id	sid	tid	token	lemma	pos
fcis-25121	1	1	frontiers	frontier	NOUN
fcis-25121	1	2	in	in	ADP
fcis-25121	1	3	computing	computing	NOUN
fcis-25121	1	4	and	and	CCONJ
fcis-25121	1	5	intelligent	intelligent	ADJ
fcis-25121	1	6	systems	system	NOUN
fcis-25121	1	7	issn	issn	VERB
fcis-25121	1	8	:	:	PUNCT
fcis-25121	1	9	2832	2832	NUM
fcis-25121	1	10	-	-	SYM
fcis-25121	1	11	6024	6024	NUM
fcis-25121	1	12	|	|	NOUN
fcis-25121	1	13	vol	vol	NOUN
fcis-25121	1	14	.	.	PROPN
fcis-25121	2	1	9	9	NUM
fcis-25121	2	2	,	,	PUNCT
fcis-25121	2	3	no	no	INTJ
fcis-25121	2	4	.	.	NOUN
fcis-25121	2	5	2	2	NUM
fcis-25121	2	6	,	,	PUNCT
fcis-25121	2	7	2024	2024	NUM
fcis-25121	2	8	70	70	NUM
fcis-25121	2	9	research	research	NOUN
fcis-25121	2	10	on	on	ADP
fcis-25121	2	11	dynasty	dynasty	ADJ
fcis-25121	2	12	classification	classification	NOUN
fcis-25121	2	13	of	of	ADP
fcis-25121	2	14	dunhuang	dunhuang	NOUN
fcis-25121	2	15	murals	mural	NOUN
fcis-25121	2	16	based	base	VERB
fcis-25121	2	17	on	on	ADP
fcis-25121	2	18	convolutional	convolutional	ADJ
fcis-25121	2	19	neural	neural	ADJ
fcis-25121	2	20	networks	network	NOUN
fcis-25121	2	21	yingdi	yingdi	PROPN
fcis-25121	2	22	wan	wan	PROPN
fcis-25121	2	23	*	*	PUNCT
fcis-25121	2	24	academy	academy	PROPN
fcis-25121	2	25	of	of	ADP
fcis-25121	2	26	arts	art	NOUN
fcis-25121	2	27	and	and	CCONJ
fcis-25121	2	28	design	design	NOUN
fcis-25121	2	29	,	,	PUNCT
fcis-25121	2	30	qingdao	qingdao	PROPN
fcis-25121	2	31	city	city	PROPN
fcis-25121	2	32	university	university	PROPN
fcis-25121	2	33	,	,	PUNCT
fcis-25121	2	34	qingdao	qingdao	PROPN
fcis-25121	2	35	,	,	PUNCT
fcis-25121	2	36	shandong	shandong	PROPN
fcis-25121	2	37	,	,	PUNCT
fcis-25121	2	38	266106	266106	NUM
fcis-25121	2	39	,	,	PUNCT
fcis-25121	2	40	china	china	PROPN
fcis-25121	2	41	*	*	PUNCT
fcis-25121	2	42	corresponding	correspond	VERB
fcis-25121	2	43	author	author	NOUN
fcis-25121	2	44	email	email	NOUN
fcis-25121	2	45	:	:	PUNCT
fcis-25121	2	46	tinawan0821@163.com	tinawan0821@163.com	X
fcis-25121	3	1	abstract	abstract	ADJ
fcis-25121	3	2	:	:	PUNCT
fcis-25121	3	3	dunhuang	dunhuang	PROPN
fcis-25121	3	4	murals	mural	NOUN
fcis-25121	3	5	are	be	AUX
fcis-25121	3	6	one	one	NUM
fcis-25121	3	7	of	of	ADP
fcis-25121	3	8	china	china	PROPN
fcis-25121	3	9	's	's	PART
fcis-25121	3	10	intangible	intangible	ADJ
fcis-25121	3	11	cultural	cultural	ADJ
fcis-25121	3	12	heritages	heritage	NOUN
fcis-25121	3	13	,	,	PUNCT
fcis-25121	3	14	which	which	PRON
fcis-25121	3	15	have	have	AUX
fcis-25121	3	16	received	receive	VERB
fcis-25121	3	17	extensive	extensive	ADJ
fcis-25121	3	18	attention	attention	NOUN
fcis-25121	3	19	.	.	PUNCT
fcis-25121	4	1	identifying	identify	VERB
fcis-25121	4	2	and	and	CCONJ
fcis-25121	4	3	classifying	classify	VERB
fcis-25121	4	4	the	the	DET
fcis-25121	4	5	dynasties	dynasty	NOUN
fcis-25121	4	6	of	of	ADP
fcis-25121	4	7	ancient	ancient	ADJ
fcis-25121	4	8	murals	mural	NOUN
fcis-25121	4	9	quickly	quickly	ADV
fcis-25121	4	10	and	and	CCONJ
fcis-25121	4	11	accurately	accurately	ADV
fcis-25121	4	12	is	be	AUX
fcis-25121	4	13	extremely	extremely	ADV
fcis-25121	4	14	important	important	ADJ
fcis-25121	4	15	for	for	ADP
fcis-25121	4	16	the	the	DET
fcis-25121	4	17	study	study	NOUN
fcis-25121	4	18	of	of	ADP
fcis-25121	4	19	dunhuang	dunhuang	PROPN
fcis-25121	4	20	murals	mural	NOUN
fcis-25121	4	21	and	and	CCONJ
fcis-25121	4	22	the	the	DET
fcis-25121	4	23	digital	digital	ADJ
fcis-25121	4	24	protection	protection	NOUN
fcis-25121	4	25	and	and	CCONJ
fcis-25121	4	26	inheritance	inheritance	NOUN
fcis-25121	4	27	.	.	PUNCT
fcis-25121	5	1	a	a	DET
fcis-25121	5	2	method	method	NOUN
fcis-25121	5	3	to	to	PART
fcis-25121	5	4	classify	classify	VERB
fcis-25121	5	5	dunhuang	dunhuang	NOUN
fcis-25121	5	6	murals	mural	NOUN
fcis-25121	5	7	dynasties	dynasty	NOUN
fcis-25121	5	8	based	base	VERB
fcis-25121	5	9	on	on	ADP
fcis-25121	5	10	convolutional	convolutional	ADJ
fcis-25121	5	11	neural	neural	ADJ
fcis-25121	5	12	network	network	NOUN
fcis-25121	5	13	(	(	PUNCT
fcis-25121	5	14	cnn	cnn	PROPN
fcis-25121	5	15	)	)	PUNCT
fcis-25121	5	16	is	be	AUX
fcis-25121	5	17	proposed	propose	VERB
fcis-25121	5	18	in	in	ADP
fcis-25121	5	19	this	this	DET
fcis-25121	5	20	paper	paper	NOUN
fcis-25121	5	21	.	.	PUNCT
fcis-25121	6	1	first	first	ADV
fcis-25121	6	2	,	,	PUNCT
fcis-25121	6	3	the	the	DET
fcis-25121	6	4	dunhuang	dunhuang	NOUN
fcis-25121	6	5	murals	mural	VERB
fcis-25121	6	6	data	datum	NOUN
fcis-25121	6	7	set	set	VERB
fcis-25121	6	8	is	be	AUX
fcis-25121	6	9	constructed	construct	VERB
fcis-25121	6	10	from	from	ADP
fcis-25121	6	11	the	the	DET
fcis-25121	6	12	mural	mural	ADJ
fcis-25121	6	13	materials	material	NOUN
fcis-25121	6	14	;	;	PUNCT
fcis-25121	6	15	then	then	ADV
fcis-25121	6	16	,	,	PUNCT
fcis-25121	6	17	four	four	NUM
fcis-25121	6	18	convolutional	convolutional	ADJ
fcis-25121	6	19	neural	neural	ADJ
fcis-25121	6	20	network	network	NOUN
fcis-25121	6	21	models	model	NOUN
fcis-25121	6	22	with	with	ADP
fcis-25121	6	23	different	different	ADJ
fcis-25121	6	24	depths	depth	NOUN
fcis-25121	6	25	and	and	CCONJ
fcis-25121	6	26	structures	structure	NOUN
fcis-25121	6	27	are	be	AUX
fcis-25121	6	28	constructed	construct	VERB
fcis-25121	6	29	and	and	CCONJ
fcis-25121	6	30	trained	train	VERB
fcis-25121	6	31	;	;	PUNCT
fcis-25121	6	32	finally	finally	ADV
fcis-25121	6	33	,	,	PUNCT
fcis-25121	6	34	the	the	DET
fcis-25121	6	35	network	network	NOUN
fcis-25121	6	36	model	model	NOUN
fcis-25121	6	37	is	be	AUX
fcis-25121	6	38	tested	test	VERB
fcis-25121	6	39	using	use	VERB
fcis-25121	6	40	the	the	DET
fcis-25121	6	41	test	test	NOUN
fcis-25121	6	42	set	set	VERB
fcis-25121	6	43	and	and	CCONJ
fcis-25121	6	44	an	an	DET
fcis-25121	6	45	appropriate	appropriate	ADJ
fcis-25121	6	46	classification	classification	NOUN
fcis-25121	6	47	model	model	NOUN
fcis-25121	6	48	is	be	AUX
fcis-25121	6	49	selected	select	VERB
fcis-25121	6	50	.	.	PUNCT
fcis-25121	7	1	the	the	DET
fcis-25121	7	2	experimental	experimental	ADJ
fcis-25121	7	3	results	result	NOUN
fcis-25121	7	4	show	show	VERB
fcis-25121	7	5	that	that	SCONJ
fcis-25121	7	6	for	for	ADP
fcis-25121	7	7	the	the	DET
fcis-25121	7	8	dunhuang	dunhuang	PROPN
fcis-25121	7	9	mural	mural	ADJ
fcis-25121	7	10	image	image	NOUN
fcis-25121	7	11	data	datum	NOUN
fcis-25121	7	12	sets	set	NOUN
fcis-25121	7	13	of	of	ADP
fcis-25121	7	14	five	five	NUM
fcis-25121	7	15	different	different	ADJ
fcis-25121	7	16	dynasties	dynasty	NOUN
fcis-25121	7	17	in	in	ADP
fcis-25121	7	18	this	this	DET
fcis-25121	7	19	paper	paper	NOUN
fcis-25121	7	20	,	,	PUNCT
fcis-25121	7	21	the	the	DET
fcis-25121	7	22	four	four	NUM
fcis-25121	7	23	models	model	NOUN
fcis-25121	7	24	obtain	obtain	VERB
fcis-25121	7	25	high	high	ADJ
fcis-25121	7	26	classification	classification	NOUN
fcis-25121	7	27	accuracy	accuracy	NOUN
fcis-25121	7	28	,	,	PUNCT
fcis-25121	7	29	and	and	CCONJ
fcis-25121	7	30	the	the	DET
fcis-25121	7	31	accuracy	accuracy	NOUN
fcis-25121	7	32	of	of	ADP
fcis-25121	7	33	vgg11	vgg11	PROPN
fcis-25121	7	34	and	and	CCONJ
fcis-25121	7	35	vgg19	vgg19	PROPN
fcis-25121	7	36	reach	reach	VERB
fcis-25121	7	37	96	96	NUM
fcis-25121	7	38	%	%	NOUN
fcis-25121	7	39	.	.	PUNCT
fcis-25121	8	1	among	among	ADP
fcis-25121	8	2	them	they	PRON
fcis-25121	8	3	,	,	PUNCT
fcis-25121	8	4	the	the	DET
fcis-25121	8	5	classification	classification	NOUN
fcis-25121	8	6	accuracy	accuracy	NOUN
fcis-25121	8	7	of	of	ADP
fcis-25121	8	8	murals	mural	NOUN
fcis-25121	8	9	in	in	ADP
fcis-25121	8	10	the	the	DET
fcis-25121	8	11	sui	sui	PROPN
fcis-25121	8	12	dynasty	dynasty	NOUN
fcis-25121	8	13	and	and	CCONJ
fcis-25121	8	14	the	the	DET
fcis-25121	8	15	five	five	NUM
fcis-25121	8	16	dynasties	dynasty	NOUN
fcis-25121	8	17	and	and	CCONJ
fcis-25121	8	18	song	song	NOUN
fcis-25121	8	19	dynasties	dynasty	NOUN
fcis-25121	8	20	is	be	AUX
fcis-25121	8	21	lower	low	ADJ
fcis-25121	8	22	than	than	ADP
fcis-25121	8	23	that	that	PRON
fcis-25121	8	24	of	of	ADP
fcis-25121	8	25	other	other	ADJ
fcis-25121	8	26	dynasties	dynasty	NOUN
fcis-25121	8	27	in	in	ADP
fcis-25121	8	28	this	this	DET
fcis-25121	8	29	paper	paper	NOUN
fcis-25121	8	30	.	.	PUNCT
fcis-25121	9	1	keywords	keyword	NOUN
fcis-25121	9	2	:	:	PUNCT
fcis-25121	9	3	intangible	intangible	ADJ
fcis-25121	9	4	cultural	cultural	ADJ
fcis-25121	9	5	heritage	heritage	NOUN
fcis-25121	9	6	;	;	PUNCT
fcis-25121	9	7	digital	digital	ADJ
fcis-25121	9	8	protection	protection	NOUN
fcis-25121	9	9	;	;	PUNCT
fcis-25121	9	10	convolutional	convolutional	ADJ
fcis-25121	9	11	neural	neural	ADJ
fcis-25121	9	12	network	network	NOUN
fcis-25121	9	13	(	(	PUNCT
fcis-25121	9	14	cnn	cnn	PROPN
fcis-25121	9	15	)	)	PUNCT
fcis-25121	9	16	;	;	PUNCT
fcis-25121	9	17	image	image	NOUN
fcis-25121	9	18	classification	classification	NOUN
fcis-25121	9	19	;	;	PUNCT
fcis-25121	9	20	dynasty	dynasty	ADJ
fcis-25121	9	21	identification	identification	NOUN
fcis-25121	9	22	.	.	PUNCT
fcis-25121	10	1	1	1	X
fcis-25121	10	2	.	.	X
fcis-25121	10	3	introduction	introduction	NOUN
fcis-25121	10	4	the	the	DET
fcis-25121	10	5	mogao	mogao	NOUN
fcis-25121	10	6	grottoes	grotto	NOUN
fcis-25121	10	7	,	,	PUNCT
fcis-25121	10	8	as	as	ADP
fcis-25121	10	9	one	one	NUM
fcis-25121	10	10	of	of	ADP
fcis-25121	10	11	the	the	DET
fcis-25121	10	12	world	world	NOUN
fcis-25121	10	13	cultural	cultural	ADJ
fcis-25121	10	14	heritages	heritage	NOUN
fcis-25121	10	15	,	,	PUNCT
fcis-25121	10	16	located	locate	VERB
fcis-25121	10	17	in	in	ADP
fcis-25121	10	18	dunhuang	dunhuang	PROPN
fcis-25121	10	19	city	city	PROPN
fcis-25121	10	20	,	,	PUNCT
fcis-25121	10	21	gansu	gansu	PROPN
fcis-25121	10	22	province	province	PROPN
fcis-25121	10	23	,	,	PUNCT
fcis-25121	10	24	china	china	PROPN
fcis-25121	10	25	,	,	PUNCT
fcis-25121	10	26	are	be	AUX
fcis-25121	10	27	the	the	DET
fcis-25121	10	28	largest	large	ADJ
fcis-25121	10	29	and	and	CCONJ
fcis-25121	10	30	most	most	ADV
fcis-25121	10	31	abundant	abundant	ADJ
fcis-25121	10	32	chinese	chinese	ADJ
fcis-25121	10	33	buddhist	buddhist	ADJ
fcis-25121	10	34	art	art	NOUN
fcis-25121	10	35	sanctuary	sanctuary	ADJ
fcis-25121	10	36	in	in	ADP
fcis-25121	10	37	the	the	DET
fcis-25121	10	38	world	world	NOUN
fcis-25121	10	39	.	.	PUNCT
fcis-25121	11	1	the	the	DET
fcis-25121	11	2	grottoes	grotto	NOUN
fcis-25121	11	3	contain	contain	VERB
fcis-25121	11	4	some	some	PRON
fcis-25121	11	5	of	of	ADP
fcis-25121	11	6	the	the	DET
fcis-25121	11	7	finest	fine	ADJ
fcis-25121	11	8	examples	example	NOUN
fcis-25121	11	9	of	of	ADP
fcis-25121	11	10	buddhist	buddhist	ADJ
fcis-25121	11	11	art	art	NOUN
fcis-25121	11	12	spanning	span	VERB
fcis-25121	11	13	a	a	DET
fcis-25121	11	14	period	period	NOUN
fcis-25121	11	15	of	of	ADP
fcis-25121	11	16	1000	1000	NUM
fcis-25121	11	17	years	year	NOUN
fcis-25121	11	18	.	.	PUNCT
fcis-25121	12	1	these	these	DET
fcis-25121	12	2	art	art	NOUN
fcis-25121	12	3	works	work	NOUN
fcis-25121	12	4	are	be	AUX
fcis-25121	12	5	treasures	treasure	NOUN
fcis-25121	12	6	of	of	ADP
fcis-25121	12	7	human	human	ADJ
fcis-25121	12	8	civilazation	civilazation	NOUN
fcis-25121	12	9	,	,	PUNCT
fcis-25121	12	10	providing	provide	VERB
fcis-25121	12	11	valuable	valuable	ADJ
fcis-25121	12	12	material	material	NOUN
fcis-25121	12	13	for	for	ADP
fcis-25121	12	14	studies	study	NOUN
fcis-25121	12	15	of	of	ADP
fcis-25121	12	16	the	the	DET
fcis-25121	12	17	foreign	foreign	ADJ
fcis-25121	12	18	exchanges	exchange	NOUN
fcis-25121	12	19	of	of	ADP
fcis-25121	12	20	china	china	PROPN
fcis-25121	12	21	in	in	ADP
fcis-25121	12	22	different	different	ADJ
fcis-25121	12	23	dynasties	dynasty	NOUN
fcis-25121	12	24	.	.	PUNCT
fcis-25121	13	1	but	but	CCONJ
fcis-25121	13	2	the	the	DET
fcis-25121	13	3	dynasties	dynasty	NOUN
fcis-25121	13	4	of	of	ADP
fcis-25121	13	5	dunhuang	dunhuang	NOUN
fcis-25121	13	6	murals	mural	NOUN
fcis-25121	13	7	are	be	AUX
fcis-25121	13	8	difficult	difficult	ADJ
fcis-25121	13	9	to	to	PART
fcis-25121	13	10	classify	classify	VERB
fcis-25121	13	11	for	for	ADP
fcis-25121	13	12	ordinary	ordinary	ADJ
fcis-25121	13	13	people	people	NOUN
fcis-25121	13	14	as	as	SCONJ
fcis-25121	13	15	the	the	DET
fcis-25121	13	16	murals	mural	NOUN
fcis-25121	13	17	seem	seem	VERB
fcis-25121	13	18	to	to	PART
fcis-25121	13	19	be	be	AUX
fcis-25121	13	20	so	so	ADV
fcis-25121	13	21	much	much	ADV
fcis-25121	13	22	alike	alike	ADV
fcis-25121	13	23	.	.	PUNCT
fcis-25121	14	1	even	even	ADV
fcis-25121	14	2	the	the	DET
fcis-25121	14	3	experts	expert	NOUN
fcis-25121	14	4	need	need	VERB
fcis-25121	14	5	analysis	analysis	NOUN
fcis-25121	14	6	of	of	ADP
fcis-25121	14	7	a	a	DET
fcis-25121	14	8	vast	vast	ADJ
fcis-25121	14	9	amount	amount	NOUN
fcis-25121	14	10	of	of	ADP
fcis-25121	14	11	historical	historical	ADJ
fcis-25121	14	12	documents	document	NOUN
fcis-25121	14	13	.	.	PUNCT
fcis-25121	15	1	therefore	therefore	ADV
fcis-25121	15	2	,	,	PUNCT
fcis-25121	15	3	murals	mural	NOUN
fcis-25121	15	4	classification	classification	NOUN
fcis-25121	15	5	has	have	VERB
fcis-25121	15	6	many	many	ADJ
fcis-25121	15	7	challenges	challenge	NOUN
fcis-25121	15	8	compared	compare	VERB
fcis-25121	15	9	with	with	ADP
fcis-25121	15	10	other	other	ADJ
fcis-25121	15	11	types	type	NOUN
fcis-25121	15	12	of	of	ADP
fcis-25121	15	13	images	image	NOUN
fcis-25121	15	14	classification	classification	NOUN
fcis-25121	15	15	.	.	PUNCT
fcis-25121	16	1	1	1	NUM
fcis-25121	16	2	)	)	PUNCT
fcis-25121	16	3	compared	compare	VERB
fcis-25121	16	4	with	with	ADP
fcis-25121	16	5	general	general	ADJ
fcis-25121	16	6	images	image	NOUN
fcis-25121	16	7	,	,	PUNCT
fcis-25121	16	8	mural	mural	ADJ
fcis-25121	16	9	images	image	NOUN
fcis-25121	16	10	have	have	VERB
fcis-25121	16	11	huge	huge	ADJ
fcis-25121	16	12	size	size	NOUN
fcis-25121	16	13	,	,	PUNCT
fcis-25121	16	14	rich	rich	ADJ
fcis-25121	16	15	content	content	NOUN
fcis-25121	16	16	and	and	CCONJ
fcis-25121	16	17	complex	complex	ADJ
fcis-25121	16	18	composition	composition	NOUN
fcis-25121	16	19	.	.	PUNCT
fcis-25121	17	1	fig	fig	NOUN
fcis-25121	17	2	.	.	PUNCT
fcis-25121	18	1	1(a	1(a	NUM
fcis-25121	18	2	)	)	PUNCT
fcis-25121	18	3	shows	show	VERB
fcis-25121	18	4	that	that	SCONJ
fcis-25121	18	5	every	every	DET
fcis-25121	18	6	bodhisattva	bodhisattva	NOUN
fcis-25121	18	7	has	have	VERB
fcis-25121	18	8	different	different	ADJ
fcis-25121	18	9	facial	facial	ADJ
fcis-25121	18	10	expressions	expression	NOUN
fcis-25121	18	11	,	,	PUNCT
fcis-25121	18	12	actions	action	NOUN
fcis-25121	18	13	,	,	PUNCT
fcis-25121	18	14	clothes	clothe	NOUN
fcis-25121	18	15	and	and	CCONJ
fcis-25121	18	16	accessories	accessory	NOUN
fcis-25121	18	17	in	in	ADP
fcis-25121	18	18	the	the	DET
fcis-25121	18	19	mural	mural	NOUN
fcis-25121	18	20	of	of	ADP
fcis-25121	18	21	the	the	DET
fcis-25121	18	22	late	late	ADJ
fcis-25121	18	23	tang	tang	PROPN
fcis-25121	18	24	dynasty	dynasty	NOUN
fcis-25121	18	25	.	.	PUNCT
fcis-25121	19	1	in	in	ADP
fcis-25121	19	2	addition	addition	NOUN
fcis-25121	19	3	to	to	ADP
fcis-25121	19	4	the	the	DET
fcis-25121	19	5	figures	figure	NOUN
fcis-25121	19	6	,	,	PUNCT
fcis-25121	19	7	a	a	DET
fcis-25121	19	8	large	large	ADJ
fcis-25121	19	9	number	number	NOUN
fcis-25121	19	10	of	of	ADP
fcis-25121	19	11	buildings	building	NOUN
fcis-25121	19	12	,	,	PUNCT
fcis-25121	19	13	moires	moire	NOUN
fcis-25121	19	14	,	,	PUNCT
fcis-25121	19	15	animals	animal	NOUN
fcis-25121	19	16	and	and	CCONJ
fcis-25121	19	17	flowers	flower	NOUN
fcis-25121	19	18	also	also	ADV
fcis-25121	19	19	can	can	AUX
fcis-25121	19	20	be	be	AUX
fcis-25121	19	21	seen	see	VERB
fcis-25121	19	22	in	in	ADP
fcis-25121	19	23	the	the	DET
fcis-25121	19	24	murals	mural	NOUN
fcis-25121	19	25	,	,	PUNCT
fcis-25121	19	26	which	which	PRON
fcis-25121	19	27	can	can	AUX
fcis-25121	19	28	create	create	VERB
fcis-25121	19	29	high	high	ADJ
fcis-25121	19	30	background	background	NOUN
fcis-25121	19	31	noise	noise	NOUN
fcis-25121	19	32	and	and	CCONJ
fcis-25121	19	33	make	make	VERB
fcis-25121	19	34	the	the	DET
fcis-25121	19	35	images	image	NOUN
fcis-25121	19	36	classification	classification	NOUN
fcis-25121	19	37	more	more	ADV
fcis-25121	19	38	difficult	difficult	ADJ
fcis-25121	19	39	.	.	PUNCT
fcis-25121	20	1	2	2	X
fcis-25121	20	2	)	)	PUNCT
fcis-25121	20	3	the	the	DET
fcis-25121	20	4	features	feature	NOUN
fcis-25121	20	5	of	of	ADP
fcis-25121	20	6	the	the	DET
fcis-25121	20	7	same	same	ADJ
fcis-25121	20	8	kind	kind	NOUN
fcis-25121	20	9	of	of	ADP
fcis-25121	20	10	objects	object	NOUN
fcis-25121	20	11	are	be	AUX
fcis-25121	20	12	quite	quite	ADV
fcis-25121	20	13	different	different	ADJ
fcis-25121	20	14	in	in	ADP
fcis-25121	20	15	the	the	DET
fcis-25121	20	16	same	same	ADJ
fcis-25121	20	17	dynasty	dynasty	NOUN
fcis-25121	20	18	.	.	PUNCT
fcis-25121	21	1	as	as	SCONJ
fcis-25121	21	2	shown	show	VERB
fcis-25121	21	3	in	in	ADP
fcis-25121	21	4	fig	fig	NOUN
fcis-25121	21	5	.	.	PUNCT
fcis-25121	22	1	1(b	1(b	NUM
fcis-25121	22	2	)	)	PUNCT
fcis-25121	22	3	and	and	CCONJ
fcis-25121	22	4	figure	figure	NOUN
fcis-25121	22	5	.	.	PUNCT
fcis-25121	23	1	1(c	1(c	NUM
fcis-25121	23	2	)	)	PUNCT
fcis-25121	24	1	,	,	PUNCT
fcis-25121	24	2	the	the	DET
fcis-25121	24	3	figures	figure	NOUN
fcis-25121	24	4	in	in	ADP
fcis-25121	24	5	the	the	DET
fcis-25121	24	6	murals	mural	NOUN
fcis-25121	24	7	are	be	AUX
fcis-25121	24	8	the	the	DET
fcis-25121	24	9	patrons	patron	NOUN
fcis-25121	24	10	in	in	ADP
fcis-25121	24	11	the	the	DET
fcis-25121	24	12	yuan	yuan	NOUN
fcis-25121	24	13	dynasty	dynasty	NOUN
fcis-25121	24	14	,	,	PUNCT
fcis-25121	24	15	but	but	CCONJ
fcis-25121	24	16	they	they	PRON
fcis-25121	24	17	look	look	VERB
fcis-25121	24	18	so	so	ADV
fcis-25121	24	19	different	different	ADJ
fcis-25121	24	20	,	,	PUNCT
fcis-25121	24	21	because	because	SCONJ
fcis-25121	24	22	one	one	NUM
fcis-25121	24	23	was	be	AUX
fcis-25121	24	24	painted	paint	VERB
fcis-25121	24	25	with	with	ADP
fcis-25121	24	26	clear	clear	ADJ
fcis-25121	24	27	lines	line	NOUN
fcis-25121	24	28	and	and	CCONJ
fcis-25121	24	29	the	the	DET
fcis-25121	24	30	other	other	ADJ
fcis-25121	24	31	was	be	AUX
fcis-25121	24	32	painted	paint	VERB
fcis-25121	24	33	with	with	ADP
fcis-25121	24	34	bright	bright	ADJ
fcis-25121	24	35	colors	color	NOUN
fcis-25121	24	36	.	.	PUNCT
fcis-25121	25	1	3	3	X
fcis-25121	25	2	)	)	PUNCT
fcis-25121	25	3	the	the	DET
fcis-25121	25	4	features	feature	NOUN
fcis-25121	25	5	of	of	ADP
fcis-25121	25	6	the	the	DET
fcis-25121	25	7	same	same	ADJ
fcis-25121	25	8	object	object	NOUN
fcis-25121	25	9	have	have	VERB
fcis-25121	25	10	a	a	DET
fcis-25121	25	11	high	high	ADJ
fcis-25121	25	12	similarity	similarity	NOUN
fcis-25121	25	13	from	from	ADP
fcis-25121	25	14	different	different	ADJ
fcis-25121	25	15	dynasties	dynasty	NOUN
fcis-25121	25	16	.	.	PUNCT
fcis-25121	26	1	fig	fig	NOUN
fcis-25121	26	2	.	.	PUNCT
fcis-25121	27	1	1(d	1(d	NUM
fcis-25121	27	2	)	)	PUNCT
fcis-25121	27	3	shows	show	VERB
fcis-25121	27	4	the	the	DET
fcis-25121	27	5	dancers	dancer	NOUN
fcis-25121	27	6	in	in	ADP
fcis-25121	27	7	the	the	DET
fcis-25121	27	8	middle	middle	PROPN
fcis-25121	27	9	tang	tang	PROPN
fcis-25121	27	10	dynasty	dynasty	NOUN
fcis-25121	27	11	,	,	PUNCT
fcis-25121	27	12	and	and	CCONJ
fcis-25121	27	13	fig	fig	NOUN
fcis-25121	27	14	.	.	PUNCT
fcis-25121	28	1	1(e	1(e	NUM
fcis-25121	28	2	)	)	PUNCT
fcis-25121	28	3	shows	show	VERB
fcis-25121	28	4	the	the	DET
fcis-25121	28	5	dancers	dancer	NOUN
fcis-25121	28	6	in	in	ADP
fcis-25121	28	7	the	the	DET
fcis-25121	28	8	song	song	NOUN
fcis-25121	28	9	dynasty	dynasty	NOUN
fcis-25121	28	10	,	,	PUNCT
fcis-25121	28	11	but	but	CCONJ
fcis-25121	28	12	they	they	PRON
fcis-25121	28	13	have	have	VERB
fcis-25121	28	14	great	great	ADJ
fcis-25121	28	15	similarities	similarity	NOUN
fcis-25121	28	16	in	in	ADP
fcis-25121	28	17	face	face	NOUN
fcis-25121	28	18	features	feature	NOUN
fcis-25121	28	19	.	.	PUNCT
fcis-25121	29	1	4	4	X
fcis-25121	29	2	)	)	PUNCT
fcis-25121	29	3	fig	fig	NOUN
fcis-25121	29	4	.	.	PUNCT
fcis-25121	30	1	1(f	1(f	NUM
fcis-25121	30	2	)	)	PUNCT
fcis-25121	30	3	shows	show	VERB
fcis-25121	30	4	the	the	DET
fcis-25121	30	5	murals	mural	NOUN
fcis-25121	30	6	have	have	AUX
fcis-25121	30	7	been	be	AUX
fcis-25121	30	8	discoloured	discolour	VERB
fcis-25121	30	9	with	with	ADP
fcis-25121	30	10	the	the	DET
fcis-25121	30	11	passage	passage	NOUN
fcis-25121	30	12	of	of	ADP
fcis-25121	30	13	time	time	NOUN
fcis-25121	30	14	,	,	PUNCT
fcis-25121	30	15	that	that	PRON
fcis-25121	30	16	will	will	AUX
fcis-25121	30	17	also	also	ADV
fcis-25121	30	18	make	make	VERB
fcis-25121	30	19	image	image	NOUN
fcis-25121	30	20	classification	classification	NOUN
fcis-25121	30	21	more	more	ADV
fcis-25121	30	22	difficult	difficult	ADJ
fcis-25121	30	23	.	.	PUNCT
fcis-25121	31	1	(	(	PUNCT
fcis-25121	31	2	a	a	DET
fcis-25121	31	3	)	)	PUNCT
fcis-25121	31	4	cave12	cave12	NOUN
fcis-25121	31	5	(	(	PUNCT
fcis-25121	31	6	b	b	NOUN
fcis-25121	31	7	)	)	PUNCT
fcis-25121	31	8	patron	patron	NOUN
fcis-25121	31	9	(	(	PUNCT
fcis-25121	31	10	c	c	NOUN
fcis-25121	31	11	)	)	PUNCT
fcis-25121	31	12	patron	patron	NOUN
fcis-25121	31	13	(	(	PUNCT
fcis-25121	31	14	d	d	NOUN
fcis-25121	31	15	)	)	PUNCT
fcis-25121	31	16	dancers	dancer	NOUN
fcis-25121	31	17	(	(	PUNCT
fcis-25121	31	18	e	e	NOUN
fcis-25121	31	19	)	)	PUNCT
fcis-25121	31	20	dancers	dancer	NOUN
fcis-25121	31	21	(	(	PUNCT
fcis-25121	31	22	f	f	X
fcis-25121	31	23	)	)	PUNCT
fcis-25121	31	24	cave12	cave12	NOUN
fcis-25121	31	25	fig	fig	NOUN
fcis-25121	31	26	1	1	NUM
fcis-25121	31	27	.	.	PUNCT
fcis-25121	32	1	examples	example	NOUN
fcis-25121	32	2	of	of	ADP
fcis-25121	32	3	some	some	DET
fcis-25121	32	4	images	image	NOUN
fcis-25121	32	5	of	of	ADP
fcis-25121	32	6	dunhuang	dunhuang	NOUN
fcis-25121	32	7	murals	mural	NOUN
fcis-25121	32	8	the	the	DET
fcis-25121	32	9	main	main	ADJ
fcis-25121	32	10	contributions	contribution	NOUN
fcis-25121	32	11	are	be	AUX
fcis-25121	32	12	as	as	SCONJ
fcis-25121	32	13	follows	follow	VERB
fcis-25121	32	14	:	:	PUNCT
fcis-25121	32	15	first	first	ADV
fcis-25121	32	16	,	,	PUNCT
fcis-25121	32	17	the	the	DET
fcis-25121	32	18	dunhuang	dunhuang	NOUN
fcis-25121	32	19	murals	mural	VERB
fcis-25121	32	20	data	datum	NOUN
fcis-25121	32	21	set	set	VERB
fcis-25121	32	22	is	be	AUX
fcis-25121	32	23	constructed	construct	VERB
fcis-25121	32	24	from	from	ADP
fcis-25121	32	25	the	the	DET
fcis-25121	32	26	mural	mural	ADJ
fcis-25121	32	27	materials	material	NOUN
fcis-25121	32	28	;	;	PUNCT
fcis-25121	32	29	second	second	X
fcis-25121	32	30	,	,	PUNCT
fcis-25121	32	31	four	four	NUM
fcis-25121	32	32	convolutional	convolutional	ADJ
fcis-25121	32	33	neural	neural	ADJ
fcis-25121	32	34	network	network	NOUN
fcis-25121	32	35	models	model	NOUN
fcis-25121	32	36	with	with	ADP
fcis-25121	32	37	different	different	ADJ
fcis-25121	32	38	depths	depth	NOUN
fcis-25121	32	39	and	and	CCONJ
fcis-25121	32	40	structures	structure	NOUN
fcis-25121	32	41	are	be	AUX
fcis-25121	32	42	constructed	construct	VERB
fcis-25121	32	43	and	and	CCONJ
fcis-25121	32	44	trained	train	VERB
fcis-25121	32	45	;	;	PUNCT
fcis-25121	32	46	finally	finally	ADV
fcis-25121	32	47	,	,	PUNCT
fcis-25121	32	48	the	the	DET
fcis-25121	32	49	network	network	NOUN
fcis-25121	32	50	model	model	NOUN
fcis-25121	32	51	is	be	AUX
fcis-25121	32	52	tested	test	VERB
fcis-25121	32	53	using	use	VERB
fcis-25121	32	54	the	the	DET
fcis-25121	32	55	test	test	NOUN
fcis-25121	32	56	set	set	VERB
fcis-25121	32	57	and	and	CCONJ
fcis-25121	32	58	an	an	DET
fcis-25121	32	59	appropriate	appropriate	ADJ
fcis-25121	32	60	classification	classification	NOUN
fcis-25121	32	61	model	model	NOUN
fcis-25121	32	62	is	be	AUX
fcis-25121	32	63	selected	select	VERB
fcis-25121	32	64	.	.	PUNCT
fcis-25121	33	1	it	it	PRON
fcis-25121	33	2	affords	afford	VERB
fcis-25121	33	3	advantages	advantage	NOUN
fcis-25121	33	4	for	for	ADP
fcis-25121	33	5	researchers	researcher	NOUN
fcis-25121	33	6	on	on	ADP
fcis-25121	33	7	classifying	classify	VERB
fcis-25121	33	8	dunhuang	dunhuang	NOUN
fcis-25121	33	9	murals	mural	NOUN
fcis-25121	33	10	dynasties	dynasty	NOUN
fcis-25121	33	11	which	which	PRON
fcis-25121	33	12	can	can	AUX
fcis-25121	33	13	bring	bring	VERB
fcis-25121	33	14	the	the	DET
fcis-25121	33	15	digital	digital	ADJ
fcis-25121	33	16	protection	protection	NOUN
fcis-25121	33	17	on	on	ADP
fcis-25121	33	18	dunhuang	dunhuang	NOUN
fcis-25121	33	19	murals	mural	NOUN
fcis-25121	33	20	to	to	ADP
fcis-25121	33	21	a	a	DET
fcis-25121	33	22	wider	wide	ADJ
fcis-25121	33	23	audience	audience	NOUN
fcis-25121	33	24	.	.	PUNCT
fcis-25121	34	1	2	2	X
fcis-25121	34	2	.	.	X
fcis-25121	34	3	background	background	NOUN
fcis-25121	34	4	in	in	ADP
fcis-25121	34	5	recent	recent	ADJ
fcis-25121	34	6	years	year	NOUN
fcis-25121	34	7	,	,	PUNCT
fcis-25121	34	8	a	a	DET
fcis-25121	34	9	large	large	ADJ
fcis-25121	34	10	number	number	NOUN
fcis-25121	34	11	of	of	ADP
fcis-25121	34	12	studies	study	NOUN
fcis-25121	34	13	have	have	AUX
fcis-25121	34	14	used	use	VERB
fcis-25121	34	15	computer	computer	NOUN
fcis-25121	34	16	technology	technology	NOUN
fcis-25121	34	17	to	to	PART
fcis-25121	34	18	aid	aid	VERB
fcis-25121	34	19	mural	mural	ADJ
fcis-25121	34	20	classification	classification	NOUN
fcis-25121	34	21	.	.	PUNCT
fcis-25121	35	1	the	the	DET
fcis-25121	35	2	research	research	NOUN
fcis-25121	35	3	methods	method	NOUN
fcis-25121	35	4	in	in	ADP
fcis-25121	35	5	this	this	DET
fcis-25121	35	6	field	field	NOUN
fcis-25121	35	7	are	be	AUX
fcis-25121	35	8	mainly	mainly	ADV
fcis-25121	35	9	divided	divide	VERB
fcis-25121	35	10	into	into	ADP
fcis-25121	35	11	traditional	traditional	ADJ
fcis-25121	35	12	methods	method	NOUN
fcis-25121	35	13	and	and	CCONJ
fcis-25121	35	14	deep	deep	ADJ
fcis-25121	35	15	learning	learning	NOUN
fcis-25121	35	16	-	-	PUNCT
fcis-25121	35	17	based	base	VERB
fcis-25121	35	18	methods	method	NOUN
fcis-25121	35	19	.	.	PUNCT
fcis-25121	36	1	most	most	ADJ
fcis-25121	36	2	of	of	ADP
fcis-25121	36	3	the	the	DET
fcis-25121	36	4	traditional	traditional	ADJ
fcis-25121	36	5	classification	classification	NOUN
fcis-25121	36	6	methods	method	NOUN
fcis-25121	36	7	are	be	AUX
fcis-25121	36	8	based	base	VERB
fcis-25121	36	9	on	on	ADP
fcis-25121	36	10	the	the	DET
fcis-25121	36	11	bag	bag	NOUN
fcis-25121	36	12	-	-	PUNCT
fcis-25121	36	13	of	of	ADP
fcis-25121	36	14	-	-	PUNCT
fcis-25121	36	15	words	word	NOUN
fcis-25121	36	16	(	(	PUNCT
fcis-25121	36	17	bow	bow	NOUN
fcis-25121	36	18	)	)	PUNCT
fcis-25121	36	19	model	model	NOUN
fcis-25121	36	20	which	which	PRON
fcis-25121	36	21	,	,	PUNCT
fcis-25121	36	22	as	as	ADP
fcis-25121	36	23	a	a	DET
fcis-25121	36	24	middle	middle	ADJ
fcis-25121	36	25	level	level	NOUN
fcis-25121	36	26	feature	feature	NOUN
fcis-25121	36	27	,	,	PUNCT
fcis-25121	36	28	can	can	AUX
fcis-25121	36	29	narrow	narrow	VERB
fcis-25121	36	30	the	the	DET
fcis-25121	36	31	gap	gap	NOUN
fcis-25121	36	32	between	between	ADP
fcis-25121	36	33	low	low	ADJ
fcis-25121	36	34	-	-	PUNCT
fcis-25121	36	35	level	level	NOUN
fcis-25121	36	36	visual	visual	ADJ
fcis-25121	36	37	features	feature	NOUN
fcis-25121	36	38	and	and	CCONJ
fcis-25121	36	39	high	high	ADJ
fcis-25121	36	40	-	-	PUNCT
fcis-25121	36	41	level	level	NOUN
fcis-25121	36	42	semantic	semantic	ADJ
fcis-25121	36	43	features	feature	NOUN
fcis-25121	36	44	[	[	X
fcis-25121	36	45	1	1	NUM
fcis-25121	36	46	]	]	PUNCT
fcis-25121	36	47	.	.	PUNCT
fcis-25121	37	1	how	how	SCONJ
fcis-25121	37	2	to	to	PART
fcis-25121	37	3	choose	choose	VERB
fcis-25121	37	4	the	the	DET
fcis-25121	37	5	keywords	keyword	NOUN
fcis-25121	37	6	of	of	ADP
fcis-25121	37	7	feature	feature	NOUN
fcis-25121	37	8	extraction	extraction	NOUN
fcis-25121	37	9	affect	affect	VERB
fcis-25121	37	10	the	the	DET
fcis-25121	37	11	final	final	ADJ
fcis-25121	37	12	classification	classification	NOUN
fcis-25121	37	13	results	result	NOUN
fcis-25121	37	14	.	.	PUNCT
fcis-25121	38	1	in	in	ADP
fcis-25121	38	2	2013	2013	NUM
fcis-25121	38	3	,	,	PUNCT
fcis-25121	38	4	a	a	DET
fcis-25121	38	5	method	method	NOUN
fcis-25121	38	6	was	be	AUX
fcis-25121	38	7	introduced	introduce	VERB
fcis-25121	38	8	based	base	VERB
fcis-25121	38	9	on	on	ADP
fcis-25121	38	10	the	the	DET
fcis-25121	38	11	71	71	NUM
fcis-25121	38	12	essential	essential	ADJ
fcis-25121	38	13	contour	contour	NOUN
fcis-25121	38	14	structure	structure	NOUN
fcis-25121	38	15	of	of	ADP
fcis-25121	38	16	the	the	DET
fcis-25121	38	17	mural	mural	ADJ
fcis-25121	38	18	images	image	NOUN
fcis-25121	38	19	,	,	PUNCT
fcis-25121	38	20	which	which	PRON
fcis-25121	38	21	leads	lead	VERB
fcis-25121	38	22	to	to	ADP
fcis-25121	38	23	high	high	ADJ
fcis-25121	38	24	classification	classification	NOUN
fcis-25121	38	25	accuracy	accuracy	NOUN
fcis-25121	39	1	[	[	X
fcis-25121	39	2	2	2	NUM
fcis-25121	39	3	]	]	PUNCT
fcis-25121	39	4	.	.	PUNCT
fcis-25121	40	1	in	in	ADP
fcis-25121	40	2	2017	2017	NUM
fcis-25121	40	3	,	,	PUNCT
fcis-25121	40	4	another	another	DET
fcis-25121	40	5	method	method	NOUN
fcis-25121	40	6	was	be	AUX
fcis-25121	40	7	introduced	introduce	VERB
fcis-25121	40	8	based	base	VERB
fcis-25121	40	9	on	on	ADP
fcis-25121	40	10	the	the	DET
fcis-25121	40	11	mean	mean	ADJ
fcis-25121	40	12	,	,	PUNCT
fcis-25121	40	13	variance	variance	NOUN
fcis-25121	40	14	,	,	PUNCT
fcis-25121	40	15	etc	etc	X
fcis-25121	40	16	of	of	ADP
fcis-25121	40	17	the	the	DET
fcis-25121	40	18	portrait	portrait	NOUN
fcis-25121	40	19	character	character	NOUN
fcis-25121	40	20	eigenvalues	eigenvalue	NOUN
fcis-25121	40	21	,	,	PUNCT
fcis-25121	40	22	which	which	PRON
fcis-25121	40	23	can	can	AUX
fcis-25121	40	24	classify	classify	VERB
fcis-25121	40	25	the	the	DET
fcis-25121	40	26	murals	mural	NOUN
fcis-25121	40	27	into	into	ADP
fcis-25121	40	28	different	different	ADJ
fcis-25121	40	29	dynasties	dynasty	NOUN
fcis-25121	40	30	[	[	X
fcis-25121	40	31	3	3	NUM
fcis-25121	40	32	]	]	PUNCT
fcis-25121	40	33	.	.	PUNCT
fcis-25121	41	1	since	since	SCONJ
fcis-25121	41	2	the	the	DET
fcis-25121	41	3	concept	concept	NOUN
fcis-25121	41	4	of	of	ADP
fcis-25121	41	5	deep	deep	ADJ
fcis-25121	41	6	learning	learning	NOUN
fcis-25121	41	7	was	be	AUX
fcis-25121	41	8	put	put	VERB
fcis-25121	41	9	forward	forward	ADV
fcis-25121	41	10	in	in	ADP
fcis-25121	41	11	2006	2006	NUM
fcis-25121	41	12	,	,	PUNCT
fcis-25121	41	13	it	it	PRON
fcis-25121	41	14	has	have	AUX
fcis-25121	41	15	been	be	AUX
fcis-25121	41	16	widely	widely	ADV
fcis-25121	41	17	used	use	VERB
fcis-25121	41	18	in	in	ADP
fcis-25121	41	19	computer	computer	NOUN
fcis-25121	41	20	vision	vision	NOUN
fcis-25121	41	21	technology	technology	NOUN
fcis-25121	41	22	.	.	PUNCT
fcis-25121	42	1	convolutional	convolutional	ADJ
fcis-25121	42	2	neural	neural	ADJ
fcis-25121	42	3	network	network	NOUN
fcis-25121	42	4	(	(	PUNCT
fcis-25121	42	5	cnn	cnn	PROPN
fcis-25121	42	6	)	)	PUNCT
fcis-25121	42	7	based	base	VERB
fcis-25121	42	8	on	on	ADP
fcis-25121	42	9	deep	deep	ADJ
fcis-25121	42	10	learning	learning	NOUN
fcis-25121	42	11	models	model	NOUN
fcis-25121	42	12	has	have	VERB
fcis-25121	42	13	a	a	DET
fcis-25121	42	14	good	good	ADJ
fcis-25121	42	15	performance	performance	NOUN
fcis-25121	42	16	in	in	ADP
fcis-25121	42	17	natural	natural	ADJ
fcis-25121	42	18	image	image	NOUN
fcis-25121	42	19	classification	classification	NOUN
fcis-25121	42	20	,	,	PUNCT
fcis-25121	42	21	and	and	CCONJ
fcis-25121	42	22	has	have	AUX
fcis-25121	42	23	also	also	ADV
fcis-25121	42	24	been	be	AUX
fcis-25121	42	25	applied	apply	VERB
fcis-25121	42	26	to	to	ADP
fcis-25121	42	27	the	the	DET
fcis-25121	42	28	classification	classification	NOUN
fcis-25121	42	29	of	of	ADP
fcis-25121	42	30	murals	mural	NOUN
fcis-25121	42	31	.	.	PUNCT
fcis-25121	43	1	in	in	ADP
fcis-25121	43	2	2012	2012	NUM
fcis-25121	43	3	,	,	PUNCT
fcis-25121	43	4	alex	alex	PROPN
fcis-25121	43	5	krezewski	krezewski	PROPN
fcis-25121	43	6	and	and	CCONJ
fcis-25121	43	7	others	other	NOUN
fcis-25121	43	8	proposed	propose	VERB
fcis-25121	43	9	alexnet	alexnet	ADJ
fcis-25121	43	10	network	network	NOUN
fcis-25121	43	11	in	in	ADP
fcis-25121	43	12	the	the	DET
fcis-25121	43	13	image	image	NOUN
fcis-25121	43	14	net	net	ADJ
fcis-25121	43	15	large	large	ADJ
fcis-25121	43	16	scale	scale	NOUN
fcis-25121	43	17	visual	visual	ADJ
fcis-25121	43	18	recognition	recognition	NOUN
fcis-25121	43	19	challenge	challenge	NOUN
fcis-25121	43	20	(	(	PUNCT
fcis-25121	43	21	ilsvrc	ilsvrc	PROPN
fcis-25121	43	22	)	)	PUNCT
fcis-25121	43	23	which	which	PRON
fcis-25121	43	24	is	be	AUX
fcis-25121	43	25	widely	widely	ADV
fcis-25121	43	26	used	use	VERB
fcis-25121	43	27	in	in	ADP
fcis-25121	43	28	natural	natural	ADJ
fcis-25121	43	29	image	image	NOUN
fcis-25121	43	30	classification	classification	NOUN
fcis-25121	43	31	[	[	X
fcis-25121	43	32	4	4	NUM
fcis-25121	43	33	]	]	PUNCT
fcis-25121	43	34	.	.	PUNCT
fcis-25121	44	1	in	in	ADP
fcis-25121	44	2	[	[	X
fcis-25121	44	3	5	5	NUM
fcis-25121	44	4	]	]	PUNCT
fcis-25121	44	5	,	,	PUNCT
fcis-25121	44	6	the	the	DET
fcis-25121	44	7	res	re	NOUN
fcis-25121	44	8	net50	net50	PROPN
fcis-25121	44	9	was	be	AUX
fcis-25121	44	10	selected	select	VERB
fcis-25121	44	11	as	as	ADP
fcis-25121	44	12	the	the	DET
fcis-25121	44	13	basic	basic	ADJ
fcis-25121	44	14	framework	framework	NOUN
fcis-25121	44	15	and	and	CCONJ
fcis-25121	44	16	a	a	DET
fcis-25121	44	17	new	new	ADJ
fcis-25121	44	18	network	network	NOUN
fcis-25121	44	19	model	model	NOUN
fcis-25121	44	20	was	be	AUX
fcis-25121	44	21	proposed	propose	VERB
fcis-25121	44	22	.	.	PUNCT
fcis-25121	45	1	the	the	DET
fcis-25121	45	2	new	new	ADJ
fcis-25121	45	3	network	network	NOUN
fcis-25121	45	4	layer	layer	NOUN
fcis-25121	45	5	was	be	AUX
fcis-25121	45	6	used	use	VERB
fcis-25121	45	7	to	to	PART
fcis-25121	45	8	replace	replace	VERB
fcis-25121	45	9	the	the	DET
fcis-25121	45	10	49th	49th	ADJ
fcis-25121	45	11	layer	layer	NOUN
fcis-25121	45	12	of	of	ADP
fcis-25121	45	13	res	re	NOUN
fcis-25121	45	14	net50	net50	PROPN
fcis-25121	45	15	,	,	PUNCT
fcis-25121	45	16	which	which	PRON
fcis-25121	45	17	improved	improve	VERB
fcis-25121	45	18	the	the	DET
fcis-25121	45	19	dynasty	dynasty	NOUN
fcis-25121	45	20	recognition	recognition	NOUN
fcis-25121	45	21	accuracy	accuracy	NOUN
fcis-25121	45	22	of	of	ADP
fcis-25121	45	23	dunhuang	dunhuang	NOUN
fcis-25121	45	24	murals	mural	NOUN
fcis-25121	45	25	.	.	PUNCT
fcis-25121	46	1	in	in	ADP
fcis-25121	46	2	[	[	X
fcis-25121	46	3	6	6	NUM
fcis-25121	46	4	]	]	PUNCT
fcis-25121	46	5	,	,	PUNCT
fcis-25121	46	6	a	a	DET
fcis-25121	46	7	prototype	prototype	NOUN
fcis-25121	46	8	system	system	NOUN
fcis-25121	46	9	of	of	ADP
fcis-25121	46	10	ancient	ancient	ADJ
fcis-25121	46	11	mural	mural	ADJ
fcis-25121	46	12	intelligent	intelligent	ADJ
fcis-25121	46	13	classification	classification	NOUN
fcis-25121	46	14	based	base	VERB
fcis-25121	46	15	on	on	ADP
fcis-25121	46	16	alexnet	alexnet	NOUN
fcis-25121	46	17	and	and	CCONJ
fcis-25121	46	18	tfnet	tfnet	NOUN
fcis-25121	46	19	was	be	AUX
fcis-25121	46	20	developed	develop	VERB
fcis-25121	46	21	.	.	PUNCT
fcis-25121	47	1	on	on	ADP
fcis-25121	47	2	the	the	DET
fcis-25121	47	3	other	other	ADJ
fcis-25121	47	4	hand	hand	NOUN
fcis-25121	47	5	,	,	PUNCT
fcis-25121	47	6	it	it	PRON
fcis-25121	47	7	completed	complete	VERB
fcis-25121	47	8	the	the	DET
fcis-25121	47	9	automatic	automatic	ADJ
fcis-25121	47	10	classification	classification	NOUN
fcis-25121	47	11	function	function	NOUN
fcis-25121	47	12	of	of	ADP
fcis-25121	47	13	mural	mural	ADJ
fcis-25121	47	14	images	image	NOUN
fcis-25121	47	15	,	,	PUNCT
fcis-25121	47	16	and	and	CCONJ
fcis-25121	47	17	saved	save	VERB
fcis-25121	47	18	the	the	DET
fcis-25121	47	19	classification	classification	NOUN
fcis-25121	47	20	results	result	NOUN
fcis-25121	47	21	to	to	ADP
fcis-25121	47	22	the	the	DET
fcis-25121	47	23	database	database	NOUN
fcis-25121	47	24	to	to	PART
fcis-25121	47	25	upload	upload	VERB
fcis-25121	47	26	murals	mural	NOUN
fcis-25121	47	27	,	,	PUNCT
fcis-25121	47	28	which	which	PRON
fcis-25121	47	29	greatly	greatly	ADV
fcis-25121	47	30	facilitated	facilitate	VERB
fcis-25121	47	31	the	the	DET
fcis-25121	47	32	unified	unified	ADJ
fcis-25121	47	33	management	management	NOUN
fcis-25121	47	34	of	of	ADP
fcis-25121	47	35	murals	mural	NOUN
fcis-25121	47	36	.	.	PUNCT
fcis-25121	48	1	in	in	ADP
fcis-25121	48	2	2021	2021	NUM
fcis-25121	48	3	,	,	PUNCT
fcis-25121	48	4	an	an	DET
fcis-25121	48	5	inception	inception	NOUN
fcis-25121	48	6	-	-	PUNCT
fcis-25121	48	7	v3	v3	NOUN
fcis-25121	48	8	model	model	NOUN
fcis-25121	48	9	integrated	integrate	VERB
fcis-25121	48	10	with	with	ADP
fcis-25121	48	11	transfer	transfer	NOUN
fcis-25121	48	12	learning	learning	NOUN
fcis-25121	48	13	was	be	AUX
fcis-25121	48	14	proposed	propose	VERB
fcis-25121	48	15	to	to	PART
fcis-25121	48	16	identify	identify	VERB
fcis-25121	48	17	and	and	CCONJ
fcis-25121	48	18	classify	classify	VERB
fcis-25121	48	19	the	the	DET
fcis-25121	48	20	dynasties	dynasty	NOUN
fcis-25121	48	21	of	of	ADP
fcis-25121	48	22	ancient	ancient	ADJ
fcis-25121	48	23	murals	mural	NOUN
fcis-25121	48	24	which	which	PRON
fcis-25121	48	25	could	could	AUX
fcis-25121	48	26	effectively	effectively	ADV
fcis-25121	48	27	identify	identify	VERB
fcis-25121	48	28	the	the	DET
fcis-25121	48	29	dynasty	dynasty	NOUN
fcis-25121	48	30	to	to	PART
fcis-25121	48	31	which	which	PRON
fcis-25121	48	32	the	the	DET
fcis-25121	48	33	murals	mural	NOUN
fcis-25121	48	34	belong	belong	VERB
fcis-25121	48	35	[	[	X
fcis-25121	48	36	7	7	NUM
fcis-25121	48	37	]	]	PUNCT
fcis-25121	48	38	.	.	PUNCT
fcis-25121	49	1	3	3	X
fcis-25121	49	2	.	.	X
fcis-25121	49	3	materials	material	NOUN
fcis-25121	49	4	and	and	CCONJ
fcis-25121	49	5	methods	method	NOUN
fcis-25121	49	6	3.1	3.1	NUM
fcis-25121	49	7	.	.	PUNCT
fcis-25121	50	1	experimental	experimental	ADJ
fcis-25121	50	2	data	datum	NOUN
fcis-25121	50	3	(	(	PUNCT
fcis-25121	50	4	a	a	X
fcis-25121	50	5	)	)	PUNCT
fcis-25121	50	6	the	the	DET
fcis-25121	50	7	northern	northern	ADJ
fcis-25121	50	8	dynasties	dynasty	NOUN
fcis-25121	50	9	(	(	PUNCT
fcis-25121	50	10	b	b	X
fcis-25121	50	11	)	)	PUNCT
fcis-25121	50	12	the	the	DET
fcis-25121	50	13	sui	sui	PROPN
fcis-25121	50	14	dynasty	dynasty	NOUN
fcis-25121	50	15	(	(	PUNCT
fcis-25121	50	16	c	c	NOUN
fcis-25121	50	17	)	)	PUNCT
fcis-25121	50	18	the	the	DET
fcis-25121	50	19	tang	tang	PROPN
fcis-25121	50	20	dynasty	dynasty	NOUN
fcis-25121	50	21	(	(	PUNCT
fcis-25121	50	22	d	d	NOUN
fcis-25121	50	23	)	)	PUNCT
fcis-25121	50	24	the	the	DET
fcis-25121	50	25	five	five	NUM
fcis-25121	50	26	dynasties	dynasty	NOUN
fcis-25121	50	27	and	and	CCONJ
fcis-25121	50	28	song	song	NOUN
fcis-25121	50	29	daynasty	daynasty	NOUN
fcis-25121	50	30	(	(	PUNCT
fcis-25121	50	31	e	e	NOUN
fcis-25121	50	32	)	)	PUNCT
fcis-25121	50	33	the	the	DET
fcis-25121	50	34	western	western	ADJ
fcis-25121	50	35	xia	xia	PROPN
fcis-25121	50	36	regime	regime	NOUN
fcis-25121	50	37	and	and	CCONJ
fcis-25121	50	38	the	the	DET
fcis-25121	50	39	yuan	yuan	NOUN
fcis-25121	50	40	dynasty	dynasty	PROPN
fcis-25121	50	41	fig	fig	PROPN
fcis-25121	50	42	2	2	NUM
fcis-25121	50	43	.	.	PUNCT
fcis-25121	51	1	examples	example	NOUN
fcis-25121	51	2	of	of	ADP
fcis-25121	51	3	dunhuang	dunhuang	PROPN
fcis-25121	51	4	mural	mural	PROPN
fcis-25121	51	5	dataset	dataset	VERB
fcis-25121	51	6	the	the	DET
fcis-25121	51	7	data	datum	NOUN
fcis-25121	51	8	set	set	VERB
fcis-25121	51	9	is	be	AUX
fcis-25121	51	10	obtained	obtain	VERB
fcis-25121	51	11	from	from	ADP
fcis-25121	51	12	a	a	DET
fcis-25121	51	13	series	series	NOUN
fcis-25121	51	14	of	of	ADP
fcis-25121	51	15	books	book	NOUN
fcis-25121	51	16	for	for	ADP
fcis-25121	51	17	appreciating	appreciate	VERB
fcis-25121	51	18	dunhuang	dunhuang	NOUN
fcis-25121	51	19	art	art	NOUN
fcis-25121	51	20	and	and	CCONJ
fcis-25121	51	21	the	the	DET
fcis-25121	51	22	entire	entire	ADJ
fcis-25121	51	23	works	work	NOUN
fcis-25121	51	24	of	of	ADP
fcis-25121	51	25	dunhuang	dunhuang	NOUN
fcis-25121	51	26	murals	mural	NOUN
fcis-25121	51	27	in	in	ADP
fcis-25121	51	28	china	china	PROPN
fcis-25121	51	29	.	.	PUNCT
fcis-25121	52	1	by	by	ADP
fcis-25121	52	2	the	the	DET
fcis-25121	52	3	content	content	NOUN
fcis-25121	52	4	depicted	depict	VERB
fcis-25121	52	5	,	,	PUNCT
fcis-25121	52	6	these	these	DET
fcis-25121	52	7	murals	mural	NOUN
fcis-25121	52	8	,	,	PUNCT
fcis-25121	52	9	spanning	span	VERB
fcis-25121	52	10	a	a	DET
fcis-25121	52	11	period	period	NOUN
fcis-25121	52	12	of	of	ADP
fcis-25121	52	13	1000	1000	NUM
fcis-25121	52	14	years	year	NOUN
fcis-25121	52	15	are	be	AUX
fcis-25121	52	16	divided	divide	VERB
fcis-25121	52	17	into	into	ADP
fcis-25121	52	18	buddha	buddha	NOUN
fcis-25121	52	19	statue	statue	NOUN
fcis-25121	52	20	paintings	painting	NOUN
fcis-25121	52	21	,	,	PUNCT
fcis-25121	52	22	sutra	sutra	ADJ
fcis-25121	52	23	converted	convert	VERB
fcis-25121	52	24	paintings	painting	NOUN
fcis-25121	52	25	,	,	PUNCT
fcis-25121	52	26	narrative	narrative	ADJ
fcis-25121	52	27	paintings	painting	NOUN
fcis-25121	52	28	and	and	CCONJ
fcis-25121	52	29	donor	donor	NOUN
fcis-25121	52	30	portraits	portrait	NOUN
fcis-25121	52	31	,	,	PUNCT
fcis-25121	52	32	etc	etc	X
fcis-25121	52	33	.	.	X
fcis-25121	53	1	according	accord	VERB
fcis-25121	53	2	to	to	ADP
fcis-25121	53	3	the	the	DET
fcis-25121	53	4	historical	historical	ADJ
fcis-25121	53	5	documents	document	NOUN
fcis-25121	53	6	,	,	PUNCT
fcis-25121	53	7	five	five	NUM
fcis-25121	53	8	data	data	NOUN
fcis-25121	53	9	sets	set	NOUN
fcis-25121	53	10	are	be	AUX
fcis-25121	53	11	adopted	adopt	VERB
fcis-25121	53	12	to	to	PART
fcis-25121	53	13	validate	validate	VERB
fcis-25121	53	14	the	the	DET
fcis-25121	53	15	proposed	propose	VERB
fcis-25121	53	16	methods	method	NOUN
fcis-25121	53	17	.	.	PUNCT
fcis-25121	54	1	they	they	PRON
fcis-25121	54	2	are	be	AUX
fcis-25121	54	3	the	the	DET
fcis-25121	54	4	northern	northern	ADJ
fcis-25121	54	5	dynasties	dynasty	NOUN
fcis-25121	54	6	,	,	PUNCT
fcis-25121	54	7	the	the	DET
fcis-25121	54	8	sui	sui	PROPN
fcis-25121	54	9	dynasty	dynasty	NOUN
fcis-25121	54	10	,	,	PUNCT
fcis-25121	54	11	the	the	DET
fcis-25121	54	12	tang	tang	PROPN
fcis-25121	54	13	dynasty	dynasty	NOUN
fcis-25121	54	14	,	,	PUNCT
fcis-25121	54	15	the	the	DET
fcis-25121	54	16	five	five	NUM
fcis-25121	54	17	dynasties	dynasty	NOUN
fcis-25121	54	18	and	and	CCONJ
fcis-25121	54	19	song	song	NOUN
fcis-25121	54	20	daynasty	daynasty	NOUN
fcis-25121	54	21	,	,	PUNCT
fcis-25121	54	22	the	the	DET
fcis-25121	54	23	western	western	ADJ
fcis-25121	54	24	xia	xia	PROPN
fcis-25121	54	25	regime	regime	NOUN
fcis-25121	54	26	and	and	CCONJ
fcis-25121	54	27	the	the	DET
fcis-25121	54	28	yuan	yuan	NOUN
fcis-25121	54	29	dynasty	dynasty	NOUN
fcis-25121	54	30	.	.	PUNCT
fcis-25121	55	1	as	as	SCONJ
fcis-25121	55	2	shown	show	VERB
fcis-25121	55	3	in	in	ADP
fcis-25121	55	4	fig	fig	NOUN
fcis-25121	55	5	.	.	PUNCT
fcis-25121	56	1	2	2	NUM
fcis-25121	56	2	,	,	PUNCT
fcis-25121	56	3	the	the	DET
fcis-25121	56	4	images	image	NOUN
fcis-25121	56	5	in	in	ADP
fcis-25121	56	6	each	each	DET
fcis-25121	56	7	row	row	NOUN
fcis-25121	56	8	belong	belong	VERB
fcis-25121	56	9	to	to	ADP
fcis-25121	56	10	the	the	DET
fcis-25121	56	11	same	same	ADJ
fcis-25121	56	12	dynasty	dynasty	NOUN
fcis-25121	56	13	.	.	PUNCT
fcis-25121	57	1	to	to	PART
fcis-25121	57	2	improve	improve	VERB
fcis-25121	57	3	the	the	DET
fcis-25121	57	4	quality	quality	NOUN
fcis-25121	57	5	of	of	ADP
fcis-25121	57	6	the	the	DET
fcis-25121	57	7	data	data	NOUN
fcis-25121	57	8	sets	set	NOUN
fcis-25121	57	9	,	,	PUNCT
fcis-25121	57	10	we	we	PRON
fcis-25121	57	11	intercept	intercept	VERB
fcis-25121	57	12	881	881	NUM
fcis-25121	57	13	clear	clear	ADJ
fcis-25121	57	14	images	image	NOUN
fcis-25121	57	15	of	of	ADP
fcis-25121	57	16	faces	face	NOUN
fcis-25121	57	17	from	from	ADP
fcis-25121	57	18	buddha	buddha	PROPN
fcis-25121	57	19	,	,	PUNCT
fcis-25121	57	20	bodhisattva	bodhisattva	PROPN
fcis-25121	57	21	,	,	PUNCT
fcis-25121	57	22	buddhist	buddhist	ADJ
fcis-25121	57	23	monk	monk	NOUN
fcis-25121	57	24	,	,	PUNCT
fcis-25121	57	25	supporters	supporter	NOUN
fcis-25121	57	26	.	.	PUNCT
fcis-25121	58	1	for	for	ADP
fcis-25121	58	2	all	all	DET
fcis-25121	58	3	five	five	NUM
fcis-25121	58	4	data	datum	NOUN
fcis-25121	58	5	sets	set	NOUN
fcis-25121	58	6	,	,	PUNCT
fcis-25121	58	7	we	we	PRON
fcis-25121	58	8	split	split	VERB
fcis-25121	58	9	the	the	DET
fcis-25121	58	10	labeled	label	VERB
fcis-25121	58	11	samples	sample	NOUN
fcis-25121	58	12	into	into	ADP
fcis-25121	58	13	two	two	NUM
fcis-25121	58	14	subsets	subset	NOUN
fcis-25121	58	15	,	,	PUNCT
fcis-25121	58	16	i.e.	i.e.	X
fcis-25121	58	17	,	,	PUNCT
fcis-25121	58	18	831	831	NUM
fcis-25121	58	19	samples	sample	NOUN
fcis-25121	58	20	for	for	ADP
fcis-25121	58	21	training	training	NOUN
fcis-25121	58	22	set	set	VERB
fcis-25121	58	23	and	and	CCONJ
fcis-25121	58	24	50	50	NUM
fcis-25121	58	25	samples	sample	NOUN
fcis-25121	58	26	for	for	ADP
fcis-25121	58	27	testing	testing	NOUN
fcis-25121	58	28	set	set	VERB
fcis-25121	58	29	.	.	PUNCT
fcis-25121	59	1	3.2	3.2	NUM
fcis-25121	59	2	.	.	PUNCT
fcis-25121	60	1	experimental	experimental	ADJ
fcis-25121	60	2	method	method	NOUN
fcis-25121	60	3	3.2.1	3.2.1	NUM
fcis-25121	60	4	.	.	PUNCT
fcis-25121	60	5	convolutional	convolutional	ADJ
fcis-25121	60	6	neural	neural	ADJ
fcis-25121	60	7	network	network	NOUN
fcis-25121	60	8	(	(	PUNCT
fcis-25121	60	9	cnn	cnn	PROPN
fcis-25121	60	10	)	)	PUNCT
fcis-25121	60	11	convolutional	convolutional	ADJ
fcis-25121	60	12	neural	neural	ADJ
fcis-25121	60	13	network	network	NOUN
fcis-25121	60	14	(	(	PUNCT
fcis-25121	60	15	cnn	cnn	PROPN
fcis-25121	60	16	)	)	PUNCT
fcis-25121	60	17	based	base	VERB
fcis-25121	60	18	on	on	ADP
fcis-25121	60	19	deep	deep	ADJ
fcis-25121	60	20	learning	learning	NOUN
fcis-25121	60	21	method	method	NOUN
fcis-25121	60	22	has	have	VERB
fcis-25121	60	23	many	many	ADJ
fcis-25121	60	24	advantages	advantage	NOUN
fcis-25121	60	25	,	,	PUNCT
fcis-25121	60	26	including	include	VERB
fcis-25121	60	27	local	local	ADJ
fcis-25121	60	28	connections	connection	NOUN
fcis-25121	60	29	and	and	CCONJ
fcis-25121	60	30	shared	share	VERB
fcis-25121	60	31	weights	weight	NOUN
fcis-25121	60	32	,	,	PUNCT
fcis-25121	60	33	which	which	PRON
fcis-25121	60	34	make	make	VERB
fcis-25121	60	35	it	it	PRON
fcis-25121	60	36	easier	easy	ADJ
fcis-25121	60	37	to	to	PART
fcis-25121	60	38	train	train	VERB
fcis-25121	60	39	with	with	ADP
fcis-25121	60	40	lower	low	ADJ
fcis-25121	60	41	complexity	complexity	NOUN
fcis-25121	60	42	[	[	X
fcis-25121	60	43	8	8	NUM
fcis-25121	60	44	]	]	PUNCT
fcis-25121	60	45	.	.	PUNCT
fcis-25121	61	1	additionally	additionally	ADV
fcis-25121	61	2	,	,	PUNCT
fcis-25121	61	3	cnn	cnn	PROPN
fcis-25121	61	4	is	be	AUX
fcis-25121	61	5	invariant	invariant	ADJ
fcis-25121	61	6	with	with	ADP
fcis-25121	61	7	respect	respect	NOUN
fcis-25121	61	8	to	to	ADP
fcis-25121	61	9	translation	translation	NOUN
fcis-25121	61	10	,	,	PUNCT
fcis-25121	61	11	scale	scale	NOUN
fcis-25121	61	12	and	and	CCONJ
fcis-25121	61	13	rotation	rotation	NOUN
fcis-25121	61	14	,	,	PUNCT
fcis-25121	61	15	which	which	PRON
fcis-25121	61	16	make	make	VERB
fcis-25121	61	17	the	the	DET
fcis-25121	61	18	network	network	NOUN
fcis-25121	61	19	with	with	ADP
fcis-25121	61	20	better	well	ADJ
fcis-25121	61	21	robustness	robustness	NOUN
fcis-25121	61	22	and	and	CCONJ
fcis-25121	61	23	fault	fault	NOUN
fcis-25121	61	24	-	-	PUNCT
fcis-25121	61	25	tolerance	tolerance	NOUN
fcis-25121	61	26	.	.	PUNCT
fcis-25121	62	1	a	a	DET
fcis-25121	62	2	cnn	cnn	PROPN
fcis-25121	62	3	layer	layer	NOUN
fcis-25121	62	4	contains	contain	VERB
fcis-25121	62	5	a	a	DET
fcis-25121	62	6	convolution	convolution	NOUN
fcis-25121	62	7	layer	layer	NOUN
fcis-25121	62	8	and	and	CCONJ
fcis-25121	62	9	a	a	DET
fcis-25121	62	10	pooling	pool	VERB
fcis-25121	62	11	layer	layer	NOUN
fcis-25121	62	12	.	.	PUNCT
fcis-25121	63	1	the	the	DET
fcis-25121	63	2	convolution	convolution	NOUN
fcis-25121	63	3	layer	layer	NOUN
fcis-25121	63	4	use	use	VERB
fcis-25121	63	5	filters	filter	NOUN
fcis-25121	63	6	to	to	PART
fcis-25121	63	7	filter	filter	VERB
fcis-25121	63	8	the	the	DET
fcis-25121	63	9	matrix	matrix	NOUN
fcis-25121	63	10	information	information	NOUN
fcis-25121	63	11	to	to	PART
fcis-25121	63	12	extract	extract	VERB
fcis-25121	63	13	features	feature	NOUN
fcis-25121	63	14	from	from	ADP
fcis-25121	63	15	the	the	DET
fcis-25121	63	16	images	image	NOUN
fcis-25121	63	17	.	.	PUNCT
fcis-25121	64	1	the	the	DET
fcis-25121	64	2	pooling	pool	VERB
fcis-25121	64	3	layer	layer	NOUN
fcis-25121	64	4	always	always	ADV
fcis-25121	64	5	appears	appear	VERB
fcis-25121	64	6	between	between	ADP
fcis-25121	64	7	two	two	NUM
fcis-25121	64	8	convolution	convolution	NOUN
fcis-25121	64	9	layers	layer	NOUN
fcis-25121	64	10	,	,	PUNCT
fcis-25121	64	11	and	and	CCONJ
fcis-25121	64	12	can	can	AUX
fcis-25121	64	13	effectively	effectively	ADV
fcis-25121	64	14	reduces	reduce	VERB
fcis-25121	64	15	the	the	DET
fcis-25121	64	16	size	size	NOUN
fcis-25121	64	17	of	of	ADP
fcis-25121	64	18	the	the	DET
fcis-25121	64	19	matrix	matrix	NOUN
fcis-25121	64	20	to	to	PART
fcis-25121	64	21	achieve	achieve	VERB
fcis-25121	64	22	sampling	sampling	NOUN
fcis-25121	64	23	.	.	PUNCT
fcis-25121	65	1	in	in	ADP
fcis-25121	65	2	addition	addition	NOUN
fcis-25121	65	3	,	,	PUNCT
fcis-25121	65	4	it	it	PRON
fcis-25121	65	5	can	can	AUX
fcis-25121	65	6	also	also	ADV
fcis-25121	65	7	reduces	reduce	VERB
fcis-25121	65	8	the	the	DET
fcis-25121	65	9	number	number	NOUN
fcis-25121	65	10	of	of	ADP
fcis-25121	65	11	parameters	parameter	NOUN
fcis-25121	65	12	in	in	ADP
fcis-25121	65	13	the	the	DET
fcis-25121	65	14	input	input	NOUN
fcis-25121	65	15	layer	layer	NOUN
fcis-25121	65	16	of	of	ADP
fcis-25121	65	17	the	the	DET
fcis-25121	65	18	fully	fully	ADV
fcis-25121	65	19	connected	connected	ADJ
fcis-25121	65	20	layer	layer	NOUN
fcis-25121	65	21	,	,	PUNCT
fcis-25121	65	22	and	and	CCONJ
fcis-25121	65	23	reduces	reduce	VERB
fcis-25121	65	24	the	the	DET
fcis-25121	65	25	amount	amount	NOUN
fcis-25121	65	26	of	of	ADP
fcis-25121	65	27	data	datum	NOUN
fcis-25121	65	28	operations	operation	NOUN
fcis-25121	65	29	to	to	PART
fcis-25121	65	30	avoid	avoid	VERB
fcis-25121	65	31	over	over	ADV
fcis-25121	65	32	-	-	PUNCT
fcis-25121	65	33	fitting	fitting	ADJ
fcis-25121	65	34	.	.	PUNCT
fcis-25121	66	1	the	the	DET
fcis-25121	66	2	fully	fully	ADV
fcis-25121	66	3	connected	connect	VERB
fcis-25121	66	4	layer	layer	NOUN
fcis-25121	66	5	is	be	AUX
fcis-25121	66	6	similar	similar	ADJ
fcis-25121	66	7	to	to	ADP
fcis-25121	66	8	a	a	DET
fcis-25121	66	9	part	part	NOUN
fcis-25121	66	10	of	of	ADP
fcis-25121	66	11	the	the	DET
fcis-25121	66	12	traditional	traditional	ADJ
fcis-25121	66	13	neural	neural	ADJ
fcis-25121	66	14	network	network	NOUN
fcis-25121	66	15	and	and	CCONJ
fcis-25121	66	16	is	be	AUX
fcis-25121	66	17	used	use	VERB
fcis-25121	66	18	to	to	ADP
fcis-25121	66	19	exports	export	NOUN
fcis-25121	66	20	result	result	VERB
fcis-25121	66	21	.	.	PUNCT
fcis-25121	67	1	the	the	DET
fcis-25121	67	2	data	datum	NOUN
fcis-25121	67	3	processed	process	VERB
fcis-25121	67	4	by	by	ADP
fcis-25121	67	5	convolution	convolution	NOUN
fcis-25121	67	6	layer	layer	NOUN
fcis-25121	67	7	and	and	CCONJ
fcis-25121	67	8	pooling	pool	VERB
fcis-25121	67	9	layer	layer	NOUN
fcis-25121	67	10	can	can	AUX
fcis-25121	67	11	be	be	AUX
fcis-25121	67	12	input	input	ADJ
fcis-25121	67	13	to	to	ADP
fcis-25121	67	14	the	the	DET
fcis-25121	67	15	fully	fully	ADV
fcis-25121	67	16	connected	connect	VERB
fcis-25121	67	17	layer	layer	NOUN
fcis-25121	67	18	.	.	PUNCT
fcis-25121	68	1	the	the	DET
fcis-25121	68	2	output	output	NOUN
fcis-25121	68	3	result	result	NOUN
fcis-25121	68	4	is	be	AUX
fcis-25121	68	5	to	to	PART
fcis-25121	68	6	combine	combine	VERB
fcis-25121	68	7	all	all	DET
fcis-25121	68	8	data	datum	NOUN
fcis-25121	68	9	into	into	ADP
fcis-25121	68	10	global	global	ADJ
fcis-25121	68	11	features	feature	NOUN
fcis-25121	68	12	to	to	PART
fcis-25121	68	13	obtain	obtain	VERB
fcis-25121	68	14	the	the	DET
fcis-25121	68	15	final	final	ADJ
fcis-25121	68	16	classification	classification	NOUN
fcis-25121	68	17	result	result	NOUN
fcis-25121	68	18	.	.	PUNCT
fcis-25121	69	1	the	the	DET
fcis-25121	69	2	activation	activation	NOUN
fcis-25121	69	3	function	function	NOUN
fcis-25121	69	4	is	be	AUX
fcis-25121	69	5	used	use	VERB
fcis-25121	69	6	to	to	PART
fcis-25121	69	7	provide	provide	VERB
fcis-25121	69	8	nonlinear	nonlinear	ADJ
fcis-25121	69	9	modeling	modeling	NOUN
fcis-25121	69	10	for	for	ADP
fcis-25121	69	11	the	the	DET
fcis-25121	69	12	network	network	NOUN
fcis-25121	69	13	by	by	ADP
fcis-25121	69	14	introducing	introduce	VERB
fcis-25121	69	15	the	the	DET
fcis-25121	69	16	nonlinear	nonlinear	ADJ
fcis-25121	69	17	elements	element	NOUN
fcis-25121	69	18	into	into	ADP
fcis-25121	69	19	the	the	DET
fcis-25121	69	20	neural	neural	ADJ
fcis-25121	69	21	network	network	NOUN
fcis-25121	69	22	,	,	PUNCT
fcis-25121	69	23	so	so	SCONJ
fcis-25121	69	24	that	that	SCONJ
fcis-25121	69	25	the	the	DET
fcis-25121	69	26	neural	neural	ADJ
fcis-25121	69	27	network	network	NOUN
fcis-25121	69	28	can	can	AUX
fcis-25121	69	29	complete	complete	VERB
fcis-25121	69	30	the	the	DET
fcis-25121	69	31	nonlinear	nonlinear	ADJ
fcis-25121	69	32	mapping	mapping	NOUN
fcis-25121	69	33	,	,	PUNCT
fcis-25121	69	34	and	and	CCONJ
fcis-25121	69	35	allow	allow	VERB
fcis-25121	69	36	the	the	DET
fcis-25121	69	37	deep	deep	ADJ
fcis-25121	69	38	neural	neural	ADJ
fcis-25121	69	39	network	network	NOUN
fcis-25121	69	40	to	to	PART
fcis-25121	69	41	have	have	VERB
fcis-25121	69	42	the	the	DET
fcis-25121	69	43	ability	ability	NOUN
fcis-25121	69	44	to	to	PART
fcis-25121	69	45	learn	learn	VERB
fcis-25121	69	46	nonlinearity	nonlinearity	NOUN
fcis-25121	69	47	.	.	PUNCT
fcis-25121	70	1	therefore	therefore	ADV
fcis-25121	70	2	,	,	PUNCT
fcis-25121	70	3	it	it	PRON
fcis-25121	70	4	is	be	AUX
fcis-25121	70	5	very	very	ADV
fcis-25121	70	6	important	important	ADJ
fcis-25121	70	7	to	to	PART
fcis-25121	70	8	add	add	VERB
fcis-25121	70	9	activation	activation	NOUN
fcis-25121	70	10	function	function	NOUN
fcis-25121	70	11	in	in	ADP
fcis-25121	70	12	the	the	DET
fcis-25121	70	13	process	process	NOUN
fcis-25121	70	14	of	of	ADP
fcis-25121	70	15	deep	deep	ADJ
fcis-25121	70	16	neural	neural	ADJ
fcis-25121	70	17	network	network	NOUN
fcis-25121	70	18	learning	learning	NOUN
fcis-25121	70	19	.	.	PUNCT
fcis-25121	71	1	the	the	DET
fcis-25121	71	2	common	common	ADJ
fcis-25121	71	3	activation	activation	NOUN
fcis-25121	71	4	functions	function	NOUN
fcis-25121	71	5	include	include	VERB
fcis-25121	71	6	sigmoid	sigmoid	NOUN
fcis-25121	71	7	,	,	PUNCT
fcis-25121	71	8	elu	elu	PROPN
fcis-25121	71	9	,	,	PUNCT
fcis-25121	71	10	tanh	tanh	PROPN
fcis-25121	71	11	and	and	CCONJ
fcis-25121	71	12	relu	relu	NOUN
fcis-25121	72	1	[	[	X
fcis-25121	72	2	9	9	NUM
fcis-25121	72	3	]	]	PUNCT
fcis-25121	72	4	.	.	PUNCT
fcis-25121	73	1	in	in	ADP
fcis-25121	73	2	this	this	DET
fcis-25121	73	3	paper	paper	NOUN
fcis-25121	73	4	,	,	PUNCT
fcis-25121	73	5	the	the	DET
fcis-25121	73	6	adopted	adopt	VERB
fcis-25121	73	7	relu	relu	NOUN
fcis-25121	73	8	is	be	AUX
fcis-25121	73	9	a	a	DET
fcis-25121	73	10	simple	simple	ADJ
fcis-25121	73	11	nonlinear	nonlinear	ADJ
fcis-25121	73	12	operation	operation	NOUN
fcis-25121	73	13	that	that	PRON
fcis-25121	73	14	take	take	VERB
fcis-25121	73	15	the	the	DET
fcis-25121	73	16	maximum	maximum	ADJ
fcis-25121	73	17	value	value	NOUN
fcis-25121	73	18	of	of	ADP
fcis-25121	73	19	0	0	NUM
fcis-25121	73	20	to	to	PART
fcis-25121	73	21	x	x	PRON
fcis-25121	73	22	which	which	PRON
fcis-25121	73	23	can	can	AUX
fcis-25121	73	24	improve	improve	VERB
fcis-25121	73	25	the	the	DET
fcis-25121	73	26	performances	performance	NOUN
fcis-25121	73	27	.	.	PUNCT
fcis-25121	74	1	the	the	DET
fcis-25121	74	2	relu	relu	NOUN
fcis-25121	74	3	function	function	NOUN
fcis-25121	74	4	is	be	AUX
fcis-25121	74	5	shown	show	VERB
fcis-25121	74	6	in	in	ADP
fcis-25121	74	7	formula	formula	NOUN
fcis-25121	74	8	(	(	PUNCT
fcis-25121	74	9	1	1	NUM
fcis-25121	74	10	):	):	PUNCT
fcis-25121	74	11	f(x	f(x	PROPN
fcis-25121	74	12	)	)	PUNCT
fcis-25121	75	1	=	=	SYM
fcis-25121	75	2	max(0	max(0	NOUN
fcis-25121	75	3	,	,	PUNCT
fcis-25121	75	4	x	x	X
fcis-25121	75	5	)	)	PUNCT
fcis-25121	75	6	(	(	PUNCT
fcis-25121	75	7	1	1	NUM
fcis-25121	75	8	)	)	PUNCT
fcis-25121	75	9	3.2.2	3.2.2	NUM
fcis-25121	75	10	.	.	PUNCT
fcis-25121	76	1	dynasty	dynasty	ADJ
fcis-25121	76	2	classification	classification	NOUN
fcis-25121	76	3	model	model	NOUN
fcis-25121	76	4	structure	structure	NOUN
fcis-25121	76	5	the	the	DET
fcis-25121	76	6	vgg	vgg	ADJ
fcis-25121	76	7	network	network	NOUN
fcis-25121	76	8	architecture	architecture	NOUN
fcis-25121	76	9	was	be	AUX
fcis-25121	76	10	introduced	introduce	VERB
fcis-25121	76	11	by	by	ADP
fcis-25121	76	12	simonyan	simonyan	ADJ
fcis-25121	76	13	and	and	CCONJ
fcis-25121	76	14	zisserman	zisserman	NOUN
fcis-25121	76	15	from	from	ADP
fcis-25121	76	16	the	the	DET
fcis-25121	76	17	university	university	NOUN
fcis-25121	76	18	of	of	ADP
fcis-25121	76	19	oxford	oxford	PROPN
fcis-25121	76	20	in	in	ADP
fcis-25121	76	21	2014	2014	NUM
fcis-25121	76	22	[	[	X
fcis-25121	76	23	10	10	NUM
fcis-25121	76	24	]	]	PUNCT
fcis-25121	76	25	.	.	PUNCT
fcis-25121	77	1	this	this	DET
fcis-25121	77	2	network	network	NOUN
fcis-25121	77	3	use	use	VERB
fcis-25121	77	4	3×3	3×3	NUM
fcis-25121	77	5	small	small	ADJ
fcis-25121	77	6	convolutional	convolutional	ADJ
fcis-25121	77	7	kernels	kernel	NOUN
fcis-25121	77	8	stacked	stack	VERB
fcis-25121	77	9	on	on	ADP
fcis-25121	77	10	top	top	NOUN
fcis-25121	77	11	of	of	ADP
fcis-25121	77	12	each	each	DET
fcis-25121	77	13	other	other	ADJ
fcis-25121	77	14	and	and	CCONJ
fcis-25121	77	15	2×2	2×2	NUM
fcis-25121	77	16	max	max	PROPN
fcis-25121	77	17	pooling	pool	VERB
fcis-25121	77	18	layers	layer	NOUN
fcis-25121	77	19	to	to	PART
fcis-25121	77	20	build	build	VERB
fcis-25121	77	21	the	the	DET
fcis-25121	77	22	cnn	cnn	PROPN
fcis-25121	77	23	with16	with16	NOUN
fcis-25121	77	24	to	to	ADP
fcis-25121	77	25	19	19	NUM
fcis-25121	77	26	layers	layer	NOUN
fcis-25121	77	27	.	.	PUNCT
fcis-25121	78	1	it	it	PRON
fcis-25121	78	2	is	be	AUX
fcis-25121	78	3	generally	generally	ADV
fcis-25121	78	4	believed	believe	VERB
fcis-25121	78	5	that	that	SCONJ
fcis-25121	78	6	deeper	deep	ADJ
fcis-25121	78	7	networks	network	NOUN
fcis-25121	78	8	have	have	VERB
fcis-25121	78	9	better	well	ADJ
fcis-25121	78	10	expression	expression	NOUN
fcis-25121	78	11	ability	ability	NOUN
fcis-25121	78	12	than	than	ADP
fcis-25121	78	13	shallow	shallow	ADJ
fcis-25121	78	14	networks	network	NOUN
fcis-25121	78	15	,	,	PUNCT
fcis-25121	78	16	and	and	CCONJ
fcis-25121	78	17	can	can	AUX
fcis-25121	78	18	complete	complete	VERB
fcis-25121	78	19	more	more	ADJ
fcis-25121	78	20	complex	complex	ADJ
fcis-25121	78	21	tasks	task	NOUN
fcis-25121	78	22	.	.	PUNCT
fcis-25121	79	1	72	72	NUM
fcis-25121	79	2	table	table	NOUN
fcis-25121	79	3	1	1	NUM
fcis-25121	79	4	.	.	PUNCT
fcis-25121	80	1	the	the	DET
fcis-25121	80	2	structural	structural	ADJ
fcis-25121	80	3	parameters	parameter	NOUN
fcis-25121	80	4	of	of	ADP
fcis-25121	80	5	different	different	ADJ
fcis-25121	80	6	vgg	vgg	PROPN
fcis-25121	80	7	networks	network	NOUN
fcis-25121	80	8	vgg11	vgg11	PRON
fcis-25121	80	9	vgg13	vgg13	VERB
fcis-25121	80	10	vgg16	vgg16	PROPN
fcis-25121	80	11	vgg19	vgg19	PROPN
fcis-25121	80	12	layer	layer	NOUN
fcis-25121	80	13	number	number	NOUN
fcis-25121	80	14	size	size	NOUN
fcis-25121	80	15	layer	layer	NOUN
fcis-25121	80	16	number	number	NOUN
fcis-25121	80	17	size	size	NOUN
fcis-25121	80	18	layer	layer	NOUN
fcis-25121	80	19	number	number	NOUN
fcis-25121	80	20	size	size	NOUN
fcis-25121	80	21	layer	layer	NOUN
fcis-25121	80	22	number	number	NOUN
fcis-25121	80	23	size	size	NOUN
fcis-25121	80	24	c1	c1	NOUN
fcis-25121	80	25	-	-	PUNCT
fcis-25121	80	26	1	1	NUM
fcis-25121	80	27	224	224	NUM
fcis-25121	80	28	*	*	NUM
fcis-25121	80	29	224	224	NUM
fcis-25121	80	30	*	*	SYM
fcis-25121	80	31	64	64	NUM
fcis-25121	80	32	c1	c1	NOUN
fcis-25121	80	33	-	-	PUNCT
fcis-25121	80	34	1	1	NUM
fcis-25121	80	35	224	224	NUM
fcis-25121	80	36	*	*	NUM
fcis-25121	80	37	224	224	NUM
fcis-25121	80	38	*	*	SYM
fcis-25121	80	39	64	64	NUM
fcis-25121	80	40	c1	c1	NOUN
fcis-25121	80	41	-	-	PUNCT
fcis-25121	80	42	1	1	NUM
fcis-25121	80	43	224	224	NUM
fcis-25121	80	44	*	*	NUM
fcis-25121	80	45	224	224	NUM
fcis-25121	80	46	*	*	SYM
fcis-25121	80	47	64	64	NUM
fcis-25121	80	48	c1	c1	NOUN
fcis-25121	80	49	-	-	PUNCT
fcis-25121	80	50	1	1	NUM
fcis-25121	80	51	224	224	NUM
fcis-25121	80	52	*	*	NUM
fcis-25121	80	53	224	224	NUM
fcis-25121	80	54	*	*	SYM
fcis-25121	80	55	64	64	NUM
fcis-25121	80	56	p1	p1	NOUN
fcis-25121	80	57	112	112	NUM
fcis-25121	80	58	*	*	NUM
fcis-25121	80	59	112	112	NUM
fcis-25121	80	60	*	*	SYM
fcis-25121	80	61	64	64	NUM
fcis-25121	80	62	c1	c1	NOUN
fcis-25121	80	63	-	-	PUNCT
fcis-25121	80	64	2	2	NUM
fcis-25121	80	65	224	224	NUM
fcis-25121	80	66	*	*	NUM
fcis-25121	80	67	224	224	NUM
fcis-25121	80	68	*	*	SYM
fcis-25121	80	69	64	64	NUM
fcis-25121	80	70	c1	c1	NOUN
fcis-25121	80	71	-	-	PUNCT
fcis-25121	80	72	2	2	NUM
fcis-25121	80	73	224	224	NUM
fcis-25121	80	74	*	*	NUM
fcis-25121	80	75	224	224	NUM
fcis-25121	80	76	*	*	SYM
fcis-25121	80	77	64	64	NUM
fcis-25121	80	78	c1	c1	NOUN
fcis-25121	80	79	-	-	PUNCT
fcis-25121	80	80	2	2	NUM
fcis-25121	80	81	224	224	NUM
fcis-25121	80	82	*	*	NUM
fcis-25121	80	83	224	224	NUM
fcis-25121	80	84	*	*	SYM
fcis-25121	80	85	64	64	NUM
fcis-25121	80	86	c2	c2	PROPN
fcis-25121	80	87	-	-	PUNCT
fcis-25121	80	88	1	1	NUM
fcis-25121	80	89	112	112	NUM
fcis-25121	80	90	*	*	NUM
fcis-25121	80	91	112	112	NUM
fcis-25121	80	92	*	*	SYM
fcis-25121	80	93	128	128	NUM
fcis-25121	80	94	p1	p1	NOUN
fcis-25121	80	95	112	112	NUM
fcis-25121	80	96	*	*	NUM
fcis-25121	80	97	112	112	NUM
fcis-25121	80	98	*	*	SYM
fcis-25121	80	99	64	64	NUM
fcis-25121	80	100	p1	p1	NOUN
fcis-25121	80	101	112	112	NUM
fcis-25121	80	102	*	*	NUM
fcis-25121	80	103	112	112	NUM
fcis-25121	80	104	*	*	SYM
fcis-25121	80	105	64	64	NUM
fcis-25121	80	106	p1	p1	NOUN
fcis-25121	80	107	112	112	NUM
fcis-25121	80	108	*	*	NUM
fcis-25121	80	109	112	112	NUM
fcis-25121	80	110	*	*	SYM
fcis-25121	80	111	64	64	NUM
fcis-25121	80	112	p2	p2	NOUN
fcis-25121	80	113	56	56	NUM
fcis-25121	80	114	*	*	SYM
fcis-25121	80	115	56	56	NUM
fcis-25121	80	116	*	*	SYM
fcis-25121	80	117	128	128	NUM
fcis-25121	80	118	c2	c2	PROPN
fcis-25121	80	119	-	-	PUNCT
fcis-25121	80	120	1	1	NUM
fcis-25121	80	121	112	112	NUM
fcis-25121	80	122	*	*	NUM
fcis-25121	80	123	112	112	NUM
fcis-25121	80	124	*	*	SYM
fcis-25121	80	125	128	128	NUM
fcis-25121	80	126	c2	c2	PROPN
fcis-25121	80	127	-	-	PUNCT
fcis-25121	80	128	1	1	NUM
fcis-25121	80	129	112	112	NUM
fcis-25121	80	130	*	*	NUM
fcis-25121	80	131	112	112	NUM
fcis-25121	80	132	*	*	SYM
fcis-25121	80	133	128	128	NUM
fcis-25121	80	134	c2	c2	PROPN
fcis-25121	80	135	-	-	PUNCT
fcis-25121	80	136	1	1	NUM
fcis-25121	80	137	112	112	NUM
fcis-25121	80	138	*	*	NUM
fcis-25121	80	139	112	112	NUM
fcis-25121	80	140	*	*	SYM
fcis-25121	80	141	128	128	NUM
fcis-25121	80	142	c3	c3	NOUN
fcis-25121	80	143	-	-	PROPN
fcis-25121	80	144	1	1	NUM
fcis-25121	80	145	56	56	NUM
fcis-25121	80	146	*	*	SYM
fcis-25121	80	147	56	56	NUM
fcis-25121	80	148	*	*	SYM
fcis-25121	80	149	256	256	NUM
fcis-25121	80	150	c2	c2	PROPN
fcis-25121	80	151	-	-	PUNCT
fcis-25121	80	152	2	2	NUM
fcis-25121	80	153	112	112	NUM
fcis-25121	80	154	*	*	NUM
fcis-25121	80	155	112	112	NUM
fcis-25121	80	156	*	*	SYM
fcis-25121	80	157	128	128	NUM
fcis-25121	80	158	c2	c2	PROPN
fcis-25121	80	159	-	-	PUNCT
fcis-25121	80	160	2	2	NUM
fcis-25121	80	161	112	112	NUM
fcis-25121	80	162	*	*	NUM
fcis-25121	80	163	112	112	NUM
fcis-25121	80	164	*	*	SYM
fcis-25121	80	165	128	128	NUM
fcis-25121	80	166	c2	c2	PROPN
fcis-25121	80	167	-	-	PUNCT
fcis-25121	80	168	2	2	NUM
fcis-25121	80	169	112	112	NUM
fcis-25121	80	170	*	*	NUM
fcis-25121	80	171	112	112	NUM
fcis-25121	80	172	*	*	SYM
fcis-25121	80	173	128	128	NUM
fcis-25121	80	174	c3	c3	NOUN
fcis-25121	80	175	-	-	PROPN
fcis-25121	80	176	2	2	NUM
fcis-25121	80	177	56	56	NUM
fcis-25121	80	178	*	*	SYM
fcis-25121	80	179	56	56	NUM
fcis-25121	80	180	*	*	SYM
fcis-25121	80	181	256	256	NUM
fcis-25121	80	182	p2	p2	NOUN
fcis-25121	80	183	56	56	NUM
fcis-25121	80	184	*	*	SYM
fcis-25121	80	185	56	56	NUM
fcis-25121	80	186	*	*	SYM
fcis-25121	80	187	128	128	NUM
fcis-25121	80	188	p2	p2	PROPN
fcis-25121	80	189	56	56	NUM
fcis-25121	80	190	*	*	SYM
fcis-25121	80	191	56	56	NUM
fcis-25121	80	192	*	*	SYM
fcis-25121	80	193	128	128	NUM
fcis-25121	80	194	p2	p2	PROPN
fcis-25121	80	195	56	56	NUM
fcis-25121	80	196	*	*	SYM
fcis-25121	80	197	56	56	NUM
fcis-25121	80	198	*	*	SYM
fcis-25121	80	199	128	128	NUM
fcis-25121	80	200	p3	p3	PROPN
fcis-25121	80	201	28	28	NUM
fcis-25121	80	202	*	*	SYM
fcis-25121	80	203	28	28	NUM
fcis-25121	80	204	*	*	SYM
fcis-25121	80	205	512	512	NUM
fcis-25121	80	206	c3	c3	NOUN
fcis-25121	80	207	-	-	PROPN
fcis-25121	80	208	1	1	NUM
fcis-25121	80	209	56	56	NUM
fcis-25121	80	210	*	*	SYM
fcis-25121	80	211	56	56	NUM
fcis-25121	80	212	*	*	SYM
fcis-25121	80	213	256	256	NUM
fcis-25121	80	214	c3	c3	NOUN
fcis-25121	80	215	-	-	PROPN
fcis-25121	80	216	1	1	NUM
fcis-25121	80	217	56	56	NUM
fcis-25121	80	218	*	*	SYM
fcis-25121	80	219	56	56	NUM
fcis-25121	80	220	*	*	SYM
fcis-25121	80	221	256	256	NUM
fcis-25121	80	222	c3	c3	NOUN
fcis-25121	80	223	-	-	PROPN
fcis-25121	80	224	1	1	NUM
fcis-25121	80	225	56	56	NUM
fcis-25121	80	226	*	*	SYM
fcis-25121	80	227	56	56	NUM
fcis-25121	80	228	*	*	SYM
fcis-25121	80	229	256	256	NUM
fcis-25121	80	230	c4	c4	NOUN
fcis-25121	80	231	-	-	PUNCT
fcis-25121	80	232	1	1	NUM
fcis-25121	80	233	28	28	NUM
fcis-25121	80	234	*	*	SYM
fcis-25121	80	235	28	28	NUM
fcis-25121	80	236	*	*	SYM
fcis-25121	80	237	512	512	NUM
fcis-25121	80	238	c3	c3	NOUN
fcis-25121	80	239	-	-	PROPN
fcis-25121	80	240	2	2	NUM
fcis-25121	80	241	56	56	NUM
fcis-25121	80	242	*	*	SYM
fcis-25121	80	243	56	56	NUM
fcis-25121	80	244	*	*	SYM
fcis-25121	80	245	256	256	NUM
fcis-25121	80	246	c3	c3	NOUN
fcis-25121	80	247	-	-	PROPN
fcis-25121	80	248	2	2	NUM
fcis-25121	80	249	56	56	NUM
fcis-25121	80	250	*	*	SYM
fcis-25121	80	251	56	56	NUM
fcis-25121	80	252	*	*	SYM
fcis-25121	80	253	256	256	NUM
fcis-25121	80	254	c3	c3	NOUN
fcis-25121	80	255	-	-	PROPN
fcis-25121	80	256	2	2	NUM
fcis-25121	80	257	56	56	NUM
fcis-25121	80	258	*	*	SYM
fcis-25121	80	259	56	56	NUM
fcis-25121	80	260	*	*	SYM
fcis-25121	80	261	256	256	NUM
fcis-25121	80	262	c4	c4	NOUN
fcis-25121	80	263	-	-	PUNCT
fcis-25121	80	264	2	2	NUM
fcis-25121	80	265	28	28	NUM
fcis-25121	80	266	*	*	SYM
fcis-25121	80	267	28	28	NUM
fcis-25121	80	268	*	*	SYM
fcis-25121	80	269	512	512	NUM
fcis-25121	80	270	p3	p3	PROPN
fcis-25121	80	271	28	28	NUM
fcis-25121	80	272	*	*	SYM
fcis-25121	80	273	28	28	NUM
fcis-25121	80	274	*	*	SYM
fcis-25121	80	275	256	256	NUM
fcis-25121	80	276	c3	c3	NOUN
fcis-25121	80	277	-	-	PROPN
fcis-25121	80	278	3	3	NUM
fcis-25121	80	279	56	56	NUM
fcis-25121	80	280	*	*	SYM
fcis-25121	80	281	56	56	NUM
fcis-25121	80	282	*	*	SYM
fcis-25121	80	283	256	256	NUM
fcis-25121	80	284	c3	c3	NOUN
fcis-25121	80	285	-	-	PROPN
fcis-25121	80	286	3	3	NUM
fcis-25121	80	287	56	56	NUM
fcis-25121	80	288	*	*	SYM
fcis-25121	80	289	56	56	NUM
fcis-25121	80	290	*	*	SYM
fcis-25121	80	291	256	256	NUM
fcis-25121	80	292	p4	p4	ADJ
fcis-25121	80	293	14	14	NUM
fcis-25121	80	294	*	*	SYM
fcis-25121	80	295	14	14	NUM
fcis-25121	80	296	*	*	SYM
fcis-25121	80	297	512	512	NUM
fcis-25121	80	298	c4	c4	NOUN
fcis-25121	80	299	-	-	PUNCT
fcis-25121	80	300	1	1	NUM
fcis-25121	80	301	28	28	NUM
fcis-25121	80	302	*	*	SYM
fcis-25121	80	303	28	28	NUM
fcis-25121	80	304	*	*	SYM
fcis-25121	80	305	512	512	NUM
fcis-25121	80	306	p3	p3	PROPN
fcis-25121	80	307	28	28	NUM
fcis-25121	80	308	*	*	SYM
fcis-25121	80	309	28	28	NUM
fcis-25121	80	310	*	*	SYM
fcis-25121	80	311	256	256	NUM
fcis-25121	80	312	c3	c3	NOUN
fcis-25121	80	313	-	-	PUNCT
fcis-25121	80	314	4	4	NUM
fcis-25121	80	315	56	56	NUM
fcis-25121	80	316	*	*	SYM
fcis-25121	80	317	56	56	NUM
fcis-25121	80	318	*	*	SYM
fcis-25121	80	319	256	256	NUM
fcis-25121	80	320	c5	c5	PROPN
fcis-25121	80	321	-	-	PUNCT
fcis-25121	80	322	1	1	NUM
fcis-25121	80	323	14	14	NUM
fcis-25121	80	324	*	*	SYM
fcis-25121	80	325	14	14	NUM
fcis-25121	80	326	*	*	SYM
fcis-25121	80	327	512	512	NUM
fcis-25121	80	328	c4	c4	NOUN
fcis-25121	80	329	-	-	PUNCT
fcis-25121	80	330	2	2	NUM
fcis-25121	80	331	28	28	NUM
fcis-25121	80	332	*	*	SYM
fcis-25121	80	333	28	28	NUM
fcis-25121	80	334	*	*	SYM
fcis-25121	80	335	512	512	NUM
fcis-25121	80	336	c4	c4	NOUN
fcis-25121	80	337	-	-	PUNCT
fcis-25121	80	338	1	1	NUM
fcis-25121	80	339	28	28	NUM
fcis-25121	80	340	*	*	SYM
fcis-25121	80	341	28	28	NUM
fcis-25121	80	342	*	*	SYM
fcis-25121	80	343	512	512	NUM
fcis-25121	80	344	p3	p3	PROPN
fcis-25121	80	345	28	28	NUM
fcis-25121	80	346	*	*	SYM
fcis-25121	80	347	28	28	NUM
fcis-25121	80	348	*	*	SYM
fcis-25121	80	349	256	256	NUM
fcis-25121	80	350	c-2	c-2	NOUN
fcis-25121	80	351	14	14	NUM
fcis-25121	80	352	*	*	SYM
fcis-25121	80	353	14	14	NUM
fcis-25121	80	354	*	*	SYM
fcis-25121	80	355	512	512	NUM
fcis-25121	80	356	p4	p4	ADJ
fcis-25121	80	357	14	14	NUM
fcis-25121	80	358	*	*	SYM
fcis-25121	80	359	14	14	NUM
fcis-25121	80	360	*	*	SYM
fcis-25121	80	361	512	512	NUM
fcis-25121	80	362	c4	c4	NOUN
fcis-25121	80	363	-	-	PUNCT
fcis-25121	80	364	2	2	NUM
fcis-25121	80	365	28	28	NUM
fcis-25121	80	366	*	*	SYM
fcis-25121	80	367	28	28	NUM
fcis-25121	80	368	*	*	SYM
fcis-25121	80	369	512	512	NUM
fcis-25121	80	370	c4	c4	NOUN
fcis-25121	80	371	-	-	PUNCT
fcis-25121	80	372	1	1	NUM
fcis-25121	80	373	28	28	NUM
fcis-25121	80	374	*	*	SYM
fcis-25121	80	375	28	28	NUM
fcis-25121	80	376	*	*	SYM
fcis-25121	80	377	512	512	NUM
fcis-25121	80	378	p5	p5	ADJ
fcis-25121	80	379	7	7	NUM
fcis-25121	80	380	*	*	SYM
fcis-25121	80	381	7	7	NUM
fcis-25121	80	382	*	*	SYM
fcis-25121	80	383	512	512	NUM
fcis-25121	80	384	c5	c5	PROPN
fcis-25121	80	385	-	-	PUNCT
fcis-25121	80	386	1	1	NUM
fcis-25121	80	387	14	14	NUM
fcis-25121	80	388	*	*	SYM
fcis-25121	80	389	14	14	NUM
fcis-25121	80	390	*	*	SYM
fcis-25121	80	391	512	512	NUM
fcis-25121	80	392	c4	c4	NOUN
fcis-25121	80	393	-	-	PUNCT
fcis-25121	80	394	3	3	NUM
fcis-25121	80	395	28	28	NUM
fcis-25121	80	396	*	*	SYM
fcis-25121	80	397	28	28	NUM
fcis-25121	80	398	*	*	SYM
fcis-25121	80	399	512	512	NUM
fcis-25121	80	400	c4	c4	NOUN
fcis-25121	80	401	-	-	PUNCT
fcis-25121	80	402	2	2	NUM
fcis-25121	80	403	28	28	NUM
fcis-25121	80	404	*	*	SYM
fcis-25121	80	405	28	28	NUM
fcis-25121	80	406	*	*	SYM
fcis-25121	80	407	512	512	NUM
fcis-25121	80	408	fc-1	fc-1	NOUN
fcis-25121	80	409	1	1	NUM
fcis-25121	80	410	*	*	NUM
fcis-25121	80	411	1	1	NUM
fcis-25121	80	412	*	*	SYM
fcis-25121	80	413	4096	4096	NUM
fcis-25121	80	414	c5	c5	PROPN
fcis-25121	80	415	-	-	PUNCT
fcis-25121	80	416	2	2	NUM
fcis-25121	80	417	14	14	NUM
fcis-25121	80	418	*	*	SYM
fcis-25121	80	419	14	14	NUM
fcis-25121	80	420	*	*	SYM
fcis-25121	80	421	512	512	NUM
fcis-25121	80	422	p4	p4	ADJ
fcis-25121	80	423	14	14	NUM
fcis-25121	80	424	*	*	SYM
fcis-25121	80	425	14	14	NUM
fcis-25121	80	426	*	*	SYM
fcis-25121	80	427	512	512	NUM
fcis-25121	80	428	c4	c4	NOUN
fcis-25121	80	429	-	-	PUNCT
fcis-25121	80	430	3	3	NUM
fcis-25121	80	431	28	28	NUM
fcis-25121	80	432	*	*	SYM
fcis-25121	80	433	28	28	NUM
fcis-25121	80	434	*	*	SYM
fcis-25121	80	435	512	512	NUM
fcis-25121	80	436	fc-2	fc-2	ADP
fcis-25121	80	437	1	1	NUM
fcis-25121	80	438	*	*	SYM
fcis-25121	80	439	1	1	NUM
fcis-25121	80	440	*	*	SYM
fcis-25121	80	441	4096	4096	NUM
fcis-25121	80	442	p5	p5	ADJ
fcis-25121	80	443	7	7	NUM
fcis-25121	80	444	*	*	SYM
fcis-25121	80	445	7	7	NUM
fcis-25121	80	446	*	*	SYM
fcis-25121	80	447	512	512	NUM
fcis-25121	80	448	c5	c5	PROPN
fcis-25121	80	449	-	-	PUNCT
fcis-25121	80	450	1	1	NUM
fcis-25121	80	451	14	14	NUM
fcis-25121	80	452	*	*	SYM
fcis-25121	80	453	14	14	NUM
fcis-25121	80	454	*	*	SYM
fcis-25121	80	455	512	512	NUM
fcis-25121	80	456	c4	c4	NOUN
fcis-25121	80	457	-	-	PUNCT
fcis-25121	80	458	4	4	NUM
fcis-25121	80	459	28	28	NUM
fcis-25121	80	460	*	*	SYM
fcis-25121	80	461	28	28	NUM
fcis-25121	80	462	*	*	SYM
fcis-25121	80	463	512	512	NUM
fcis-25121	80	464	fc-3	fc-3	NUM
fcis-25121	80	465	1	1	NUM
fcis-25121	80	466	*	*	NUM
fcis-25121	80	467	1	1	NUM
fcis-25121	80	468	*	*	SYM
fcis-25121	80	469	6	6	NUM
fcis-25121	80	470	fc-1	fc-1	NOUN
fcis-25121	80	471	1	1	NUM
fcis-25121	80	472	*	*	SYM
fcis-25121	80	473	1	1	NUM
fcis-25121	80	474	*	*	SYM
fcis-25121	80	475	4096	4096	NUM
fcis-25121	80	476	c5	c5	PROPN
fcis-25121	80	477	-	-	PUNCT
fcis-25121	80	478	2	2	NUM
fcis-25121	80	479	14	14	NUM
fcis-25121	80	480	*	*	SYM
fcis-25121	80	481	14	14	NUM
fcis-25121	80	482	*	*	SYM
fcis-25121	80	483	512	512	NUM
fcis-25121	80	484	p4	p4	ADJ
fcis-25121	80	485	14	14	NUM
fcis-25121	80	486	*	*	SYM
fcis-25121	80	487	14	14	NUM
fcis-25121	80	488	*	*	SYM
fcis-25121	80	489	512	512	NUM
fcis-25121	80	490	sm	sm	NOUN
fcis-25121	80	491	1	1	NUM
fcis-25121	80	492	*	*	SYM
fcis-25121	80	493	1	1	NUM
fcis-25121	80	494	*	*	SYM
fcis-25121	80	495	6	6	NUM
fcis-25121	80	496	fc-2	fc-2	NUM
fcis-25121	80	497	1	1	NUM
fcis-25121	80	498	*	*	SYM
fcis-25121	80	499	1	1	NUM
fcis-25121	80	500	*	*	SYM
fcis-25121	80	501	4096	4096	NUM
fcis-25121	80	502	c5	c5	PROPN
fcis-25121	80	503	-	-	PUNCT
fcis-25121	80	504	3	3	NUM
fcis-25121	80	505	14	14	NUM
fcis-25121	80	506	*	*	SYM
fcis-25121	80	507	14	14	NUM
fcis-25121	80	508	*	*	SYM
fcis-25121	80	509	512	512	NUM
fcis-25121	80	510	c5	c5	PROPN
fcis-25121	80	511	-	-	PUNCT
fcis-25121	80	512	1	1	NUM
fcis-25121	80	513	14	14	NUM
fcis-25121	80	514	*	*	SYM
fcis-25121	80	515	14	14	NUM
fcis-25121	80	516	*	*	SYM
fcis-25121	80	517	512	512	NUM
fcis-25121	80	518	fc-3	fc-3	NUM
fcis-25121	80	519	1	1	NUM
fcis-25121	80	520	*	*	NUM
fcis-25121	80	521	1	1	NUM
fcis-25121	80	522	*	*	SYM
fcis-25121	80	523	6	6	NUM
fcis-25121	80	524	p5	p5	ADJ
fcis-25121	80	525	7	7	NUM
fcis-25121	80	526	*	*	SYM
fcis-25121	80	527	7	7	NUM
fcis-25121	80	528	*	*	SYM
fcis-25121	80	529	512	512	NUM
fcis-25121	80	530	c5	c5	PROPN
fcis-25121	80	531	-	-	PUNCT
fcis-25121	80	532	2	2	NUM
fcis-25121	80	533	14	14	NUM
fcis-25121	80	534	*	*	SYM
fcis-25121	80	535	14	14	NUM
fcis-25121	80	536	*	*	SYM
fcis-25121	80	537	512	512	NUM
fcis-25121	80	538	sm	sm	NOUN
fcis-25121	80	539	1	1	NUM
fcis-25121	80	540	*	*	SYM
fcis-25121	80	541	1	1	NUM
fcis-25121	80	542	*	*	SYM
fcis-25121	80	543	6	6	NUM
fcis-25121	80	544	fc-1	fc-1	NOUN
fcis-25121	80	545	1	1	NUM
fcis-25121	80	546	*	*	SYM
fcis-25121	80	547	1	1	NUM
fcis-25121	80	548	*	*	SYM
fcis-25121	80	549	4096	4096	NUM
fcis-25121	80	550	c5	c5	PROPN
fcis-25121	80	551	-	-	PUNCT
fcis-25121	80	552	3	3	NUM
fcis-25121	80	553	14	14	NUM
fcis-25121	80	554	*	*	SYM
fcis-25121	80	555	14	14	NUM
fcis-25121	80	556	*	*	SYM
fcis-25121	80	557	512	512	NUM
fcis-25121	80	558	fc-2	fc-2	ADP
fcis-25121	80	559	1	1	NUM
fcis-25121	80	560	*	*	SYM
fcis-25121	80	561	1	1	NUM
fcis-25121	80	562	*	*	SYM
fcis-25121	80	563	4096	4096	NUM
fcis-25121	80	564	c5	c5	PROPN
fcis-25121	80	565	-	-	PUNCT
fcis-25121	80	566	4	4	NUM
fcis-25121	80	567	14	14	NUM
fcis-25121	80	568	*	*	SYM
fcis-25121	80	569	14	14	NUM
fcis-25121	80	570	*	*	SYM
fcis-25121	80	571	512	512	NUM
fcis-25121	80	572	fc-3	fc-3	NUM
fcis-25121	80	573	1	1	NUM
fcis-25121	80	574	*	*	NUM
fcis-25121	80	575	1	1	NUM
fcis-25121	80	576	*	*	SYM
fcis-25121	80	577	6	6	NUM
fcis-25121	80	578	p5	p5	ADJ
fcis-25121	80	579	7	7	NUM
fcis-25121	80	580	*	*	SYM
fcis-25121	80	581	7	7	NUM
fcis-25121	80	582	*	*	SYM
fcis-25121	80	583	512	512	NUM
fcis-25121	80	584	sm	sm	NOUN
fcis-25121	80	585	1	1	NUM
fcis-25121	80	586	*	*	SYM
fcis-25121	80	587	1	1	NUM
fcis-25121	80	588	*	*	SYM
fcis-25121	80	589	6	6	NUM
fcis-25121	80	590	fc-1	fc-1	NOUN
fcis-25121	80	591	1	1	NUM
fcis-25121	80	592	*	*	SYM
fcis-25121	80	593	1	1	NUM
fcis-25121	80	594	*	*	SYM
fcis-25121	80	595	4096	4096	NUM
fcis-25121	80	596	fc-2	fc-2	ADP
fcis-25121	80	597	1	1	NUM
fcis-25121	80	598	*	*	SYM
fcis-25121	80	599	1	1	NUM
fcis-25121	80	600	*	*	SYM
fcis-25121	80	601	4096	4096	NUM
fcis-25121	80	602	fc-3	fc-3	NOUN
fcis-25121	80	603	1	1	NUM
fcis-25121	80	604	*	*	NUM
fcis-25121	80	605	1	1	NUM
fcis-25121	80	606	*	*	NUM
fcis-25121	80	607	6	6	NUM
fcis-25121	80	608	sm	sm	NOUN
fcis-25121	80	609	1	1	NUM
fcis-25121	80	610	*	*	SYM
fcis-25121	80	611	1	1	NUM
fcis-25121	80	612	*	*	SYM
fcis-25121	80	613	6	6	NUM
fcis-25121	80	614	because	because	SCONJ
fcis-25121	80	615	the	the	DET
fcis-25121	80	616	choice	choice	NOUN
fcis-25121	80	617	of	of	ADP
fcis-25121	80	618	network	network	NOUN
fcis-25121	80	619	architecture	architecture	NOUN
fcis-25121	80	620	has	have	VERB
fcis-25121	80	621	a	a	DET
fcis-25121	80	622	great	great	ADJ
fcis-25121	80	623	influence	influence	NOUN
fcis-25121	80	624	on	on	ADP
fcis-25121	80	625	the	the	DET
fcis-25121	80	626	effect	effect	NOUN
fcis-25121	80	627	and	and	CCONJ
fcis-25121	80	628	efficiency	efficiency	NOUN
fcis-25121	80	629	of	of	ADP
fcis-25121	80	630	image	image	NOUN
fcis-25121	80	631	classification	classification	NOUN
fcis-25121	80	632	.	.	PUNCT
fcis-25121	81	1	therefore	therefore	ADV
fcis-25121	81	2	,	,	PUNCT
fcis-25121	81	3	this	this	DET
fcis-25121	81	4	dynasty	dynasty	ADJ
fcis-25121	81	5	classification	classification	NOUN
fcis-25121	81	6	model	model	NOUN
fcis-25121	81	7	adopts	adopt	VERB
fcis-25121	81	8	the	the	DET
fcis-25121	81	9	network	network	NOUN
fcis-25121	81	10	architecture	architecture	NOUN
fcis-25121	81	11	to	to	ADP
fcis-25121	81	12	the	the	DET
fcis-25121	81	13	depth	depth	NOUN
fcis-25121	81	14	of	of	ADP
fcis-25121	81	15	11	11	NUM
fcis-25121	81	16	to	to	PART
fcis-25121	81	17	19	19	NUM
fcis-25121	81	18	layers	layer	NOUN
fcis-25121	81	19	on	on	ADP
fcis-25121	81	20	the	the	DET
fcis-25121	81	21	basis	basis	NOUN
fcis-25121	81	22	of	of	ADP
fcis-25121	81	23	vgg	vgg	ADJ
fcis-25121	81	24	network	network	NOUN
fcis-25121	81	25	.	.	PUNCT
fcis-25121	82	1	the	the	DET
fcis-25121	82	2	structural	structural	ADJ
fcis-25121	82	3	parameters	parameter	NOUN
fcis-25121	82	4	of	of	ADP
fcis-25121	82	5	vgg	vgg	PROPN
fcis-25121	82	6	networks	network	NOUN
fcis-25121	82	7	are	be	AUX
fcis-25121	82	8	shown	show	VERB
fcis-25121	82	9	in	in	ADP
fcis-25121	82	10	table	table	NOUN
fcis-25121	82	11	1	1	NUM
fcis-25121	82	12	.	.	PUNCT
fcis-25121	83	1	we	we	PRON
fcis-25121	83	2	choose	choose	VERB
fcis-25121	83	3	rgb	rgb	PROPN
fcis-25121	83	4	3	3	NUM
fcis-25121	83	5	-	-	PUNCT
fcis-25121	83	6	channels	channel	NOUN
fcis-25121	83	7	mural	mural	ADJ
fcis-25121	83	8	images	image	NOUN
fcis-25121	83	9	with	with	ADP
fcis-25121	83	10	the	the	DET
fcis-25121	83	11	size	size	NOUN
fcis-25121	83	12	of	of	ADP
fcis-25121	83	13	224	224	NUM
fcis-25121	83	14	×224	×224	NOUN
fcis-25121	83	15	as	as	ADP
fcis-25121	83	16	the	the	DET
fcis-25121	83	17	input	input	NOUN
fcis-25121	83	18	to	to	ADP
fcis-25121	83	19	the	the	DET
fcis-25121	83	20	model	model	NOUN
fcis-25121	83	21	.	.	PUNCT
fcis-25121	84	1	in	in	ADP
fcis-25121	84	2	order	order	NOUN
fcis-25121	84	3	to	to	PART
fcis-25121	84	4	retain	retain	VERB
fcis-25121	84	5	the	the	DET
fcis-25121	84	6	image	image	NOUN
fcis-25121	84	7	information	information	NOUN
fcis-25121	84	8	more	more	ADV
fcis-25121	84	9	completely	completely	ADV
fcis-25121	84	10	and	and	CCONJ
fcis-25121	84	11	improve	improve	VERB
fcis-25121	84	12	the	the	DET
fcis-25121	84	13	accuracy	accuracy	NOUN
fcis-25121	84	14	of	of	ADP
fcis-25121	84	15	mural	mural	ADJ
fcis-25121	84	16	classification	classification	NOUN
fcis-25121	84	17	,	,	PUNCT
fcis-25121	84	18	3×3	3×3	NUM
fcis-25121	84	19	kernel	kernel	NOUN
fcis-25121	84	20	and	and	CCONJ
fcis-25121	84	21	1step	1step	NUM
fcis-25121	84	22	size	size	NOUN
fcis-25121	84	23	are	be	AUX
fcis-25121	84	24	selected	select	VERB
fcis-25121	84	25	to	to	PART
fcis-25121	84	26	run	run	VERB
fcis-25121	84	27	convolution	convolution	NOUN
fcis-25121	84	28	,	,	PUNCT
fcis-25121	84	29	2×2	2×2	NUM
fcis-25121	84	30	kernel	kernel	NOUN
fcis-25121	84	31	and	and	CCONJ
fcis-25121	84	32	2	2	NUM
fcis-25121	84	33	step	step	NOUN
fcis-25121	84	34	size	size	NOUN
fcis-25121	84	35	are	be	AUX
fcis-25121	84	36	selected	select	VERB
fcis-25121	84	37	for	for	ADP
fcis-25121	84	38	pooling.the	pooling.the	DET
fcis-25121	84	39	rectified	rectify	VERB
fcis-25121	84	40	linear	linear	NOUN
fcis-25121	84	41	unit	unit	NOUN
fcis-25121	84	42	(	(	PUNCT
fcis-25121	84	43	relu)is	relu)is	PROPN
fcis-25121	84	44	applied	apply	VERB
fcis-25121	84	45	as	as	ADP
fcis-25121	84	46	activation	activation	NOUN
fcis-25121	84	47	function	function	NOUN
fcis-25121	84	48	.	.	PUNCT
fcis-25121	85	1	dropout	dropout	NOUN
fcis-25121	85	2	is	be	AUX
fcis-25121	85	3	applied	apply	VERB
fcis-25121	85	4	in	in	ADP
fcis-25121	85	5	the	the	DET
fcis-25121	85	6	fully	fully	ADV
fcis-25121	85	7	connected	connect	VERB
fcis-25121	85	8	layer	layer	NOUN
fcis-25121	85	9	by	by	ADP
fcis-25121	85	10	dropping	drop	VERB
fcis-25121	85	11	neurons	neuron	NOUN
fcis-25121	85	12	randomly	randomly	ADV
fcis-25121	85	13	to	to	PART
fcis-25121	85	14	avoid	avoid	VERB
fcis-25121	85	15	overfitting	overfitte	VERB
fcis-25121	85	16	.	.	PUNCT
fcis-25121	86	1	softmax	softmax	NOUN
fcis-25121	86	2	is	be	AUX
fcis-25121	86	3	used	use	VERB
fcis-25121	86	4	in	in	ADP
fcis-25121	86	5	the	the	DET
fcis-25121	86	6	fully	fully	ADV
fcis-25121	86	7	connected	connect	VERB
fcis-25121	86	8	layer	layer	NOUN
fcis-25121	86	9	to	to	PART
fcis-25121	86	10	classify	classify	VERB
fcis-25121	86	11	the	the	DET
fcis-25121	86	12	mural	mural	ADJ
fcis-25121	86	13	images	image	NOUN
fcis-25121	86	14	,	,	PUNCT
fcis-25121	86	15	which	which	PRON
fcis-25121	86	16	can	can	AUX
fcis-25121	86	17	map	map	VERB
fcis-25121	86	18	the	the	DET
fcis-25121	86	19	fully	fully	ADV
fcis-25121	86	20	connected	connect	VERB
fcis-25121	86	21	layer	layer	NOUN
fcis-25121	86	22	to	to	ADP
fcis-25121	86	23	values	value	NOUN
fcis-25121	86	24	from	from	ADP
fcis-25121	86	25	0	0	NUM
fcis-25121	86	26	to1	to1	PROPN
fcis-25121	86	27	,	,	PUNCT
fcis-25121	86	28	that	that	PRON
fcis-25121	86	29	can	can	AUX
fcis-25121	86	30	represent	represent	VERB
fcis-25121	86	31	the	the	DET
fcis-25121	86	32	probability	probability	NOUN
fcis-25121	86	33	value	value	NOUN
fcis-25121	86	34	of	of	ADP
fcis-25121	86	35	each	each	DET
fcis-25121	86	36	category	category	NOUN
fcis-25121	86	37	.	.	PUNCT
fcis-25121	87	1	its	its	PRON
fcis-25121	87	2	definition	definition	NOUN
fcis-25121	87	3	function	function	NOUN
fcis-25121	87	4	is	be	AUX
fcis-25121	87	5	shown	show	VERB
fcis-25121	87	6	in	in	ADP
fcis-25121	87	7	equation	equation	NOUN
fcis-25121	87	8	(	(	PUNCT
fcis-25121	87	9	2	2	NUM
fcis-25121	87	10	):	):	PUNCT
fcis-25121	87	11	vi	vi	VERB
fcis-25121	88	1	i	i	PRON
fcis-25121	88	2	c	c	NOUN
fcis-25121	88	3	vi	vi	VERB
fcis-25121	89	1	i	i	NOUN
fcis-25121	89	2	e	e	NOUN
fcis-25121	89	3	s	s	PROPN
fcis-25121	89	4	e	e	X
fcis-25121	89	5			NOUN
fcis-25121	89	6			X
fcis-25121	89	7	(	(	PUNCT
fcis-25121	89	8	2	2	X
fcis-25121	89	9	)	)	PUNCT
fcis-25121	89	10	the	the	DET
fcis-25121	89	11	mural	mural	ADJ
fcis-25121	89	12	classification	classification	NOUN
fcis-25121	89	13	network	network	NOUN
fcis-25121	89	14	models	model	NOUN
fcis-25121	89	15	use	use	VERB
fcis-25121	89	16	cross	cross	NOUN
fcis-25121	89	17	entropy	entropy	NOUN
fcis-25121	89	18	as	as	ADP
fcis-25121	89	19	the	the	DET
fcis-25121	89	20	loss	loss	NOUN
fcis-25121	89	21	function	function	NOUN
fcis-25121	89	22	.	.	PUNCT
fcis-25121	90	1	as	as	SCONJ
fcis-25121	90	2	shown	show	VERB
fcis-25121	90	3	in	in	ADP
fcis-25121	90	4	formula	formula	NOUN
fcis-25121	90	5	(	(	PUNCT
fcis-25121	90	6	3	3	NUM
fcis-25121	90	7	)	)	PUNCT
fcis-25121	90	8	,	,	PUNCT
fcis-25121	90	9	the	the	DET
fcis-25121	90	10	loss	loss	NOUN
fcis-25121	90	11	function	function	NOUN
fcis-25121	90	12	use	use	VERB
fcis-25121	90	13	the	the	DET
fcis-25121	90	14	distribution	distribution	NOUN
fcis-25121	90	15	probability	probability	NOUN
fcis-25121	90	16	q	q	NOUN
fcis-25121	90	17	to	to	PART
fcis-25121	90	18	express	express	VERB
fcis-25121	90	19	the	the	DET
fcis-25121	90	20	accuracy	accuracy	NOUN
fcis-25121	90	21	of	of	ADP
fcis-25121	90	22	probability	probability	NOUN
fcis-25121	90	23	distribution	distribution	NOUN
fcis-25121	90	24	p	p	NOUN
fcis-25121	90	25	.	.	PUNCT
fcis-25121	91	1	the	the	DET
fcis-25121	91	2	proximity	proximity	NOUN
fcis-25121	91	3	of	of	ADP
fcis-25121	91	4	the	the	DET
fcis-25121	91	5	two	two	NUM
fcis-25121	91	6	probability	probability	NOUN
fcis-25121	91	7	distributions	distribution	NOUN
fcis-25121	91	8	is	be	AUX
fcis-25121	91	9	expressed	express	VERB
fcis-25121	91	10	by	by	ADP
fcis-25121	91	11	calculating	calculate	VERB
fcis-25121	91	12	the	the	DET
fcis-25121	91	13	distance	distance	NOUN
fcis-25121	91	14	between	between	ADP
fcis-25121	91	15	the	the	DET
fcis-25121	91	16	two	two	NUM
fcis-25121	91	17	probability	probability	NOUN
fcis-25121	91	18	distributions	distribution	NOUN
fcis-25121	91	19	.	.	PUNCT
fcis-25121	92	1	where	where	SCONJ
fcis-25121	92	2	p	p	NOUN
fcis-25121	92	3	is	be	AUX
fcis-25121	92	4	the	the	DET
fcis-25121	92	5	tag	tag	NOUN
fcis-25121	92	6	value	value	NOUN
fcis-25121	92	7	and	and	CCONJ
fcis-25121	92	8	q	q	NOUN
fcis-25121	92	9	is	be	AUX
fcis-25121	92	10	predicted	predict	VERB
fcis-25121	92	11	value	value	NOUN
fcis-25121	92	12	.	.	PUNCT
fcis-25121	93	1	the	the	DET
fcis-25121	93	2	shorter	short	ADJ
fcis-25121	93	3	the	the	DET
fcis-25121	93	4	distance	distance	NOUN
fcis-25121	93	5	between	between	ADP
fcis-25121	93	6	the	the	DET
fcis-25121	93	7	two	two	NUM
fcis-25121	93	8	probability	probability	NOUN
fcis-25121	93	9	distributions	distribution	NOUN
fcis-25121	93	10	,	,	PUNCT
fcis-25121	93	11	the	the	PRON
fcis-25121	93	12	closer	close	ADV
fcis-25121	93	13	the	the	DET
fcis-25121	93	14	predicted	predict	VERB
fcis-25121	93	15	value	value	NOUN
fcis-25121	93	16	is	be	AUX
fcis-25121	93	17	to	to	ADP
fcis-25121	93	18	the	the	DET
fcis-25121	93	19	tag	tag	NOUN
fcis-25121	93	20	value	value	NOUN
fcis-25121	93	21	.	.	PUNCT
fcis-25121	94	1	(	(	PUNCT
fcis-25121	94	2	,	,	PUNCT
fcis-25121	94	3	)	)	PUNCT
fcis-25121	94	4	(	(	PUNCT
fcis-25121	94	5	)	)	PUNCT
fcis-25121	94	6	log	log	NOUN
fcis-25121	94	7	(	(	PUNCT
fcis-25121	94	8	)	)	PUNCT
fcis-25121	94	9	x	x	SYM
fcis-25121	94	10	h	h	NOUN
fcis-25121	95	1	p	p	NOUN
fcis-25121	95	2	q	q	X
fcis-25121	95	3	p	p	X
fcis-25121	95	4	x	x	X
fcis-25121	95	5	q	q	X
fcis-25121	95	6	x	x	PUNCT
fcis-25121	96	1			X
fcis-25121	96	2	(	(	PUNCT
fcis-25121	96	3	3	3	NUM
fcis-25121	96	4	)	)	PUNCT
fcis-25121	96	5	4	4	NUM
fcis-25121	96	6	.	.	PUNCT
fcis-25121	96	7	experimental	experimental	ADJ
fcis-25121	96	8	verification	verification	NOUN
fcis-25121	96	9	and	and	CCONJ
fcis-25121	96	10	result	result	VERB
fcis-25121	96	11	analysis	analysis	NOUN
fcis-25121	96	12	4.1	4.1	NUM
fcis-25121	96	13	.	.	PUNCT
fcis-25121	97	1	experimental	experimental	ADJ
fcis-25121	97	2	environment	environment	NOUN
fcis-25121	97	3	model	model	NOUN
fcis-25121	97	4	training	training	NOUN
fcis-25121	97	5	and	and	CCONJ
fcis-25121	97	6	testing	testing	NOUN
fcis-25121	97	7	are	be	AUX
fcis-25121	97	8	carried	carry	VERB
fcis-25121	97	9	out	out	ADP
fcis-25121	97	10	on	on	ADP
fcis-25121	97	11	windows10	windows10	NOUN
fcis-25121	97	12	operating	operating	NOUN
fcis-25121	97	13	system	system	NOUN
fcis-25121	97	14	,	,	PUNCT
fcis-25121	97	15	and	and	CCONJ
fcis-25121	97	16	python	python	NOUN
fcis-25121	97	17	is	be	AUX
fcis-25121	97	18	selected	select	VERB
fcis-25121	97	19	as	as	ADP
fcis-25121	97	20	the	the	DET
fcis-25121	97	21	programming	programming	NOUN
fcis-25121	97	22	language	language	NOUN
fcis-25121	97	23	.	.	PUNCT
fcis-25121	98	1	i7	i7	NOUN
fcis-25121	98	2	-	-	PUNCT
fcis-25121	98	3	10700	10700	NUM
fcis-25121	98	4	processor	processor	NOUN
fcis-25121	98	5	,	,	PUNCT
fcis-25121	98	6	nvidia	nvidia	PROPN
fcis-25121	98	7	gtx3060	gtx3060	NOUN
fcis-25121	98	8	gpu	gpu	PROPN
fcis-25121	98	9	,	,	PUNCT
fcis-25121	98	10	memory	memory	NOUN
fcis-25121	98	11	16	16	NUM
fcis-25121	98	12	gb	gb	NOUN
fcis-25121	98	13	.	.	PUNCT
fcis-25121	99	1	4.2	4.2	NUM
fcis-25121	99	2	.	.	PUNCT
fcis-25121	100	1	model	model	NOUN
fcis-25121	100	2	comparison	comparison	NOUN
fcis-25121	100	3	experiment	experiment	NOUN
fcis-25121	100	4	831	831	NUM
fcis-25121	100	5	images	image	NOUN
fcis-25121	100	6	of	of	ADP
fcis-25121	100	7	each	each	DET
fcis-25121	100	8	dynasty	dynasty	NOUN
fcis-25121	100	9	of	of	ADP
fcis-25121	100	10	murals	mural	NOUN
fcis-25121	100	11	are	be	AUX
fcis-25121	100	12	in	in	ADP
fcis-25121	100	13	the	the	DET
fcis-25121	100	14	training	training	NOUN
fcis-25121	100	15	set	set	NOUN
fcis-25121	100	16	.	.	PUNCT
fcis-25121	101	1	after	after	ADP
fcis-25121	101	2	training	training	NOUN
fcis-25121	101	3	and	and	CCONJ
fcis-25121	101	4	testing	test	VERB
fcis-25121	101	5	the	the	DET
fcis-25121	101	6	models	model	NOUN
fcis-25121	101	7	for	for	ADP
fcis-25121	101	8	many	many	ADJ
fcis-25121	101	9	epochs	epoch	NOUN
fcis-25121	101	10	,	,	PUNCT
fcis-25121	101	11	the	the	DET
fcis-25121	101	12	learning	learning	NOUN
fcis-25121	101	13	rate	rate	NOUN
fcis-25121	101	14	is	be	AUX
fcis-25121	101	15	set	set	VERB
fcis-25121	101	16	to	to	ADP
fcis-25121	101	17	0.0001	0.0001	NUM
fcis-25121	101	18	,	,	PUNCT
fcis-25121	101	19	the	the	DET
fcis-25121	101	20	batch	batch	NOUN
fcis-25121	101	21	size	size	NOUN
fcis-25121	101	22	is	be	AUX
fcis-25121	101	23	set	set	VERB
fcis-25121	101	24	to	to	ADP
fcis-25121	101	25	20	20	NUM
fcis-25121	101	26	and	and	CCONJ
fcis-25121	101	27	the	the	DET
fcis-25121	101	28	number	number	NOUN
fcis-25121	101	29	of	of	ADP
fcis-25121	101	30	epochs	epoch	NOUN
fcis-25121	101	31	is	be	AUX
fcis-25121	101	32	set	set	VERB
fcis-25121	101	33	to	to	ADP
fcis-25121	101	34	200	200	NUM
fcis-25121	101	35	.	.	PUNCT
fcis-25121	102	1	four	four	NUM
fcis-25121	102	2	models	model	NOUN
fcis-25121	102	3	are	be	AUX
fcis-25121	102	4	trained	train	VERB
fcis-25121	102	5	to	to	PART
fcis-25121	102	6	classify	classify	VERB
fcis-25121	102	7	the	the	DET
fcis-25121	102	8	mural	mural	ADJ
fcis-25121	102	9	data	data	NOUN
fcis-25121	102	10	sets	set	NOUN
fcis-25121	102	11	of	of	ADP
fcis-25121	102	12	five	five	NUM
fcis-25121	102	13	dynasties	dynasty	NOUN
fcis-25121	102	14	,	,	PUNCT
fcis-25121	102	15	and	and	CCONJ
fcis-25121	102	16	the	the	DET
fcis-25121	102	17	loss	loss	NOUN
fcis-25121	102	18	value	value	NOUN
fcis-25121	102	19	and	and	CCONJ
fcis-25121	102	20	accuracy	accuracy	NOUN
fcis-25121	102	21	are	be	AUX
fcis-25121	102	22	visualized	visualize	VERB
fcis-25121	102	23	to	to	PART
fcis-25121	102	24	analyze	analyze	VERB
fcis-25121	102	25	the	the	DET
fcis-25121	102	26	convergence	convergence	NOUN
fcis-25121	102	27	performance	performance	NOUN
fcis-25121	102	28	of	of	ADP
fcis-25121	102	29	the	the	DET
fcis-25121	102	30	model	model	NOUN
fcis-25121	102	31	,	,	PUNCT
fcis-25121	102	32	as	as	SCONJ
fcis-25121	102	33	shown	show	VERB
fcis-25121	102	34	in	in	ADP
fcis-25121	102	35	fig	fig	NOUN
fcis-25121	102	36	.	.	PUNCT
fcis-25121	103	1	3(a	3(a	NUM
fcis-25121	103	2	)	)	PUNCT
fcis-25121	104	1	,	,	PUNCT
fcis-25121	104	2	x	x	X
fcis-25121	104	3	-	-	NOUN
fcis-25121	104	4	axis	axis	NOUN
fcis-25121	104	5	is	be	AUX
fcis-25121	104	6	the	the	DET
fcis-25121	104	7	number	number	NOUN
fcis-25121	104	8	of	of	ADP
fcis-25121	104	9	epochs	epoch	NOUN
fcis-25121	104	10	,	,	PUNCT
fcis-25121	104	11	y	y	NOUN
fcis-25121	104	12	-	-	PUNCT
fcis-25121	104	13	axis	axis	NOUN
fcis-25121	104	14	is	be	AUX
fcis-25121	104	15	the	the	DET
fcis-25121	104	16	training	training	NOUN
fcis-25121	104	17	loss	loss	NOUN
fcis-25121	104	18	.	.	PUNCT
fcis-25121	105	1	in	in	ADP
fcis-25121	105	2	fig	fig	NOUN
fcis-25121	105	3	.	.	PUNCT
fcis-25121	106	1	3(b	3(b	NUM
fcis-25121	106	2	)	)	PUNCT
fcis-25121	106	3	,	,	PUNCT
fcis-25121	106	4	x	x	X
fcis-25121	106	5	-	-	NOUN
fcis-25121	106	6	axis	axis	NOUN
fcis-25121	106	7	is	be	AUX
fcis-25121	106	8	the	the	DET
fcis-25121	106	9	number	number	NOUN
fcis-25121	106	10	of	of	ADP
fcis-25121	106	11	epochs	epoch	NOUN
fcis-25121	106	12	,	,	PUNCT
fcis-25121	106	13	y	y	NOUN
fcis-25121	106	14	-	-	PUNCT
fcis-25121	106	15	axis	axis	NOUN
fcis-25121	106	16	is	be	AUX
fcis-25121	106	17	the	the	DET
fcis-25121	106	18	training	training	NOUN
fcis-25121	106	19	accuracy	accuracy	NOUN
fcis-25121	106	20	.	.	PUNCT
fcis-25121	107	1	it	it	PRON
fcis-25121	107	2	can	can	AUX
fcis-25121	107	3	be	be	AUX
fcis-25121	107	4	seen	see	VERB
fcis-25121	107	5	from	from	ADP
fcis-25121	107	6	fig	fig	NOUN
fcis-25121	107	7	.	.	PUNCT
fcis-25121	108	1	3(a	3(a	NUM
fcis-25121	108	2	)	)	PUNCT
fcis-25121	108	3	that	that	SCONJ
fcis-25121	108	4	as	as	ADP
fcis-25121	108	5	the	the	DET
fcis-25121	108	6	number	number	NOUN
fcis-25121	108	7	of	of	ADP
fcis-25121	108	8	epochs	epoch	NOUN
fcis-25121	108	9	increases	increase	NOUN
fcis-25121	108	10	,	,	PUNCT
fcis-25121	108	11	the	the	DET
fcis-25121	108	12	vgg11	vgg11	PROPN
fcis-25121	108	13	model	model	NOUN
fcis-25121	108	14	training	training	NOUN
fcis-25121	108	15	loss	loss	NOUN
fcis-25121	108	16	is	be	AUX
fcis-25121	108	17	gradually	gradually	ADV
fcis-25121	108	18	decreasing	decrease	VERB
fcis-25121	108	19	.	.	PUNCT
fcis-25121	109	1	after	after	ADP
fcis-25121	109	2	roughly	roughly	ADV
fcis-25121	109	3	100	100	NUM
fcis-25121	109	4	epochs	epoch	NOUN
fcis-25121	109	5	,	,	PUNCT
fcis-25121	109	6	the	the	DET
fcis-25121	109	7	training	training	NOUN
fcis-25121	109	8	loss	loss	NOUN
fcis-25121	109	9	curve	curve	NOUN
fcis-25121	109	10	become	become	VERB
fcis-25121	109	11	stable	stable	ADJ
fcis-25121	109	12	.	.	PUNCT
fcis-25121	110	1	in	in	ADP
fcis-25121	110	2	fig	fig	NOUN
fcis-25121	110	3	.	.	PUNCT
fcis-25121	111	1	3(b	3(b	NUM
fcis-25121	111	2	)	)	PUNCT
fcis-25121	111	3	,	,	PUNCT
fcis-25121	111	4	the	the	DET
fcis-25121	111	5	vgg11	vgg11	PROPN
fcis-25121	111	6	model	model	NOUN
fcis-25121	111	7	training	training	NOUN
fcis-25121	111	8	accuracy	accuracy	NOUN
fcis-25121	111	9	is	be	AUX
fcis-25121	111	10	gradually	gradually	ADV
fcis-25121	111	11	increasing	increase	VERB
fcis-25121	111	12	and	and	CCONJ
fcis-25121	111	13	converges	converge	NOUN
fcis-25121	111	14	between	between	ADP
fcis-25121	111	15	0.85	0.85	NUM
fcis-25121	111	16	and	and	CCONJ
fcis-25121	111	17	0.9	0.9	NUM
fcis-25121	111	18	.	.	PUNCT
fcis-25121	112	1	after	after	ADP
fcis-25121	112	2	100	100	NUM
fcis-25121	112	3	epochs	epoch	NOUN
fcis-25121	112	4	,	,	PUNCT
fcis-25121	112	5	the	the	DET
fcis-25121	112	6	curve	curve	NOUN
fcis-25121	112	7	stabilizes	stabilize	VERB
fcis-25121	112	8	,	,	PUNCT
fcis-25121	112	9	and	and	CCONJ
fcis-25121	112	10	its	its	PRON
fcis-25121	112	11	accuracy	accuracy	NOUN
fcis-25121	112	12	is	be	AUX
fcis-25121	112	13	about	about	ADV
fcis-25121	112	14	0.9	0.9	NUM
fcis-25121	112	15	.	.	PUNCT
fcis-25121	113	1	73	73	NUM
fcis-25121	113	2	(	(	PUNCT
fcis-25121	113	3	a	a	NOUN
fcis-25121	113	4	)	)	PUNCT
fcis-25121	113	5	comparison	comparison	NOUN
fcis-25121	113	6	of	of	ADP
fcis-25121	113	7	traning	trane	VERB
fcis-25121	113	8	loss	loss	NOUN
fcis-25121	113	9	(	(	PUNCT
fcis-25121	113	10	b	b	NOUN
fcis-25121	113	11	)	)	PUNCT
fcis-25121	113	12	comparison	comparison	NOUN
fcis-25121	113	13	of	of	ADP
fcis-25121	113	14	classification	classification	NOUN
fcis-25121	113	15	accuracy	accuracy	NOUN
fcis-25121	113	16	fig	fig	NOUN
fcis-25121	113	17	3	3	X
fcis-25121	113	18	.	.	PUNCT
fcis-25121	114	1	experimental	experimental	ADJ
fcis-25121	114	2	results	result	NOUN
fcis-25121	114	3	of	of	ADP
fcis-25121	114	4	network	network	NOUN
fcis-25121	114	5	training	training	NOUN
fcis-25121	114	6	compared	compare	VERB
fcis-25121	114	7	to	to	ADP
fcis-25121	114	8	vgg11	vgg11	PROPN
fcis-25121	114	9	,	,	PUNCT
fcis-25121	114	10	although	although	SCONJ
fcis-25121	114	11	vgg16	vgg16	NOUN
fcis-25121	114	12	model	model	NOUN
fcis-25121	114	13	and	and	CCONJ
fcis-25121	114	14	vgg19	vgg19	PROPN
fcis-25121	114	15	model	model	NOUN
fcis-25121	114	16	have	have	VERB
fcis-25121	114	17	deeper	deep	ADJ
fcis-25121	114	18	layers	layer	NOUN
fcis-25121	114	19	,	,	PUNCT
fcis-25121	114	20	but	but	CCONJ
fcis-25121	114	21	they	they	PRON
fcis-25121	114	22	have	have	VERB
fcis-25121	114	23	similar	similar	ADJ
fcis-25121	114	24	convergence	convergence	NOUN
fcis-25121	114	25	trend	trend	NOUN
fcis-25121	114	26	of	of	ADP
fcis-25121	114	27	loss	loss	NOUN
fcis-25121	114	28	curve	curve	NOUN
fcis-25121	114	29	and	and	CCONJ
fcis-25121	114	30	higher	high	ADJ
fcis-25121	114	31	training	training	NOUN
fcis-25121	114	32	loss	loss	NOUN
fcis-25121	114	33	.	.	PUNCT
fcis-25121	115	1	the	the	DET
fcis-25121	115	2	convergence	convergence	NOUN
fcis-25121	115	3	values	value	NOUN
fcis-25121	115	4	and	and	CCONJ
fcis-25121	115	5	the	the	DET
fcis-25121	115	6	accuracy	accuracy	NOUN
fcis-25121	115	7	of	of	ADP
fcis-25121	115	8	vgg13	vgg13	PROPN
fcis-25121	115	9	model	model	NOUN
fcis-25121	115	10	are	be	AUX
fcis-25121	115	11	basically	basically	ADV
fcis-25121	115	12	the	the	DET
fcis-25121	115	13	same	same	ADJ
fcis-25121	115	14	as	as	ADP
fcis-25121	115	15	the	the	DET
fcis-25121	115	16	vgg11	vgg11	PROPN
fcis-25121	115	17	model	model	NOUN
fcis-25121	115	18	.	.	PUNCT
fcis-25121	116	1	in	in	ADP
fcis-25121	116	2	the	the	DET
fcis-25121	116	3	training	training	NOUN
fcis-25121	116	4	process	process	NOUN
fcis-25121	116	5	,	,	PUNCT
fcis-25121	116	6	when	when	SCONJ
fcis-25121	116	7	the	the	DET
fcis-25121	116	8	loss	loss	NOUN
fcis-25121	116	9	curve	curve	NOUN
fcis-25121	116	10	fluctuates	fluctuate	NOUN
fcis-25121	116	11	,	,	PUNCT
fcis-25121	116	12	its	its	PRON
fcis-25121	116	13	training	training	NOUN
fcis-25121	116	14	accuracy	accuracy	NOUN
fcis-25121	116	15	will	will	AUX
fcis-25121	116	16	also	also	ADV
fcis-25121	116	17	decline	decline	VERB
fcis-25121	116	18	.	.	PUNCT
fcis-25121	117	1	by	by	ADP
fcis-25121	117	2	comparing	compare	VERB
fcis-25121	117	3	four	four	NUM
fcis-25121	117	4	networks	network	NOUN
fcis-25121	117	5	with	with	ADP
fcis-25121	117	6	different	different	ADJ
fcis-25121	117	7	structures	structure	NOUN
fcis-25121	117	8	,	,	PUNCT
fcis-25121	117	9	it	it	PRON
fcis-25121	117	10	is	be	AUX
fcis-25121	117	11	known	know	VERB
fcis-25121	117	12	that	that	SCONJ
fcis-25121	117	13	the	the	PRON
fcis-25121	117	14	deeper	deep	ADJ
fcis-25121	117	15	the	the	DET
fcis-25121	117	16	network	network	NOUN
fcis-25121	117	17	is	be	AUX
fcis-25121	117	18	,	,	PUNCT
fcis-25121	117	19	the	the	PRON
fcis-25121	117	20	greater	great	ADJ
fcis-25121	117	21	the	the	DET
fcis-25121	117	22	convergence	convergence	NOUN
fcis-25121	117	23	value	value	NOUN
fcis-25121	117	24	of	of	ADP
fcis-25121	117	25	the	the	DET
fcis-25121	117	26	loss	loss	NOUN
fcis-25121	117	27	function	function	NOUN
fcis-25121	117	28	is	be	AUX
fcis-25121	117	29	and	and	CCONJ
fcis-25121	117	30	the	the	PRON
fcis-25121	117	31	lower	low	ADJ
fcis-25121	117	32	the	the	DET
fcis-25121	117	33	training	training	NOUN
fcis-25121	117	34	accuracy	accuracy	NOUN
fcis-25121	117	35	is	be	AUX
fcis-25121	117	36	.	.	PUNCT
fcis-25121	118	1	therefore	therefore	ADV
fcis-25121	118	2	,	,	PUNCT
fcis-25121	118	3	when	when	SCONJ
fcis-25121	118	4	the	the	DET
fcis-25121	118	5	network	network	NOUN
fcis-25121	118	6	depth	depth	NOUN
fcis-25121	118	7	is	be	AUX
fcis-25121	118	8	enough	enough	ADJ
fcis-25121	118	9	,	,	PUNCT
fcis-25121	118	10	continuing	continue	VERB
fcis-25121	118	11	to	to	PART
fcis-25121	118	12	increase	increase	VERB
fcis-25121	118	13	the	the	DET
fcis-25121	118	14	network	network	NOUN
fcis-25121	118	15	depth	depth	NOUN
fcis-25121	118	16	does	do	AUX
fcis-25121	118	17	not	not	PART
fcis-25121	118	18	help	help	VERB
fcis-25121	118	19	the	the	DET
fcis-25121	118	20	network	network	NOUN
fcis-25121	118	21	convergence	convergence	NOUN
fcis-25121	118	22	.	.	PUNCT
fcis-25121	119	1	4.3	4.3	NUM
fcis-25121	119	2	.	.	PUNCT
fcis-25121	119	3	model	model	NOUN
fcis-25121	119	4	accuracy	accuracy	NOUN
fcis-25121	119	5	analysis	analysis	NOUN
fcis-25121	119	6	in	in	ADP
fcis-25121	119	7	this	this	DET
fcis-25121	119	8	experiment	experiment	NOUN
fcis-25121	119	9	,	,	PUNCT
fcis-25121	119	10	we	we	PRON
fcis-25121	119	11	compare	compare	VERB
fcis-25121	119	12	and	and	CCONJ
fcis-25121	119	13	analyze	analyze	VERB
fcis-25121	119	14	the	the	DET
fcis-25121	119	15	accuracy	accuracy	NOUN
fcis-25121	119	16	of	of	ADP
fcis-25121	119	17	vgg	vgg	PROPN
fcis-25121	119	18	model	model	NOUN
fcis-25121	119	19	in	in	ADP
fcis-25121	119	20	the	the	DET
fcis-25121	119	21	category	category	NOUN
fcis-25121	119	22	recognition	recognition	NOUN
fcis-25121	119	23	of	of	ADP
fcis-25121	119	24	each	each	DET
fcis-25121	119	25	dynasty	dynasty	NOUN
fcis-25121	119	26	on	on	ADP
fcis-25121	119	27	the	the	DET
fcis-25121	119	28	11th	11th	NOUN
fcis-25121	119	29	to	to	ADP
fcis-25121	119	30	19th	19th	ADJ
fcis-25121	119	31	layers	layer	NOUN
fcis-25121	119	32	of	of	ADP
fcis-25121	119	33	network	network	NOUN
fcis-25121	119	34	structure	structure	NOUN
fcis-25121	119	35	.	.	PUNCT
fcis-25121	120	1	the	the	DET
fcis-25121	120	2	accuracy	accuracy	NOUN
fcis-25121	120	3	of	of	ADP
fcis-25121	120	4	each	each	DET
fcis-25121	120	5	model	model	NOUN
fcis-25121	120	6	is	be	AUX
fcis-25121	120	7	shown	show	VERB
fcis-25121	120	8	in	in	ADP
fcis-25121	120	9	table	table	NOUN
fcis-25121	120	10	2	2	NUM
fcis-25121	120	11	.	.	PUNCT
fcis-25121	121	1	the	the	DET
fcis-25121	121	2	classification	classification	NOUN
fcis-25121	121	3	accuracy	accuracy	NOUN
fcis-25121	121	4	of	of	ADP
fcis-25121	121	5	vgg11	vgg11	PROPN
fcis-25121	121	6	in	in	ADP
fcis-25121	121	7	the	the	DET
fcis-25121	121	8	sui	sui	PROPN
fcis-25121	121	9	dynasty	dynasty	NOUN
fcis-25121	121	10	is	be	AUX
fcis-25121	121	11	lower	low	ADJ
fcis-25121	121	12	than	than	ADP
fcis-25121	121	13	that	that	PRON
fcis-25121	121	14	in	in	ADP
fcis-25121	121	15	the	the	DET
fcis-25121	121	16	five	five	NUM
fcis-25121	121	17	dynasties	dynasty	NOUN
fcis-25121	121	18	and	and	CCONJ
fcis-25121	121	19	song	song	NOUN
fcis-25121	121	20	dynasty	dynasty	NOUN
fcis-25121	121	21	and	and	CCONJ
fcis-25121	121	22	the	the	DET
fcis-25121	121	23	classification	classification	NOUN
fcis-25121	121	24	accuracy	accuracy	NOUN
fcis-25121	121	25	of	of	ADP
fcis-25121	121	26	other	other	ADJ
fcis-25121	121	27	dynasties	dynasty	NOUN
fcis-25121	121	28	is	be	AUX
fcis-25121	121	29	more	more	ADJ
fcis-25121	121	30	than	than	ADP
fcis-25121	121	31	95	95	NUM
fcis-25121	121	32	%	%	NOUN
fcis-25121	121	33	.	.	PUNCT
fcis-25121	122	1	in	in	ADP
fcis-25121	122	2	vgg13	vgg13	PROPN
fcis-25121	122	3	,	,	PUNCT
fcis-25121	122	4	only	only	ADV
fcis-25121	122	5	the	the	DET
fcis-25121	122	6	classification	classification	NOUN
fcis-25121	122	7	accuracy	accuracy	NOUN
fcis-25121	122	8	of	of	ADP
fcis-25121	122	9	the	the	DET
fcis-25121	122	10	northern	northern	ADJ
fcis-25121	122	11	dynasties	dynasty	NOUN
fcis-25121	122	12	is	be	AUX
fcis-25121	122	13	more	more	ADJ
fcis-25121	122	14	than	than	ADP
fcis-25121	122	15	95	95	NUM
fcis-25121	122	16	%	%	NOUN
fcis-25121	122	17	,	,	PUNCT
fcis-25121	122	18	and	and	CCONJ
fcis-25121	122	19	that	that	PRON
fcis-25121	122	20	of	of	ADP
fcis-25121	122	21	other	other	ADJ
fcis-25121	122	22	dynasties	dynasty	NOUN
fcis-25121	122	23	is	be	AUX
fcis-25121	122	24	about	about	ADV
fcis-25121	122	25	90	90	NUM
fcis-25121	122	26	%	%	NOUN
fcis-25121	122	27	.	.	PUNCT
fcis-25121	123	1	the	the	DET
fcis-25121	123	2	lowest	low	ADJ
fcis-25121	123	3	classification	classification	NOUN
fcis-25121	123	4	accuracy	accuracy	NOUN
fcis-25121	123	5	of	of	ADP
fcis-25121	123	6	vgg16	vgg16	NOUN
fcis-25121	123	7	is	be	AUX
fcis-25121	123	8	about	about	ADV
fcis-25121	123	9	60	60	NUM
fcis-25121	123	10	%	%	NOUN
fcis-25121	123	11	in	in	ADP
fcis-25121	123	12	the	the	DET
fcis-25121	123	13	five	five	NUM
fcis-25121	123	14	dynasties	dynasty	NOUN
fcis-25121	123	15	and	and	CCONJ
fcis-25121	123	16	song	song	NOUN
fcis-25121	123	17	dynasty	dynasty	NOUN
fcis-25121	123	18	,	,	PUNCT
fcis-25121	123	19	and	and	CCONJ
fcis-25121	123	20	the	the	DET
fcis-25121	123	21	classification	classification	NOUN
fcis-25121	123	22	accuracy	accuracy	NOUN
fcis-25121	123	23	of	of	ADP
fcis-25121	123	24	other	other	ADJ
fcis-25121	123	25	dynasties	dynasty	NOUN
fcis-25121	123	26	is	be	AUX
fcis-25121	123	27	more	more	ADJ
fcis-25121	123	28	than	than	ADP
fcis-25121	123	29	90	90	NUM
fcis-25121	123	30	%	%	NOUN
fcis-25121	123	31	.	.	PUNCT
fcis-25121	124	1	the	the	DET
fcis-25121	124	2	classification	classification	NOUN
fcis-25121	124	3	accuracy	accuracy	NOUN
fcis-25121	124	4	of	of	ADP
fcis-25121	124	5	vgg19	vgg19	NOUN
fcis-25121	124	6	in	in	ADP
fcis-25121	124	7	the	the	DET
fcis-25121	124	8	sui	sui	PROPN
fcis-25121	124	9	dynasty	dynasty	NOUN
fcis-25121	124	10	is	be	AUX
fcis-25121	124	11	lower	low	ADJ
fcis-25121	124	12	than	than	ADP
fcis-25121	124	13	that	that	PRON
fcis-25121	124	14	in	in	ADP
fcis-25121	124	15	the	the	DET
fcis-25121	124	16	five	five	NUM
fcis-25121	124	17	dynasties	dynasty	NOUN
fcis-25121	124	18	and	and	CCONJ
fcis-25121	124	19	song	song	NOUN
fcis-25121	124	20	dynasty	dynasty	NOUN
fcis-25121	124	21	,	,	PUNCT
fcis-25121	124	22	and	and	CCONJ
fcis-25121	124	23	the	the	DET
fcis-25121	124	24	classification	classification	NOUN
fcis-25121	124	25	accuracy	accuracy	NOUN
fcis-25121	124	26	of	of	ADP
fcis-25121	124	27	other	other	ADJ
fcis-25121	124	28	dynasties	dynasty	NOUN
fcis-25121	124	29	is	be	AUX
fcis-25121	124	30	more	more	ADJ
fcis-25121	124	31	than	than	ADP
fcis-25121	124	32	90	90	NUM
fcis-25121	124	33	%	%	NOUN
fcis-25121	124	34	.	.	PUNCT
fcis-25121	125	1	these	these	DET
fcis-25121	125	2	results	result	NOUN
fcis-25121	125	3	indicate	indicate	VERB
fcis-25121	125	4	that	that	SCONJ
fcis-25121	125	5	vgg11	vgg11	PROPN
fcis-25121	125	6	achieves	achieve	VERB
fcis-25121	125	7	a	a	DET
fcis-25121	125	8	high	high	ADJ
fcis-25121	125	9	accuracy	accuracy	NOUN
fcis-25121	125	10	of	of	ADP
fcis-25121	125	11	96	96	NUM
fcis-25121	125	12	%	%	NOUN
fcis-25121	125	13	on	on	ADP
fcis-25121	125	14	fewer	few	ADJ
fcis-25121	125	15	training	training	NOUN
fcis-25121	125	16	parameters	parameter	NOUN
fcis-25121	125	17	.	.	PUNCT
fcis-25121	126	1	table	table	NOUN
fcis-25121	126	2	2	2	NUM
fcis-25121	126	3	.	.	PUNCT
fcis-25121	126	4	comparison	comparison	NOUN
fcis-25121	126	5	of	of	ADP
fcis-25121	126	6	accuracy	accuracy	NOUN
fcis-25121	126	7	under	under	ADP
fcis-25121	126	8	different	different	ADJ
fcis-25121	126	9	dynasties	dynasty	NOUN
fcis-25121	126	10	/%	/%	PRON
fcis-25121	126	11	model	model	VERB
fcis-25121	126	12	the	the	DET
fcis-25121	126	13	northern	northern	ADJ
fcis-25121	126	14	dynasties	dynasty	NOUN
fcis-25121	126	15	the	the	DET
fcis-25121	126	16	sui	sui	PROPN
fcis-25121	126	17	dynasty	dynasty	PROPN
fcis-25121	126	18	the	the	DET
fcis-25121	126	19	tang	tang	PROPN
fcis-25121	126	20	dynasty	dynasty	VERB
fcis-25121	126	21	the	the	DET
fcis-25121	126	22	five	five	NUM
fcis-25121	126	23	dynasties	dynasty	NOUN
fcis-25121	126	24	and	and	CCONJ
fcis-25121	126	25	song	song	NOUN
fcis-25121	126	26	daynasty	daynasty	NOUN
fcis-25121	126	27	the	the	DET
fcis-25121	126	28	western	western	ADJ
fcis-25121	126	29	xia	xia	PROPN
fcis-25121	126	30	regime	regime	NOUN
fcis-25121	126	31	and	and	CCONJ
fcis-25121	127	1	the	the	DET
fcis-25121	127	2	yuan	yuan	NOUN
fcis-25121	127	3	dynasty	dynasty	PROPN
fcis-25121	127	4	avg	avg	NOUN
fcis-25121	127	5	vgg11	vgg11	VERB
fcis-25121	127	6	100	100	NUM
fcis-25121	127	7	90	90	NUM
fcis-25121	127	8	100	100	NUM
fcis-25121	127	9	90	90	NUM
fcis-25121	127	10	100	100	NUM
fcis-25121	127	11	96	96	NUM
fcis-25121	127	12	vgg13	vgg13	VERB
fcis-25121	127	13	100	100	NUM
fcis-25121	127	14	90	90	NUM
fcis-25121	127	15	90	90	NUM
fcis-25121	127	16	90	90	NUM
fcis-25121	127	17	90	90	NUM
fcis-25121	127	18	92	92	NUM
fcis-25121	127	19	vgg16	vgg16	NOUN
fcis-25121	127	20	100	100	NUM
fcis-25121	127	21	90	90	NUM
fcis-25121	127	22	100	100	NUM
fcis-25121	127	23	60	60	NUM
fcis-25121	127	24	90	90	NUM
fcis-25121	127	25	88	88	NUM
fcis-25121	127	26	vgg19	vgg19	NOUN
fcis-25121	127	27	100	100	NUM
fcis-25121	127	28	90	90	NUM
fcis-25121	127	29	100	100	NUM
fcis-25121	127	30	90	90	NUM
fcis-25121	127	31	100	100	NUM
fcis-25121	127	32	96	96	NUM
fcis-25121	127	33	(	(	PUNCT
fcis-25121	127	34	a	a	NOUN
fcis-25121	127	35	)	)	PUNCT
fcis-25121	127	36	bodhisattva	bodhisattva	PROPN
fcis-25121	127	37	(	(	PUNCT
fcis-25121	127	38	b	b	NOUN
fcis-25121	127	39	)	)	PUNCT
fcis-25121	127	40	bodhisattva	bodhisattva	PROPN
fcis-25121	127	41	fig	fig	PROPN
fcis-25121	127	42	4	4	NUM
fcis-25121	127	43	.	.	PUNCT
fcis-25121	127	44	murals	mural	NOUN
fcis-25121	127	45	of	of	ADP
fcis-25121	127	46	the	the	DET
fcis-25121	127	47	northern	northern	ADJ
fcis-25121	127	48	dynasties	dynasty	NOUN
fcis-25121	127	49	a	a	DET
fcis-25121	127	50	b	b	NOUN
fcis-25121	127	51	c	c	NOUN
fcis-25121	127	52	d	d	PROPN
fcis-25121	127	53	(	(	PUNCT
fcis-25121	127	54	a)the	a)the	DET
fcis-25121	127	55	sui	sui	PROPN
fcis-25121	127	56	dynasty	dynasty	NOUN
fcis-25121	127	57	(	(	PUNCT
fcis-25121	127	58	b	b	NOUN
fcis-25121	127	59	)	)	PUNCT
fcis-25121	127	60	the	the	DET
fcis-25121	127	61	northern	northern	ADJ
fcis-25121	127	62	dynasties	dynasty	NOUN
fcis-25121	127	63	(	(	PUNCT
fcis-25121	127	64	c	c	X
fcis-25121	127	65	)	)	PUNCT
fcis-25121	127	66	the	the	DET
fcis-25121	127	67	five	five	NUM
fcis-25121	127	68	dynasties	dynasty	NOUN
fcis-25121	127	69	and	and	CCONJ
fcis-25121	127	70	song	song	NOUN
fcis-25121	127	71	dynasty	dynasty	NOUN
fcis-25121	127	72	(	(	PUNCT
fcis-25121	127	73	d	d	NOUN
fcis-25121	127	74	)	)	PUNCT
fcis-25121	127	75	the	the	DET
fcis-25121	127	76	tang	tang	PROPN
fcis-25121	127	77	dynasty	dynasty	PROPN
fcis-25121	127	78	fig	fig	PROPN
fcis-25121	127	79	5	5	NUM
fcis-25121	127	80	.	.	PUNCT
fcis-25121	128	1	examples	example	NOUN
fcis-25121	128	2	of	of	ADP
fcis-25121	128	3	misclassified	misclassified	ADJ
fcis-25121	128	4	murals	mural	NOUN
fcis-25121	128	5	through	through	ADP
fcis-25121	128	6	comparison	comparison	NOUN
fcis-25121	129	1	,	,	PUNCT
fcis-25121	129	2	it	it	PRON
fcis-25121	129	3	can	can	AUX
fcis-25121	129	4	be	be	AUX
fcis-25121	129	5	seen	see	VERB
fcis-25121	129	6	that	that	SCONJ
fcis-25121	129	7	the	the	DET
fcis-25121	129	8	classification	classification	NOUN
fcis-25121	129	9	accuracy	accuracy	NOUN
fcis-25121	129	10	of	of	ADP
fcis-25121	129	11	four	four	NUM
fcis-25121	129	12	models	model	NOUN
fcis-25121	129	13	in	in	ADP
fcis-25121	129	14	the	the	DET
fcis-25121	129	15	northern	northern	ADJ
fcis-25121	129	16	dynasty	dynasty	NOUN
fcis-25121	129	17	is	be	AUX
fcis-25121	129	18	generally	generally	ADV
fcis-25121	129	19	high	high	ADJ
fcis-25121	129	20	.	.	PUNCT
fcis-25121	130	1	as	as	SCONJ
fcis-25121	130	2	shown	show	VERB
fcis-25121	130	3	in	in	ADP
fcis-25121	130	4	fig	fig	NOUN
fcis-25121	130	5	.	.	PUNCT
fcis-25121	131	1	4	4	NUM
fcis-25121	131	2	,	,	PUNCT
fcis-25121	131	3	compared	compare	VERB
fcis-25121	131	4	with	with	ADP
fcis-25121	131	5	other	other	ADJ
fcis-25121	131	6	dynasties	dynasty	NOUN
fcis-25121	131	7	,	,	PUNCT
fcis-25121	131	8	the	the	DET
fcis-25121	131	9	characters	character	NOUN
fcis-25121	131	10	in	in	ADP
fcis-25121	131	11	dunhuang	dunhuang	NOUN
fcis-25121	131	12	murals	mural	NOUN
fcis-25121	131	13	in	in	ADP
fcis-25121	131	14	the	the	DET
fcis-25121	131	15	northern	northern	ADJ
fcis-25121	131	16	dynasty	dynasty	NOUN
fcis-25121	131	17	are	be	AUX
fcis-25121	131	18	showned	showne	VERB
fcis-25121	131	19	with	with	ADP
fcis-25121	131	20	white	white	PROPN
fcis-25121	131	21	nose	nose	NOUN
fcis-25121	131	22	and	and	CCONJ
fcis-25121	131	23	eyes	eye	NOUN
fcis-25121	131	24	which	which	PRON
fcis-25121	131	25	made	make	VERB
fcis-25121	131	26	the	the	DET
fcis-25121	131	27	characteristics	characteristic	NOUN
fcis-25121	131	28	of	of	ADP
fcis-25121	131	29	the	the	DET
fcis-25121	131	30	characters	character	NOUN
fcis-25121	131	31	are	be	AUX
fcis-25121	131	32	significant	significant	ADJ
fcis-25121	131	33	.	.	PUNCT
fcis-25121	132	1	therefore	therefore	ADV
fcis-25121	132	2	,	,	PUNCT
fcis-25121	132	3	all	all	DET
fcis-25121	132	4	the	the	DET
fcis-25121	132	5	models	model	NOUN
fcis-25121	132	6	in	in	ADP
fcis-25121	132	7	the	the	DET
fcis-25121	132	8	experiment	experiment	NOUN
fcis-25121	132	9	can	can	AUX
fcis-25121	132	10	recognize	recognize	VERB
fcis-25121	132	11	the	the	DET
fcis-25121	132	12	mural	mural	ADJ
fcis-25121	132	13	images	image	NOUN
fcis-25121	132	14	of	of	ADP
fcis-25121	132	15	the	the	DET
fcis-25121	132	16	northern	northern	ADJ
fcis-25121	132	17	dynasty	dynasty	NOUN
fcis-25121	132	18	.	.	PUNCT
fcis-25121	133	1	the	the	DET
fcis-25121	133	2	data	data	NOUN
fcis-25121	133	3	sets	set	VERB
fcis-25121	133	4	with	with	ADP
fcis-25121	133	5	the	the	DET
fcis-25121	133	6	most	most	ADJ
fcis-25121	133	7	misclassification	misclassification	NOUN
fcis-25121	133	8	examples	example	NOUN
fcis-25121	133	9	are	be	AUX
fcis-25121	133	10	in	in	ADP
fcis-25121	133	11	the	the	DET
fcis-25121	133	12	sui	sui	PROPN
fcis-25121	133	13	and	and	CCONJ
fcis-25121	133	14	tang	tang	ADJ
fcis-25121	133	15	dynasties	dynasty	NOUN
fcis-25121	133	16	and	and	CCONJ
fcis-25121	133	17	the	the	DET
fcis-25121	133	18	song	song	NOUN
fcis-25121	133	19	and	and	CCONJ
fcis-25121	133	20	yuan	yuan	NOUN
fcis-25121	133	21	dynasties	dynasty	NOUN
fcis-25121	133	22	.	.	PUNCT
fcis-25121	134	1	as	as	SCONJ
fcis-25121	134	2	shown	show	VERB
fcis-25121	134	3	in	in	ADP
fcis-25121	134	4	fig	fig	NOUN
fcis-25121	134	5	.	.	PUNCT
fcis-25121	135	1	5	5	NUM
fcis-25121	135	2	,	,	PUNCT
fcis-25121	135	3	the	the	DET
fcis-25121	135	4	feminine	feminine	ADJ
fcis-25121	135	5	characteristics	characteristic	NOUN
fcis-25121	135	6	of	of	ADP
fcis-25121	135	7	the	the	DET
fcis-25121	135	8	characters	character	NOUN
fcis-25121	135	9	are	be	AUX
fcis-25121	135	10	quite	quite	ADV
fcis-25121	135	11	obviously	obviously	ADV
fcis-25121	135	12	.	.	PUNCT
fcis-25121	136	1	due	due	ADP
fcis-25121	136	2	to	to	ADP
fcis-25121	136	3	the	the	DET
fcis-25121	136	4	similarity	similarity	NOUN
fcis-25121	136	5	of	of	ADP
fcis-25121	136	6	colors	color	NOUN
fcis-25121	136	7	and	and	CCONJ
fcis-25121	136	8	painting	painting	NOUN
fcis-25121	136	9	styles	style	NOUN
fcis-25121	136	10	,	,	PUNCT
fcis-25121	136	11	it	it	PRON
fcis-25121	136	12	is	be	AUX
fcis-25121	136	13	difficult	difficult	ADJ
fcis-25121	136	14	to	to	PART
fcis-25121	136	15	extract	extract	VERB
fcis-25121	136	16	features	feature	NOUN
fcis-25121	136	17	in	in	ADP
fcis-25121	136	18	dynasty	dynasty	ADJ
fcis-25121	136	19	classification	classification	NOUN
fcis-25121	136	20	.	.	PUNCT
fcis-25121	137	1	as	as	ADP
fcis-25121	137	2	a	a	DET
fcis-25121	137	3	whole	whole	NOUN
fcis-25121	137	4	,	,	PUNCT
fcis-25121	137	5	vgg11	vgg11	PROPN
fcis-25121	137	6	and	and	CCONJ
fcis-25121	137	7	vgg19	vgg19	PROPN
fcis-25121	137	8	have	have	VERB
fcis-25121	137	9	the	the	DET
fcis-25121	137	10	highest	high	ADJ
fcis-25121	137	11	classification	classification	NOUN
fcis-25121	137	12	accuracy	accuracy	NOUN
fcis-25121	137	13	,	,	PUNCT
fcis-25121	137	14	followed	follow	VERB
fcis-25121	137	15	by	by	ADP
fcis-25121	137	16	vgg13	vgg13	NOUN
fcis-25121	137	17	,	,	PUNCT
fcis-25121	137	18	and	and	CCONJ
fcis-25121	137	19	vgg16	vgg16	PROPN
fcis-25121	137	20	has	have	VERB
fcis-25121	137	21	the	the	DET
fcis-25121	137	22	lowest	low	ADJ
fcis-25121	137	23	classification	classification	NOUN
fcis-25121	137	24	accuracy	accuracy	NOUN
fcis-25121	137	25	.	.	PUNCT
fcis-25121	138	1	in	in	ADP
fcis-25121	138	2	the	the	DET
fcis-25121	138	3	misclassification	misclassification	NOUN
fcis-25121	138	4	examples	example	NOUN
fcis-25121	138	5	,	,	PUNCT
fcis-25121	138	6	fig	fig	NOUN
fcis-25121	138	7	.	.	PUNCT
fcis-25121	138	8	5(a	5(a	NUM
fcis-25121	138	9	)	)	PUNCT
fcis-25121	138	10	shows	show	VERB
fcis-25121	138	11	that	that	SCONJ
fcis-25121	138	12	the	the	DET
fcis-25121	138	13	mural	mural	NOUN
fcis-25121	138	14	of	of	ADP
fcis-25121	138	15	the	the	DET
fcis-25121	138	16	sui	sui	PROPN
fcis-25121	138	17	dynasty	dynasty	NOUN
fcis-25121	138	18	is	be	AUX
fcis-25121	138	19	wrongly	wrongly	ADV
fcis-25121	138	20	divided	divide	VERB
fcis-25121	138	21	into	into	ADP
fcis-25121	138	22	the	the	DET
fcis-25121	138	23	northern	northern	ADJ
fcis-25121	138	24	dynasty	dynasty	NOUN
fcis-25121	138	25	.	.	PUNCT
fcis-25121	139	1	we	we	PRON
fcis-25121	139	2	can	can	AUX
fcis-25121	139	3	see	see	VERB
fcis-25121	139	4	the	the	DET
fcis-25121	139	5	color	color	NOUN
fcis-25121	139	6	fades	fade	VERB
fcis-25121	139	7	to	to	PART
fcis-25121	139	8	gray	gray	VERB
fcis-25121	139	9	which	which	PRON
fcis-25121	139	10	is	be	AUX
fcis-25121	139	11	similar	similar	ADJ
fcis-25121	139	12	to	to	ADP
fcis-25121	139	13	the	the	DET
fcis-25121	139	14	color	color	NOUN
fcis-25121	139	15	of	of	ADP
fcis-25121	139	16	the	the	DET
fcis-25121	139	17	mural	mural	NOUN
fcis-25121	139	18	in	in	ADP
fcis-25121	139	19	the	the	DET
fcis-25121	139	20	northern	northern	ADJ
fcis-25121	139	21	dynasty	dynasty	NOUN
fcis-25121	139	22	.	.	PUNCT
fcis-25121	140	1	in	in	ADP
fcis-25121	140	2	addition	addition	NOUN
fcis-25121	140	3	,	,	PUNCT
fcis-25121	140	4	the	the	DET
fcis-25121	140	5	character	character	NOUN
fcis-25121	140	6	's	's	PART
fcis-25121	140	7	beard	beard	NOUN
fcis-25121	140	8	is	be	AUX
fcis-25121	140	9	easy	easy	ADJ
fcis-25121	140	10	to	to	PART
fcis-25121	140	11	be	be	AUX
fcis-25121	140	12	confused	confuse	VERB
fcis-25121	140	13	with	with	ADP
fcis-25121	140	14	the	the	DET
fcis-25121	140	15	rough	rough	ADJ
fcis-25121	140	16	line	line	NOUN
fcis-25121	140	17	drawing	draw	VERB
fcis-25121	140	18	style	style	NOUN
fcis-25121	140	19	in	in	ADP
fcis-25121	140	20	the	the	DET
fcis-25121	140	21	mural	mural	NOUN
fcis-25121	140	22	of	of	ADP
fcis-25121	140	23	the	the	DET
fcis-25121	140	24	northern	northern	ADJ
fcis-25121	140	25	dynasty	dynasty	NOUN
fcis-25121	140	26	.	.	PUNCT
fcis-25121	141	1	fig	fig	NOUN
fcis-25121	141	2	.	.	PUNCT
fcis-25121	142	1	5(c	5(c	NUM
fcis-25121	142	2	)	)	PUNCT
fcis-25121	142	3	shows	show	VERB
fcis-25121	142	4	that	that	SCONJ
fcis-25121	142	5	the	the	DET
fcis-25121	142	6	mural	mural	NOUN
fcis-25121	142	7	of	of	ADP
fcis-25121	142	8	the	the	DET
fcis-25121	142	9	five	five	NUM
fcis-25121	142	10	dynasties	dynasty	NOUN
fcis-25121	142	11	and	and	CCONJ
fcis-25121	142	12	song	song	NOUN
fcis-25121	142	13	dynasty	dynasty	NOUN
fcis-25121	142	14	is	be	AUX
fcis-25121	142	15	wrongly	wrongly	ADV
fcis-25121	142	16	divided	divide	VERB
fcis-25121	142	17	into	into	ADP
fcis-25121	142	18	the	the	DET
fcis-25121	142	19	tang	tang	PROPN
fcis-25121	142	20	dynasty	dynasty	NOUN
fcis-25121	142	21	.	.	PUNCT
fcis-25121	143	1	the	the	DET
fcis-25121	143	2	facial	facial	ADJ
fcis-25121	143	3	features	feature	NOUN
fcis-25121	143	4	,	,	PUNCT
fcis-25121	143	5	the	the	DET
fcis-25121	143	6	colors	color	NOUN
fcis-25121	143	7	and	and	CCONJ
fcis-25121	143	8	the	the	DET
fcis-25121	143	9	painting	painting	NOUN
fcis-25121	143	10	styles	style	NOUN
fcis-25121	143	11	of	of	ADP
fcis-25121	143	12	the	the	DET
fcis-25121	143	13	the	the	DET
fcis-25121	143	14	mural	mural	ADJ
fcis-25121	143	15	look	look	VERB
fcis-25121	143	16	very	very	ADV
fcis-25121	143	17	similar	similar	ADJ
fcis-25121	143	18	and	and	CCONJ
fcis-25121	143	19	are	be	AUX
fcis-25121	143	20	difficult	difficult	ADJ
fcis-25121	143	21	to	to	PART
fcis-25121	143	22	distinguish	distinguish	VERB
fcis-25121	143	23	.	.	PUNCT
fcis-25121	144	1	to	to	ADP
fcis-25121	144	2	a	a	DET
fcis-25121	144	3	certain	certain	ADJ
fcis-25121	144	4	extent	extent	NOUN
fcis-25121	144	5	,	,	PUNCT
fcis-25121	144	6	the	the	DET
fcis-25121	144	7	above	above	ADJ
fcis-25121	144	8	misclassification	misclassification	NOUN
fcis-25121	144	9	examples	example	NOUN
fcis-25121	144	10	are	be	AUX
fcis-25121	144	11	due	due	ADJ
fcis-25121	144	12	to	to	ADP
fcis-25121	144	13	the	the	DET
fcis-25121	144	14	similarity	similarity	NOUN
fcis-25121	144	15	of	of	ADP
fcis-25121	144	16	colors	color	NOUN
fcis-25121	144	17	and	and	CCONJ
fcis-25121	144	18	painting	painting	NOUN
fcis-25121	144	19	styles	style	NOUN
fcis-25121	144	20	,	,	PUNCT
fcis-25121	144	21	which	which	PRON
fcis-25121	144	22	lead	lead	VERB
fcis-25121	144	23	to	to	ADP
fcis-25121	144	24	the	the	DET
fcis-25121	144	25	reduction	reduction	NOUN
fcis-25121	144	26	of	of	ADP
fcis-25121	144	27	accuracy	accuracy	NOUN
fcis-25121	144	28	in	in	ADP
fcis-25121	144	29	the	the	DET
fcis-25121	144	30	classification	classification	NOUN
fcis-25121	144	31	of	of	ADP
fcis-25121	144	32	murals	mural	NOUN
fcis-25121	144	33	in	in	ADP
fcis-25121	144	34	the	the	DET
fcis-25121	144	35	sui	sui	PROPN
fcis-25121	144	36	dynasty	dynasty	NOUN
fcis-25121	144	37	,	,	PUNCT
fcis-25121	144	38	the	the	DET
fcis-25121	144	39	five	five	NUM
fcis-25121	144	40	dynasties	dynasty	NOUN
fcis-25121	144	41	and	and	CCONJ
fcis-25121	144	42	the	the	DET
fcis-25121	144	43	song	song	NOUN
fcis-25121	144	44	dynasty	dynasty	NOUN
fcis-25121	144	45	.	.	PUNCT
fcis-25121	145	1	74	74	NUM
fcis-25121	145	2	5	5	NUM
fcis-25121	145	3	.	.	PUNCT
fcis-25121	145	4	conclusion	conclusion	NOUN
fcis-25121	145	5	in	in	ADP
fcis-25121	145	6	this	this	DET
fcis-25121	145	7	paper	paper	NOUN
fcis-25121	145	8	,	,	PUNCT
fcis-25121	145	9	we	we	PRON
fcis-25121	145	10	propose	propose	VERB
fcis-25121	145	11	a	a	DET
fcis-25121	145	12	method	method	NOUN
fcis-25121	145	13	based	base	VERB
fcis-25121	145	14	on	on	ADP
fcis-25121	145	15	convolutional	convolutional	ADJ
fcis-25121	145	16	neural	neural	ADJ
fcis-25121	145	17	network	network	NOUN
fcis-25121	145	18	to	to	PART
fcis-25121	145	19	classify	classify	VERB
fcis-25121	145	20	the	the	DET
fcis-25121	145	21	dynasties	dynasty	NOUN
fcis-25121	145	22	of	of	ADP
fcis-25121	145	23	ancient	ancient	ADJ
fcis-25121	145	24	murals	mural	NOUN
fcis-25121	145	25	,	,	PUNCT
fcis-25121	145	26	which	which	PRON
fcis-25121	145	27	is	be	AUX
fcis-25121	145	28	important	important	ADJ
fcis-25121	145	29	for	for	ADP
fcis-25121	145	30	the	the	DET
fcis-25121	145	31	research	research	NOUN
fcis-25121	145	32	of	of	ADP
fcis-25121	145	33	dunhuang	dunhuang	NOUN
fcis-25121	145	34	murals	mural	NOUN
fcis-25121	145	35	and	and	CCONJ
fcis-25121	145	36	the	the	DET
fcis-25121	145	37	digital	digital	ADJ
fcis-25121	145	38	protection	protection	NOUN
fcis-25121	145	39	of	of	ADP
fcis-25121	145	40	inheritance	inheritance	NOUN
fcis-25121	145	41	.	.	PUNCT
fcis-25121	146	1	from	from	ADP
fcis-25121	146	2	the	the	DET
fcis-25121	146	3	results	result	NOUN
fcis-25121	146	4	,	,	PUNCT
fcis-25121	146	5	we	we	PRON
fcis-25121	146	6	can	can	AUX
fcis-25121	146	7	see	see	VERB
fcis-25121	146	8	that	that	SCONJ
fcis-25121	146	9	the	the	DET
fcis-25121	146	10	classification	classification	NOUN
fcis-25121	146	11	accuracy	accuracy	NOUN
fcis-25121	146	12	of	of	ADP
fcis-25121	146	13	vgg11	vgg11	PROPN
fcis-25121	146	14	and	and	CCONJ
fcis-25121	146	15	vgg19	vgg19	PRON
fcis-25121	146	16	are	be	AUX
fcis-25121	146	17	both	both	PRON
fcis-25121	146	18	96	96	NUM
fcis-25121	146	19	%	%	NOUN
fcis-25121	146	20	,	,	PUNCT
fcis-25121	146	21	which	which	PRON
fcis-25121	146	22	are	be	AUX
fcis-25121	146	23	higher	high	ADJ
fcis-25121	146	24	than	than	ADP
fcis-25121	146	25	the	the	DET
fcis-25121	146	26	traditional	traditional	ADJ
fcis-25121	146	27	artificial	artificial	ADJ
fcis-25121	146	28	classification	classification	NOUN
fcis-25121	146	29	.	.	PUNCT
fcis-25121	147	1	the	the	DET
fcis-25121	147	2	classification	classification	NOUN
fcis-25121	147	3	accuracy	accuracy	NOUN
fcis-25121	147	4	of	of	ADP
fcis-25121	147	5	murals	mural	NOUN
fcis-25121	147	6	in	in	ADP
fcis-25121	147	7	the	the	DET
fcis-25121	147	8	sui	sui	PROPN
fcis-25121	147	9	dynasty	dynasty	NOUN
fcis-25121	147	10	and	and	CCONJ
fcis-25121	147	11	the	the	DET
fcis-25121	147	12	five	five	NUM
fcis-25121	147	13	dynasties	dynasty	NOUN
fcis-25121	147	14	and	and	CCONJ
fcis-25121	147	15	song	song	NOUN
fcis-25121	147	16	dynasties	dynasty	NOUN
fcis-25121	147	17	is	be	AUX
fcis-25121	147	18	lower	low	ADJ
fcis-25121	147	19	than	than	ADP
fcis-25121	147	20	that	that	PRON
fcis-25121	147	21	of	of	ADP
fcis-25121	147	22	other	other	ADJ
fcis-25121	147	23	dynasties	dynasty	NOUN
fcis-25121	147	24	.	.	PUNCT
fcis-25121	148	1	although	although	SCONJ
fcis-25121	148	2	the	the	DET
fcis-25121	148	3	model	model	NOUN
fcis-25121	148	4	in	in	ADP
fcis-25121	148	5	this	this	DET
fcis-25121	148	6	paper	paper	NOUN
fcis-25121	148	7	has	have	AUX
fcis-25121	148	8	been	be	AUX
fcis-25121	148	9	implemented	implement	VERB
fcis-25121	148	10	with	with	ADP
fcis-25121	148	11	a	a	DET
fcis-25121	148	12	high	high	ADJ
fcis-25121	148	13	calssification	calssification	NOUN
fcis-25121	148	14	accuracy	accuracy	NOUN
fcis-25121	148	15	.	.	PUNCT
fcis-25121	149	1	the	the	DET
fcis-25121	149	2	examples	example	NOUN
fcis-25121	149	3	of	of	ADP
fcis-25121	149	4	misclassification	misclassification	NOUN
fcis-25121	149	5	also	also	ADV
fcis-25121	149	6	show	show	VERB
fcis-25121	149	7	that	that	SCONJ
fcis-25121	149	8	the	the	DET
fcis-25121	149	9	models	model	NOUN
fcis-25121	149	10	need	need	VERB
fcis-25121	149	11	further	further	ADJ
fcis-25121	149	12	reseach	reseach	NOUN
fcis-25121	149	13	.	.	PUNCT
fcis-25121	150	1	this	this	DET
fcis-25121	150	2	method	method	NOUN
fcis-25121	150	3	can	can	AUX
fcis-25121	150	4	be	be	AUX
fcis-25121	150	5	applied	apply	VERB
fcis-25121	150	6	to	to	ADP
fcis-25121	150	7	the	the	DET
fcis-25121	150	8	calssification	calssification	NOUN
fcis-25121	150	9	of	of	ADP
fcis-25121	150	10	patterns	pattern	NOUN
fcis-25121	150	11	,	,	PUNCT
fcis-25121	150	12	clothing	clothing	NOUN
fcis-25121	150	13	and	and	CCONJ
fcis-25121	150	14	headwear	headwear	NOUN
fcis-25121	150	15	in	in	ADP
fcis-25121	150	16	ancient	ancient	ADJ
fcis-25121	150	17	murals	mural	NOUN
fcis-25121	150	18	which	which	PRON
fcis-25121	150	19	can	can	AUX
fcis-25121	150	20	save	save	VERB
fcis-25121	150	21	the	the	DET
fcis-25121	150	22	human	human	ADJ
fcis-25121	150	23	and	and	CCONJ
fcis-25121	150	24	material	material	NOUN
fcis-25121	150	25	resources	resource	NOUN
fcis-25121	150	26	spent	spend	VERB
fcis-25121	150	27	on	on	ADP
fcis-25121	150	28	manual	manual	ADJ
fcis-25121	150	29	classification	classification	NOUN
fcis-25121	150	30	.	.	PUNCT
fcis-25121	151	1	references	reference	NOUN
fcis-25121	151	2	[	[	X
fcis-25121	151	3	1	1	NUM
fcis-25121	151	4	]	]	X
fcis-25121	151	5	zhao	zhao	PROPN
fcis-25121	151	6	lijun	lijun	PROPN
fcis-25121	151	7	,	,	PUNCT
fcis-25121	151	8	tang	tang	PROPN
fcis-25121	151	9	ping	ping	PROPN
fcis-25121	151	10	,	,	PUNCT
fcis-25121	151	11	huo	huo	PROPN
fcis-25121	151	12	lianzhi	lianzhi	PROPN
fcis-25121	151	13	,	,	PUNCT
fcis-25121	151	14	et	et	PROPN
fcis-25121	151	15	al	al	PROPN
fcis-25121	151	16	.	.	PROPN
fcis-25121	152	1	review	review	NOUN
fcis-25121	152	2	of	of	ADP
fcis-25121	152	3	the	the	DET
fcis-25121	152	4	bagof	bagof	ADJ
fcis-25121	152	5	visual	visual	ADJ
fcis-25121	152	6	-	-	PUNCT
fcis-25121	152	7	word	word	NOUN
fcis-25121	152	8	models	model	NOUN
fcis-25121	152	9	in	in	ADP
fcis-25121	152	10	image	image	NOUN
fcis-25121	152	11	scene	scene	NOUN
fcis-25121	152	12	classification[j	classification[j	PROPN
fcis-25121	152	13	]	]	PUNCT
fcis-25121	152	14	.	.	PUNCT
fcis-25121	153	1	journal	journal	PROPN
fcis-25121	153	2	of	of	ADP
fcis-25121	153	3	image	image	NOUN
fcis-25121	153	4	and	and	CCONJ
fcis-25121	153	5	graphics	graphic	NOUN
fcis-25121	153	6	,	,	PUNCT
fcis-25121	153	7	2014	2014	NUM
fcis-25121	153	8	,	,	PUNCT
fcis-25121	153	9	19(3	19(3	NUM
fcis-25121	153	10	):	):	PUNCT
fcis-25121	153	11	11	11	NUM
fcis-25121	153	12	.	.	PUNCT
fcis-25121	154	1	[	[	X
fcis-25121	154	2	2	2	NUM
fcis-25121	154	3	]	]	X
fcis-25121	154	4	tang	tang	PROPN
fcis-25121	154	5	dawei	dawei	PROPN
fcis-25121	154	6	,	,	PUNCT
fcis-25121	154	7	lu	lu	PROPN
fcis-25121	154	8	dongming	dongming	NOUN
fcis-25121	154	9	,	,	PUNCT
fcis-25121	154	10	yangbing	yangbing	NOUN
fcis-25121	154	11	,	,	PUNCT
fcis-25121	154	12	et	et	PROPN
fcis-25121	155	1	al	al	PROPN
fcis-25121	155	2	.	.	PROPN
fcis-25121	155	3	similarity	similarity	NOUN
fcis-25121	155	4	metrics	metric	NOUN
fcis-25121	155	5	between	between	ADP
fcis-25121	155	6	mural	mural	ADJ
fcis-25121	155	7	images	image	NOUN
fcis-25121	155	8	with	with	ADP
fcis-25121	155	9	constraints	constraint	NOUN
fcis-25121	155	10	of	of	ADP
fcis-25121	155	11	the	the	DET
fcis-25121	155	12	overall	overall	ADJ
fcis-25121	155	13	structure	structure	NOUN
fcis-25121	155	14	of	of	ADP
fcis-25121	155	15	contours[j	contours[j	NOUN
fcis-25121	155	16	]	]	PUNCT
fcis-25121	155	17	.	.	PUNCT
fcis-25121	156	1	journal	journal	PROPN
fcis-25121	156	2	of	of	ADP
fcis-25121	156	3	image	image	NOUN
fcis-25121	156	4	and	and	CCONJ
fcis-25121	156	5	graphics	graphic	NOUN
fcis-25121	156	6	,	,	PUNCT
fcis-25121	156	7	2013	2013	NUM
fcis-25121	156	8	,	,	PUNCT
fcis-25121	156	9	18	18	NUM
fcis-25121	156	10	(	(	PUNCT
fcis-25121	156	11	8)	8)	NUM
fcis-25121	156	12	:	:	PUNCT
fcis-25121	156	13	968	968	NUM
fcis-25121	156	14	-	-	SYM
fcis-25121	156	15	975	975	NUM
fcis-25121	156	16	.	.	PUNCT
fcis-25121	156	17	2014	2014	NUM
fcis-25121	156	18	,	,	PUNCT
fcis-25121	156	19	19(5	19(5	NUM
fcis-25121	156	20	):	):	PUNCT
fcis-25121	156	21	708	708	NUM
fcis-25121	156	22	-	-	SYM
fcis-25121	156	23	715	715	NUM
fcis-25121	156	24	.	.	PUNCT
fcis-25121	157	1	[	[	X
fcis-25121	157	2	3	3	X
fcis-25121	157	3	]	]	X
fcis-25121	157	4	hao	hao	PROPN
fcis-25121	157	5	yabo	yabo	PROPN
fcis-25121	157	6	.	.	PUNCT
fcis-25121	158	1	research	research	NOUN
fcis-25121	158	2	and	and	CCONJ
fcis-25121	158	3	implementation	implementation	NOUN
fcis-25121	158	4	on	on	ADP
fcis-25121	158	5	classification	classification	NOUN
fcis-25121	158	6	algorthm	algorthm	NOUN
fcis-25121	158	7	with	with	ADP
fcis-25121	158	8	people	people	NOUN
fcis-25121	158	9	of	of	ADP
fcis-25121	158	10	acient	acient	NOUN
fcis-25121	158	11	chinese	chinese	ADJ
fcis-25121	158	12	murals	mural	NOUN
fcis-25121	158	13	based	base	VERB
fcis-25121	158	14	on	on	ADP
fcis-25121	158	15	style	style	NOUN
fcis-25121	158	16	characteristics[d	characteristics[d	PROPN
fcis-25121	158	17	]	]	PUNCT
fcis-25121	158	18	.	.	PUNCT
fcis-25121	159	1	tianjing	tianje	VERB
fcis-25121	159	2	:	:	PUNCT
fcis-25121	160	1	tianjin	tianjin	PROPN
fcis-25121	160	2	university	university	PROPN
fcis-25121	160	3	,	,	PUNCT
fcis-25121	160	4	2017	2017	NUM
fcis-25121	160	5	.	.	PUNCT
fcis-25121	161	1	[	[	X
fcis-25121	161	2	4	4	NUM
fcis-25121	161	3	]	]	PUNCT
fcis-25121	161	4	a.	a.	NOUN
fcis-25121	161	5	krizhevsky	krizhevsky	PROPN
fcis-25121	161	6	,	,	PUNCT
fcis-25121	161	7	i.	i.	PROPN
fcis-25121	161	8	sutskever	sutskever	PROPN
fcis-25121	161	9	,	,	PUNCT
fcis-25121	161	10	and	and	CCONJ
fcis-25121	161	11	g.e	g.e	PROPN
fcis-25121	161	12	.	.	PROPN
fcis-25121	161	13	hinton	hinton	PROPN
fcis-25121	161	14	.	.	PUNCT
fcis-25121	162	1	image	image	NOUN
fcis-25121	162	2	net	net	ADJ
fcis-25121	162	3	classification	classification	NOUN
fcis-25121	162	4	with	with	ADP
fcis-25121	162	5	deep	deep	ADJ
fcis-25121	162	6	convolutional	convolutional	ADJ
fcis-25121	162	7	neural	neural	ADJ
fcis-25121	162	8	networks[j	networks[j	NOUN
fcis-25121	162	9	]	]	X
fcis-25121	162	10	.	.	PUNCT
fcis-25121	163	1	advances	advance	NOUN
fcis-25121	163	2	in	in	ADP
fcis-25121	163	3	neural	neural	ADJ
fcis-25121	163	4	information	information	NOUN
fcis-25121	163	5	processing	processing	NOUN
fcis-25121	163	6	,	,	PUNCT
fcis-25121	163	7	2012	2012	NUM
fcis-25121	163	8	,	,	PUNCT
fcis-25121	163	9	25(2	25(2	NUM
fcis-25121	163	10	)	)	PUNCT
fcis-25121	163	11	,	,	PUNCT
fcis-25121	163	12	1097–1105	1097–1105	NUM
fcis-25121	163	13	.	.	PUNCT
fcis-25121	164	1	[	[	X
fcis-25121	164	2	5	5	NUM
fcis-25121	164	3	]	]	PUNCT
fcis-25121	164	4	li	li	PROPN
fcis-25121	164	5	xiangyu	xiangyu	PROPN
fcis-25121	164	6	.	.	PUNCT
fcis-25121	165	1	based	base	VERB
fcis-25121	165	2	on	on	ADP
fcis-25121	165	3	in	in	ADP
fcis-25121	165	4	-	-	PUNCT
fcis-25121	165	5	depth	depth	NOUN
fcis-25121	165	6	learning	learning	NOUN
fcis-25121	165	7	,	,	PUNCT
fcis-25121	165	8	dunhuang	dunhuang	PROPN
fcis-25121	165	9	mural	mural	PROPN
fcis-25121	165	10	era	era	NOUN
fcis-25121	165	11	identification	identification	NOUN
fcis-25121	165	12	and	and	CCONJ
fcis-25121	165	13	damage	damage	NOUN
fcis-25121	165	14	detection[d].harbin	detection[d].harbin	NOUN
fcis-25121	165	15	:	:	PUNCT
fcis-25121	165	16	harbin	harbin	PROPN
fcis-25121	165	17	institute	institute	PROPN
fcis-25121	165	18	of	of	ADP
fcis-25121	165	19	techonology	techonology	NOUN
fcis-25121	165	20	,	,	PUNCT
fcis-25121	165	21	2019	2019	NUM
fcis-25121	165	22	.	.	PUNCT
fcis-25121	166	1	[	[	X
fcis-25121	166	2	6	6	NUM
fcis-25121	166	3	]	]	PUNCT
fcis-25121	166	4	cui	cui	PROPN
fcis-25121	166	5	hongyan	hongyan	PROPN
fcis-25121	166	6	.	.	PUNCT
fcis-25121	167	1	research	research	NOUN
fcis-25121	167	2	on	on	ADP
fcis-25121	167	3	acient	acient	NOUN
fcis-25121	167	4	mural	mural	ADJ
fcis-25121	167	5	classification	classification	NOUN
fcis-25121	167	6	method	method	NOUN
fcis-25121	167	7	based	base	VERB
fcis-25121	167	8	on	on	ADP
fcis-25121	167	9	deep	deep	ADJ
fcis-25121	167	10	learning[d	learning[d	PROPN
fcis-25121	167	11	]	]	PUNCT
fcis-25121	167	12	.	.	PUNCT
fcis-25121	168	1	taiyuan	taiyuan	PROPN
fcis-25121	168	2	:	:	PUNCT
fcis-25121	168	3	taiyuan	taiyuan	PROPN
fcis-25121	168	4	university	university	PROPN
fcis-25121	168	5	of	of	ADP
fcis-25121	168	6	science	science	NOUN
fcis-25121	168	7	and	and	CCONJ
fcis-25121	168	8	technology	technology	NOUN
fcis-25121	168	9	,	,	PUNCT
fcis-25121	168	10	2020	2020	NUM
fcis-25121	168	11	.	.	PUNCT
fcis-25121	169	1	[	[	X
fcis-25121	169	2	7	7	X
fcis-25121	169	3	]	]	X
fcis-25121	169	4	cao	cao	PROPN
fcis-25121	169	5	jianfang	jianfang	PROPN
fcis-25121	169	6	,	,	PUNCT
fcis-25121	169	7	yan	yan	PROPN
fcis-25121	169	8	minmin	minmin	PROPN
fcis-25121	169	9	,	,	PUNCT
fcis-25121	169	10	jia	jia	PROPN
fcis-25121	169	11	yinming	yinming	PROPN
fcis-25121	169	12	,	,	PUNCT
fcis-25121	169	13	tian	tian	PROPN
fcis-25121	169	14	xiaodong	xiaodong	PROPN
fcis-25121	169	15	.	.	PUNCT
fcis-25121	170	1	application	application	NOUN
fcis-25121	170	2	of	of	ADP
fcis-25121	170	3	inception	inception	PROPN
fcis-25121	170	4	-	-	PUNCT
fcis-25121	170	5	v3	v3	NOUN
fcis-25121	170	6	model	model	NOUN
fcis-25121	170	7	integrated	integrate	VERB
fcis-25121	170	8	with	with	ADP
fcis-25121	170	9	transfer	transfer	NOUN
fcis-25121	170	10	learning	learning	NOUN
fcis-25121	170	11	in	in	ADP
fcis-25121	170	12	dynasty	dynasty	ADJ
fcis-25121	170	13	identification	identification	NOUN
fcis-25121	170	14	of	of	ADP
fcis-25121	170	15	ancient	ancient	ADJ
fcis-25121	170	16	murals[j	murals[j	PROPN
fcis-25121	170	17	]	]	PUNCT
fcis-25121	170	18	.	.	PUNCT
fcis-25121	171	1	journal	journal	PROPN
fcis-25121	171	2	of	of	ADP
fcis-25121	171	3	computer	computer	NOUN
fcis-25121	171	4	applications	application	NOUN
fcis-25121	171	5	,	,	PUNCT
fcis-25121	171	6	2021	2021	NUM
fcis-25121	171	7	,	,	PUNCT
fcis-25121	171	8	41(11	41(11	NUM
fcis-25121	171	9	):	):	PUNCT
fcis-25121	171	10	3219	3219	NUM
fcis-25121	171	11	-	-	SYM
fcis-25121	171	12	3227	3227	NUM
fcis-25121	171	13	.	.	PUNCT
fcis-25121	172	1	[	[	X
fcis-25121	172	2	8	8	NUM
fcis-25121	172	3	]	]	X
fcis-25121	172	4	cheng	cheng	PROPN
fcis-25121	172	5	bing	bing	PROPN
fcis-25121	172	6	.	.	PUNCT
fcis-25121	173	1	research	research	NOUN
fcis-25121	173	2	on	on	ADP
fcis-25121	173	3	automatic	automatic	ADJ
fcis-25121	173	4	image	image	NOUN
fcis-25121	173	5	annotation	annotation	NOUN
fcis-25121	173	6	technology	technology	NOUN
fcis-25121	173	7	based	base	VERB
fcis-25121	173	8	on	on	ADP
fcis-25121	173	9	convolutional	convolutional	ADJ
fcis-25121	173	10	neural	neural	ADJ
fcis-25121	173	11	network[j	network[j	NOUN
fcis-25121	173	12	]	]	PUNCT
fcis-25121	173	13	.	.	PUNCT
fcis-25121	174	1	electronics	electronic	NOUN
fcis-25121	174	2	world	world	NOUN
fcis-25121	174	3	,	,	PUNCT
fcis-25121	174	4	2019(16	2019(16	NUM
fcis-25121	174	5	):	):	PUNCT
fcis-25121	174	6	124	124	NUM
fcis-25121	174	7	-	-	SYM
fcis-25121	174	8	126	126	NUM
fcis-25121	174	9	.	.	PUNCT
fcis-25121	175	1	[	[	X
fcis-25121	175	2	9	9	NUM
fcis-25121	175	3	]	]	X
fcis-25121	175	4	zhang	zhang	PROPN
fcis-25121	175	5	lulu	lulu	PROPN
fcis-25121	175	6	.	.	PUNCT
fcis-25121	176	1	research	research	NOUN
fcis-25121	176	2	on	on	ADP
fcis-25121	176	3	facial	facial	ADJ
fcis-25121	176	4	expression	expression	NOUN
fcis-25121	176	5	recognition	recognition	NOUN
fcis-25121	176	6	method	method	NOUN
fcis-25121	176	7	based	base	VERB
fcis-25121	176	8	on	on	ADP
fcis-25121	176	9	convolutional	convolutional	ADJ
fcis-25121	176	10	neural	neural	ADJ
fcis-25121	176	11	network[d	network[d	NOUN
fcis-25121	176	12	]	]	PUNCT
fcis-25121	176	13	.	.	PUNCT
fcis-25121	177	1	shijiazhuang	shijiazhuang	PROPN
fcis-25121	177	2	:	:	PUNCT
fcis-25121	177	3	hebei	hebei	PROPN
fcis-25121	177	4	university	university	PROPN
fcis-25121	177	5	of	of	ADP
fcis-25121	177	6	science	science	PROPN
fcis-25121	177	7	&	&	CCONJ
fcis-25121	177	8	technology	technology	PROPN
fcis-25121	177	9	,	,	PUNCT
fcis-25121	177	10	2019	2019	NUM
fcis-25121	177	11	.	.	PUNCT
fcis-25121	178	1	[	[	X
fcis-25121	178	2	10	10	NUM
fcis-25121	178	3	]	]	SYM
fcis-25121	178	4	simonyan	simonyan	PROPN
fcis-25121	178	5	k	k	NOUN
fcis-25121	178	6	,	,	PUNCT
fcis-25121	178	7	zisserman	zisserman	NOUN
fcis-25121	178	8	a.	a.	NOUN
fcis-25121	178	9	very	very	ADV
fcis-25121	178	10	deep	deep	ADJ
fcis-25121	178	11	convolutional	convolutional	ADJ
fcis-25121	178	12	networks	network	NOUN
fcis-25121	178	13	for	for	ADP
fcis-25121	178	14	large	large	ADJ
fcis-25121	178	15	-	-	PUNCT
fcis-25121	178	16	scale	scale	NOUN
fcis-25121	178	17	image	image	NOUN
fcis-25121	178	18	recognition[j	recognition[j	NOUN
fcis-25121	178	19	]	]	PUNCT
fcis-25121	178	20	.	.	PUNCT
fcis-25121	179	1	computer	computer	NOUN
fcis-25121	179	2	science	science	NOUN
fcis-25121	179	3	,	,	PUNCT
fcis-25121	179	4	2014	2014	NUM
fcis-25121	179	5	.	.	PUNCT