of	of	ADP
fcis-18739	59	23	each	each	DET
fcis-18739	59	24	convolutional	convolutional	ADJ
fcis-18739	59	25	layer	layer	NOUN
fcis-18739	59	26	is	be	AUX
fcis-18739	59	27	obtained	obtain	VERB
fcis-18739	59	28	by	by	ADP
fcis-18739	59	29	convolving	convolve	VERB
fcis-18739	59	30	the	the	DET
fcis-18739	59	31	kernel	kernel	NOUN
fcis-18739	59	32	with	with	ADP
fcis-18739	59	33	the	the	DET
fcis-18739	59	34	input	input	NOUN
fcis-18739	59	35	image	image	NOUN
fcis-18739	59	36	,	,	PUNCT
fcis-18739	59	37	adding	add	VERB
fcis-18739	59	38	a	a	DET
fcis-18739	59	39	bias	bias	NOUN
fcis-18739	59	40	term	term	NOUN
fcis-18739	59	41	,	,	PUNCT
fcis-18739	59	42	and	and	CCONJ
fcis-18739	59	43	then	then	ADV
fcis-18739	59	44	applying	apply	VERB
fcis-18739	59	45	an	an	DET
fcis-18739	59	46	activation	activation	NOUN
fcis-18739	59	47	function	function	NOUN
fcis-18739	59	48	.	.	PUNCT
fcis-18739	60	1	the	the	DET
fcis-18739	60	2	size	size	NOUN
fcis-18739	60	3	of	of	ADP
fcis-18739	60	4	the	the	DET
fcis-18739	60	5	input	input	NOUN
fcis-18739	60	6	image	image	NOUN
fcis-18739	60	7	for	for	ADP
fcis-18739	60	8	each	each	DET
fcis-18739	60	9	convolutional	convolutional	ADJ
fcis-18739	60	10	layer	layer	NOUN
fcis-18739	60	11	is	be	AUX
fcis-18739	60	12	cwh	cwh	VERB
fcis-18739	60	13			NOUN
fcis-18739	60	14	,	,	PUNCT
fcis-18739	60	15	where	where	SCONJ
fcis-18739	60	16	h	h	NOUN
fcis-18739	60	17	and	and	CCONJ
fcis-18739	60	18	w	w	PROPN
fcis-18739	60	19	represent	represent	VERB
fcis-18739	60	20	the	the	DET
fcis-18739	60	21	height	height	NOUN
fcis-18739	60	22	and	and	CCONJ
fcis-18739	60	23	width	width	NOUN
fcis-18739	60	24	of	of	ADP
fcis-18739	60	25	the	the	DET
fcis-18739	60	26	image	image	NOUN
fcis-18739	60	27	,	,	PUNCT
fcis-18739	60	28	c	c	PROPN
fcis-18739	60	29	represents	represent	VERB
fcis-18739	60	30	the	the	DET
fcis-18739	60	31	number	number	NOUN
fcis-18739	60	32	of	of	ADP
fcis-18739	60	33	channels	channel	NOUN
fcis-18739	60	34	,	,	PUNCT
fcis-18739	60	35	iw	iw	PROPN
fcis-18739	60	36	is	be	AUX
fcis-18739	60	37	the	the	DET
fcis-18739	60	38	weight	weight	NOUN
fcis-18739	60	39	of	of	ADP
fcis-18739	60	40	the	the	DET
fcis-18739	60	41	i	i	NOUN
fcis-18739	60	42	th	th	X
fcis-18739	60	43	convolutional	convolutional	ADJ
fcis-18739	60	44	kernel	kernel	NOUN
fcis-18739	60	45	,	,	PUNCT
fcis-18739	60	46	ib	ib	PROPN
fcis-18739	60	47	is	be	AUX
fcis-18739	60	48	the	the	DET
fcis-18739	60	49	weight	weight	NOUN
fcis-18739	60	50	bias	bias	NOUN
fcis-18739	60	51	of	of	ADP
fcis-18739	60	52	the	the	DET
fcis-18739	60	53	i	i	NOUN
fcis-18739	60	54	th	th	X
fcis-18739	60	55	convolutional	convolutional	ADJ
fcis-18739	60	56	kernel	kernel	NOUN
fcis-18739	60	57	,	,	PUNCT
fcis-18739	60	58	ix	ix	ADV
fcis-18739	60	59	is	be	AUX
fcis-18739	60	60	the	the	DET
fcis-18739	60	61	input	input	NOUN
fcis-18739	60	62	of	of	ADP
fcis-18739	60	63	the	the	DET
fcis-18739	60	64	i	i	NOUN
fcis-18739	60	65	th	th	X
fcis-18739	60	66	convolutional	convolutional	ADJ
fcis-18739	60	67	kernel	kernel	NOUN
fcis-18739	60	68	,	,	PUNCT
fcis-18739	60	69	and	and	CCONJ
fcis-18739	60	70	f	f	PROPN
fcis-18739	60	71	is	be	AUX
fcis-18739	60	72	the	the	DET
fcis-18739	60	73	activation	activation	NOUN
fcis-18739	60	74	function	function	NOUN
fcis-18739	60	75	.	.	PUNCT
fcis-18739	61	1	the	the	DET
fcis-18739	61	2	formula	formula	NOUN
fcis-18739	61	3	for	for	ADP
fcis-18739	61	4	the	the	DET
fcis-18739	61	5	output	output	NOUN
fcis-18739	61	6	y	y	PROPN
fcis-18739	61	7	of	of	ADP
fcis-18739	61	8	the	the	DET
fcis-18739	61	9	i	i	NOUN
fcis-18739	61	10	th	th	NOUN
fcis-18739	61	11	layer	layer	NOUN
fcis-18739	61	12	is	be	AUX
fcis-18739	61	13	:	:	PUNCT
fcis-18739	61	14	19	19	NUM
fcis-18739	61	15			NOUN
fcis-18739	61	16	iii	iii	PUNCT
fcis-18739	61	17	bxwfy	bxwfy	PROPN
fcis-18739	61	18			PROPN
fcis-18739	61	19	(	(	PUNCT
fcis-18739	61	20	2	2	NUM
fcis-18739	61	21	)	)	PUNCT
fcis-18739	61	22	in	in	ADP
fcis-18739	61	23	cnn	cnn	PROPN
fcis-18739	61	24	,	,	PUNCT
fcis-18739	61	25	each	each	DET
fcis-18739	61	26	layer	layer	NOUN
fcis-18739	61	27	of	of	ADP
fcis-18739	61	28	the	the	DET
fcis-18739	61	29	network	network	NOUN
fcis-18739	61	30	receives	receive	VERB
fcis-18739	61	31	the	the	DET
fcis-18739	61	32	output	output	NOUN
fcis-18739	61	33	of	of	ADP
fcis-18739	61	34	the	the	DET
fcis-18739	61	35	previous	previous	ADJ
fcis-18739	61	36	layer	layer	NOUN
fcis-18739	61	37	as	as	ADP
fcis-18739	61	38	its	its	PRON
fcis-18739	61	39	input	input	NOUN
fcis-18739	61	40	.	.	PUNCT
fcis-18739	62	1	this	this	DET
fcis-18739	62	2	hierarchical	hierarchical	ADJ
fcis-18739	62	3	structure	structure	NOUN
fcis-18739	62	4	enables	enable	VERB
fcis-18739	62	5	information	information	NOUN
fcis-18739	62	6	to	to	PART
fcis-18739	62	7	be	be	AUX
fcis-18739	62	8	transmitted	transmit	VERB
fcis-18739	62	9	and	and	CCONJ
fcis-18739	62	10	processed	process	VERB
fcis-18739	62	11	layer	layer	NOUN
fcis-18739	62	12	by	by	ADP
fcis-18739	62	13	layer	layer	NOUN
fcis-18739	62	14	in	in	ADP
fcis-18739	62	15	the	the	DET
fcis-18739	62	16	network	network	NOUN
fcis-18739	62	17	,	,	PUNCT
fcis-18739	62	18	ultimately	ultimately	ADV
fcis-18739	62	19	resulting	result	VERB
fcis-18739	62	20	in	in	ADP
fcis-18739	62	21	a	a	DET
fcis-18739	62	22	specific	specific	ADJ
fcis-18739	62	23	representation	representation	NOUN
fcis-18739	62	24	of	of	ADP
fcis-18739	62	25	input	input	NOUN
fcis-18739	62	26	data	datum	NOUN
fcis-18739	62	27	.	.	PUNCT
fcis-18739	63	1	the	the	DET
fcis-18739	63	2	activation	activation	NOUN
fcis-18739	63	3	function	function	NOUN
fcis-18739	63	4	plays	play	VERB
fcis-18739	63	5	a	a	DET
fcis-18739	63	6	crucial	crucial	ADJ
fcis-18739	63	7	role	role	NOUN
fcis-18739	63	8	in	in	ADP
fcis-18739	63	9	cnn	cnn	PROPN
fcis-18739	63	10	.	.	PUNCT
fcis-18739	64	1	relu	relu	NOUN
fcis-18739	64	2	is	be	AUX
fcis-18739	64	3	one	one	NUM
fcis-18739	64	4	of	of	ADP
fcis-18739	64	5	the	the	DET
fcis-18739	64	6	commonly	commonly	ADV
fcis-18739	64	7	used	use	VERB
fcis-18739	64	8	activation	activation	NOUN
fcis-18739	64	9	functions	function	NOUN
fcis-18739	64	10	.	.	PUNCT
fcis-18739	65	1	the	the	DET
fcis-18739	65	2	activation	activation	NOUN
fcis-18739	65	3	function	function	NOUN
fcis-18739	65	4	relu	relu	NOUN
fcis-18739	65	5	is	be	AUX
fcis-18739	65	6	:	:	PUNCT
fcis-18739	65	7			NOUN
fcis-18739	65	8	ni	ni	NUM
fcis-18739	66	1	xx	xx	NUM
fcis-18739	66	2	x	x	SYM
fcis-18739	66	3	y	y	NOUN
fcis-18739	66	4	ii	ii	PROPN
fcis-18739	67	1	i	i	PRON
fcis-18739	67	2	i	i	PRON
fcis-18739	67	3	,	,	PUNCT
fcis-18739	67	4	,	,	PUNCT
fcis-18739	67	5	1	1	NUM
fcis-18739	67	6	0	0	NUM
fcis-18739	67	7	,	,	PUNCT
fcis-18739	67	8	0,0	0,0	NOUN
fcis-18739	67	9			NUM
fcis-18739	67	10			PROPN
fcis-18739	67	11			NUM
fcis-18739	67	12			ADP
fcis-18739	67	13			NUM
fcis-18739	67	14			PROPN
fcis-18739	68	1			NUM
fcis-18739	68	2	(	(	PUNCT
fcis-18739	68	3	3	3	NUM
fcis-18739	68	4	)	)	PUNCT
fcis-18739	68	5	among	among	ADP
fcis-18739	68	6	them	they	PRON
fcis-18739	68	7	,	,	PUNCT
fcis-18739	68	8	n	n	X
fcis-18739	68	9	is	be	AUX
fcis-18739	68	10	the	the	DET
fcis-18739	68	11	size	size	NOUN
fcis-18739	68	12	of	of	ADP
fcis-18739	68	13	the	the	DET
fcis-18739	68	14	input	input	NOUN
fcis-18739	68	15	feature	feature	NOUN
fcis-18739	68	16	map	map	NOUN
fcis-18739	68	17	;	;	PUNCT
fcis-18739	68	18	ix	ix	ADV
fcis-18739	68	19	is	be	AUX
fcis-18739	68	20	the	the	DET
fcis-18739	68	21	i	i	NOUN
fcis-18739	68	22	th	th	NOUN
fcis-18739	68	23	value	value	NOUN
fcis-18739	68	24	of	of	ADP
fcis-18739	68	25	the	the	DET
fcis-18739	68	26	input	input	NOUN
fcis-18739	68	27	feature	feature	NOUN
fcis-18739	68	28	map	map	NOUN
fcis-18739	68	29	;	;	PUNCT
fcis-18739	68	30	iy	iy	PROPN
fcis-18739	68	31	is	be	AUX
fcis-18739	68	32	the	the	DET
fcis-18739	68	33	corresponding	corresponding	ADJ
fcis-18739	68	34	output	output	NOUN
fcis-18739	68	35	.	.	PUNCT
fcis-18739	69	1	4	4	X
fcis-18739	69	2	.	.	X
fcis-18739	69	3	experimental	experimental	ADJ
fcis-18739	69	4	result	result	NOUN
fcis-18739	69	5	experiment	experiment	NOUN
fcis-18739	69	6	on	on	ADP
fcis-18739	69	7	image	image	NOUN
fcis-18739	69	8	segmentation	segmentation	NOUN
fcis-18739	69	9	using	use	VERB
fcis-18739	69	10	matlab	matlab	PROPN
fcis-18739	69	11	,	,	PUNCT
fcis-18739	69	12	using	use	VERB
fcis-18739	69	13	the	the	DET
fcis-18739	69	14	cityscapes	cityscape	NOUN
fcis-18739	69	15	dataset	dataset	NOUN
fcis-18739	69	16	,	,	PUNCT
fcis-18739	69	17	which	which	PRON
fcis-18739	69	18	is	be	AUX
fcis-18739	69	19	a	a	DET
fcis-18739	69	20	well	well	ADV
fcis-18739	69	21	-	-	PUNCT
fcis-18739	69	22	known	know	VERB
fcis-18739	69	23	dataset	dataset	NOUN
fcis-18739	69	24	focused	focus	VERB
fcis-18739	69	25	on	on	ADP
fcis-18739	69	26	semantic	semantic	ADJ
fcis-18739	69	27	segmentation	segmentation	NOUN
fcis-18739	69	28	and	and	CCONJ
fcis-18739	69	29	scene	scene	NOUN
fcis-18739	69	30	understanding	understand	VERB
fcis-18739	69	31	tasks	task	NOUN
fcis-18739	69	32	in	in	ADP
fcis-18739	69	33	urban	urban	ADJ
fcis-18739	69	34	street	street	NOUN
fcis-18739	69	35	scenes	scene	NOUN
fcis-18739	69	36	.	.	PUNCT
fcis-18739	70	1	this	this	DET
fcis-18739	70	2	dataset	dataset	NOUN
fcis-18739	70	3	is	be	AUX
fcis-18739	70	4	crucial	crucial	ADJ
fcis-18739	70	5	for	for	ADP
fcis-18739	70	6	researching	research	VERB
fcis-18739	70	7	and	and	CCONJ
fcis-18739	70	8	developing	develop	VERB
fcis-18739	70	9	technologies	technology	NOUN
fcis-18739	70	10	in	in	ADP
fcis-18739	70	11	fields	field	NOUN
fcis-18739	70	12	such	such	ADJ
fcis-18739	70	13	as	as	ADP
fcis-18739	70	14	autonomous	autonomous	ADJ
fcis-18739	70	15	driving	driving	NOUN
fcis-18739	70	16	,	,	PUNCT
fcis-18739	70	17	robot	robot	NOUN
fcis-18739	70	18	navigation	navigation	NOUN
fcis-18739	70	19	,	,	PUNCT
fcis-18739	70	20	and	and	CCONJ
fcis-18739	70	21	video	video	NOUN
fcis-18739	70	22	surveillance	surveillance	NOUN
fcis-18739	70	23	.	.	PUNCT
fcis-18739	71	1	the	the	DET
fcis-18739	71	2	cityscapes	cityscape	NOUN
fcis-18739	71	3	dataset	dataset	NOUN
fcis-18739	71	4	contains	contain	VERB
fcis-18739	71	5	approximately	approximately	ADV
fcis-18739	71	6	25000	25000	NUM
fcis-18739	71	7	images	image	NOUN
fcis-18739	71	8	,	,	PUNCT
fcis-18739	71	9	which	which	PRON
fcis-18739	71	10	mainly	mainly	ADV
fcis-18739	71	11	capture	capture	VERB
fcis-18739	71	12	street	street	NOUN
fcis-18739	71	13	scenes	scene	NOUN
fcis-18739	71	14	in	in	ADP
fcis-18739	71	15	different	different	ADJ
fcis-18739	71	16	cities	city	NOUN
fcis-18739	71	17	,	,	PUNCT
fcis-18739	71	18	weather	weather	NOUN
fcis-18739	71	19	,	,	PUNCT
fcis-18739	71	20	and	and	CCONJ
fcis-18739	71	21	time	time	NOUN
fcis-18739	71	22	conditions	condition	NOUN
fcis-18739	71	23	.	.	PUNCT
fcis-18739	72	1	the	the	DET
fcis-18739	72	2	dataset	dataset	NOUN
fcis-18739	72	3	defines	define	VERB
fcis-18739	72	4	8	8	NUM
fcis-18739	72	5	major	major	ADJ
fcis-18739	72	6	categories	category	NOUN
fcis-18739	72	7	(	(	PUNCT
fcis-18739	72	8	such	such	ADJ
fcis-18739	72	9	as	as	ADP
fcis-18739	72	10	roads	road	NOUN
fcis-18739	72	11	,	,	PUNCT
fcis-18739	72	12	sidewalks	sidewalk	NOUN
fcis-18739	72	13	,	,	PUNCT
fcis-18739	72	14	buildings	building	NOUN
fcis-18739	72	15	,	,	PUNCT
fcis-18739	72	16	vehicles	vehicle	NOUN
fcis-18739	72	17	,	,	PUNCT
fcis-18739	72	18	trees	tree	NOUN
fcis-18739	72	19	,	,	PUNCT
fcis-18739	72	20	etc	etc	X
fcis-18739	72	21	.	.	X
fcis-18739	72	22	)	)	PUNCT
fcis-18739	72	23	and	and	CCONJ
fcis-18739	72	24	finer	fine	ADJ
fcis-18739	72	25	subclasses	subclass	NOUN
fcis-18739	72	26	for	for	ADP
fcis-18739	72	27	more	more	ADV
fcis-18739	72	28	precise	precise	ADJ
fcis-18739	72	29	semantic	semantic	ADJ
fcis-18739	72	30	segmentation	segmentation	NOUN
fcis-18739	72	31	tasks	task	NOUN
fcis-18739	72	32	.	.	PUNCT
fcis-18739	73	1	figure	figure	NOUN
fcis-18739	73	2	2	2	NUM
fcis-18739	73	3	shows	show	VERB
fcis-18739	73	4	the	the	DET
fcis-18739	73	5	comparison	comparison	NOUN
fcis-18739	73	6	of	of	ADP
fcis-18739	73	7	image	image	NOUN
fcis-18739	73	8	segmentation	segmentation	NOUN
fcis-18739	73	9	precision	precision	NOUN
fcis-18739	73	10	between	between	ADP
fcis-18739	73	11	different	different	ADJ
fcis-18739	73	12	algorithms	algorithm	NOUN
fcis-18739	73	13	,	,	PUNCT
fcis-18739	73	14	which	which	PRON
fcis-18739	73	15	fully	fully	ADV
fcis-18739	73	16	demonstrates	demonstrate	VERB
fcis-18739	73	17	the	the	DET
fcis-18739	73	18	advantages	advantage	NOUN
fcis-18739	73	19	of	of	ADP
fcis-18739	73	20	using	use	VERB
fcis-18739	73	21	dl	dl	PROPN
fcis-18739	73	22	technology	technology	NOUN
fcis-18739	73	23	in	in	ADP
fcis-18739	73	24	our	our	PRON
fcis-18739	73	25	algorithm	algorithm	NOUN
fcis-18739	73	26	.	.	PUNCT
fcis-18739	74	1	by	by	ADP
fcis-18739	74	2	introducing	introduce	VERB
fcis-18739	74	3	dl	dl	PROPN
fcis-18739	74	4	technology	technology	NOUN
fcis-18739	74	5	,	,	PUNCT
fcis-18739	74	6	our	our	PRON
fcis-18739	74	7	algorithm	algorithm	NOUN
fcis-18739	74	8	has	have	AUX
fcis-18739	74	9	achieved	achieve	VERB
fcis-18739	74	10	significantly	significantly	ADV
fcis-18739	74	11	better	well	ADJ
fcis-18739	74	12	performance	performance	NOUN
fcis-18739	74	13	than	than	ADP
fcis-18739	74	14	traditional	traditional	ADJ
fcis-18739	74	15	algorithms	algorithm	NOUN
fcis-18739	74	16	in	in	ADP
fcis-18739	74	17	image	image	NOUN
fcis-18739	74	18	segmentation	segmentation	NOUN
fcis-18739	74	19	tasks	task	NOUN
fcis-18739	74	20	.	.	PUNCT
fcis-18739	75	1	dl	dl	PROPN
fcis-18739	75	2	technology	technology	PROPN
fcis-18739	75	3	can	can	AUX
fcis-18739	75	4	automatically	automatically	ADV
fcis-18739	75	5	learn	learn	VERB
fcis-18739	75	6	to	to	PART
fcis-18739	75	7	extract	extract	VERB
fcis-18739	75	8	useful	useful	ADJ
fcis-18739	75	9	features	feature	NOUN
fcis-18739	75	10	from	from	ADP
fcis-18739	75	11	raw	raw	ADJ
fcis-18739	75	12	data	datum	NOUN
fcis-18739	75	13	,	,	PUNCT
fcis-18739	75	14	which	which	PRON
fcis-18739	75	15	is	be	AUX
fcis-18739	75	16	a	a	DET
fcis-18739	75	17	major	major	ADJ
fcis-18739	75	18	advantage	advantage	NOUN
fcis-18739	75	19	compared	compare	VERB
fcis-18739	75	20	to	to	ADP
fcis-18739	75	21	traditional	traditional	ADJ
fcis-18739	75	22	algorithms	algorithm	NOUN
fcis-18739	75	23	.	.	PUNCT
fcis-18739	76	1	by	by	ADP
fcis-18739	76	2	adopting	adopt	VERB
fcis-18739	76	3	a	a	DET
fcis-18739	76	4	cnn	cnn	PROPN
fcis-18739	76	5	structure	structure	NOUN
fcis-18739	76	6	,	,	PUNCT
fcis-18739	76	7	this	this	DET
fcis-18739	76	8	algorithm	algorithm	NOUN
fcis-18739	76	9	can	can	AUX
fcis-18739	76	10	automatically	automatically	ADV
fcis-18739	76	11	learn	learn	VERB
fcis-18739	76	12	hierarchical	hierarchical	ADJ
fcis-18739	76	13	features	feature	NOUN
fcis-18739	76	14	in	in	ADP
fcis-18739	76	15	images	image	NOUN
fcis-18739	76	16	.	.	PUNCT
fcis-18739	77	1	by	by	ADP
fcis-18739	77	2	constructing	construct	VERB
fcis-18739	77	3	a	a	DET
fcis-18739	77	4	deep	deep	ADJ
fcis-18739	77	5	neural	neural	ADJ
fcis-18739	77	6	network	network	NOUN
fcis-18739	77	7	structure	structure	NOUN
fcis-18739	77	8	,	,	PUNCT
fcis-18739	77	9	algorithms	algorithm	NOUN
fcis-18739	77	10	can	can	AUX
fcis-18739	77	11	abstract	abstract	VERB
fcis-18739	77	12	and	and	CCONJ
fcis-18739	77	13	extract	extract	VERB
fcis-18739	77	14	feature	feature	NOUN
fcis-18739	77	15	information	information	NOUN
fcis-18739	77	16	from	from	ADP
fcis-18739	77	17	images	image	NOUN
fcis-18739	77	18	layer	layer	NOUN
fcis-18739	77	19	by	by	ADP
fcis-18739	77	20	layer	layer	NOUN
fcis-18739	77	21	,	,	PUNCT
fcis-18739	77	22	thereby	thereby	ADV
fcis-18739	77	23	more	more	ADV
fcis-18739	77	24	accurately	accurately	ADV
fcis-18739	77	25	identifying	identify	VERB
fcis-18739	77	26	different	different	ADJ
fcis-18739	77	27	regions	region	NOUN
fcis-18739	77	28	in	in	ADP
fcis-18739	77	29	the	the	DET
fcis-18739	77	30	image	image	NOUN
fcis-18739	77	31	.	.	PUNCT
fcis-18739	78	1	the	the	DET
fcis-18739	78	2	ability	ability	NOUN
fcis-18739	78	3	of	of	ADP
fcis-18739	78	4	automatic	automatic	ADJ
fcis-18739	78	5	feature	feature	NOUN
fcis-18739	78	6	learning	learn	VERB
fcis-18739	78	7	enables	enable	VERB
fcis-18739	78	8	dl	dl	PROPN
fcis-18739	78	9	algorithms	algorithm	NOUN
fcis-18739	78	10	to	to	PART
fcis-18739	78	11	have	have	VERB
fcis-18739	78	12	higher	high	ADJ
fcis-18739	78	13	flexibility	flexibility	NOUN
fcis-18739	78	14	and	and	CCONJ
fcis-18739	78	15	adaptability	adaptability	NOUN
fcis-18739	78	16	when	when	SCONJ
fcis-18739	78	17	dealing	deal	VERB
fcis-18739	78	18	with	with	ADP
fcis-18739	78	19	complex	complex	ADJ
fcis-18739	78	20	image	image	NOUN
fcis-18739	78	21	data	datum	NOUN
fcis-18739	78	22	.	.	PUNCT
fcis-18739	79	1	figure	figure	NOUN
fcis-18739	79	2	2	2	NUM
fcis-18739	79	3	.	.	PUNCT
fcis-18739	80	1	comparison	comparison	NOUN
fcis-18739	80	2	of	of	ADP
fcis-18739	80	3	segmentation	segmentation	NOUN
fcis-18739	80	4	precision	precision	NOUN
fcis-18739	80	5	figure	figure	NOUN
fcis-18739	80	6	3	3	NUM
fcis-18739	80	7	shows	show	VERB
fcis-18739	80	8	the	the	DET
fcis-18739	80	9	loss	loss	NOUN
fcis-18739	80	10	function	function	NOUN
fcis-18739	80	11	curve	curve	NOUN
fcis-18739	80	12	of	of	ADP
fcis-18739	80	13	the	the	DET
fcis-18739	80	14	algorithm	algorithm	NOUN
fcis-18739	80	15	in	in	ADP
fcis-18739	80	16	this	this	DET
fcis-18739	80	17	article	article	NOUN
fcis-18739	80	18	.	.	PUNCT
fcis-18739	81	1	it	it	PRON
fcis-18739	81	2	can	can	AUX
fcis-18739	81	3	be	be	AUX
fcis-18739	81	4	seen	see	VERB
fcis-18739	81	5	from	from	ADP
fcis-18739	81	6	the	the	DET
fcis-18739	81	7	figure	figure	NOUN
fcis-18739	81	8	that	that	SCONJ
fcis-18739	81	9	the	the	DET
fcis-18739	81	10	loss	loss	NOUN
fcis-18739	81	11	function	function	NOUN
fcis-18739	81	12	value	value	NOUN
fcis-18739	81	13	gradually	gradually	ADV
fcis-18739	81	14	decreases	decrease	VERB
fcis-18739	81	15	with	with	ADP
fcis-18739	81	16	the	the	DET
fcis-18739	81	17	increase	increase	NOUN
fcis-18739	81	18	of	of	ADP
fcis-18739	81	19	iteration	iteration	NOUN
fcis-18739	81	20	times	time	NOUN
fcis-18739	81	21	,	,	PUNCT
fcis-18739	81	22	indicating	indicate	VERB
fcis-18739	81	23	that	that	SCONJ
fcis-18739	81	24	the	the	DET
fcis-18739	81	25	algorithm	algorithm	NOUN
fcis-18739	81	26	continuously	continuously	ADV
fcis-18739	81	27	learns	learn	VERB
fcis-18739	81	28	and	and	CCONJ
fcis-18739	81	29	optimizes	optimize	VERB
fcis-18739	81	30	during	during	ADP
fcis-18739	81	31	the	the	DET
fcis-18739	81	32	training	training	NOUN
fcis-18739	81	33	process	process	NOUN
fcis-18739	81	34	.	.	PUNCT
fcis-18739	82	1	secondly	secondly	ADV
fcis-18739	82	2	,	,	PUNCT
fcis-18739	82	3	the	the	DET
fcis-18739	82	4	speed	speed	NOUN
fcis-18739	82	5	of	of	ADP
fcis-18739	82	6	curve	curve	NOUN
fcis-18739	82	7	descent	descent	NOUN
fcis-18739	82	8	is	be	AUX
fcis-18739	82	9	relatively	relatively	ADV
fcis-18739	82	10	rapid	rapid	ADJ
fcis-18739	82	11	in	in	ADP
fcis-18739	82	12	the	the	DET
fcis-18739	82	13	initial	initial	ADJ
fcis-18739	82	14	stage	stage	NOUN
fcis-18739	82	15	,	,	PUNCT
fcis-18739	82	16	and	and	CCONJ
fcis-18739	82	17	then	then	ADV
fcis-18739	82	18	gradually	gradually	ADV
fcis-18739	82	19	stabilizes	stabilize	VERB
fcis-18739	82	20	,	,	PUNCT
fcis-18739	82	21	which	which	PRON
fcis-18739	82	22	reflects	reflect	VERB
fcis-18739	82	23	the	the	DET
fcis-18739	82	24	algorithm	algorithm	NOUN
fcis-18739	82	25	's	's	PART
fcis-18739	82	26	ability	ability	NOUN
fcis-18739	82	27	to	to	PART
fcis-18739	82	28	converge	converge	VERB
fcis-18739	82	29	quickly	quickly	ADV
fcis-18739	82	30	in	in	ADP
fcis-18739	82	31	the	the	DET
fcis-18739	82	32	early	early	ADJ
fcis-18739	82	33	training	training	NOUN
fcis-18739	82	34	stage	stage	NOUN
fcis-18739	82	35	and	and	CCONJ
fcis-18739	82	36	gradually	gradually	ADV
fcis-18739	82	37	reach	reach	VERB
fcis-18739	82	38	a	a	DET
fcis-18739	82	39	stable	stable	ADJ
fcis-18739	82	40	state	state	NOUN
fcis-18739	82	41	in	in	ADP
fcis-18739	82	42	the	the	DET
fcis-18739	82	43	later	later	ADJ
fcis-18739	82	44	stage	stage	NOUN
fcis-18739	82	45	.	.	PUNCT
fcis-18739	83	1	this	this	PRON
fcis-18739	83	2	means	mean	VERB
fcis-18739	83	3	that	that	SCONJ
fcis-18739	83	4	the	the	DET
fcis-18739	83	5	algorithm	algorithm	NOUN
fcis-18739	83	6	proposed	propose	VERB
fcis-18739	83	7	in	in	ADP
fcis-18739	83	8	this	this	DET
fcis-18739	83	9	article	article	NOUN
fcis-18739	83	10	can	can	AUX
fcis-18739	83	11	better	well	ADV
fcis-18739	83	12	cope	cope	VERB
fcis-18739	83	13	with	with	ADP
fcis-18739	83	14	complex	complex	ADJ
fcis-18739	83	15	scenarios	scenario	NOUN
fcis-18739	83	16	and	and	CCONJ
fcis-18739	83	17	changes	change	NOUN
fcis-18739	83	18	,	,	PUNCT
fcis-18739	83	19	and	and	CCONJ
fcis-18739	83	20	has	have	VERB
fcis-18739	83	21	stronger	strong	ADJ
fcis-18739	83	22	robustness	robustness	NOUN
fcis-18739	83	23	and	and	CCONJ
fcis-18739	83	24	adaptability	adaptability	NOUN
fcis-18739	83	25	.	.	PUNCT
fcis-18739	84	1	5	5	X
fcis-18739	84	2	.	.	X
fcis-18739	84	3	conclusion	conclusion	NOUN
fcis-18739	84	4	with	with	ADP
fcis-18739	84	5	the	the	DET
fcis-18739	84	6	continuous	continuous	ADJ
fcis-18739	84	7	progress	progress	NOUN
fcis-18739	84	8	of	of	ADP
fcis-18739	84	9	computer	computer	NOUN
fcis-18739	84	10	technology	technology	NOUN
fcis-18739	84	11	and	and	CCONJ
fcis-18739	84	12	the	the	DET
fcis-18739	84	13	increasing	increase	VERB
fcis-18739	84	14	demand	demand	NOUN
fcis-18739	84	15	for	for	ADP
fcis-18739	84	16	image	image	NOUN
fcis-18739	84	17	recognition	recognition	NOUN
fcis-18739	84	18	precision	precision	NOUN
fcis-18739	84	19	by	by	ADP
fcis-18739	84	20	humans	human	NOUN
fcis-18739	84	21	,	,	PUNCT
fcis-18739	84	22	ai	ai	AUX
fcis-18739	84	23	algorithms	algorithm	NOUN
fcis-18739	84	24	are	be	AUX
fcis-18739	84	25	gradually	gradually	ADV
fcis-18739	84	26	integrating	integrate	VERB
fcis-18739	84	27	into	into	ADP
fcis-18739	84	28	various	various	ADJ
fcis-18739	84	29	industries	industry	NOUN
fcis-18739	84	30	,	,	PUNCT
fcis-18739	84	31	providing	provide	VERB
fcis-18739	84	32	effective	effective	ADJ
fcis-18739	84	33	solutions	solution	NOUN
fcis-18739	84	34	for	for	ADP
fcis-18739	84	35	various	various	ADJ
fcis-18739	84	36	problems	problem	NOUN
fcis-18739	84	37	.	.	PUNCT
fcis-18739	85	1	especially	especially	ADV
fcis-18739	85	2	in	in	ADP
fcis-18739	85	3	the	the	DET
fcis-18739	85	4	field	field	NOUN
fcis-18739	85	5	of	of	ADP
fcis-18739	85	6	image	image	NOUN
fcis-18739	85	7	segmentation	segmentation	NOUN
fcis-18739	85	8	,	,	PUNCT
fcis-18739	85	9	the	the	DET
fcis-18739	85	10	introduction	introduction	NOUN
fcis-18739	85	11	of	of	ADP
fcis-18739	85	12	dl	dl	PROPN
fcis-18739	85	13	has	have	AUX
fcis-18739	85	14	brought	bring	VERB
fcis-18739	85	15	unprecedented	unprecedented	ADJ
fcis-18739	85	16	development	development	NOUN
fcis-18739	85	17	opportunities	opportunity	NOUN
fcis-18739	85	18	in	in	ADP
fcis-18739	85	19	this	this	DET
fcis-18739	85	20	field	field	NOUN
fcis-18739	85	21	.	.	PUNCT
fcis-18739	86	1	image	image	NOUN
fcis-18739	86	2	segmentation	segmentation	NOUN
fcis-18739	86	3	,	,	PUNCT
fcis-18739	86	4	as	as	SCONJ
fcis-18739	86	5	one	one	NUM
fcis-18739	86	6	of	of	ADP
fcis-18739	86	7	the	the	DET
fcis-18739	86	8	key	key	ADJ
fcis-18739	86	9	technologies	technology	NOUN
fcis-18739	86	10	in	in	ADP
fcis-18739	86	11	cv	cv	PROPN
fcis-18739	86	12	,	,	PUNCT
fcis-18739	86	13	is	be	AUX
fcis-18739	86	14	of	of	ADP
fcis-18739	86	15	great	great	ADJ
fcis-18739	86	16	significance	significance	NOUN
fcis-18739	86	17	for	for	ADP
fcis-18739	86	18	the	the	DET
fcis-18739	86	19	understanding	understanding	NOUN
fcis-18739	86	20	and	and	CCONJ
fcis-18739	86	21	analysis	analysis	NOUN
fcis-18739	86	22	of	of	ADP
fcis-18739	86	23	images	image	NOUN
fcis-18739	86	24	.	.	PUNCT
fcis-18739	87	1	with	with	ADP
fcis-18739	87	2	the	the	DET
fcis-18739	87	3	rapid	rapid	ADJ
fcis-18739	87	4	development	development	NOUN
fcis-18739	87	5	of	of	ADP
fcis-18739	87	6	dl	dl	PROPN
fcis-18739	87	7	technology	technology	PROPN
fcis-18739	87	8	,	,	PUNCT
fcis-18739	87	9	especially	especially	ADV
fcis-18739	87	10	the	the	DET
fcis-18739	87	11	continuous	continuous	ADJ
fcis-18739	87	12	improvement	improvement	NOUN
fcis-18739	87	13	of	of	ADP
fcis-18739	87	14	models	model	NOUN
fcis-18739	87	15	such	such	ADJ
fcis-18739	87	16	as	as	ADP
fcis-18739	87	17	cnn	cnn	PROPN
fcis-18739	87	18	,	,	PUNCT
fcis-18739	87	19	the	the	DET
fcis-18739	87	20	performance	performance	NOUN
fcis-18739	87	21	of	of	ADP
fcis-18739	87	22	image	image	NOUN
fcis-18739	87	23	segmentation	segmentation	NOUN
fcis-18739	87	24	algorithms	algorithm	NOUN
fcis-18739	87	25	has	have	AUX
fcis-18739	87	26	been	be	AUX
fcis-18739	87	27	significantly	significantly	ADV
fcis-18739	87	28	improved	improve	VERB
fcis-18739	87	29	.	.	PUNCT
fcis-18739	88	1	these	these	DET
fcis-18739	88	2	algorithms	algorithm	NOUN
fcis-18739	88	3	can	can	AUX
fcis-18739	88	4	automatically	automatically	ADV
fcis-18739	88	5	learn	learn	VERB
fcis-18739	88	6	and	and	CCONJ
fcis-18739	88	7	extract	extract	VERB
fcis-18739	88	8	complex	complex	ADJ
fcis-18739	88	9	features	feature	NOUN
fcis-18739	88	10	from	from	ADP
fcis-18739	88	11	images	image	NOUN
fcis-18739	88	12	,	,	PUNCT
fcis-18739	88	13	achieving	achieve	VERB
fcis-18739	88	14	more	more	ADV
fcis-18739	88	15	accurate	accurate	ADJ
fcis-18739	88	16	and	and	CCONJ
fcis-18739	88	17	efficient	efficient	ADJ
fcis-18739	88	18	segmentation	segmentation	NOUN
fcis-18739	88	19	.	.	PUNCT
fcis-18739	89	1	with	with	ADP
fcis-18739	89	2	the	the	DET
fcis-18739	89	3	widespread	widespread	ADJ
fcis-18739	89	4	application	application	NOUN
fcis-18739	89	5	of	of	ADP
fcis-18739	89	6	dl	dl	PROPN
fcis-18739	89	7	,	,	PUNCT
fcis-18739	89	8	image	image	NOUN
fcis-18739	89	9	segmentation	segmentation	NOUN
fcis-18739	89	10	algorithms	algorithm	NOUN
fcis-18739	89	11	have	have	AUX
fcis-18739	89	12	become	become	VERB
fcis-18739	89	13	a	a	DET
fcis-18739	89	14	research	research	NOUN
fcis-18739	89	15	hotspot	hotspot	NOUN
fcis-18739	89	16	in	in	ADP
fcis-18739	89	17	the	the	DET
fcis-18739	89	18	field	field	NOUN
fcis-18739	89	19	of	of	ADP
fcis-18739	89	20	cv	cv	PROPN
fcis-18739	89	21	.	.	PUNCT
fcis-18739	90	1	this	this	DET
fcis-18739	90	2	trend	trend	NOUN
fcis-18739	90	3	has	have	AUX
fcis-18739	90	4	driven	drive	VERB
fcis-18739	90	5	the	the	DET
fcis-18739	90	6	rapid	rapid	ADJ
fcis-18739	90	7	development	development	NOUN
fcis-18739	90	8	of	of	ADP
fcis-18739	90	9	image	image	NOUN
fcis-18739	90	10	segmentation	segmentation	NOUN
fcis-18739	90	11	,	,	PUNCT
fcis-18739	90	12	providing	provide	VERB
fcis-18739	90	13	strong	strong	ADJ
fcis-18739	90	14	support	support	NOUN
fcis-18739	90	15	for	for	ADP
fcis-18739	90	16	many	many	ADJ
fcis-18739	90	17	fields	field	NOUN
fcis-18739	90	18	such	such	ADJ
fcis-18739	90	19	as	as	ADP
fcis-18739	90	20	autonomous	autonomous	ADJ
fcis-18739	90	21	20	20	NUM
fcis-18739	90	22	driving	driving	NOUN
fcis-18739	90	23	,	,	PUNCT
fcis-18739	90	24	medical	medical	ADJ
fcis-18739	90	25	image	image	NOUN
fcis-18739	90	26	analysis	analysis	NOUN
fcis-18739	90	27	,	,	PUNCT
fcis-18739	90	28	and	and	CCONJ
fcis-18739	90	29	security	security	NOUN
fcis-18739	90	30	monitoring	monitoring	NOUN
fcis-18739	90	31	.	.	PUNCT
fcis-18739	91	1	with	with	ADP
fcis-18739	91	2	the	the	DET
fcis-18739	91	3	further	further	ADJ
fcis-18739	91	4	development	development	NOUN
fcis-18739	91	5	of	of	ADP
fcis-18739	91	6	computer	computer	NOUN
fcis-18739	91	7	technology	technology	NOUN
fcis-18739	91	8	and	and	CCONJ
fcis-18739	91	9	the	the	DET
fcis-18739	91	10	continuous	continuous	ADJ
fcis-18739	91	11	enrichment	enrichment	NOUN
fcis-18739	91	12	of	of	ADP
fcis-18739	91	13	data	datum	NOUN
fcis-18739	91	14	resources	resource	NOUN
fcis-18739	91	15	,	,	PUNCT
fcis-18739	91	16	the	the	DET
fcis-18739	91	17	performance	performance	NOUN
fcis-18739	91	18	of	of	ADP
fcis-18739	91	19	image	image	NOUN
fcis-18739	91	20	segmentation	segmentation	NOUN
fcis-18739	91	21	algorithms	algorithm	NOUN
fcis-18739	91	22	will	will	AUX
fcis-18739	91	23	be	be	AUX
fcis-18739	91	24	further	far	ADV
fcis-18739	91	25	improved	improve	VERB
fcis-18739	91	26	.	.	PUNCT
fcis-18739	92	1	figure	figure	VERB
fcis-18739	92	2	3	3	NUM
fcis-18739	92	3	.	.	PUNCT
fcis-18739	92	4	loss	loss	NOUN
fcis-18739	92	5	function	function	NOUN
fcis-18739	92	6	curve	curve	NOUN
fcis-18739	92	7	references	reference	NOUN
fcis-18739	92	8	[	[	X
fcis-18739	92	9	1	1	NUM
fcis-18739	92	10	]	]	X
fcis-18739	92	11	liu	liu	PROPN
fcis-18739	92	12	siting	siting	PROPN
fcis-18739	92	13	,	,	PUNCT
fcis-18739	92	14	han	han	PROPN
fcis-18739	92	15	xiaoxia	xiaoxia	PROPN
fcis-18739	92	16	,	,	PUNCT
fcis-18739	92	17	zhao	zhao	PROPN
fcis-18739	92	18	xiaoning	xiaoning	PROPN
fcis-18739	92	19	,	,	PUNCT
fcis-18739	92	20	et	et	PROPN
fcis-18739	92	21	al	al	PROPN
fcis-18739	92	22	.	.	PUNCT
fcis-18739	93	1	image	image	NOUN
fcis-18739	93	2	segmentation	segmentation	NOUN
fcis-18739	93	3	based	base	VERB
fcis-18739	93	4	on	on	ADP
fcis-18739	93	5	superpixel	superpixel	NOUN
fcis-18739	93	6	and	and	CCONJ
fcis-18739	93	7	density	density	NOUN
fcis-18739	93	8	peak	peak	NOUN
fcis-18739	93	9	clustering	cluster	VERB
fcis-18739	93	10	[	[	X
fcis-18739	93	11	j	j	X
fcis-18739	93	12	]	]	X
fcis-18739	93	13	.	.	PUNCT
fcis-18739	94	1	journal	journal	PROPN
fcis-18739	94	2	of	of	ADP
fcis-18739	94	3	shanghai	shanghai	PROPN
fcis-18739	94	4	electric	electric	PROPN
fcis-18739	94	5	power	power	PROPN
fcis-18739	94	6	university	university	PROPN
fcis-18739	94	7	,	,	PUNCT
fcis-18739	94	8	2023	2023	NUM
fcis-18739	94	9	,	,	PUNCT
fcis-18739	94	10	39	39	NUM
fcis-18739	94	11	(	(	PUNCT
fcis-18739	94	12	2	2	NUM
fcis-18739	94	13	):	):	PUNCT
fcis-18739	94	14	182	182	NUM
fcis-18739	94	15	-	-	SYM
fcis-18739	94	16	188	188	NUM
fcis-18739	94	17	.	.	PUNCT
fcis-18739	95	1	[	[	X
fcis-18739	95	2	2	2	X
fcis-18739	95	3	]	]	X
fcis-18739	95	4	liu	liu	PROPN
fcis-18739	95	5	yi	yi	PROPN
fcis-18739	95	6	,	,	PUNCT
fcis-18739	95	7	zhang	zhang	PROPN
fcis-18739	95	8	xiaofeng	xiaofeng	PROPN
fcis-18739	95	9	,	,	PUNCT
fcis-18739	95	10	sun	sun	PROPN
fcis-18739	95	11	yujuan	yujuan	PROPN
fcis-18739	95	12	,	,	PUNCT
fcis-18739	95	13	et	et	PROPN
fcis-18739	95	14	al	al	PROPN
fcis-18739	95	15	.	.	PUNCT
fcis-18739	95	16	robust	robust	ADJ
fcis-18739	95	17	image	image	NOUN
fcis-18739	95	18	segmentation	segmentation	NOUN
fcis-18739	95	19	algorithm	algorithm	NOUN
fcis-18739	95	20	based	base	VERB
fcis-18739	95	21	on	on	ADP
fcis-18739	95	22	weighted	weight	VERB
fcis-18739	95	23	filtering	filtering	NOUN
fcis-18739	95	24	and	and	CCONJ
fcis-18739	95	25	kernel	kernel	PROPN
fcis-18739	95	26	metric	metric	PROPN
fcis-18739	96	1	[	[	X
fcis-18739	96	2	j	j	X
fcis-18739	96	3	]	]	X
fcis-18739	96	4	.	.	PUNCT
fcis-18739	97	1	progress	progress	NOUN
fcis-18739	97	2	of	of	ADP
fcis-18739	97	3	laser	laser	NOUN
fcis-18739	97	4	and	and	CCONJ
fcis-18739	97	5	optoelectronics	optoelectronic	NOUN
fcis-18739	97	6	,	,	PUNCT
fcis-18739	97	7	2024	2024	NUM
fcis-18739	97	8	,	,	PUNCT
fcis-18739	97	9	61	61	NUM
fcis-18739	97	10	(	(	PUNCT
fcis-18739	97	11	08	08	NUM
fcis-18739	97	12	):	):	PUNCT
fcis-18739	97	13	09	09	NUM
fcis-18739	97	14	.	.	PUNCT
fcis-18739	98	1	[	[	X
fcis-18739	98	2	3	3	X
fcis-18739	98	3	]	]	X
fcis-18739	98	4	tan	tan	PROPN
fcis-18739	98	5	zhiguo	zhiguo	PROPN
fcis-18739	98	6	,	,	PUNCT
fcis-18739	98	7	ou	ou	ADP
fcis-18739	98	8	jianping	jianping	NOUN
fcis-18739	98	9	,	,	PUNCT
fcis-18739	98	10	zhang	zhang	PROPN
fcis-18739	98	11	jun	jun	PROPN
fcis-18739	98	12	,	,	PUNCT
fcis-18739	98	13	et	et	PROPN
fcis-18739	98	14	al	al	PROPN
fcis-18739	98	15	.	.	PUNCT
fcis-18739	98	16	multi	multi	ADJ
fcis-18739	98	17	-	-	ADJ
fcis-18739	98	18	feature	feature	ADJ
fcis-18739	98	19	combination	combination	NOUN
fcis-18739	98	20	depth	depth	NOUN
fcis-18739	98	21	image	image	NOUN
fcis-18739	98	22	segmentation	segmentation	NOUN
fcis-18739	98	23	algorithm	algorithm	NOUN
fcis-18739	99	1	[	[	X
fcis-18739	99	2	j	j	X
fcis-18739	99	3	]	]	X
fcis-18739	99	4	.	.	PUNCT
fcis-18739	100	1	computer	computer	NOUN
fcis-18739	100	2	engineering	engineering	NOUN
fcis-18739	100	3	and	and	CCONJ
fcis-18739	100	4	science	science	NOUN
fcis-18739	100	5	,	,	PUNCT
fcis-18739	100	6	2018	2018	NUM
fcis-18739	100	7	,	,	PUNCT
fcis-18739	100	8	040(008):14291434	040(008):14291434	NUM
fcis-18739	100	9	.	.	PUNCT
fcis-18739	101	1	[	[	X
fcis-18739	101	2	4	4	X
fcis-18739	101	3	]	]	PUNCT
fcis-18739	101	4	gu	gu	NOUN
fcis-18739	101	5	pan	pan	PROPN
fcis-18739	101	6	,	,	PUNCT
fcis-18739	101	7	zhang	zhang	PROPN
fcis-18739	101	8	fengdong	fengdong	PROPN
fcis-18739	101	9	.	.	PUNCT
fcis-18739	101	10	weak	weak	ADJ
fcis-18739	101	11	supervised	supervised	ADJ
fcis-18739	101	12	semantic	semantic	ADJ
fcis-18739	101	13	image	image	NOUN
fcis-18739	101	14	segmentation	segmentation	NOUN
fcis-18739	101	15	algorithm	algorithm	NOUN
fcis-18739	101	16	based	base	VERB
fcis-18739	101	17	on	on	ADP
fcis-18739	101	18	neural	neural	ADJ
fcis-18739	101	19	network	network	NOUN
fcis-18739	102	1	[	[	X
fcis-18739	102	2	j	j	X
fcis-18739	102	3	]	]	X
fcis-18739	102	4	.	.	PUNCT
fcis-18739	103	1	computer	computer	NOUN
fcis-18739	103	2	applications	application	NOUN
fcis-18739	103	3	and	and	CCONJ
fcis-18739	103	4	software	software	NOUN
fcis-18739	103	5	,	,	PUNCT
fcis-18739	103	6	2018	2018	NUM
fcis-18739	103	7	,	,	PUNCT
fcis-18739	103	8	35(2):5	35(2):5	NUM
fcis-18739	103	9	.	.	PUNCT
fcis-18739	104	1	[	[	X
fcis-18739	104	2	5	5	NUM
fcis-18739	104	3	]	]	SYM
fcis-18739	104	4	li	li	PROPN
fcis-18739	104	5	xuan	xuan	PROPN
fcis-18739	104	6	,	,	PUNCT
fcis-18739	104	7	sun	sun	PROPN
fcis-18739	104	8	xinnan	xinnan	PROPN
fcis-18739	104	9	.	.	PUNCT
fcis-18739	105	1	image	image	NOUN
fcis-18739	105	2	segmentation	segmentation	NOUN
fcis-18739	105	3	algorithm	algorithm	NOUN
fcis-18739	105	4	based	base	VERB
fcis-18739	105	5	on	on	ADP
fcis-18739	105	6	convolutional	convolutional	ADJ
fcis-18739	105	7	neural	neural	ADJ
fcis-18739	105	8	network	network	NOUN
fcis-18739	106	1	[	[	X
fcis-18739	106	2	j	j	X
fcis-18739	106	3	]	]	X
fcis-18739	106	4	.	.	PUNCT
fcis-18739	107	1	journal	journal	PROPN
fcis-18739	107	2	of	of	ADP
fcis-18739	107	3	shenyang	shenyang	PROPN
fcis-18739	107	4	university	university	PROPN
fcis-18739	107	5	of	of	ADP
fcis-18739	107	6	aeronautics	aeronautics	PROPN
fcis-18739	107	7	and	and	CCONJ
fcis-18739	107	8	astronautics	astronautic	NOUN
fcis-18739	107	9	,	,	PUNCT
fcis-18739	107	10	2020(1):50	2020(1):50	NUM
fcis-18739	107	11	-	-	SYM
fcis-18739	107	12	57	57	NUM
fcis-18739	107	13	.	.	PUNCT
fcis-18739	108	1	[	[	X
fcis-18739	108	2	6	6	NUM
fcis-18739	108	3	]	]	X
fcis-18739	108	4	guan	guan	PROPN
fcis-18739	108	5	shenke	shenke	PROPN
fcis-18739	108	6	,	,	PUNCT
fcis-18739	108	7	lin	lin	PROPN
fcis-18739	108	8	xiao	xiao	PROPN
fcis-18739	108	9	,	,	PUNCT
fcis-18739	108	10	zheng	zheng	PROPN
fcis-18739	108	11	xiaomei	xiaomei	PROPN
fcis-18739	108	12	,	,	PUNCT
fcis-18739	108	13	zhu	zhu	PROPN
fcis-18739	108	14	yuanyuan	yuanyuan	PROPN
fcis-18739	108	15	,	,	PUNCT
fcis-18739	108	16	malizhuang	malizhuang	PROPN
fcis-18739	108	17	.	.	PUNCT
fcis-18739	109	1	multi	multi	ADJ
fcis-18739	109	2	-	-	ADJ
fcis-18739	109	3	scale	scale	ADJ
fcis-18739	109	4	feature	feature	NOUN
fcis-18739	109	5	fusion	fusion	NOUN
fcis-18739	109	6	image	image	NOUN
fcis-18739	109	7	semantic	semantic	ADJ
fcis-18739	109	8	segmentation	segmentation	NOUN
fcis-18739	109	9	algorithm	algorithm	NOUN
fcis-18739	109	10	combined	combine	VERB
fcis-18739	109	11	with	with	ADP
fcis-18739	109	12	superpixel	superpixel	ADJ
fcis-18739	109	13	segmentation	segmentation	NOUN
fcis-18739	110	1	[	[	X
fcis-18739	110	2	j	j	X
fcis-18739	110	3	]	]	X
fcis-18739	110	4	.	.	PUNCT
fcis-18739	111	1	acta	acta	PROPN
fcis-18739	111	2	graphics	graphics	PROPN
fcis-18739	111	3	,	,	PUNCT
fcis-18739	111	4	2021	2021	NUM
fcis-18739	111	5	,	,	PUNCT
fcis-18739	111	6	042(003):406	042(003):406	VERB
fcis-18739	111	7	-	-	PUNCT
fcis-18739	111	8	413	413	NUM
fcis-18739	111	9	.	.	PUNCT
fcis-18739	112	1	[	[	X
fcis-18739	112	2	7	7	X
fcis-18739	112	3	]	]	X
fcis-18739	112	4	zhang	zhang	PROPN
fcis-18739	112	5	xin	xin	PROPN
fcis-18739	112	6	,	,	PUNCT
fcis-18739	112	7	zhang	zhang	PROPN
fcis-18739	112	8	xuyang	xuyang	PROPN
fcis-18739	112	9	,	,	PUNCT
fcis-18739	112	10	mao	mao	PROPN
fcis-18739	112	11	yuxin	yuxin	PROPN
fcis-18739	112	12	,	,	PUNCT
fcis-18739	112	13	et	et	PROPN
fcis-18739	112	14	al	al	PROPN
fcis-18739	112	15	.	.	PROPN
fcis-18739	112	16	overview	overview	NOUN
fcis-18739	112	17	of	of	ADP
fcis-18739	112	18	image	image	NOUN
fcis-18739	112	19	semantic	semantic	ADJ
fcis-18739	112	20	segmentation	segmentation	NOUN
fcis-18739	112	21	algorithms	algorithm	NOUN
fcis-18739	112	22	based	base	VERB
fcis-18739	112	23	on	on	ADP
fcis-18739	112	24	weak	weak	ADJ
fcis-18739	112	25	supervised	supervised	ADJ
fcis-18739	112	26	learning	learning	NOUN
fcis-18739	112	27	[	[	X
fcis-18739	112	28	j	j	X
fcis-18739	112	29	]	]	X
fcis-18739	112	30	.	.	PUNCT
fcis-18739	113	1	changjiang	changjiang	PROPN
fcis-18739	113	2	information	information	NOUN
fcis-18739	113	3	communication	communication	NOUN
fcis-18739	113	4	,	,	PUNCT
fcis-18739	113	5	2023	2023	NUM
fcis-18739	113	6	,	,	PUNCT
fcis-18739	113	7	36(5):105	36(5):105	NUM
fcis-18739	113	8	-	-	PUNCT
fcis-18739	113	9	108	108	NUM
fcis-18739	113	10	.	.	PUNCT
fcis-18739	114	1	[	[	X
fcis-18739	114	2	8	8	NUM
fcis-18739	114	3	]	]	X
fcis-18739	114	4	lv	lv	PROPN
fcis-18739	114	5	xin	xin	PROPN
fcis-18739	114	6	,	,	PUNCT
fcis-18739	114	7	mu	mu	PROPN
fcis-18739	114	8	xiaodong	xiaodong	PROPN
fcis-18739	114	9	,	,	PUNCT
fcis-18739	114	10	zhang	zhang	PROPN
fcis-18739	114	11	jun	jun	PROPN
fcis-18739	114	12	.	.	PROPN
fcis-18739	114	13	improved	improve	VERB
fcis-18739	114	14	fuzzy	fuzzy	ADJ
fcis-18739	114	15	c	c	NOUN
fcis-18739	114	16	-	-	PUNCT
fcis-18739	114	17	means	mean	VERB
fcis-18739	114	18	image	image	NOUN
fcis-18739	114	19	segmentation	segmentation	NOUN
fcis-18739	114	20	algorithm	algorithm	NOUN
fcis-18739	114	21	based	base	VERB
fcis-18739	114	22	on	on	ADP
fcis-18739	114	23	spatial	spatial	ADJ
fcis-18739	114	24	information	information	NOUN
fcis-18739	115	1	[	[	X
fcis-18739	115	2	j	j	X
fcis-18739	115	3	]	]	X
fcis-18739	115	4	.	.	PUNCT
fcis-18739	116	1	journal	journal	PROPN
fcis-18739	116	2	of	of	ADP
fcis-18739	116	3	rocket	rocket	NOUN
fcis-18739	116	4	force	force	NOUN
fcis-18739	116	5	university	university	PROPN
fcis-18739	116	6	of	of	ADP
fcis-18739	116	7	engineering	engineering	NOUN
fcis-18739	116	8	(	(	PUNCT
fcis-18739	116	9	natural	natural	ADJ
fcis-18739	116	10	science	science	NOUN
fcis-18739	116	11	edition	edition	NOUN
fcis-18739	116	12	)	)	PUNCT
fcis-18739	116	13	,	,	PUNCT
fcis-18739	116	14	2020(2):8	2020(2):8	PROPN
fcis-18739	116	15	.	.	PUNCT
fcis-18739	117	1	[	[	X
fcis-18739	117	2	9	9	NUM
fcis-18739	117	3	]	]	X
fcis-18739	117	4	chen	chen	PROPN
fcis-18739	117	5	xingzhi	xingzhi	PROPN
fcis-18739	117	6	,	,	PUNCT
fcis-18739	117	7	wang	wang	PROPN
fcis-18739	117	8	daiwen	daiwen	PROPN
fcis-18739	117	9	,	,	PUNCT
fcis-18739	117	10	liu	liu	PROPN
fcis-18739	117	11	naiyao	naiyao	PROPN
fcis-18739	117	12	,	,	PUNCT
fcis-18739	117	13	et	et	PROPN
fcis-18739	117	14	al	al	PROPN
fcis-18739	117	15	.	.	PUNCT
fcis-18739	117	16	application	application	NOUN
fcis-18739	117	17	of	of	ADP
fcis-18739	117	18	fast	fast	ADJ
fcis-18739	117	19	image	image	NOUN
fcis-18739	117	20	segmentation	segmentation	NOUN
fcis-18739	117	21	algorithm	algorithm	NOUN
fcis-18739	117	22	model	model	NOUN
fcis-18739	117	23	based	base	VERB
fcis-18739	117	24	on	on	ADP
fcis-18739	117	25	psokmeans	psokmean	NOUN
fcis-18739	117	26	[	[	X
fcis-18739	117	27	j	j	X
fcis-18739	117	28	]	]	X
fcis-18739	117	29	.	.	PUNCT
fcis-18739	118	1	modern	modern	ADJ
fcis-18739	118	2	information	information	NOUN
fcis-18739	118	3	technology	technology	NOUN
fcis-18739	118	4	,	,	PUNCT
fcis-18739	118	5	2020	2020	NUM
fcis-18739	118	6	,	,	PUNCT
fcis-18739	118	7	4(5):4	4(5):4	PROPN
fcis-18739	118	8	.	.	PUNCT
fcis-18739	119	1	[	[	X
fcis-18739	119	2	10	10	NUM
fcis-18739	119	3	]	]	X
fcis-18739	119	4	wu	wu	PROPN
fcis-18739	119	5	shangzhi	shangzhi	PROPN
fcis-18739	119	6	,	,	PUNCT
fcis-18739	119	7	duan	duan	PROPN
fcis-18739	119	8	chao	chao	PROPN
fcis-18739	119	9	,	,	PUNCT
fcis-18739	119	10	she	she	PRON
fcis-18739	119	11	zhiyong	zhiyong	NOUN
fcis-18739	119	12	.	.	PUNCT
fcis-18739	120	1	image	image	PROPN
fcis-18739	120	2	multithreshold	multithreshold	PROPN
fcis-18739	120	3	segmentation	segmentation	NOUN
fcis-18739	120	4	algorithm	algorithm	NOUN
fcis-18739	120	5	based	base	VERB
fcis-18739	120	6	on	on	ADP
fcis-18739	120	7	variable	variable	ADJ
fcis-18739	120	8	precision	precision	NOUN
fcis-18739	120	9	rough	rough	ADJ
fcis-18739	120	10	set	set	NOUN
fcis-18739	120	11	and	and	CCONJ
fcis-18739	120	12	particle	particle	NOUN
fcis-18739	120	13	swarm	swarm	NOUN
fcis-18739	120	14	optimization	optimization	NOUN
fcis-18739	121	1	[	[	X
fcis-18739	121	2	j	j	X
fcis-18739	121	3	]	]	X
fcis-18739	121	4	.	.	PUNCT
fcis-18739	122	1	journal	journal	PROPN
fcis-18739	122	2	of	of	ADP
fcis-18739	122	3	northwest	northwest	PROPN
fcis-18739	122	4	normal	normal	ADJ
fcis-18739	122	5	university	university	NOUN
fcis-18739	122	6	:	:	PUNCT
fcis-18739	122	7	natural	natural	ADJ
fcis-18739	122	8	science	science	NOUN
fcis-18739	122	9	edition	edition	NOUN
fcis-18739	122	10	,	,	PUNCT
fcis-18739	122	11	2019	2019	NUM
fcis-18739	122	12	,	,	PUNCT
fcis-18739	122	13	55(2):8	55(2):8	PROPN
fcis-18739	122	14	.	.	PUNCT