id	sid	tid	token	lemma	pos
ajst-32001	1	1	95	95	NUM
ajst-32001	1	2	review	review	NOUN
ajst-32001	1	3	of	of	ADP
ajst-32001	1	4	emotion	emotion	NOUN
ajst-32001	1	5	recognition	recognition	NOUN
ajst-32001	1	6	technology	technology	NOUN
ajst-32001	1	7	methods	method	NOUN
ajst-32001	1	8	based	base	VERB
ajst-32001	1	9	on	on	ADP
ajst-32001	1	10	deep	deep	ADJ
ajst-32001	1	11	recognition	recognition	NOUN
ajst-32001	1	12	zhizhou	zhizhou	PROPN
ajst-32001	1	13	lu	lu	PROPN
ajst-32001	1	14	*	*	PUNCT
ajst-32001	1	15	nanjing	nanjing	PROPN
ajst-32001	1	16	university	university	PROPN
ajst-32001	1	17	of	of	ADP
ajst-32001	1	18	posts	post	NOUN
ajst-32001	1	19	and	and	CCONJ
ajst-32001	1	20	telecommunications	telecommunication	NOUN
ajst-32001	1	21	,	,	PUNCT
ajst-32001	1	22	nanjing	nanjing	PROPN
ajst-32001	1	23	,	,	PUNCT
ajst-32001	1	24	210023	210023	NUM
ajst-32001	1	25	,	,	PUNCT
ajst-32001	1	26	jiangsu	jiangsu	PROPN
ajst-32001	1	27	,	,	PUNCT
ajst-32001	1	28	china	china	PROPN
ajst-32001	1	29	*	*	PUNCT
ajst-32001	1	30	corresponding	correspond	VERB
ajst-32001	1	31	author	author	NOUN
ajst-32001	1	32	email	email	NOUN
ajst-32001	1	33	:	:	PUNCT
ajst-32001	1	34	2897694110@qq.com	2897694110@qq.com	NUM
ajst-32001	1	35	abstract	abstract	NOUN
ajst-32001	1	36	.	.	PUNCT
ajst-32001	2	1	with	with	ADP
ajst-32001	2	2	the	the	DET
ajst-32001	2	3	rapid	rapid	ADJ
ajst-32001	2	4	advancement	advancement	NOUN
ajst-32001	2	5	of	of	ADP
ajst-32001	2	6	artificial	artificial	ADJ
ajst-32001	2	7	intelligence	intelligence	NOUN
ajst-32001	2	8	,	,	PUNCT
ajst-32001	2	9	sensor	sensor	NOUN
ajst-32001	2	10	technology	technology	NOUN
ajst-32001	2	11	,	,	PUNCT
ajst-32001	2	12	and	and	CCONJ
ajst-32001	2	13	big	big	ADJ
ajst-32001	2	14	data	datum	NOUN
ajst-32001	2	15	analytics	analytic	NOUN
ajst-32001	2	16	,	,	PUNCT
ajst-32001	2	17	the	the	DET
ajst-32001	2	18	way	way	NOUN
ajst-32001	2	19	humans	human	NOUN
ajst-32001	2	20	interact	interact	VERB
ajst-32001	2	21	with	with	ADP
ajst-32001	2	22	machines	machine	NOUN
ajst-32001	2	23	is	be	AUX
ajst-32001	2	24	shifting	shift	VERB
ajst-32001	2	25	from	from	ADP
ajst-32001	2	26	an	an	DET
ajst-32001	2	27	"	"	PUNCT
ajst-32001	2	28	instruction	instruction	NOUN
ajst-32001	2	29	-	-	PUNCT
ajst-32001	2	30	based	base	VERB
ajst-32001	2	31	"	"	PUNCT
ajst-32001	2	32	mode	mode	NOUN
ajst-32001	2	33	to	to	ADP
ajst-32001	2	34	a	a	DET
ajst-32001	2	35	"	"	PUNCT
ajst-32001	2	36	perception	perception	NOUN
ajst-32001	2	37	-	-	PUNCT
ajst-32001	2	38	based	base	VERB
ajst-32001	2	39	"	"	PUNCT
ajst-32001	2	40	one	one	NUM
ajst-32001	2	41	.	.	PUNCT
ajst-32001	3	1	as	as	SCONJ
ajst-32001	3	2	emotions	emotion	NOUN
ajst-32001	3	3	are	be	AUX
ajst-32001	3	4	the	the	DET
ajst-32001	3	5	core	core	NOUN
ajst-32001	3	6	driving	drive	VERB
ajst-32001	3	7	force	force	NOUN
ajst-32001	3	8	behind	behind	ADP
ajst-32001	3	9	human	human	ADJ
ajst-32001	3	10	behaviors	behavior	NOUN
ajst-32001	3	11	and	and	CCONJ
ajst-32001	3	12	decisions	decision	NOUN
ajst-32001	3	13	,	,	PUNCT
ajst-32001	3	14	the	the	DET
ajst-32001	3	15	objective	objective	ADJ
ajst-32001	3	16	and	and	CCONJ
ajst-32001	3	17	quantitative	quantitative	ADJ
ajst-32001	3	18	detection	detection	NOUN
ajst-32001	3	19	of	of	ADP
ajst-32001	3	20	emotions	emotion	NOUN
ajst-32001	3	21	has	have	AUX
ajst-32001	3	22	become	become	VERB
ajst-32001	3	23	a	a	DET
ajst-32001	3	24	focus	focus	NOUN
ajst-32001	3	25	of	of	ADP
ajst-32001	3	26	interdisciplinary	interdisciplinary	ADJ
ajst-32001	3	27	research	research	NOUN
ajst-32001	3	28	.	.	PUNCT
ajst-32001	4	1	studies	study	NOUN
ajst-32001	4	2	have	have	AUX
ajst-32001	4	3	shown	show	VERB
ajst-32001	4	4	that	that	SCONJ
ajst-32001	4	5	emotions	emotion	NOUN
ajst-32001	4	6	can	can	AUX
ajst-32001	4	7	be	be	AUX
ajst-32001	4	8	detected	detect	VERB
ajst-32001	4	9	and	and	CCONJ
ajst-32001	4	10	analyzed	analyze	VERB
ajst-32001	4	11	through	through	ADP
ajst-32001	4	12	four	four	NUM
ajst-32001	4	13	methods	method	NOUN
ajst-32001	4	14	:	:	PUNCT
ajst-32001	4	15	facial	facial	ADJ
ajst-32001	4	16	expression	expression	NOUN
ajst-32001	4	17	recognition	recognition	NOUN
ajst-32001	4	18	,	,	PUNCT
ajst-32001	4	19	speech	speech	NOUN
ajst-32001	4	20	emotion	emotion	NOUN
ajst-32001	4	21	recognition	recognition	NOUN
ajst-32001	4	22	,	,	PUNCT
ajst-32001	4	23	text	text	NOUN
ajst-32001	4	24	sentiment	sentiment	NOUN
ajst-32001	4	25	analysis	analysis	NOUN
ajst-32001	4	26	,	,	PUNCT
ajst-32001	4	27	and	and	CCONJ
ajst-32001	4	28	physiological	physiological	ADJ
ajst-32001	4	29	signal	signal	NOUN
ajst-32001	4	30	recognition	recognition	NOUN
ajst-32001	4	31	.	.	PUNCT
ajst-32001	5	1	reducing	reduce	VERB
ajst-32001	5	2	errors	error	NOUN
ajst-32001	5	3	in	in	ADP
ajst-32001	5	4	emotion	emotion	NOUN
ajst-32001	5	5	recognition	recognition	NOUN
ajst-32001	5	6	is	be	AUX
ajst-32001	5	7	of	of	ADP
ajst-32001	5	8	great	great	ADJ
ajst-32001	5	9	help	help	NOUN
ajst-32001	5	10	to	to	ADP
ajst-32001	5	11	the	the	DET
ajst-32001	5	12	development	development	NOUN
ajst-32001	5	13	and	and	CCONJ
ajst-32001	5	14	widespread	widespread	ADJ
ajst-32001	5	15	application	application	NOUN
ajst-32001	5	16	of	of	ADP
ajst-32001	5	17	human	human	ADJ
ajst-32001	5	18	-	-	PUNCT
ajst-32001	5	19	computer	computer	NOUN
ajst-32001	5	20	interaction	interaction	NOUN
ajst-32001	5	21	.	.	PUNCT
ajst-32001	6	1	the	the	DET
ajst-32001	6	2	development	development	NOUN
ajst-32001	6	3	and	and	CCONJ
ajst-32001	6	4	research	research	NOUN
ajst-32001	6	5	on	on	ADP
ajst-32001	6	6	emotion	emotion	NOUN
ajst-32001	6	7	detection	detection	NOUN
ajst-32001	6	8	is	be	AUX
ajst-32001	6	9	not	not	PART
ajst-32001	6	10	only	only	ADV
ajst-32001	6	11	an	an	DET
ajst-32001	6	12	inevitable	inevitable	ADJ
ajst-32001	6	13	outcome	outcome	NOUN
ajst-32001	6	14	of	of	ADP
ajst-32001	6	15	technological	technological	ADJ
ajst-32001	6	16	development	development	NOUN
ajst-32001	6	17	but	but	CCONJ
ajst-32001	6	18	also	also	ADV
ajst-32001	6	19	a	a	DET
ajst-32001	6	20	key	key	NOUN
ajst-32001	6	21	to	to	ADP
ajst-32001	6	22	addressing	address	VERB
ajst-32001	6	23	practical	practical	ADJ
ajst-32001	6	24	needs	need	NOUN
ajst-32001	6	25	.	.	PUNCT
ajst-32001	7	1	this	this	DET
ajst-32001	7	2	paper	paper	NOUN
ajst-32001	7	3	will	will	AUX
ajst-32001	7	4	analyze	analyze	VERB
ajst-32001	7	5	deep	deep	ADJ
ajst-32001	7	6	learning	learning	NOUN
ajst-32001	7	7	models	model	NOUN
ajst-32001	7	8	such	such	ADJ
ajst-32001	7	9	as	as	ADP
ajst-32001	7	10	cnn	cnn	PROPN
ajst-32001	7	11	,	,	PUNCT
ajst-32001	7	12	lstm	lstm	PROPN
ajst-32001	7	13	,	,	PUNCT
ajst-32001	7	14	and	and	CCONJ
ajst-32001	7	15	senn	senn	PROPN
ajst-32001	7	16	,	,	PUNCT
ajst-32001	7	17	and	and	CCONJ
ajst-32001	7	18	summarize	summarize	VERB
ajst-32001	7	19	their	their	PRON
ajst-32001	7	20	advantages	advantage	NOUN
ajst-32001	7	21	and	and	CCONJ
ajst-32001	7	22	disadvantages	disadvantage	NOUN
ajst-32001	7	23	.	.	PUNCT
ajst-32001	8	1	it	it	PRON
ajst-32001	8	2	breaks	break	VERB
ajst-32001	8	3	the	the	DET
ajst-32001	8	4	traditional	traditional	ADJ
ajst-32001	8	5	perception	perception	NOUN
ajst-32001	8	6	that	that	SCONJ
ajst-32001	8	7	"	"	PUNCT
ajst-32001	8	8	emotions	emotion	NOUN
ajst-32001	8	9	can	can	AUX
ajst-32001	8	10	not	not	PART
ajst-32001	8	11	be	be	AUX
ajst-32001	8	12	quantified	quantify	VERB
ajst-32001	8	13	,	,	PUNCT
ajst-32001	8	14	"	"	PUNCT
ajst-32001	8	15	enabling	enable	VERB
ajst-32001	8	16	machines	machine	NOUN
ajst-32001	8	17	to	to	PART
ajst-32001	8	18	move	move	VERB
ajst-32001	8	19	from	from	ADP
ajst-32001	8	20	"	"	PUNCT
ajst-32001	8	21	understanding	understand	VERB
ajst-32001	8	22	language	language	NOUN
ajst-32001	8	23	"	"	PUNCT
ajst-32001	8	24	to	to	ADP
ajst-32001	8	25	"	"	PUNCT
ajst-32001	8	26	understanding	understand	VERB
ajst-32001	8	27	the	the	DET
ajst-32001	8	28	human	human	ADJ
ajst-32001	8	29	heart	heart	NOUN
ajst-32001	8	30	,	,	PUNCT
ajst-32001	8	31	"	"	PUNCT
ajst-32001	8	32	and	and	CCONJ
ajst-32001	8	33	ultimately	ultimately	ADV
ajst-32001	8	34	promoting	promote	VERB
ajst-32001	8	35	the	the	DET
ajst-32001	8	36	harmonious	harmonious	ADJ
ajst-32001	8	37	coexistence	coexistence	NOUN
ajst-32001	8	38	of	of	ADP
ajst-32001	8	39	technology	technology	NOUN
ajst-32001	8	40	and	and	CCONJ
ajst-32001	8	41	human	human	ADJ
ajst-32001	8	42	society	society	NOUN
ajst-32001	8	43	.	.	PUNCT
ajst-32001	9	1	keywords	keyword	NOUN
ajst-32001	9	2	:	:	PUNCT
ajst-32001	9	3	emotion	emotion	NOUN
ajst-32001	9	4	recognition	recognition	NOUN
ajst-32001	9	5	;	;	PUNCT
ajst-32001	9	6	reducing	reduce	VERB
ajst-32001	9	7	recognition	recognition	NOUN
ajst-32001	9	8	errors	error	NOUN
ajst-32001	9	9	;	;	PUNCT
ajst-32001	9	10	human	human	ADJ
ajst-32001	9	11	-	-	PUNCT
ajst-32001	9	12	computer	computer	NOUN
ajst-32001	9	13	interaction	interaction	NOUN
ajst-32001	9	14	;	;	PUNCT
ajst-32001	9	15	deep	deep	ADJ
ajst-32001	9	16	learning	learn	VERB
ajst-32001	9	17	technology	technology	NOUN
ajst-32001	9	18	.	.	PUNCT
ajst-32001	10	1	1	1	X
ajst-32001	10	2	.	.	X
ajst-32001	10	3	introduction	introduction	NOUN
ajst-32001	10	4	the	the	DET
ajst-32001	10	5	emergence	emergence	NOUN
ajst-32001	10	6	and	and	CCONJ
ajst-32001	10	7	development	development	NOUN
ajst-32001	10	8	of	of	ADP
ajst-32001	10	9	emotion	emotion	NOUN
ajst-32001	10	10	recognition	recognition	NOUN
ajst-32001	10	11	technology	technology	NOUN
ajst-32001	10	12	are	be	AUX
ajst-32001	10	13	the	the	DET
ajst-32001	10	14	result	result	NOUN
ajst-32001	10	15	of	of	ADP
ajst-32001	10	16	various	various	ADJ
ajst-32001	10	17	factors	factor	NOUN
ajst-32001	10	18	working	work	VERB
ajst-32001	10	19	together	together	ADV
ajst-32001	10	20	,	,	PUNCT
ajst-32001	10	21	including	include	VERB
ajst-32001	10	22	advancements	advancement	NOUN
ajst-32001	10	23	in	in	ADP
ajst-32001	10	24	artificial	artificial	ADJ
ajst-32001	10	25	intelligence	intelligence	NOUN
ajst-32001	10	26	technology	technology	NOUN
ajst-32001	10	27	,	,	PUNCT
ajst-32001	10	28	increasing	increase	VERB
ajst-32001	10	29	demand	demand	NOUN
ajst-32001	10	30	for	for	ADP
ajst-32001	10	31	human	human	ADJ
ajst-32001	10	32	-	-	PUNCT
ajst-32001	10	33	computer	computer	NOUN
ajst-32001	10	34	interaction	interaction	NOUN
ajst-32001	10	35	,	,	PUNCT
ajst-32001	10	36	and	and	CCONJ
ajst-32001	10	37	the	the	DET
ajst-32001	10	38	integration	integration	NOUN
ajst-32001	10	39	of	of	ADP
ajst-32001	10	40	multiple	multiple	ADJ
ajst-32001	10	41	disciplines	discipline	NOUN
ajst-32001	10	42	.	.	PUNCT
ajst-32001	11	1	the	the	DET
ajst-32001	11	2	reason	reason	NOUN
ajst-32001	11	3	of	of	ADP
ajst-32001	11	4	why	why	SCONJ
ajst-32001	11	5	we	we	PRON
ajst-32001	11	6	need	need	VERB
ajst-32001	11	7	to	to	PART
ajst-32001	11	8	promote	promote	VERB
ajst-32001	11	9	the	the	DET
ajst-32001	11	10	development	development	NOUN
ajst-32001	11	11	of	of	ADP
ajst-32001	11	12	emotion	emotion	NOUN
ajst-32001	11	13	recognition	recognition	NOUN
ajst-32001	11	14	technology	technology	NOUN
ajst-32001	11	15	is	be	AUX
ajst-32001	11	16	multifaceted	multifaceted	ADJ
ajst-32001	11	17	.	.	PUNCT
ajst-32001	12	1	the	the	DET
ajst-32001	12	2	first	first	ADJ
ajst-32001	12	3	reason	reason	NOUN
ajst-32001	12	4	is	be	AUX
ajst-32001	12	5	the	the	DET
ajst-32001	12	6	growing	grow	VERB
ajst-32001	12	7	demand	demand	NOUN
ajst-32001	12	8	for	for	ADP
ajst-32001	12	9	human	human	ADJ
ajst-32001	12	10	-	-	PUNCT
ajst-32001	12	11	computer	computer	NOUN
ajst-32001	12	12	interaction	interaction	NOUN
ajst-32001	12	13	.	.	PUNCT
ajst-32001	13	1	emotion	emotion	NOUN
ajst-32001	13	2	recognition	recognition	NOUN
ajst-32001	13	3	is	be	AUX
ajst-32001	13	4	ready	ready	ADJ
ajst-32001	13	5	to	to	PART
ajst-32001	13	6	make	make	VERB
ajst-32001	13	7	significant	significant	ADJ
ajst-32001	13	8	contributions	contribution	NOUN
ajst-32001	13	9	to	to	ADP
ajst-32001	13	10	artificial	artificial	ADJ
ajst-32001	13	11	intelligence	intelligence	NOUN
ajst-32001	13	12	and	and	CCONJ
ajst-32001	13	13	human	human	NOUN
ajst-32001	13	14	-	-	PUNCT
ajst-32001	13	15	computer	computer	NOUN
ajst-32001	13	16	interaction.[1	interaction.[1	NOUN
ajst-32001	13	17	]	]	X
ajst-32001	13	18	contemporary	contemporary	ADJ
ajst-32001	13	19	users	user	NOUN
ajst-32001	13	20	increasingly	increasingly	ADV
ajst-32001	13	21	expect	expect	VERB
ajst-32001	13	22	emotionally	emotionally	ADV
ajst-32001	13	23	intelligent	intelligent	ADJ
ajst-32001	13	24	interactions	interaction	NOUN
ajst-32001	13	25	beyond	beyond	ADP
ajst-32001	13	26	functional	functional	ADJ
ajst-32001	13	27	exchanges	exchange	NOUN
ajst-32001	13	28	.	.	PUNCT
ajst-32001	14	1	[	[	X
ajst-32001	14	2	1,2	1,2	NUM
ajst-32001	14	3	]	]	PUNCT
ajst-32001	14	4	emotion	emotion	NOUN
ajst-32001	14	5	recognition	recognition	NOUN
ajst-32001	14	6	enables	enable	VERB
ajst-32001	14	7	machines	machine	NOUN
ajst-32001	14	8	to	to	PART
ajst-32001	14	9	identify	identify	VERB
ajst-32001	14	10	human	human	ADJ
ajst-32001	14	11	emotional	emotional	ADJ
ajst-32001	14	12	states[3	states[3	PROPN
ajst-32001	14	13	]	]	X
ajst-32001	14	14	,	,	PUNCT
ajst-32001	14	15	thereby	thereby	ADV
ajst-32001	14	16	achieving	achieve	VERB
ajst-32001	14	17	more	more	ADV
ajst-32001	14	18	natural	natural	ADJ
ajst-32001	14	19	and	and	CCONJ
ajst-32001	14	20	intelligent	intelligent	ADJ
ajst-32001	14	21	human	human	ADJ
ajst-32001	14	22	-	-	PUNCT
ajst-32001	14	23	computer	computer	NOUN
ajst-32001	14	24	interaction	interaction	NOUN
ajst-32001	14	25	.	.	PUNCT
ajst-32001	15	1	for	for	ADP
ajst-32001	15	2	example	example	NOUN
ajst-32001	15	3	,	,	PUNCT
ajst-32001	15	4	if	if	SCONJ
ajst-32001	15	5	a	a	DET
ajst-32001	15	6	smart	smart	ADJ
ajst-32001	15	7	speaker	speaker	NOUN
ajst-32001	15	8	detects	detect	VERB
ajst-32001	15	9	anxiety	anxiety	NOUN
ajst-32001	15	10	in	in	ADP
ajst-32001	15	11	the	the	DET
ajst-32001	15	12	user	user	NOUN
ajst-32001	15	13	's	's	PART
ajst-32001	15	14	tone	tone	NOUN
ajst-32001	15	15	,	,	PUNCT
ajst-32001	15	16	it	it	PRON
ajst-32001	15	17	will	will	AUX
ajst-32001	15	18	proactively	proactively	ADV
ajst-32001	15	19	offer	offer	VERB
ajst-32001	15	20	soothing	soothe	VERB
ajst-32001	15	21	music	music	NOUN
ajst-32001	15	22	or	or	CCONJ
ajst-32001	15	23	suggestions	suggestion	NOUN
ajst-32001	15	24	.	.	PUNCT
ajst-32001	16	1	in	in	ADP
ajst-32001	16	2	other	other	ADJ
ajst-32001	16	3	words	word	NOUN
ajst-32001	16	4	,	,	PUNCT
ajst-32001	16	5	the	the	DET
ajst-32001	16	6	promoted	promote	VERB
ajst-32001	16	7	machines	machine	NOUN
ajst-32001	16	8	can	can	AUX
ajst-32001	16	9	perceive	perceive	VERB
ajst-32001	16	10	human	human	ADJ
ajst-32001	16	11	joys	joy	NOUN
ajst-32001	16	12	,	,	PUNCT
ajst-32001	16	13	sorrows	sorrow	NOUN
ajst-32001	16	14	,	,	PUNCT
ajst-32001	16	15	anger	anger	NOUN
ajst-32001	16	16	,	,	PUNCT
ajst-32001	16	17	and	and	CCONJ
ajst-32001	16	18	happiness	happiness	NOUN
ajst-32001	16	19	,	,	PUNCT
ajst-32001	16	20	and	and	CCONJ
ajst-32001	16	21	then	then	ADV
ajst-32001	16	22	adjust	adjust	VERB
ajst-32001	16	23	interaction	interaction	NOUN
ajst-32001	16	24	strategies	strategy	NOUN
ajst-32001	16	25	.	.	PUNCT
ajst-32001	17	1	besides	besides	SCONJ
ajst-32001	17	2	,	,	PUNCT
ajst-32001	17	3	we	we	PRON
ajst-32001	17	4	can	can	AUX
ajst-32001	17	5	obviously	obviously	ADV
ajst-32001	17	6	find	find	VERB
ajst-32001	17	7	out	out	ADP
ajst-32001	17	8	that	that	SCONJ
ajst-32001	17	9	as	as	SCONJ
ajst-32001	17	10	emotion	emotion	NOUN
ajst-32001	17	11	recognition	recognition	NOUN
ajst-32001	17	12	technology	technology	NOUN
ajst-32001	17	13	developing	develop	VERB
ajst-32001	17	14	,	,	PUNCT
ajst-32001	17	15	a	a	DET
ajst-32001	17	16	new	new	ADJ
ajst-32001	17	17	dimension	dimension	NOUN
ajst-32001	17	18	of	of	ADP
ajst-32001	17	19	insight	insight	NOUN
ajst-32001	17	20	is	be	AUX
ajst-32001	17	21	provided	provide	VERB
ajst-32001	17	22	for	for	ADP
ajst-32001	17	23	various	various	ADJ
ajst-32001	17	24	industries	industry	NOUN
ajst-32001	17	25	,	,	PUNCT
ajst-32001	17	26	promote	promote	VERB
ajst-32001	17	27	the	the	DET
ajst-32001	17	28	transformation	transformation	NOUN
ajst-32001	17	29	of	of	ADP
ajst-32001	17	30	service	service	NOUN
ajst-32001	17	31	models	model	NOUN
ajst-32001	17	32	from	from	ADP
ajst-32001	17	33	"	"	PUNCT
ajst-32001	17	34	generalization	generalization	NOUN
ajst-32001	17	35	"	"	PUNCT
ajst-32001	17	36	to	to	ADP
ajst-32001	17	37	"	"	PUNCT
ajst-32001	17	38	personalization	personalization	NOUN
ajst-32001	17	39	"	"	PUNCT
ajst-32001	17	40	and	and	CCONJ
ajst-32001	17	41	"	"	PUNCT
ajst-32001	17	42	precision	precision	NOUN
ajst-32001	17	43	"	"	PUNCT
ajst-32001	17	44	.	.	PUNCT
ajst-32001	18	1	emotion	emotion	NOUN
ajst-32001	18	2	recognition	recognition	NOUN
ajst-32001	18	3	technology	technology	NOUN
ajst-32001	18	4	has	have	VERB
ajst-32001	18	5	many	many	ADJ
ajst-32001	18	6	applications	application	NOUN
ajst-32001	18	7	in	in	ADP
ajst-32001	18	8	daily	daily	ADJ
ajst-32001	18	9	life	life	NOUN
ajst-32001	18	10	.	.	PUNCT
ajst-32001	19	1	for	for	ADP
ajst-32001	19	2	instance	instance	NOUN
ajst-32001	19	3	,	,	PUNCT
ajst-32001	19	4	an	an	DET
ajst-32001	19	5	emotion	emotion	NOUN
ajst-32001	19	6	recognition	recognition	NOUN
ajst-32001	19	7	system	system	NOUN
ajst-32001	19	8	can	can	AUX
ajst-32001	19	9	be	be	AUX
ajst-32001	19	10	used	use	VERB
ajst-32001	19	11	into	into	ADP
ajst-32001	19	12	a	a	DET
ajst-32001	19	13	car	car	NOUN
ajst-32001	19	14	's	's	PART
ajst-32001	19	15	onboard	onboard	NOUN
ajst-32001	19	16	driving	driving	NOUN
ajst-32001	19	17	system	system	NOUN
ajst-32001	19	18	.	.	PUNCT
ajst-32001	20	1	this	this	PRON
ajst-32001	20	2	can	can	AUX
ajst-32001	20	3	help	help	VERB
ajst-32001	20	4	prevent	prevent	VERB
ajst-32001	20	5	accidents	accident	NOUN
ajst-32001	20	6	caused	cause	VERB
ajst-32001	20	7	by	by	ADP
ajst-32001	20	8	a	a	DET
ajst-32001	20	9	driver	driver	NOUN
ajst-32001	20	10	's	's	PART
ajst-32001	20	11	stressed	stress	VERB
ajst-32001	20	12	mental	mental	ADJ
ajst-32001	20	13	state.[4	state.[4	NOUN
ajst-32001	20	14	]	]	X
ajst-32001	20	15	this	this	DET
ajst-32001	20	16	paper	paper	NOUN
ajst-32001	20	17	aims	aim	VERB
ajst-32001	20	18	to	to	PART
ajst-32001	20	19	gather	gather	VERB
ajst-32001	20	20	and	and	CCONJ
ajst-32001	20	21	evaluate	evaluate	VERB
ajst-32001	20	22	significant	significant	ADJ
ajst-32001	20	23	emotion	emotion	NOUN
ajst-32001	20	24	recognition	recognition	NOUN
ajst-32001	20	25	techniques	technique	NOUN
ajst-32001	20	26	developed	develop	VERB
ajst-32001	20	27	in	in	ADP
ajst-32001	20	28	recent	recent	ADJ
ajst-32001	20	29	years	year	NOUN
ajst-32001	20	30	.	.	PUNCT
ajst-32001	21	1	it	it	PRON
ajst-32001	21	2	discusses	discuss	VERB
ajst-32001	21	3	three	three	NUM
ajst-32001	21	4	main	main	ADJ
ajst-32001	21	5	types	type	NOUN
ajst-32001	21	6	of	of	ADP
ajst-32001	21	7	emotion	emotion	NOUN
ajst-32001	21	8	recognition	recognition	NOUN
ajst-32001	21	9	and	and	CCONJ
ajst-32001	21	10	detection	detection	NOUN
ajst-32001	21	11	methods	method	NOUN
ajst-32001	21	12	,	,	PUNCT
ajst-32001	21	13	which	which	PRON
ajst-32001	21	14	perform	perform	VERB
ajst-32001	21	15	emotion	emotion	NOUN
ajst-32001	21	16	recognition	recognition	NOUN
ajst-32001	21	17	through	through	ADP
ajst-32001	21	18	text	text	NOUN
ajst-32001	21	19	,	,	PUNCT
ajst-32001	21	20	images	image	NOUN
ajst-32001	21	21	,	,	PUNCT
ajst-32001	21	22	and	and	CCONJ
ajst-32001	21	23	speech	speech	NOUN
ajst-32001	21	24	.	.	PUNCT
ajst-32001	22	1	the	the	DET
ajst-32001	22	2	three	three	NUM
ajst-32001	22	3	methods	method	NOUN
ajst-32001	22	4	are	be	AUX
ajst-32001	22	5	shown	show	VERB
ajst-32001	22	6	in	in	ADP
ajst-32001	22	7	figure	figure	NOUN
ajst-32001	22	8	1	1	NUM
ajst-32001	22	9	.	.	PUNCT
ajst-32001	23	1	it	it	PRON
ajst-32001	23	2	also	also	ADV
ajst-32001	23	3	collects	collect	VERB
ajst-32001	23	4	innovative	innovative	ADJ
ajst-32001	23	5	recognition	recognition	NOUN
ajst-32001	23	6	technologies	technology	NOUN
ajst-32001	23	7	and	and	CCONJ
ajst-32001	23	8	optimized	optimize	VERB
ajst-32001	23	9	models	model	NOUN
ajst-32001	23	10	,	,	PUNCT
ajst-32001	23	11	compares	compare	VERB
ajst-32001	23	12	and	and	CCONJ
ajst-32001	23	13	analyzes	analyze	VERB
ajst-32001	23	14	the	the	DET
ajst-32001	23	15	pros	pro	NOUN
ajst-32001	23	16	and	and	CCONJ
ajst-32001	23	17	cons	con	NOUN
ajst-32001	23	18	of	of	ADP
ajst-32001	23	19	various	various	ADJ
ajst-32001	23	20	models	model	NOUN
ajst-32001	23	21	,	,	PUNCT
ajst-32001	23	22	aiming	aim	VERB
ajst-32001	23	23	to	to	PART
ajst-32001	23	24	provide	provide	VERB
ajst-32001	23	25	a	a	DET
ajst-32001	23	26	systematic	systematic	ADJ
ajst-32001	23	27	contribution	contribution	NOUN
ajst-32001	23	28	to	to	ADP
ajst-32001	23	29	the	the	DET
ajst-32001	23	30	development	development	NOUN
ajst-32001	23	31	of	of	ADP
ajst-32001	23	32	emotion	emotion	NOUN
ajst-32001	23	33	recognition	recognition	NOUN
ajst-32001	23	34	technology	technology	NOUN
ajst-32001	23	35	.	.	PUNCT
ajst-32001	24	1	96	96	NUM
ajst-32001	24	2	figure	figure	NOUN
ajst-32001	24	3	1	1	NUM
ajst-32001	24	4	.	.	PUNCT
ajst-32001	24	5	three	three	NUM
ajst-32001	24	6	methods	method	NOUN
ajst-32001	24	7	of	of	ADP
ajst-32001	24	8	emotion	emotion	NOUN
ajst-32001	24	9	recognition	recognition	NOUN
ajst-32001	24	10	the	the	DET
ajst-32001	24	11	structural	structural	ADJ
ajst-32001	24	12	framework	framework	NOUN
ajst-32001	24	13	of	of	ADP
ajst-32001	24	14	this	this	DET
ajst-32001	24	15	paper	paper	NOUN
ajst-32001	24	16	is	be	AUX
ajst-32001	24	17	shown	show	VERB
ajst-32001	24	18	below	below	ADP
ajst-32001	24	19	:	:	PUNCT
ajst-32001	24	20	point	point	VERB
ajst-32001	24	21	out	out	ADP
ajst-32001	24	22	the	the	DET
ajst-32001	24	23	defects	defect	NOUN
ajst-32001	24	24	of	of	ADP
ajst-32001	24	25	current	current	ADJ
ajst-32001	24	26	text	text	NOUN
ajst-32001	24	27	sentiment	sentiment	NOUN
ajst-32001	24	28	recognition	recognition	NOUN
ajst-32001	24	29	,	,	PUNCT
ajst-32001	24	30	and	and	CCONJ
ajst-32001	24	31	introduce	introduce	VERB
ajst-32001	24	32	context	context	NOUN
ajst-32001	24	33	-	-	PUNCT
ajst-32001	24	34	aware	aware	ADJ
ajst-32001	24	35	emotion	emotion	NOUN
ajst-32001	24	36	recognition	recognition	NOUN
ajst-32001	24	37	and	and	CCONJ
ajst-32001	24	38	the	the	DET
ajst-32001	24	39	senn	senn	PROPN
ajst-32001	24	40	model	model	PROPN
ajst-32001	24	41	.	.	PUNCT
ajst-32001	25	1	compare	compare	VERB
ajst-32001	25	2	traditional	traditional	ADJ
ajst-32001	25	3	facial	facial	ADJ
ajst-32001	25	4	emotion	emotion	NOUN
ajst-32001	25	5	recognition	recognition	NOUN
ajst-32001	25	6	technologies	technology	NOUN
ajst-32001	25	7	with	with	ADP
ajst-32001	25	8	deep	deep	ADJ
ajst-32001	25	9	learning	learning	NOUN
ajst-32001	25	10	-	-	PUNCT
ajst-32001	25	11	based	base	VERB
ajst-32001	25	12	methods	method	NOUN
ajst-32001	25	13	,	,	PUNCT
ajst-32001	25	14	and	and	CCONJ
ajst-32001	25	15	highlight	highlight	VERB
ajst-32001	25	16	the	the	DET
ajst-32001	25	17	advantages	advantage	NOUN
ajst-32001	25	18	of	of	ADP
ajst-32001	25	19	deep	deep	ADJ
ajst-32001	25	20	learning	learning	NOUN
ajst-32001	25	21	in	in	ADP
ajst-32001	25	22	facial	facial	ADJ
ajst-32001	25	23	emotion	emotion	NOUN
ajst-32001	25	24	recognition	recognition	NOUN
ajst-32001	25	25	(	(	PUNCT
ajst-32001	25	26	fer	fer	PROPN
ajst-32001	25	27	)	)	PUNCT
ajst-32001	25	28	.	.	PUNCT
ajst-32001	26	1	analysis	analysis	NOUN
ajst-32001	26	2	begins	begin	VERB
ajst-32001	26	3	with	with	ADP
ajst-32001	26	4	traditional	traditional	ADJ
ajst-32001	26	5	speech	speech	NOUN
ajst-32001	26	6	emotion	emotion	NOUN
ajst-32001	26	7	recognition	recognition	NOUN
ajst-32001	26	8	(	(	PUNCT
ajst-32001	26	9	ser	ser	NOUN
ajst-32001	26	10	)	)	PUNCT
ajst-32001	26	11	technologies	technology	NOUN
ajst-32001	26	12	and	and	CCONJ
ajst-32001	26	13	provides	provide	VERB
ajst-32001	26	14	a	a	DET
ajst-32001	26	15	comprehensive	comprehensive	ADJ
ajst-32001	26	16	comparison	comparison	NOUN
ajst-32001	26	17	of	of	ADP
ajst-32001	26	18	the	the	DET
ajst-32001	26	19	pros	pro	NOUN
ajst-32001	26	20	and	and	CCONJ
ajst-32001	26	21	cons	con	NOUN
ajst-32001	26	22	of	of	ADP
ajst-32001	26	23	various	various	ADJ
ajst-32001	26	24	deep	deep	ADJ
ajst-32001	26	25	learning	learning	NOUN
ajst-32001	26	26	-	-	PUNCT
ajst-32001	26	27	based	base	VERB
ajst-32001	26	28	ser	ser	NOUN
ajst-32001	26	29	methods	method	NOUN
ajst-32001	26	30	.	.	PUNCT
ajst-32001	27	1	2	2	X
ajst-32001	27	2	.	.	X
ajst-32001	27	3	optimization	optimization	NOUN
ajst-32001	27	4	based	base	VERB
ajst-32001	27	5	on	on	ADP
ajst-32001	27	6	text	text	NOUN
ajst-32001	27	7	sentiment	sentiment	NOUN
ajst-32001	27	8	recognition	recognition	NOUN
ajst-32001	27	9	technology	technology	NOUN
ajst-32001	27	10	the	the	DET
ajst-32001	27	11	sentiment	sentiment	NOUN
ajst-32001	27	12	of	of	ADP
ajst-32001	27	13	text	text	NOUN
ajst-32001	27	14	is	be	AUX
ajst-32001	27	15	complex	complex	ADJ
ajst-32001	27	16	,	,	PUNCT
ajst-32001	27	17	with	with	ADP
ajst-32001	27	18	a	a	DET
ajst-32001	27	19	large	large	ADJ
ajst-32001	27	20	number	number	NOUN
ajst-32001	27	21	of	of	ADP
ajst-32001	27	22	ambiguous	ambiguous	ADJ
ajst-32001	27	23	expressions	expression	NOUN
ajst-32001	27	24	.	.	PUNCT
ajst-32001	28	1	for	for	ADP
ajst-32001	28	2	example	example	NOUN
ajst-32001	28	3	,	,	PUNCT
ajst-32001	28	4	the	the	DET
ajst-32001	28	5	emotional	emotional	ADJ
ajst-32001	28	6	tendency	tendency	NOUN
ajst-32001	28	7	of	of	ADP
ajst-32001	28	8	neutral	neutral	ADJ
ajst-32001	28	9	words	word	NOUN
ajst-32001	28	10	depends	depend	VERB
ajst-32001	28	11	on	on	ADP
ajst-32001	28	12	the	the	DET
ajst-32001	28	13	context	context	NOUN
ajst-32001	28	14	.	.	PUNCT
ajst-32001	29	1	the	the	DET
ajst-32001	29	2	ccim	ccim	NOUN
ajst-32001	29	3	module	module	NOUN
ajst-32001	29	4	proposed	propose	VERB
ajst-32001	29	5	by	by	ADP
ajst-32001	29	6	dingkang	dingkang	PROPN
ajst-32001	29	7	yang	yang	PROPN
ajst-32001	29	8	eliminates	eliminate	VERB
ajst-32001	29	9	background	background	NOUN
ajst-32001	29	10	interference	interference	NOUN
ajst-32001	29	11	by	by	ADP
ajst-32001	29	12	constructing	construct	VERB
ajst-32001	29	13	a	a	DET
ajst-32001	29	14	causal	causal	ADJ
ajst-32001	29	15	graph	graph	NOUN
ajst-32001	29	16	(	(	PUNCT
ajst-32001	29	17	x	x	NOUN
ajst-32001	29	18	-	-	PUNCT
ajst-32001	29	19	s	s	ADP
ajst-32001	29	20	-	-	PUNCT
ajst-32001	29	21	c	c	NOUN
ajst-32001	29	22	-	-	PUNCT
ajst-32001	29	23	z	z	NOUN
ajst-32001	29	24	-	-	PUNCT
ajst-32001	29	25	y	y	PROPN
ajst-32001	29	26	variables	variable	NOUN
ajst-32001	29	27	)	)	PUNCT
ajst-32001	30	1	[	[	X
ajst-32001	30	2	5	5	NUM
ajst-32001	30	3	]	]	PUNCT
ajst-32001	30	4	.	.	PUNCT
ajst-32001	31	1	this	this	DET
ajst-32001	31	2	idea	idea	NOUN
ajst-32001	31	3	is	be	AUX
ajst-32001	31	4	consistent	consistent	ADJ
ajst-32001	31	5	with	with	ADP
ajst-32001	31	6	the	the	DET
ajst-32001	31	7	goal	goal	NOUN
ajst-32001	31	8	of	of	ADP
ajst-32001	31	9	"	"	PUNCT
ajst-32001	31	10	improving	improve	VERB
ajst-32001	31	11	the	the	DET
ajst-32001	31	12	robustness	robustness	NOUN
ajst-32001	31	13	of	of	ADP
ajst-32001	31	14	emotion	emotion	NOUN
ajst-32001	31	15	recognition	recognition	NOUN
ajst-32001	31	16	"	"	PUNCT
ajst-32001	31	17	focused	focus	VERB
ajst-32001	31	18	on	on	ADP
ajst-32001	31	19	in	in	ADP
ajst-32001	31	20	this	this	DET
ajst-32001	31	21	paper	paper	NOUN
ajst-32001	31	22	—	—	PUNCT
ajst-32001	31	23	especially	especially	ADV
ajst-32001	31	24	in	in	ADP
ajst-32001	31	25	facial	facial	ADJ
ajst-32001	31	26	expression	expression	NOUN
ajst-32001	31	27	recognition	recognition	NOUN
ajst-32001	31	28	in	in	ADP
ajst-32001	31	29	multi	multi	ADJ
ajst-32001	31	30	-	-	ADJ
ajst-32001	31	31	person	person	NOUN
ajst-32001	31	32	scenarios	scenario	NOUN
ajst-32001	31	33	,	,	PUNCT
ajst-32001	31	34	it	it	PRON
ajst-32001	31	35	can	can	AUX
ajst-32001	31	36	effectively	effectively	ADV
ajst-32001	31	37	prevent	prevent	VERB
ajst-32001	31	38	the	the	DET
ajst-32001	31	39	model	model	NOUN
ajst-32001	31	40	from	from	ADP
ajst-32001	31	41	mistakenly	mistakenly	ADV
ajst-32001	31	42	incorporating	incorporate	VERB
ajst-32001	31	43	the	the	DET
ajst-32001	31	44	expression	expression	NOUN
ajst-32001	31	45	features	feature	NOUN
ajst-32001	31	46	of	of	ADP
ajst-32001	31	47	background	background	NOUN
ajst-32001	31	48	characters	character	NOUN
ajst-32001	31	49	into	into	ADP
ajst-32001	31	50	its	its	PRON
ajst-32001	31	51	judgment	judgment	NOUN
ajst-32001	31	52	.	.	PUNCT
ajst-32001	32	1	besides	besides	SCONJ
ajst-32001	32	2	,	,	PUNCT
ajst-32001	32	3	erdenebileg	erdenebileg	NOUN
ajst-32001	32	4	batbaatar	batbaatar	NOUN
ajst-32001	32	5	introduced	introduce	VERB
ajst-32001	32	6	a	a	DET
ajst-32001	32	7	new	new	ADJ
ajst-32001	32	8	neural	neural	ADJ
ajst-32001	32	9	network	network	NOUN
ajst-32001	32	10	architecture	architecture	NOUN
ajst-32001	32	11	named	name	VERB
ajst-32001	32	12	semantic	semantic	ADJ
ajst-32001	32	13	-	-	PUNCT
ajst-32001	32	14	emotion	emotion	NOUN
ajst-32001	32	15	neural	neural	ADJ
ajst-32001	32	16	network	network	NOUN
ajst-32001	32	17	(	(	PUNCT
ajst-32001	32	18	senn	senn	PROPN
ajst-32001	32	19	)	)	PUNCT
ajst-32001	32	20	.	.	PUNCT
ajst-32001	33	1	this	this	DET
ajst-32001	33	2	model	model	NOUN
ajst-32001	33	3	uses	use	VERB
ajst-32001	33	4	both	both	CCONJ
ajst-32001	33	5	semantic	semantic	ADJ
ajst-32001	33	6	and	and	CCONJ
ajst-32001	33	7	emotional	emotional	ADJ
ajst-32001	33	8	data	datum	NOUN
ajst-32001	33	9	through	through	ADP
ajst-32001	33	10	the	the	DET
ajst-32001	33	11	use	use	NOUN
ajst-32001	33	12	of	of	ADP
ajst-32001	33	13	pre	pre	ADJ
ajst-32001	33	14	-	-	ADJ
ajst-32001	33	15	trained	train	VERB
ajst-32001	33	16	word	word	NOUN
ajst-32001	33	17	embeddings.[1	embeddings.[1	NOUN
ajst-32001	33	18	]	]	PUNCT
ajst-32001	33	19	figure	figure	NOUN
ajst-32001	33	20	2	2	NUM
ajst-32001	33	21	.	.	PUNCT
ajst-32001	34	1	the	the	DET
ajst-32001	34	2	caer	caer	PROPN
ajst-32001	34	3	causal	causal	PROPN
ajst-32001	34	4	graph	graph	NOUN
ajst-32001	34	5	existing	exist	VERB
ajst-32001	34	6	emotion	emotion	NOUN
ajst-32001	34	7	recognition	recognition	NOUN
ajst-32001	34	8	models	model	NOUN
ajst-32001	34	9	rely	rely	VERB
ajst-32001	34	10	on	on	ADP
ajst-32001	34	11	multi	multi	ADJ
ajst-32001	34	12	-	-	ADJ
ajst-32001	34	13	modal	modal	ADJ
ajst-32001	34	14	information	information	NOUN
ajst-32001	34	15	from	from	ADP
ajst-32001	34	16	subjects	subject	NOUN
ajst-32001	34	17	and	and	CCONJ
ajst-32001	34	18	backgrounds	background	NOUN
ajst-32001	34	19	.	.	PUNCT
ajst-32001	35	1	however	however	ADV
ajst-32001	35	2	,	,	PUNCT
ajst-32001	35	3	these	these	DET
ajst-32001	35	4	methods	method	NOUN
ajst-32001	35	5	fail	fail	VERB
ajst-32001	35	6	to	to	PART
ajst-32001	35	7	consider	consider	VERB
ajst-32001	35	8	the	the	DET
ajst-32001	35	9	bias	bias	NOUN
ajst-32001	35	10	in	in	ADP
ajst-32001	35	11	background	background	NOUN
ajst-32001	35	12	information	information	NOUN
ajst-32001	35	13	,	,	PUNCT
ajst-32001	35	14	which	which	PRON
ajst-32001	35	15	may	may	AUX
ajst-32001	35	16	cause	cause	VERB
ajst-32001	35	17	the	the	DET
ajst-32001	35	18	models	model	NOUN
ajst-32001	35	19	to	to	ADP
ajst-32001	35	20	over	over	ADV
ajst-32001	35	21	-	-	PUNCT
ajst-32001	35	22	rely	rely	NOUN
ajst-32001	35	23	on	on	ADP
ajst-32001	35	24	certain	certain	ADJ
ajst-32001	35	25	contextual	contextual	ADJ
ajst-32001	35	26	features	feature	NOUN
ajst-32001	35	27	and	and	CCONJ
ajst-32001	35	28	affect	affect	VERB
ajst-32001	35	29	the	the	DET
ajst-32001	35	30	accuracy	accuracy	NOUN
ajst-32001	35	31	of	of	ADP
ajst-32001	35	32	emotion	emotion	NOUN
ajst-32001	35	33	recognition	recognition	NOUN
ajst-32001	35	34	.	.	PUNCT
ajst-32001	36	1	by	by	ADP
ajst-32001	36	2	utilizing	utilize	VERB
ajst-32001	36	3	context	context	NOUN
ajst-32001	36	4	-	-	PUNCT
ajst-32001	36	5	aware	aware	ADJ
ajst-32001	36	6	emotion	emotion	NOUN
ajst-32001	36	7	recognition	recognition	NOUN
ajst-32001	36	8	(	(	PUNCT
ajst-32001	36	9	caer	caer	PROPN
ajst-32001	36	10	)	)	PUNCT
ajst-32001	36	11	and	and	CCONJ
ajst-32001	36	12	through	through	ADP
ajst-32001	36	13	a	a	DET
ajst-32001	36	14	contextual	contextual	ADJ
ajst-32001	36	15	causal	causal	NOUN
ajst-32001	36	16	intervention	intervention	NOUN
ajst-32001	36	17	module	module	NOUN
ajst-32001	36	18	(	(	PUNCT
ajst-32001	36	19	ccim	ccim	NOUN
ajst-32001	36	20	)	)	PUNCT
ajst-32001	36	21	,	,	PUNCT
ajst-32001	36	22	as	as	SCONJ
ajst-32001	36	23	shown	show	VERB
ajst-32001	36	24	in	in	ADP
ajst-32001	36	25	figure	figure	NOUN
ajst-32001	36	26	2	2	NUM
ajst-32001	36	27	,	,	PUNCT
ajst-32001	36	28	this	this	DET
ajst-32001	36	29	method	method	NOUN
ajst-32001	36	30	constructs	construct	VERB
ajst-32001	36	31	a	a	DET
ajst-32001	36	32	causal	causal	ADJ
ajst-32001	36	33	graph	graph	NOUN
ajst-32001	36	34	for	for	ADP
ajst-32001	36	35	the	the	DET
ajst-32001	36	36	caer	caer	PROPN
ajst-32001	36	37	task	task	PROPN
ajst-32001	36	38	,	,	PUNCT
ajst-32001	36	39	involving	involve	VERB
ajst-32001	36	40	variables	variable	NOUN
ajst-32001	36	41	such	such	ADJ
ajst-32001	36	42	as	as	ADP
ajst-32001	36	43	input	input	NOUN
ajst-32001	36	44	images	image	NOUN
ajst-32001	36	45	(	(	PUNCT
ajst-32001	36	46	x	x	NOUN
ajst-32001	36	47	)	)	PUNCT
ajst-32001	36	48	,	,	PUNCT
ajst-32001	36	49	subject	subject	NOUN
ajst-32001	36	50	features	feature	NOUN
ajst-32001	36	51	(	(	PUNCT
ajst-32001	36	52	s	s	NOUN
ajst-32001	36	53	)	)	PUNCT
ajst-32001	36	54	,	,	PUNCT
ajst-32001	36	55	contextual	contextual	ADJ
ajst-32001	36	56	features	feature	NOUN
ajst-32001	36	57	(	(	PUNCT
ajst-32001	36	58	c	c	NOUN
ajst-32001	36	59	)	)	PUNCT
ajst-32001	36	60	,	,	PUNCT
ajst-32001	36	61	confounding	confound	VERB
ajst-32001	36	62	factors	factor	NOUN
ajst-32001	36	63	(	(	PUNCT
ajst-32001	36	64	z	z	NOUN
ajst-32001	36	65	)	)	PUNCT
ajst-32001	36	66	,	,	PUNCT
ajst-32001	36	67	and	and	CCONJ
ajst-32001	36	68	predictions	prediction	NOUN
ajst-32001	36	69	(	(	PUNCT
ajst-32001	36	70	y).[5	y).[5	X
ajst-32001	36	71	]	]	X
ajst-32001	36	72	97	97	NUM
ajst-32001	36	73	it	it	PRON
ajst-32001	36	74	clarifies	clarify	VERB
ajst-32001	36	75	the	the	DET
ajst-32001	36	76	causal	causal	ADJ
ajst-32001	36	77	relationships	relationship	NOUN
ajst-32001	36	78	between	between	ADP
ajst-32001	36	79	variables	variable	NOUN
ajst-32001	36	80	,	,	PUNCT
ajst-32001	36	81	effectively	effectively	ADV
ajst-32001	36	82	eliminates	eliminate	VERB
ajst-32001	36	83	the	the	DET
ajst-32001	36	84	adverse	adverse	ADJ
ajst-32001	36	85	impact	impact	NOUN
ajst-32001	36	86	of	of	ADP
ajst-32001	36	87	contextual	contextual	ADJ
ajst-32001	36	88	bias	bias	NOUN
ajst-32001	36	89	,	,	PUNCT
ajst-32001	36	90	and	and	CCONJ
ajst-32001	36	91	perceives	perceive	VERB
ajst-32001	36	92	the	the	DET
ajst-32001	36	93	emotional	emotional	ADJ
ajst-32001	36	94	state	state	NOUN
ajst-32001	36	95	of	of	ADP
ajst-32001	36	96	the	the	DET
ajst-32001	36	97	target	target	NOUN
ajst-32001	36	98	person	person	NOUN
ajst-32001	36	99	more	more	ADV
ajst-32001	36	100	accurately	accurately	ADV
ajst-32001	36	101	.	.	PUNCT
ajst-32001	37	1	meanwhile	meanwhile	ADV
ajst-32001	37	2	,	,	PUNCT
ajst-32001	37	3	the	the	DET
ajst-32001	37	4	contextual	contextual	ADJ
ajst-32001	37	5	causal	causal	NOUN
ajst-32001	37	6	intervention	intervention	NOUN
ajst-32001	37	7	module	module	NOUN
ajst-32001	37	8	(	(	PUNCT
ajst-32001	37	9	ccim	ccim	NOUN
ajst-32001	37	10	)	)	PUNCT
ajst-32001	37	11	module	module	NOUN
ajst-32001	37	12	eliminates	eliminate	VERB
ajst-32001	37	13	the	the	DET
ajst-32001	37	14	influence	influence	NOUN
ajst-32001	37	15	of	of	ADP
ajst-32001	37	16	confounding	confound	VERB
ajst-32001	37	17	factors	factor	NOUN
ajst-32001	37	18	based	base	VERB
ajst-32001	37	19	on	on	ADP
ajst-32001	37	20	backdoor	backdoor	NOUN
ajst-32001	37	21	adjustment	adjustment	NOUN
ajst-32001	37	22	and	and	CCONJ
ajst-32001	37	23	existing	exist	VERB
ajst-32001	37	24	context	context	NOUN
ajst-32001	37	25	-	-	PUNCT
ajst-32001	37	26	aware	aware	ADJ
ajst-32001	37	27	emotion	emotion	NOUN
ajst-32001	37	28	recognition	recognition	NOUN
ajst-32001	37	29	(	(	PUNCT
ajst-32001	37	30	caer	caer	PROPN
ajst-32001	37	31	)	)	PUNCT
ajst-32001	37	32	methods	method	NOUN
ajst-32001	37	33	rely	rely	VERB
ajst-32001	37	34	on	on	ADP
ajst-32001	37	35	the	the	DET
ajst-32001	37	36	likelihood	likelihood	NOUN
ajst-32001	37	37	p	p	X
ajst-32001	37	38	(	(	PUNCT
ajst-32001	37	39	y	y	PROPN
ajst-32001	37	40	|x	|x	PROPN
ajst-32001	37	41	)	)	PUNCT
ajst-32001	37	42	.	.	PUNCT
ajst-32001	38	1	this	this	DET
ajst-32001	38	2	process	process	NOUN
ajst-32001	38	3	is	be	AUX
ajst-32001	38	4	formulated	formulate	VERB
ajst-32001	38	5	by	by	ADP
ajst-32001	38	6	bayes	bayes	PROPN
ajst-32001	38	7	rule	rule	NOUN
ajst-32001	38	8	:	:	PUNCT
ajst-32001	38	9	p(y|x)=	p(y|x)=	PROPN
ajst-32001	38	10	∑p	∑p	PROPN
ajst-32001	38	11	z	z	PROPN
ajst-32001	38	12	(	(	PUNCT
ajst-32001	38	13	y|x	y|x	NOUN
ajst-32001	38	14	,	,	PUNCT
ajst-32001	38	15	s	s	PROPN
ajst-32001	38	16	=	=	PROPN
ajst-32001	38	17	f	f	X
ajst-32001	38	18	s	s	X
ajst-32001	38	19	(	(	PUNCT
ajst-32001	38	20	x),c	x),c	PROPN
ajst-32001	38	21	=	=	PROPN
ajst-32001	38	22	f	f	X
ajst-32001	38	23	c	c	X
ajst-32001	38	24	(	(	PUNCT
ajst-32001	38	25	x	x	NOUN
ajst-32001	38	26	,	,	PUNCT
ajst-32001	38	27	z))p(z|x	z))p(z|x	NOUN
ajst-32001	38	28	)	)	PUNCT
ajst-32001	38	29	(	(	PUNCT
ajst-32001	38	30	1	1	X
ajst-32001	38	31	)	)	PUNCT
ajst-32001	38	32	this	this	DET
ajst-32001	38	33	formula	formula	NOUN
ajst-32001	38	34	is	be	AUX
ajst-32001	38	35	adapted	adapt	VERB
ajst-32001	38	36	from	from	ADP
ajst-32001	38	37	[	[	X
ajst-32001	38	38	5	5	NUM
ajst-32001	38	39	]	]	PUNCT
ajst-32001	38	40	.	.	PUNCT
ajst-32001	39	1	in	in	ADP
ajst-32001	39	2	the	the	DET
ajst-32001	39	3	formula	formula	NOUN
ajst-32001	39	4	,	,	PUNCT
ajst-32001	39	5	two	two	NUM
ajst-32001	39	6	generalized	generalized	ADJ
ajst-32001	39	7	encoding	encoding	NOUN
ajst-32001	39	8	functions	function	NOUN
ajst-32001	39	9	are	be	AUX
ajst-32001	39	10	f	f	PROPN
ajst-32001	39	11	s	s	PROPN
ajst-32001	39	12	(	(	PUNCT
ajst-32001	39	13	·	·	PUNCT
ajst-32001	39	14	)	)	PUNCT
ajst-32001	39	15	and	and	CCONJ
ajst-32001	39	16	f	f	PROPN
ajst-32001	39	17	c	c	X
ajst-32001	39	18	(	(	PUNCT
ajst-32001	39	19	·	·	PUNCT
ajst-32001	39	20	)	)	PUNCT
ajst-32001	39	21	,	,	PUNCT
ajst-32001	39	22	which	which	PRON
ajst-32001	39	23	are	be	AUX
ajst-32001	39	24	used	use	VERB
ajst-32001	39	25	to	to	PART
ajst-32001	39	26	obtain	obtain	VERB
ajst-32001	39	27	the	the	DET
ajst-32001	39	28	subject	subject	ADJ
ajst-32001	39	29	feature	feature	NOUN
ajst-32001	39	30	s	s	PART
ajst-32001	39	31	and	and	CCONJ
ajst-32001	39	32	context	context	NOUN
ajst-32001	39	33	feature	feature	NOUN
ajst-32001	39	34	c	c	NOUN
ajst-32001	39	35	from	from	ADP
ajst-32001	39	36	the	the	DET
ajst-32001	39	37	input	input	NOUN
ajst-32001	39	38	image	image	NOUN
ajst-32001	39	39	x	x	PUNCT
ajst-32001	39	40	respectively	respectively	ADV
ajst-32001	39	41	;	;	PUNCT
ajst-32001	39	42	the	the	DET
ajst-32001	39	43	confounding	confound	VERB
ajst-32001	39	44	factor	factor	NOUN
ajst-32001	39	45	z	z	PROPN
ajst-32001	39	46	will	will	AUX
ajst-32001	39	47	introduce	introduce	VERB
ajst-32001	39	48	observational	observational	ADJ
ajst-32001	39	49	bias	bias	NOUN
ajst-32001	39	50	through	through	ADP
ajst-32001	39	51	the	the	DET
ajst-32001	39	52	probability	probability	NOUN
ajst-32001	39	53	p(z|x	p(z|x	NOUN
ajst-32001	39	54	)	)	PUNCT
ajst-32001	39	55	,	,	PUNCT
ajst-32001	39	56	affecting	affect	VERB
ajst-32001	39	57	the	the	DET
ajst-32001	39	58	accuracy	accuracy	NOUN
ajst-32001	39	59	of	of	ADP
ajst-32001	39	60	the	the	DET
ajst-32001	39	61	model	model	NOUN
ajst-32001	39	62	's	's	PART
ajst-32001	39	63	judgment	judgment	NOUN
ajst-32001	39	64	on	on	ADP
ajst-32001	39	65	emotion	emotion	NOUN
ajst-32001	39	66	y.	y.	PROPN
ajst-32001	39	67	to	to	PART
ajst-32001	39	68	eliminate	eliminate	VERB
ajst-32001	39	69	the	the	DET
ajst-32001	39	70	interference	interference	NOUN
ajst-32001	39	71	of	of	ADP
ajst-32001	39	72	z	z	PROPN
ajst-32001	39	73	,	,	PUNCT
ajst-32001	39	74	the	the	DET
ajst-32001	39	75	model	model	NOUN
ajst-32001	39	76	is	be	AUX
ajst-32001	39	77	enabled	enable	VERB
ajst-32001	39	78	to	to	PART
ajst-32001	39	79	accurately	accurately	ADV
ajst-32001	39	80	predict	predict	VERB
ajst-32001	39	81	emotion	emotion	NOUN
ajst-32001	39	82	y	y	PROPN
ajst-32001	39	83	based	base	VERB
ajst-32001	39	84	solely	solely	ADV
ajst-32001	39	85	on	on	ADP
ajst-32001	39	86	input	input	NOUN
ajst-32001	39	87	x	x	X
ajst-32001	39	88	,	,	PUNCT
ajst-32001	39	89	ensuring	ensure	VERB
ajst-32001	39	90	every	every	DET
ajst-32001	39	91	context	context	NOUN
ajst-32001	39	92	's	's	PART
ajst-32001	39	93	semantics	semantic	NOUN
ajst-32001	39	94	make	make	VERB
ajst-32001	39	95	equally	equally	ADV
ajst-32001	39	96	to	to	ADP
ajst-32001	39	97	emotion	emotion	NOUN
ajst-32001	39	98	prediction	prediction	NOUN
ajst-32001	39	99	,	,	PUNCT
ajst-32001	39	100	thereby	thereby	ADV
ajst-32001	39	101	letting	let	VERB
ajst-32001	39	102	model	model	NOUN
ajst-32001	39	103	's	's	PART
ajst-32001	39	104	recognition	recognition	NOUN
ajst-32001	39	105	accuracy	accuracy	NOUN
ajst-32001	39	106	higher.[5	higher.[5	X
ajst-32001	39	107	]	]	PUNCT
ajst-32001	39	108	kang	kang	PROPN
ajst-32001	39	109	yang	yang	PROPN
ajst-32001	39	110	uses	use	VERB
ajst-32001	39	111	backdoor	backdoor	NOUN
ajst-32001	39	112	adjustment	adjustment	NOUN
ajst-32001	39	113	to	to	PART
ajst-32001	39	114	stratify	stratify	VERB
ajst-32001	39	115	z	z	NOUN
ajst-32001	39	116	,	,	PUNCT
ajst-32001	39	117	achieving	achieve	VERB
ajst-32001	39	118	causal	causal	ADJ
ajst-32001	39	119	intervention	intervention	NOUN
ajst-32001	39	120	p	p	PROPN
ajst-32001	39	121	(	(	PUNCT
ajst-32001	39	122	y	y	PROPN
ajst-32001	39	123	|	|	ADV
ajst-32001	39	124	do(x	do(x	NUM
ajst-32001	39	125	)	)	PUNCT
ajst-32001	39	126	)	)	PUNCT
ajst-32001	39	127	and	and	CCONJ
ajst-32001	39	128	training	train	VERB
ajst-32001	39	129	the	the	DET
ajst-32001	39	130	model	model	NOUN
ajst-32001	39	131	with	with	ADP
ajst-32001	39	132	real	real	ADJ
ajst-32001	39	133	causal	causal	ADJ
ajst-32001	39	134	effects	effect	NOUN
ajst-32001	39	135	.	.	PUNCT
ajst-32001	40	1	word	word	NOUN
ajst-32001	40	2	embedding	embed	VERB
ajst-32001	40	3	is	be	AUX
ajst-32001	40	4	widely	widely	ADV
ajst-32001	40	5	used	use	VERB
ajst-32001	40	6	in	in	ADP
ajst-32001	40	7	many	many	ADJ
ajst-32001	40	8	natural	natural	ADJ
ajst-32001	40	9	language	language	NOUN
ajst-32001	40	10	processing	processing	NOUN
ajst-32001	40	11	tasks	task	NOUN
ajst-32001	40	12	.	.	PUNCT
ajst-32001	41	1	training	train	VERB
ajst-32001	41	2	a	a	DET
ajst-32001	41	3	single	single	ADJ
ajst-32001	41	4	model	model	NOUN
ajst-32001	41	5	with	with	ADP
ajst-32001	41	6	semantic	semantic	ADJ
ajst-32001	41	7	or	or	CCONJ
ajst-32001	41	8	emotional	emotional	ADJ
ajst-32001	41	9	word	word	NOUN
ajst-32001	41	10	embedding	embed	VERB
ajst-32001	41	11	yields	yield	NOUN
ajst-32001	41	12	good	good	ADJ
ajst-32001	41	13	results	result	NOUN
ajst-32001	41	14	.	.	PUNCT
ajst-32001	42	1	however	however	ADV
ajst-32001	42	2	,	,	PUNCT
ajst-32001	42	3	when	when	SCONJ
ajst-32001	42	4	embedding	embed	VERB
ajst-32001	42	5	words	word	NOUN
ajst-32001	42	6	with	with	ADP
ajst-32001	42	7	similar	similar	ADJ
ajst-32001	42	8	semantic	semantic	ADJ
ajst-32001	42	9	functions	function	NOUN
ajst-32001	42	10	are	be	AUX
ajst-32001	42	11	confused	confuse	VERB
ajst-32001	42	12	in	in	ADP
ajst-32001	42	13	emotional	emotional	ADJ
ajst-32001	42	14	states	state	NOUN
ajst-32001	42	15	.	.	PUNCT
ajst-32001	43	1	this	this	PRON
ajst-32001	43	2	makes	make	VERB
ajst-32001	43	3	it	it	PRON
ajst-32001	43	4	impossible	impossible	ADJ
ajst-32001	43	5	to	to	PART
ajst-32001	43	6	effectively	effectively	ADV
ajst-32001	43	7	encode	encode	VERB
ajst-32001	43	8	and	and	CCONJ
ajst-32001	43	9	learn	learn	VERB
ajst-32001	43	10	semantic	semantic	ADJ
ajst-32001	43	11	and	and	CCONJ
ajst-32001	43	12	emotional	emotional	ADJ
ajst-32001	43	13	relationships	relationship	NOUN
ajst-32001	43	14	in	in	ADP
ajst-32001	43	15	short	short	ADJ
ajst-32001	43	16	texts	text	NOUN
ajst-32001	43	17	.	.	PUNCT
ajst-32001	44	1	figure	figure	NOUN
ajst-32001	44	2	3	3	NUM
ajst-32001	44	3	.	.	PUNCT
ajst-32001	45	1	senn	senn	PROPN
ajst-32001	45	2	model	model	PROPN
ajst-32001	45	3	architecture	architecture	NOUN
ajst-32001	45	4	the	the	DET
ajst-32001	45	5	emergence	emergence	NOUN
ajst-32001	45	6	of	of	ADP
ajst-32001	45	7	the	the	DET
ajst-32001	45	8	senn	senn	PROPN
ajst-32001	45	9	model	model	PROPN
ajst-32001	45	10	breaks	break	VERB
ajst-32001	45	11	the	the	DET
ajst-32001	45	12	limitations	limitation	NOUN
ajst-32001	45	13	of	of	ADP
ajst-32001	45	14	single	single	ADJ
ajst-32001	45	15	representation	representation	NOUN
ajst-32001	45	16	and	and	CCONJ
ajst-32001	45	17	emotional	emotional	ADJ
ajst-32001	45	18	context	context	NOUN
ajst-32001	45	19	in	in	ADP
ajst-32001	45	20	traditional	traditional	ADJ
ajst-32001	45	21	research	research	NOUN
ajst-32001	45	22	.	.	PUNCT
ajst-32001	46	1	senn	senn	PROPN
ajst-32001	46	2	adopts	adopt	VERB
ajst-32001	46	3	a	a	DET
ajst-32001	46	4	dual	dual	ADJ
ajst-32001	46	5	-	-	PUNCT
ajst-32001	46	6	path	path	NOUN
ajst-32001	46	7	architecture	architecture	NOUN
ajst-32001	46	8	:	:	PUNCT
ajst-32001	46	9	a	a	DET
ajst-32001	46	10	bilstm	bilstm	NOUN
ajst-32001	46	11	-	-	PUNCT
ajst-32001	46	12	based	base	VERB
ajst-32001	46	13	semantic	semantic	ADJ
ajst-32001	46	14	encoder	encoder	NOUN
ajst-32001	46	15	and	and	CCONJ
ajst-32001	46	16	a	a	DET
ajst-32001	46	17	cnn	cnn	PROPN
ajst-32001	46	18	-	-	PUNCT
ajst-32001	46	19	based	base	VERB
ajst-32001	46	20	emotional	emotional	ADJ
ajst-32001	46	21	encoder	encoder	NOUN
ajst-32001	46	22	,	,	PUNCT
ajst-32001	46	23	enabling	enable	VERB
ajst-32001	46	24	parallel	parallel	ADJ
ajst-32001	46	25	extraction	extraction	NOUN
ajst-32001	46	26	and	and	CCONJ
ajst-32001	46	27	fusion	fusion	NOUN
ajst-32001	46	28	of	of	ADP
ajst-32001	46	29	features	feature	NOUN
ajst-32001	46	30	[	[	X
ajst-32001	46	31	1	1	NUM
ajst-32001	46	32	]	]	PUNCT
ajst-32001	46	33	,	,	PUNCT
ajst-32001	46	34	and	and	CCONJ
ajst-32001	46	35	its	its	PRON
ajst-32001	46	36	structure	structure	NOUN
ajst-32001	46	37	is	be	AUX
ajst-32001	46	38	shown	show	VERB
ajst-32001	46	39	in	in	ADP
ajst-32001	46	40	figure	figure	NOUN
ajst-32001	46	41	3	3	NUM
ajst-32001	46	42	.	.	PUNCT
ajst-32001	47	1	it	it	PRON
ajst-32001	47	2	abandons	abandon	VERB
ajst-32001	47	3	the	the	DET
ajst-32001	47	4	disadvantages	disadvantage	NOUN
ajst-32001	47	5	of	of	ADP
ajst-32001	47	6	manual	manual	ADJ
ajst-32001	47	7	feature	feature	NOUN
ajst-32001	47	8	models	model	NOUN
ajst-32001	47	9	,	,	PUNCT
ajst-32001	47	10	such	such	ADJ
ajst-32001	47	11	as	as	ADP
ajst-32001	47	12	poor	poor	ADJ
ajst-32001	47	13	adaptability	adaptability	NOUN
ajst-32001	47	14	and	and	CCONJ
ajst-32001	47	15	time	time	NOUN
ajst-32001	47	16	-	-	PUNCT
ajst-32001	47	17	consuming	consume	VERB
ajst-32001	47	18	processes	process	NOUN
ajst-32001	47	19	,	,	PUNCT
ajst-32001	47	20	and	and	CCONJ
ajst-32001	47	21	introduces	introduce	VERB
ajst-32001	47	22	a	a	DET
ajst-32001	47	23	dual	dual	ADJ
ajst-32001	47	24	-	-	PUNCT
ajst-32001	47	25	network	network	NOUN
ajst-32001	47	26	structure	structure	NOUN
ajst-32001	47	27	.	.	PUNCT
ajst-32001	48	1	by	by	ADP
ajst-32001	48	2	processing	process	VERB
ajst-32001	48	3	the	the	DET
ajst-32001	48	4	same	same	ADJ
ajst-32001	48	5	input	input	NOUN
ajst-32001	48	6	text	text	NOUN
ajst-32001	48	7	simultaneously	simultaneously	ADV
ajst-32001	48	8	,	,	PUNCT
ajst-32001	48	9	it	it	PRON
ajst-32001	48	10	extracts	extract	VERB
ajst-32001	48	11	semantic	semantic	ADJ
ajst-32001	48	12	and	and	CCONJ
ajst-32001	48	13	emotional	emotional	ADJ
ajst-32001	48	14	information	information	NOUN
ajst-32001	48	15	respectively	respectively	ADV
ajst-32001	48	16	and	and	CCONJ
ajst-32001	48	17	merges	merge	VERB
ajst-32001	48	18	the	the	DET
ajst-32001	48	19	outputs	output	NOUN
ajst-32001	48	20	,	,	PUNCT
ajst-32001	48	21	thereby	thereby	ADV
ajst-32001	48	22	achieving	achieve	VERB
ajst-32001	48	23	more	more	ADV
ajst-32001	48	24	accurate	accurate	ADJ
ajst-32001	48	25	emotion	emotion	NOUN
ajst-32001	48	26	recognition.[1	recognition.[1	NOUN
ajst-32001	48	27	]	]	PUNCT
ajst-32001	48	28	utilizing	utilize	VERB
ajst-32001	48	29	bilstm	bilstm	NOUN
ajst-32001	48	30	-	-	PUNCT
ajst-32001	48	31	based	base	VERB
ajst-32001	48	32	semantic	semantic	ADJ
ajst-32001	48	33	encoder	encoder	NOUN
ajst-32001	48	34	and	and	CCONJ
ajst-32001	48	35	cnn	cnn	PROPN
ajst-32001	48	36	-	-	PUNCT
ajst-32001	48	37	based	base	VERB
ajst-32001	48	38	emotional	emotional	ADJ
ajst-32001	48	39	encoder	encoder	NOUN
ajst-32001	48	40	,	,	PUNCT
ajst-32001	48	41	and	and	CCONJ
ajst-32001	48	42	the	the	DET
ajst-32001	48	43	output	output	NOUN
ajst-32001	48	44	of	of	ADP
ajst-32001	48	45	the	the	DET
ajst-32001	48	46	subnetwork	subnetwork	NOUN
ajst-32001	48	47	is	be	AUX
ajst-32001	48	48	used	use	VERB
ajst-32001	48	49	to	to	PART
ajst-32001	48	50	identify	identify	VERB
ajst-32001	48	51	emotions	emotion	NOUN
ajst-32001	48	52	in	in	ADP
ajst-32001	48	53	the	the	DET
ajst-32001	48	54	text	text	NOUN
ajst-32001	48	55	.	.	PUNCT
ajst-32001	49	1	this	this	DET
ajst-32001	49	2	architecture	architecture	NOUN
ajst-32001	49	3	processes	process	VERB
ajst-32001	49	4	text	text	NOUN
ajst-32001	49	5	through	through	ADP
ajst-32001	49	6	dual	dual	ADJ
ajst-32001	49	7	networks	network	NOUN
ajst-32001	49	8	in	in	ADP
ajst-32001	49	9	parallel	parallel	NOUN
ajst-32001	49	10	,	,	PUNCT
ajst-32001	49	11	extracting	extract	VERB
ajst-32001	49	12	semantic	semantic	ADJ
ajst-32001	49	13	and	and	CCONJ
ajst-32001	49	14	emotional	emotional	ADJ
ajst-32001	49	15	features	feature	NOUN
ajst-32001	49	16	separately	separately	ADV
ajst-32001	49	17	before	before	ADP
ajst-32001	49	18	fusing	fuse	VERB
ajst-32001	49	19	them	they	PRON
ajst-32001	49	20	,	,	PUNCT
ajst-32001	49	21	thus	thus	ADV
ajst-32001	49	22	addressing	address	VERB
ajst-32001	49	23	the	the	DET
ajst-32001	49	24	limitation	limitation	NOUN
ajst-32001	49	25	of	of	ADP
ajst-32001	49	26	traditional	traditional	ADJ
ajst-32001	49	27	models	model	NOUN
ajst-32001	49	28	98	98	NUM
ajst-32001	49	29	that	that	PRON
ajst-32001	49	30	rely	rely	VERB
ajst-32001	49	31	on	on	ADP
ajst-32001	49	32	a	a	DET
ajst-32001	49	33	single	single	ADJ
ajst-32001	49	34	type	type	NOUN
ajst-32001	49	35	of	of	ADP
ajst-32001	49	36	feature	feature	NOUN
ajst-32001	49	37	.	.	PUNCT
ajst-32001	50	1	what	what	PRON
ajst-32001	50	2	’s	’	VERB
ajst-32001	50	3	more	more	ADJ
ajst-32001	50	4	,	,	PUNCT
ajst-32001	50	5	senn	senn	PROPN
ajst-32001	50	6	can	can	AUX
ajst-32001	50	7	improve	improve	VERB
ajst-32001	50	8	performance	performance	NOUN
ajst-32001	50	9	by	by	ADP
ajst-32001	50	10	fine	fine	ADV
ajst-32001	50	11	-	-	PUNCT
ajst-32001	50	12	tuning	tune	VERB
ajst-32001	50	13	pre	pre	ADJ
ajst-32001	50	14	-	-	ADJ
ajst-32001	50	15	trained	train	VERB
ajst-32001	50	16	word	word	NOUN
ajst-32001	50	17	embeddings.[1	embeddings.[1	NOUN
ajst-32001	50	18	]	]	X
ajst-32001	50	19	3	3	X
ajst-32001	50	20	.	.	X
ajst-32001	50	21	optimization	optimization	NOUN
ajst-32001	50	22	based	base	VERB
ajst-32001	50	23	on	on	ADP
ajst-32001	50	24	facial	facial	ADJ
ajst-32001	50	25	and	and	CCONJ
ajst-32001	50	26	image	image	NOUN
ajst-32001	50	27	emotion	emotion	NOUN
ajst-32001	50	28	recognition	recognition	NOUN
ajst-32001	50	29	technology	technology	NOUN
ajst-32001	50	30	unlike	unlike	ADP
ajst-32001	50	31	text	text	NOUN
ajst-32001	50	32	,	,	PUNCT
ajst-32001	50	33	which	which	PRON
ajst-32001	50	34	relies	rely	VERB
ajst-32001	50	35	on	on	ADP
ajst-32001	50	36	contextual	contextual	ADJ
ajst-32001	50	37	semantics	semantic	NOUN
ajst-32001	50	38	,	,	PUNCT
ajst-32001	50	39	facial	facial	ADJ
ajst-32001	50	40	expression	expression	NOUN
ajst-32001	50	41	recognition	recognition	NOUN
ajst-32001	50	42	is	be	AUX
ajst-32001	50	43	more	more	ADV
ajst-32001	50	44	dependent	dependent	ADJ
ajst-32001	50	45	on	on	ADP
ajst-32001	50	46	the	the	DET
ajst-32001	50	47	fine	fine	ADJ
ajst-32001	50	48	extraction	extraction	NOUN
ajst-32001	50	49	of	of	ADP
ajst-32001	50	50	visual	visual	ADJ
ajst-32001	50	51	features	feature	NOUN
ajst-32001	50	52	.	.	PUNCT
ajst-32001	51	1	therefore	therefore	ADV
ajst-32001	51	2	,	,	PUNCT
ajst-32001	51	3	the	the	DET
ajst-32001	51	4	cnn	cnn	NOUN
ajst-32001	51	5	architecture	architecture	NOUN
ajst-32001	51	6	in	in	ADP
ajst-32001	51	7	deep	deep	ADJ
ajst-32001	51	8	learning	learning	NOUN
ajst-32001	51	9	,	,	PUNCT
ajst-32001	51	10	through	through	ADP
ajst-32001	51	11	the	the	DET
ajst-32001	51	12	combination	combination	NOUN
ajst-32001	51	13	of	of	ADP
ajst-32001	51	14	edge	edge	NOUN
ajst-32001	51	15	detection	detection	NOUN
ajst-32001	51	16	and	and	CCONJ
ajst-32001	51	17	convolutional	convolutional	ADJ
ajst-32001	51	18	layers	layer	NOUN
ajst-32001	51	19	,	,	PUNCT
ajst-32001	51	20	demonstrates	demonstrate	VERB
ajst-32001	51	21	advantages	advantage	NOUN
ajst-32001	51	22	in	in	ADP
ajst-32001	51	23	retaining	retain	VERB
ajst-32001	51	24	texture	texture	ADJ
ajst-32001	51	25	information	information	NOUN
ajst-32001	51	26	that	that	SCONJ
ajst-32001	51	27	traditional	traditional	ADJ
ajst-32001	51	28	methods	method	NOUN
ajst-32001	51	29	can	can	AUX
ajst-32001	51	30	hardly	hardly	ADV
ajst-32001	51	31	match	match	VERB
ajst-32001	51	32	.	.	PUNCT
ajst-32001	52	1	with	with	ADP
ajst-32001	52	2	the	the	DET
ajst-32001	52	3	rapid	rapid	ADJ
ajst-32001	52	4	development	development	NOUN
ajst-32001	52	5	of	of	ADP
ajst-32001	52	6	computer	computer	NOUN
ajst-32001	52	7	technology	technology	NOUN
ajst-32001	52	8	and	and	CCONJ
ajst-32001	52	9	the	the	DET
ajst-32001	52	10	continuous	continuous	ADJ
ajst-32001	52	11	growth	growth	NOUN
ajst-32001	52	12	of	of	ADP
ajst-32001	52	13	data	datum	NOUN
ajst-32001	52	14	volume	volume	NOUN
ajst-32001	52	15	,	,	PUNCT
ajst-32001	52	16	computer	computer	NOUN
ajst-32001	52	17	vision	vision	NOUN
ajst-32001	52	18	has	have	AUX
ajst-32001	52	19	become	become	VERB
ajst-32001	52	20	an	an	DET
ajst-32001	52	21	emerging	emerge	VERB
ajst-32001	52	22	research	research	NOUN
ajst-32001	52	23	direction	direction	NOUN
ajst-32001	52	24	.	.	PUNCT
ajst-32001	53	1	among	among	ADP
ajst-32001	53	2	them	they	PRON
ajst-32001	53	3	,	,	PUNCT
ajst-32001	53	4	face	face	VERB
ajst-32001	53	5	detection	detection	NOUN
ajst-32001	53	6	,	,	PUNCT
ajst-32001	53	7	recognition	recognition	NOUN
ajst-32001	53	8	and	and	CCONJ
ajst-32001	53	9	expression	expression	NOUN
ajst-32001	53	10	analysis	analysis	NOUN
ajst-32001	53	11	technologies	technology	NOUN
ajst-32001	53	12	are	be	AUX
ajst-32001	53	13	one	one	NUM
ajst-32001	53	14	of	of	ADP
ajst-32001	53	15	the	the	DET
ajst-32001	53	16	most	most	ADV
ajst-32001	53	17	active	active	ADJ
ajst-32001	53	18	research	research	NOUN
ajst-32001	53	19	directions	direction	NOUN
ajst-32001	53	20	in	in	ADP
ajst-32001	53	21	the	the	DET
ajst-32001	53	22	current	current	ADJ
ajst-32001	53	23	computer	computer	NOUN
ajst-32001	53	24	vision	vision	NOUN
ajst-32001	53	25	community.[6	community.[6	PRON
ajst-32001	53	26	]	]	X
ajst-32001	53	27	traditional	traditional	ADJ
ajst-32001	53	28	facial	facial	ADJ
ajst-32001	53	29	expression	expression	NOUN
ajst-32001	53	30	recognition	recognition	NOUN
ajst-32001	53	31	requires	require	VERB
ajst-32001	53	32	first	first	ADV
ajst-32001	53	33	locating	locate	VERB
ajst-32001	53	34	the	the	DET
ajst-32001	53	35	face	face	NOUN
ajst-32001	53	36	and	and	CCONJ
ajst-32001	53	37	facial	facial	ADJ
ajst-32001	53	38	features	feature	NOUN
ajst-32001	53	39	from	from	ADP
ajst-32001	53	40	the	the	DET
ajst-32001	53	41	image	image	NOUN
ajst-32001	53	42	,	,	PUNCT
ajst-32001	53	43	then	then	ADV
ajst-32001	53	44	extracting	extract	VERB
ajst-32001	53	45	spatial	spatial	ADJ
ajst-32001	53	46	and	and	CCONJ
ajst-32001	53	47	temporal	temporal	ADJ
ajst-32001	53	48	features	feature	NOUN
ajst-32001	53	49	from	from	ADP
ajst-32001	53	50	these	these	DET
ajst-32001	53	51	regions	region	NOUN
ajst-32001	53	52	,	,	PUNCT
ajst-32001	53	53	and	and	CCONJ
ajst-32001	53	54	finally	finally	ADV
ajst-32001	53	55	outputting	output	VERB
ajst-32001	53	56	the	the	DET
ajst-32001	53	57	recognition	recognition	NOUN
ajst-32001	53	58	results	result	NOUN
ajst-32001	53	59	according	accord	VERB
ajst-32001	53	60	to	to	ADP
ajst-32001	53	61	the	the	DET
ajst-32001	53	62	features	feature	NOUN
ajst-32001	53	63	by	by	ADP
ajst-32001	53	64	pre	pre	ADJ
ajst-32001	53	65	-	-	ADJ
ajst-32001	53	66	trained	train	VERB
ajst-32001	53	67	classifiers	classifier	NOUN
ajst-32001	53	68	such	such	ADJ
ajst-32001	53	69	as	as	ADP
ajst-32001	53	70	support	support	NOUN
ajst-32001	53	71	vector	vector	NOUN
ajst-32001	53	72	machine	machine	NOUN
ajst-32001	53	73	(	(	PUNCT
ajst-32001	53	74	svm	svm	PROPN
ajst-32001	53	75	)	)	PUNCT
ajst-32001	53	76	,	,	PUNCT
ajst-32001	53	77	adaboost	adaboost	ADV
ajst-32001	53	78	,	,	PUNCT
ajst-32001	53	79	and	and	CCONJ
ajst-32001	53	80	random	random	ADJ
ajst-32001	53	81	forest.[7	forest.[7	NOUN
ajst-32001	53	82	]	]	PUNCT
ajst-32001	53	83	unlike	unlike	ADP
ajst-32001	53	84	traditional	traditional	ADJ
ajst-32001	53	85	methods	method	NOUN
ajst-32001	53	86	that	that	PRON
ajst-32001	53	87	rely	rely	VERB
ajst-32001	53	88	on	on	ADP
ajst-32001	53	89	handcrafted	handcraft	VERB
ajst-32001	53	90	features	feature	NOUN
ajst-32001	53	91	,	,	PUNCT
ajst-32001	53	92	deep	deep	ADJ
ajst-32001	53	93	learning	learning	NOUN
ajst-32001	53	94	has	have	AUX
ajst-32001	53	95	become	become	VERB
ajst-32001	53	96	a	a	DET
ajst-32001	53	97	leading	lead	VERB
ajst-32001	53	98	approach	approach	NOUN
ajst-32001	53	99	in	in	ADP
ajst-32001	53	100	machine	machine	NOUN
ajst-32001	53	101	learning	learning	NOUN
ajst-32001	53	102	,	,	PUNCT
ajst-32001	53	103	achieving	achieve	VERB
ajst-32001	53	104	top	top	ADJ
ajst-32001	53	105	-	-	PUNCT
ajst-32001	53	106	tier	tier	NOUN
ajst-32001	53	107	results	result	NOUN
ajst-32001	53	108	in	in	ADP
ajst-32001	53	109	numerous	numerous	ADJ
ajst-32001	53	110	computer	computer	NOUN
ajst-32001	53	111	vision	vision	NOUN
ajst-32001	53	112	studies	study	NOUN
ajst-32001	53	113	,	,	PUNCT
ajst-32001	53	114	thanks	thank	NOUN
ajst-32001	53	115	to	to	ADP
ajst-32001	53	116	the	the	DET
ajst-32001	53	117	availability	availability	NOUN
ajst-32001	53	118	of	of	ADP
ajst-32001	53	119	big	big	ADJ
ajst-32001	53	120	data	datum	NOUN
ajst-32001	53	121	.	.	PUNCT
ajst-32001	54	1	[	[	X
ajst-32001	54	2	8	8	NUM
ajst-32001	54	3	]	]	SYM
ajst-32001	54	4	3.1	3.1	NUM
ajst-32001	54	5	fer	fer	NOUN
ajst-32001	54	6	based	base	VERB
ajst-32001	54	7	on	on	ADP
ajst-32001	54	8	deep	deep	ADJ
ajst-32001	54	9	learning	learn	VERB
ajst-32001	54	10	a	a	DET
ajst-32001	54	11	method	method	NOUN
ajst-32001	54	12	for	for	ADP
ajst-32001	54	13	recognizing	recognize	VERB
ajst-32001	54	14	facial	facial	ADJ
ajst-32001	54	15	expressions	expression	NOUN
ajst-32001	54	16	using	use	VERB
ajst-32001	54	17	convolutional	convolutional	ADJ
ajst-32001	54	18	neural	neural	ADJ
ajst-32001	54	19	networks	network	NOUN
ajst-32001	54	20	(	(	PUNCT
ajst-32001	54	21	cnn	cnn	PROPN
ajst-32001	54	22	)	)	PUNCT
ajst-32001	54	23	and	and	CCONJ
ajst-32001	54	24	image	image	NOUN
ajst-32001	54	25	edge	edge	NOUN
ajst-32001	54	26	detection	detection	NOUN
ajst-32001	54	27	:	:	PUNCT
ajst-32001	54	28	first	first	ADV
ajst-32001	54	29	,	,	PUNCT
ajst-32001	54	30	perform	perform	VERB
ajst-32001	54	31	image	image	NOUN
ajst-32001	54	32	preprocessing	preprocessing	NOUN
ajst-32001	54	33	;	;	PUNCT
ajst-32001	54	34	then	then	ADV
ajst-32001	54	35	,	,	PUNCT
ajst-32001	54	36	cnns	cnn	NOUN
ajst-32001	54	37	gradually	gradually	ADV
ajst-32001	54	38	extract	extract	VERB
ajst-32001	54	39	facial	facial	ADJ
ajst-32001	54	40	texture	texture	NOUN
ajst-32001	54	41	features	feature	NOUN
ajst-32001	54	42	through	through	ADP
ajst-32001	54	43	hierarchical	hierarchical	ADJ
ajst-32001	54	44	convolution	convolution	NOUN
ajst-32001	54	45	[	[	X
ajst-32001	54	46	9	9	NUM
ajst-32001	54	47	]	]	PUNCT
ajst-32001	54	48	,	,	PUNCT
ajst-32001	54	49	which	which	PRON
ajst-32001	54	50	is	be	AUX
ajst-32001	54	51	similar	similar	ADJ
ajst-32001	54	52	to	to	ADP
ajst-32001	54	53	a	a	DET
ajst-32001	54	54	progressive	progressive	ADJ
ajst-32001	54	55	detail	detail	NOUN
ajst-32001	54	56	magnification	magnification	NOUN
ajst-32001	54	57	process	process	NOUN
ajst-32001	54	58	;	;	PUNCT
ajst-32001	54	59	subsequently	subsequently	ADV
ajst-32001	54	60	,	,	PUNCT
ajst-32001	54	61	simplify	simplify	VERB
ajst-32001	54	62	the	the	DET
ajst-32001	54	63	features	feature	NOUN
ajst-32001	54	64	and	and	CCONJ
ajst-32001	54	65	process	process	NOUN
ajst-32001	54	66	and	and	CCONJ
ajst-32001	54	67	compress	compress	VERB
ajst-32001	54	68	the	the	DET
ajst-32001	54	69	extracted	extract	VERB
ajst-32001	54	70	features.finally	features.finally	ADV
ajst-32001	54	71	,	,	PUNCT
ajst-32001	54	72	use	use	VERB
ajst-32001	54	73	the	the	DET
ajst-32001	54	74	softmax	softmax	NOUN
ajst-32001	54	75	classifier	classifier	NOUN
ajst-32001	54	76	to	to	PART
ajst-32001	54	77	classify	classify	VERB
ajst-32001	54	78	and	and	CCONJ
ajst-32001	54	79	recognize	recognize	VERB
ajst-32001	54	80	the	the	DET
ajst-32001	54	81	expressions	expression	NOUN
ajst-32001	54	82	of	of	ADP
ajst-32001	54	83	the	the	DET
ajst-32001	54	84	test	test	NOUN
ajst-32001	54	85	sample	sample	NOUN
ajst-32001	54	86	images.[9	images.[9	NOUN
ajst-32001	54	87	]	]	X
ajst-32001	54	88	cnn	cnn	PROPN
ajst-32001	54	89	structure	structure	NOUN
ajst-32001	54	90	for	for	ADP
ajst-32001	54	91	facial	facial	ADJ
ajst-32001	54	92	expression	expression	NOUN
ajst-32001	54	93	recognition	recognition	NOUN
ajst-32001	54	94	is	be	AUX
ajst-32001	54	95	shown	show	VERB
ajst-32001	54	96	in	in	ADP
ajst-32001	54	97	figure	figure	NOUN
ajst-32001	54	98	4.[9	4.[9	X
ajst-32001	54	99	]	]	X
ajst-32001	54	100	figure	figure	NOUN
ajst-32001	55	1	4	4	NUM
ajst-32001	55	2	.	.	PUNCT
ajst-32001	55	3	cnn	cnn	PROPN
ajst-32001	55	4	architecture	architecture	NOUN
ajst-32001	55	5	for	for	ADP
ajst-32001	55	6	facial	facial	ADJ
ajst-32001	55	7	expression	expression	NOUN
ajst-32001	55	8	recognition[9	recognition[9	NUM
ajst-32001	55	9	]	]	X
ajst-32001	55	10	suci	suci	ADJ
ajst-32001	55	11	dwijayanti	dwijayanti	PROPN
ajst-32001	55	12	proposed	propose	VERB
ajst-32001	55	13	a	a	DET
ajst-32001	55	14	real	real	ADJ
ajst-32001	55	15	-	-	PUNCT
ajst-32001	55	16	time	time	NOUN
ajst-32001	55	17	application	application	NOUN
ajst-32001	55	18	scheme	scheme	NOUN
ajst-32001	55	19	combining	combine	VERB
ajst-32001	55	20	face	face	NOUN
ajst-32001	55	21	and	and	CCONJ
ajst-32001	55	22	emotion	emotion	NOUN
ajst-32001	55	23	recognition	recognition	NOUN
ajst-32001	55	24	.	.	PUNCT
ajst-32001	56	1	specifically	specifically	ADV
ajst-32001	56	2	,	,	PUNCT
ajst-32001	56	3	a	a	DET
ajst-32001	56	4	convolutional	convolutional	ADJ
ajst-32001	56	5	neural	neural	ADJ
ajst-32001	56	6	network	network	NOUN
ajst-32001	56	7	architecture	architecture	NOUN
ajst-32001	56	8	is	be	AUX
ajst-32001	56	9	used	use	VERB
ajst-32001	56	10	to	to	PART
ajst-32001	56	11	simultaneously	simultaneously	ADV
ajst-32001	56	12	develop	develop	VERB
ajst-32001	56	13	face	face	NOUN
ajst-32001	56	14	and	and	CCONJ
ajst-32001	56	15	emotion	emotion	NOUN
ajst-32001	56	16	recognition	recognition	NOUN
ajst-32001	56	17	systems	system	NOUN
ajst-32001	56	18	.	.	PUNCT
ajst-32001	57	1	this	this	DET
ajst-32001	57	2	model	model	NOUN
ajst-32001	57	3	is	be	AUX
ajst-32001	57	4	compared	compare	VERB
ajst-32001	57	5	with	with	ADP
ajst-32001	57	6	well	well	ADV
ajst-32001	57	7	-	-	PUNCT
ajst-32001	57	8	known	know	VERB
ajst-32001	57	9	architectures	architecture	NOUN
ajst-32001	57	10	such	such	ADJ
ajst-32001	57	11	as	as	ADP
ajst-32001	57	12	alexnet	alexnet	NOUN
ajst-32001	57	13	and	and	CCONJ
ajst-32001	57	14	vgg16	vgg16	NOUN
ajst-32001	57	15	to	to	PART
ajst-32001	57	16	determine	determine	VERB
ajst-32001	57	17	which	which	PRON
ajst-32001	57	18	is	be	AUX
ajst-32001	57	19	more	more	ADV
ajst-32001	57	20	suitable	suitable	ADJ
ajst-32001	57	21	for	for	ADP
ajst-32001	57	22	implementation	implementation	NOUN
ajst-32001	57	23	in	in	ADP
ajst-32001	57	24	humanoid	humanoid	NOUN
ajst-32001	57	25	robots	robot	NOUN
ajst-32001	57	26	.	.	PUNCT
ajst-32001	58	1	table	table	NOUN
ajst-32001	58	2	1	1	NUM
ajst-32001	59	1	[	[	X
ajst-32001	59	2	10	10	NUM
ajst-32001	59	3	]	]	PUNCT
ajst-32001	59	4	shows	show	VERB
ajst-32001	59	5	the	the	DET
ajst-32001	59	6	accuracy	accuracy	NOUN
ajst-32001	59	7	obtained	obtain	VERB
ajst-32001	59	8	from	from	ADP
ajst-32001	59	9	the	the	DET
ajst-32001	59	10	test	test	NOUN
ajst-32001	59	11	data	datum	NOUN
ajst-32001	59	12	.	.	PUNCT
ajst-32001	60	1	99	99	NUM
ajst-32001	60	2	table	table	NOUN
ajst-32001	60	3	1	1	NUM
ajst-32001	60	4	.	.	PUNCT
ajst-32001	61	1	emotion	emotion	NOUN
ajst-32001	61	2	recognition	recognition	NOUN
ajst-32001	61	3	adapted	adapt	VERB
ajst-32001	61	4	from	from	ADP
ajst-32001	61	5	[	[	X
ajst-32001	61	6	10	10	NUM
ajst-32001	61	7	]	]	X
ajst-32001	61	8	no	no	INTJ
ajst-32001	61	9	.	.	PUNCT
ajst-32001	62	1	emotion	emotion	NOUN
ajst-32001	62	2	model	model	NOUN
ajst-32001	62	3	a	a	DET
ajst-32001	62	4	(	(	PUNCT
ajst-32001	62	5	alexnet	alexnet	ADJ
ajst-32001	62	6	)	)	PUNCT
ajst-32001	62	7	500	500	NUM
ajst-32001	62	8	epochs	epoch	NOUN
ajst-32001	62	9	model	model	NOUN
ajst-32001	62	10	b	b	PROPN
ajst-32001	62	11	(	(	PUNCT
ajst-32001	62	12	vgg16	vgg16	PROPN
ajst-32001	62	13	)	)	PUNCT
ajst-32001	62	14	500	500	NUM
ajst-32001	62	15	epochs	epoch	NOUN
ajst-32001	62	16	model	model	NOUN
ajst-32001	62	17	c	c	PROPN
ajst-32001	62	18	(	(	PUNCT
ajst-32001	62	19	proposed	propose	VERB
ajst-32001	62	20	model	model	NOUN
ajst-32001	62	21	)	)	PUNCT
ajst-32001	62	22	500	500	NUM
ajst-32001	62	23	epochs	epoch	NOUN
ajst-32001	62	24	1	1	NUM
ajst-32001	62	25	surprise	surprise	NOUN
ajst-32001	62	26	recognized	recognize	VERB
ajst-32001	62	27	recognized	recognize	VERB
ajst-32001	62	28	recognized	recognize	VERB
ajst-32001	62	29	2	2	NUM
ajst-32001	62	30	angry	angry	ADJ
ajst-32001	62	31	unrecognized	unrecognize	VERB
ajst-32001	62	32	recognized	recognize	VERB
ajst-32001	62	33	recognized	recognize	VERB
ajst-32001	62	34	3	3	NUM
ajst-32001	62	35	neutral	neutral	ADJ
ajst-32001	62	36	recognized	recognize	VERB
ajst-32001	62	37	recognized	recognize	VERB
ajst-32001	62	38	recognized	recognize	VERB
ajst-32001	62	39	4	4	NUM
ajst-32001	62	40	sad	sad	ADJ
ajst-32001	62	41	recognized	recognize	VERB
ajst-32001	62	42	recognized	recognize	VERB
ajst-32001	62	43	recognized	recognize	VERB
ajst-32001	62	44	5	5	NUM
ajst-32001	62	45	smile	smile	NOUN
ajst-32001	62	46	recognized	recognize	VERB
ajst-32001	62	47	recognized	recognize	VERB
ajst-32001	62	48	recognized	recognize	VERB
ajst-32001	62	49	accuracy	accuracy	NOUN
ajst-32001	62	50	64	64	NUM
ajst-32001	62	51	%	%	NOUN
ajst-32001	62	52	82	82	NUM
ajst-32001	62	53	%	%	NOUN
ajst-32001	62	54	71	71	NUM
ajst-32001	62	55	%	%	NOUN
ajst-32001	62	56	after	after	ADP
ajst-32001	62	57	500	500	NUM
ajst-32001	62	58	rounds	round	NOUN
ajst-32001	62	59	of	of	ADP
ajst-32001	62	60	training	training	NOUN
ajst-32001	62	61	,	,	PUNCT
ajst-32001	62	62	the	the	DET
ajst-32001	62	63	accuracy	accuracy	NOUN
ajst-32001	62	64	rate	rate	NOUN
ajst-32001	62	65	of	of	ADP
ajst-32001	62	66	vgg16	vgg16	NOUN
ajst-32001	62	67	reaches	reach	VERB
ajst-32001	62	68	82	82	NUM
ajst-32001	62	69	%	%	NOUN
ajst-32001	62	70	,	,	PUNCT
ajst-32001	62	71	which	which	PRON
ajst-32001	62	72	is	be	AUX
ajst-32001	62	73	higher	high	ADJ
ajst-32001	62	74	than	than	ADP
ajst-32001	62	75	that	that	PRON
ajst-32001	62	76	of	of	ADP
ajst-32001	62	77	alexnet	alexnet	NOUN
ajst-32001	62	78	and	and	CCONJ
ajst-32001	62	79	the	the	DET
ajst-32001	62	80	proposed	propose	VERB
ajst-32001	62	81	model	model	NOUN
ajst-32001	62	82	,	,	PUNCT
ajst-32001	62	83	indicating	indicate	VERB
ajst-32001	62	84	that	that	SCONJ
ajst-32001	62	85	the	the	DET
ajst-32001	62	86	vgg16	vgg16	NOUN
ajst-32001	62	87	architecture	architecture	NOUN
ajst-32001	62	88	is	be	AUX
ajst-32001	62	89	more	more	ADV
ajst-32001	62	90	adaptable	adaptable	ADJ
ajst-32001	62	91	to	to	ADP
ajst-32001	62	92	the	the	DET
ajst-32001	62	93	speech	speech	NOUN
ajst-32001	62	94	emotion	emotion	NOUN
ajst-32001	62	95	recognition	recognition	NOUN
ajst-32001	62	96	task	task	NOUN
ajst-32001	62	97	.	.	PUNCT
ajst-32001	63	1	the	the	DET
ajst-32001	63	2	accuracy	accuracy	NOUN
ajst-32001	63	3	rate	rate	NOUN
ajst-32001	63	4	of	of	ADP
ajst-32001	63	5	alexnet	alexnet	NOUN
ajst-32001	63	6	is	be	AUX
ajst-32001	63	7	only	only	ADV
ajst-32001	63	8	64	64	NUM
ajst-32001	63	9	%	%	NOUN
ajst-32001	63	10	,	,	PUNCT
ajst-32001	63	11	which	which	PRON
ajst-32001	63	12	may	may	AUX
ajst-32001	63	13	be	be	AUX
ajst-32001	63	14	due	due	ADJ
ajst-32001	63	15	to	to	ADP
ajst-32001	63	16	the	the	DET
ajst-32001	63	17	relatively	relatively	ADV
ajst-32001	63	18	simple	simple	ADJ
ajst-32001	63	19	network	network	NOUN
ajst-32001	63	20	structure	structure	NOUN
ajst-32001	63	21	and	and	CCONJ
ajst-32001	63	22	insufficient	insufficient	ADJ
ajst-32001	63	23	ability	ability	NOUN
ajst-32001	63	24	to	to	PART
ajst-32001	63	25	extract	extract	VERB
ajst-32001	63	26	and	and	CCONJ
ajst-32001	63	27	distinguish	distinguish	VERB
ajst-32001	63	28	speech	speech	NOUN
ajst-32001	63	29	emotion	emotion	NOUN
ajst-32001	63	30	features.[10	features.[10	NOUN
ajst-32001	63	31	]	]	PUNCT
ajst-32001	63	32	the	the	DET
ajst-32001	63	33	accuracy	accuracy	NOUN
ajst-32001	63	34	rate	rate	NOUN
ajst-32001	63	35	of	of	ADP
ajst-32001	63	36	the	the	DET
ajst-32001	63	37	proposed	propose	VERB
ajst-32001	63	38	model	model	NOUN
ajst-32001	63	39	is	be	AUX
ajst-32001	63	40	lower	low	ADJ
ajst-32001	63	41	than	than	ADP
ajst-32001	63	42	that	that	PRON
ajst-32001	63	43	of	of	ADP
ajst-32001	63	44	model	model	NOUN
ajst-32001	63	45	b	b	PROPN
ajst-32001	63	46	,	,	PUNCT
ajst-32001	63	47	and	and	CCONJ
ajst-32001	63	48	there	there	PRON
ajst-32001	63	49	may	may	AUX
ajst-32001	63	50	be	be	AUX
ajst-32001	63	51	room	room	NOUN
ajst-32001	63	52	for	for	ADP
ajst-32001	63	53	optimization	optimization	NOUN
ajst-32001	63	54	in	in	ADP
ajst-32001	63	55	its	its	PRON
ajst-32001	63	56	network	network	NOUN
ajst-32001	63	57	design	design	NOUN
ajst-32001	63	58	(	(	PUNCT
ajst-32001	63	59	such	such	ADJ
ajst-32001	63	60	as	as	ADP
ajst-32001	63	61	structural	structural	ADJ
ajst-32001	63	62	complexity	complexity	NOUN
ajst-32001	63	63	,	,	PUNCT
ajst-32001	63	64	feature	feature	NOUN
ajst-32001	63	65	fusion	fusion	NOUN
ajst-32001	63	66	method	method	NOUN
ajst-32001	63	67	,	,	PUNCT
ajst-32001	63	68	etc	etc	X
ajst-32001	63	69	.	.	X
ajst-32001	63	70	)	)	PUNCT
ajst-32001	63	71	.	.	PUNCT
ajst-32001	64	1	all	all	ADV
ajst-32001	64	2	in	in	ADV
ajst-32001	64	3	all	all	PRON
ajst-32001	64	4	,	,	PUNCT
ajst-32001	64	5	in	in	ADP
ajst-32001	64	6	dwijayanti	dwijayanti	PROPN
ajst-32001	64	7	et	et	PROPN
ajst-32001	64	8	al	al	PROPN
ajst-32001	64	9	.	.	PROPN
ajst-32001	64	10	's	's	PART
ajst-32001	64	11	humanoid	humanoid	NOUN
ajst-32001	64	12	robot	robot	NOUN
ajst-32001	64	13	experiment	experiment	NOUN
ajst-32001	64	14	[	[	X
ajst-32001	64	15	10	10	NUM
ajst-32001	64	16	]	]	PUNCT
ajst-32001	64	17	,	,	PUNCT
ajst-32001	64	18	vgg16	vgg16	PROPN
ajst-32001	64	19	showed	show	VERB
ajst-32001	64	20	the	the	DET
ajst-32001	64	21	optimal	optimal	ADJ
ajst-32001	64	22	recognition	recognition	NOUN
ajst-32001	64	23	performance	performance	NOUN
ajst-32001	64	24	(	(	PUNCT
ajst-32001	64	25	82	82	NUM
ajst-32001	64	26	%	%	NOUN
ajst-32001	64	27	accuracy	accuracy	NOUN
ajst-32001	64	28	)	)	PUNCT
ajst-32001	64	29	.	.	PUNCT
ajst-32001	65	1	vgg16	vgg16	NOUN
ajst-32001	65	2	excels	excel	NOUN
ajst-32001	65	3	in	in	ADP
ajst-32001	65	4	recognizing	recognize	VERB
ajst-32001	65	5	faces	face	NOUN
ajst-32001	65	6	and	and	CCONJ
ajst-32001	65	7	emotions	emotion	NOUN
ajst-32001	65	8	,	,	PUNCT
ajst-32001	65	9	making	make	VERB
ajst-32001	65	10	it	it	PRON
ajst-32001	65	11	suitable	suitable	ADJ
ajst-32001	65	12	for	for	ADP
ajst-32001	65	13	implementation	implementation	NOUN
ajst-32001	65	14	in	in	ADP
ajst-32001	65	15	humanoid	humanoid	NOUN
ajst-32001	65	16	robots	robot	NOUN
ajst-32001	65	17	.	.	PUNCT
ajst-32001	66	1	[	[	X
ajst-32001	66	2	10	10	NUM
ajst-32001	66	3	]	]	SYM
ajst-32001	66	4	3.2	3.2	NUM
ajst-32001	66	5	image	image	NOUN
ajst-32001	66	6	recognition	recognition	NOUN
ajst-32001	66	7	technology	technology	NOUN
ajst-32001	66	8	nvidia	nvidia	PROPN
ajst-32001	66	9	's	's	PART
ajst-32001	66	10	jetson	jetson	PROPN
ajst-32001	66	11	nano	nano	NOUN
ajst-32001	66	12	device	device	NOUN
ajst-32001	66	13	has	have	VERB
ajst-32001	66	14	three	three	NUM
ajst-32001	66	15	functions	function	NOUN
ajst-32001	66	16	:	:	PUNCT
ajst-32001	66	17	detect	detect	VERB
ajst-32001	66	18	human	human	ADJ
ajst-32001	66	19	faces	face	NOUN
ajst-32001	66	20	,	,	PUNCT
ajst-32001	66	21	recognize	recognize	VERB
ajst-32001	66	22	human	human	ADJ
ajst-32001	66	23	faces	face	NOUN
ajst-32001	66	24	,	,	PUNCT
ajst-32001	66	25	and	and	CCONJ
ajst-32001	66	26	identify	identify	VERB
ajst-32001	66	27	facial	facial	ADJ
ajst-32001	66	28	expressions	expression	NOUN
ajst-32001	66	29	.	.	PUNCT
ajst-32001	67	1	among	among	ADP
ajst-32001	67	2	them	they	PRON
ajst-32001	67	3	,	,	PUNCT
ajst-32001	67	4	face	face	NOUN
ajst-32001	67	5	detection	detection	NOUN
ajst-32001	67	6	is	be	AUX
ajst-32001	67	7	done	do	VERB
ajst-32001	67	8	by	by	ADP
ajst-32001	67	9	the	the	DET
ajst-32001	67	10	deep	deep	ADJ
ajst-32001	67	11	learning	learning	NOUN
ajst-32001	67	12	-	-	PUNCT
ajst-32001	67	13	based	base	VERB
ajst-32001	67	14	dnn	dnn	PROPN
ajst-32001	67	15	face	face	NOUN
ajst-32001	67	16	detector	detector	NOUN
ajst-32001	67	17	in	in	ADP
ajst-32001	67	18	opencv	opencv	PROPN
ajst-32001	67	19	.	.	PUNCT
ajst-32001	68	1	this	this	DET
ajst-32001	68	2	detector	detector	NOUN
ajst-32001	68	3	uses	use	VERB
ajst-32001	68	4	the	the	DET
ajst-32001	68	5	resnet	resnet	ADJ
ajst-32001	68	6	architecture	architecture	NOUN
ajst-32001	68	7	and	and	CCONJ
ajst-32001	68	8	is	be	AUX
ajst-32001	68	9	much	much	ADV
ajst-32001	68	10	more	more	ADV
ajst-32001	68	11	accurate	accurate	ADJ
ajst-32001	68	12	than	than	ADP
ajst-32001	68	13	previous	previous	ADJ
ajst-32001	68	14	models	model	NOUN
ajst-32001	68	15	.	.	PUNCT
ajst-32001	69	1	with	with	ADP
ajst-32001	69	2	the	the	DET
ajst-32001	69	3	support	support	NOUN
ajst-32001	69	4	of	of	ADP
ajst-32001	69	5	the	the	DET
ajst-32001	69	6	aforementioned	aforementioned	ADJ
ajst-32001	69	7	hardware	hardware	NOUN
ajst-32001	69	8	,	,	PUNCT
ajst-32001	69	9	even	even	ADV
ajst-32001	69	10	if	if	SCONJ
ajst-32001	69	11	the	the	DET
ajst-32001	69	12	lighting	lighting	NOUN
ajst-32001	69	13	and	and	CCONJ
ajst-32001	69	14	shooting	shoot	VERB
ajst-32001	69	15	angles	angle	NOUN
ajst-32001	69	16	change	change	NOUN
ajst-32001	69	17	,	,	PUNCT
ajst-32001	69	18	the	the	DET
ajst-32001	69	19	calculation	calculation	NOUN
ajst-32001	69	20	results	result	NOUN
ajst-32001	69	21	of	of	ADP
ajst-32001	69	22	the	the	DET
ajst-32001	69	23	opencv	opencv	PROPN
ajst-32001	69	24	framework	framework	NOUN
ajst-32001	69	25	library	library	NOUN
ajst-32001	69	26	can	can	AUX
ajst-32001	69	27	still	still	ADV
ajst-32001	69	28	maintain	maintain	VERB
ajst-32001	69	29	reliable	reliable	ADJ
ajst-32001	69	30	accuracy	accuracy	NOUN
ajst-32001	69	31	.	.	PUNCT
ajst-32001	70	1	as	as	ADP
ajst-32001	70	2	for	for	ADP
ajst-32001	70	3	face	face	NOUN
ajst-32001	70	4	recognition	recognition	NOUN
ajst-32001	70	5	,	,	PUNCT
ajst-32001	70	6	the	the	DET
ajst-32001	70	7	deep	deep	ADJ
ajst-32001	70	8	metric	metric	ADJ
ajst-32001	70	9	learning	learning	NOUN
ajst-32001	70	10	method	method	NOUN
ajst-32001	70	11	of	of	ADP
ajst-32001	70	12	opencv	opencv	PROPN
ajst-32001	70	13	and	and	CCONJ
ajst-32001	70	14	resnet-34	resnet-34	PROPN
ajst-32001	70	15	architecture	architecture	NOUN
ajst-32001	70	16	is	be	AUX
ajst-32001	70	17	used.[6	used.[6	ADP
ajst-32001	70	18	]	]	X
ajst-32001	70	19	facial	facial	ADJ
ajst-32001	70	20	expression	expression	NOUN
ajst-32001	70	21	recognition	recognition	NOUN
ajst-32001	70	22	is	be	AUX
ajst-32001	70	23	achieved	achieve	VERB
ajst-32001	70	24	by	by	ADP
ajst-32001	70	25	analyzing	analyze	VERB
ajst-32001	70	26	the	the	DET
ajst-32001	70	27	eye	eye	NOUN
ajst-32001	70	28	and	and	CCONJ
ajst-32001	70	29	mouth	mouth	NOUN
ajst-32001	70	30	regions	region	NOUN
ajst-32001	70	31	:	:	PUNCT
ajst-32001	70	32	extract	extract	VERB
ajst-32001	70	33	features	feature	NOUN
ajst-32001	70	34	of	of	ADP
ajst-32001	70	35	key	key	ADJ
ajst-32001	70	36	parts	part	NOUN
ajst-32001	70	37	fuse	fuse	NOUN
ajst-32001	70	38	to	to	PART
ajst-32001	70	39	generate	generate	VERB
ajst-32001	70	40	a	a	DET
ajst-32001	70	41	new	new	ADJ
ajst-32001	70	42	image	image	NOUN
ajst-32001	70	43	match	match	VERB
ajst-32001	70	44	the	the	DET
ajst-32001	70	45	seven	seven	NUM
ajst-32001	70	46	basic	basic	ADJ
ajst-32001	70	47	expressions	expression	NOUN
ajst-32001	70	48	(	(	PUNCT
ajst-32001	70	49	the	the	DET
ajst-32001	70	50	seven	seven	NUM
ajst-32001	70	51	basic	basic	ADJ
ajst-32001	70	52	expressions	expression	NOUN
ajst-32001	70	53	include	include	VERB
ajst-32001	70	54	:	:	PUNCT
ajst-32001	70	55	fear	fear	NOUN
ajst-32001	70	56	,	,	PUNCT
ajst-32001	70	57	anger	anger	NOUN
ajst-32001	70	58	,	,	PUNCT
ajst-32001	70	59	disgust	disgust	NOUN
ajst-32001	70	60	,	,	PUNCT
ajst-32001	70	61	happy	happy	ADJ
ajst-32001	70	62	,	,	PUNCT
ajst-32001	70	63	sad	sad	ADJ
ajst-32001	70	64	,	,	PUNCT
ajst-32001	70	65	surprise	surprise	NOUN
ajst-32001	70	66	,	,	PUNCT
ajst-32001	70	67	neutra	neutra	PROPN
ajst-32001	70	68	)	)	PUNCT
ajst-32001	70	69	.	.	PUNCT
ajst-32001	71	1	its	its	PRON
ajst-32001	71	2	advantage	advantage	NOUN
ajst-32001	71	3	lies	lie	VERB
ajst-32001	71	4	in	in	ADP
ajst-32001	71	5	that	that	SCONJ
ajst-32001	71	6	jetson	jetson	PROPN
ajst-32001	71	7	nano	nano	PROPN
ajst-32001	71	8	is	be	AUX
ajst-32001	71	9	a	a	DET
ajst-32001	71	10	powerful	powerful	ADJ
ajst-32001	71	11	platform	platform	NOUN
ajst-32001	71	12	with	with	ADP
ajst-32001	71	13	low	low	ADJ
ajst-32001	71	14	power	power	NOUN
ajst-32001	71	15	consumption	consumption	NOUN
ajst-32001	71	16	,	,	PUNCT
ajst-32001	71	17	which	which	PRON
ajst-32001	71	18	can	can	AUX
ajst-32001	71	19	easily	easily	ADV
ajst-32001	71	20	perform	perform	VERB
ajst-32001	71	21	intensive	intensive	ADJ
ajst-32001	71	22	computing	computing	NOUN
ajst-32001	71	23	of	of	ADP
ajst-32001	71	24	algorithms	algorithm	NOUN
ajst-32001	71	25	and	and	CCONJ
ajst-32001	71	26	contribute	contribute	VERB
ajst-32001	71	27	to	to	ADP
ajst-32001	71	28	high	high	ADJ
ajst-32001	71	29	video	video	NOUN
ajst-32001	71	30	processing	processing	NOUN
ajst-32001	71	31	frames.[6	frames.[6	NOUN
ajst-32001	71	32	]	]	X
ajst-32001	71	33	4	4	X
ajst-32001	71	34	.	.	X
ajst-32001	71	35	optimization	optimization	NOUN
ajst-32001	71	36	based	base	VERB
ajst-32001	71	37	on	on	ADP
ajst-32001	71	38	speech	speech	NOUN
ajst-32001	71	39	emotion	emotion	NOUN
ajst-32001	71	40	recognition	recognition	NOUN
ajst-32001	71	41	technology	technology	NOUN
ajst-32001	71	42	speech	speech	NOUN
ajst-32001	71	43	emotion	emotion	NOUN
ajst-32001	71	44	recognition	recognition	NOUN
ajst-32001	71	45	has	have	AUX
ajst-32001	71	46	evolved	evolve	VERB
ajst-32001	71	47	from	from	ADP
ajst-32001	71	48	a	a	DET
ajst-32001	71	49	specialized	specialized	ADJ
ajst-32001	71	50	area	area	NOUN
ajst-32001	71	51	to	to	ADP
ajst-32001	71	52	an	an	DET
ajst-32001	71	53	essential	essential	ADJ
ajst-32001	71	54	component	component	NOUN
ajst-32001	71	55	of	of	ADP
ajst-32001	71	56	human	human	ADJ
ajst-32001	71	57	-	-	PUNCT
ajst-32001	71	58	computer	computer	NOUN
ajst-32001	71	59	interaction	interaction	NOUN
ajst-32001	71	60	(	(	PUNCT
ajst-32001	71	61	hci	hci	NOUN
ajst-32001	71	62	)	)	PUNCT
ajst-32001	71	63	.	.	PUNCT
ajst-32001	72	1	[	[	X
ajst-32001	72	2	2,11,12	2,11,12	NUM
ajst-32001	72	3	]	]	PUNCT
ajst-32001	72	4	voice	voice	NOUN
ajst-32001	72	5	interaction	interaction	NOUN
ajst-32001	72	6	facilitates	facilitate	VERB
ajst-32001	72	7	more	more	ADJ
ajst-32001	72	8	natural	natural	ADJ
ajst-32001	72	9	communication	communication	NOUN
ajst-32001	72	10	between	between	ADP
ajst-32001	72	11	humans	human	NOUN
ajst-32001	72	12	and	and	CCONJ
ajst-32001	72	13	machines	machine	NOUN
ajst-32001	72	14	,	,	PUNCT
ajst-32001	72	15	which	which	PRON
ajst-32001	72	16	is	be	AUX
ajst-32001	72	17	different	different	ADJ
ajst-32001	72	18	from	from	ADP
ajst-32001	72	19	the	the	DET
ajst-32001	72	20	way	way	NOUN
ajst-32001	72	21	traditional	traditional	ADJ
ajst-32001	72	22	devices	device	NOUN
ajst-32001	72	23	are	be	AUX
ajst-32001	72	24	used	use	VERB
ajst-32001	72	25	for	for	ADP
ajst-32001	72	26	input	input	NOUN
ajst-32001	72	27	to	to	PART
ajst-32001	72	28	make	make	VERB
ajst-32001	72	29	machines	machine	NOUN
ajst-32001	72	30	understand	understand	VERB
ajst-32001	72	31	verbal	verbal	ADJ
ajst-32001	72	32	content	content	NOUN
ajst-32001	72	33	.	.	PUNCT
ajst-32001	73	1	moreover	moreover	ADV
ajst-32001	73	2	,	,	PUNCT
ajst-32001	73	3	voice	voice	NOUN
ajst-32001	73	4	interaction	interaction	NOUN
ajst-32001	73	5	makes	make	VERB
ajst-32001	73	6	it	it	PRON
ajst-32001	73	7	easier	easy	ADJ
ajst-32001	73	8	for	for	SCONJ
ajst-32001	73	9	human	human	ADJ
ajst-32001	73	10	listeners	listener	NOUN
ajst-32001	73	11	to	to	PART
ajst-32001	73	12	respond	respond	VERB
ajst-32001	73	13	and	and	CCONJ
ajst-32001	73	14	comprehend	comprehend	VERB
ajst-32001	73	15	.	.	PUNCT
ajst-32001	74	1	process	process	NOUN
ajst-32001	74	2	speech	speech	NOUN
ajst-32001	74	3	signals	signal	NOUN
ajst-32001	74	4	to	to	PART
ajst-32001	74	5	identify	identify	VERB
ajst-32001	74	6	underlying	underlying	ADJ
ajst-32001	74	7	emotions	emotion	NOUN
ajst-32001	74	8	,	,	PUNCT
ajst-32001	74	9	understand	understand	VERB
ajst-32001	74	10	the	the	DET
ajst-32001	74	11	emotions	emotion	NOUN
ajst-32001	74	12	present	present	ADJ
ajst-32001	74	13	in	in	ADP
ajst-32001	74	14	speech	speech	NOUN
ajst-32001	74	15	,	,	PUNCT
ajst-32001	74	16	and	and	CCONJ
ajst-32001	74	17	synthesize	synthesize	VERB
ajst-32001	74	18	the	the	DET
ajst-32001	74	19	emotions	emotion	NOUN
ajst-32001	74	20	required	require	VERB
ajst-32001	74	21	in	in	ADP
ajst-32001	74	22	speech	speech	NOUN
ajst-32001	74	23	based	base	VERB
ajst-32001	74	24	on	on	ADP
ajst-32001	74	25	expected	expect	VERB
ajst-32001	74	26	information	information	NOUN
ajst-32001	74	27	.	.	PUNCT
ajst-32001	75	1	it	it	PRON
ajst-32001	75	2	’s	’	VERB
ajst-32001	75	3	important	important	ADJ
ajst-32001	75	4	to	to	PART
ajst-32001	75	5	establish	establish	VERB
ajst-32001	75	6	an	an	DET
ajst-32001	75	7	appropriate	appropriate	ADJ
ajst-32001	75	8	emotional	emotional	ADJ
ajst-32001	75	9	language	language	NOUN
ajst-32001	75	10	database	database	NOUN
ajst-32001	75	11	.	.	PUNCT
ajst-32001	76	1	by	by	ADP
ajst-32001	76	2	analyzing	analyze	VERB
ajst-32001	76	3	the	the	DET
ajst-32001	76	4	impact	impact	NOUN
ajst-32001	76	5	of	of	ADP
ajst-32001	76	6	text	text	NOUN
ajst-32001	76	7	information	information	NOUN
ajst-32001	76	8	on	on	ADP
ajst-32001	76	9	emotional	emotional	ADJ
ajst-32001	76	10	expression	expression	NOUN
ajst-32001	76	11	,	,	PUNCT
ajst-32001	76	12	there	there	PRON
ajst-32001	76	13	are	be	VERB
ajst-32001	76	14	three	three	NUM
ajst-32001	76	15	types	type	NOUN
ajst-32001	76	16	of	of	ADP
ajst-32001	76	17	databases	database	NOUN
ajst-32001	76	18	for	for	ADP
ajst-32001	76	19	better	well	ADJ
ajst-32001	76	20	emotion	emotion	NOUN
ajst-32001	76	21	recognition	recognition	NOUN
ajst-32001	76	22	:	:	PUNCT
ajst-32001	76	23	actor	actor	NOUN
ajst-32001	76	24	-	-	PUNCT
ajst-32001	76	25	based	base	VERB
ajst-32001	76	26	(	(	PUNCT
ajst-32001	76	27	simulated	simulated	ADJ
ajst-32001	76	28	)	)	PUNCT
ajst-32001	76	29	emotional	emotional	ADJ
ajst-32001	76	30	speech	speech	NOUN
ajst-32001	76	31	databases	database	NOUN
ajst-32001	76	32	,	,	PUNCT
ajst-32001	76	33	induced	induce	VERB
ajst-32001	76	34	emotional	emotional	ADJ
ajst-32001	76	35	speech	speech	NOUN
ajst-32001	76	36	databases	database	NOUN
ajst-32001	76	37	,	,	PUNCT
ajst-32001	76	38	and	and	CCONJ
ajst-32001	76	39	natural	natural	ADJ
ajst-32001	76	40	emotional	emotional	ADJ
ajst-32001	76	41	speech	speech	NOUN
ajst-32001	76	42	databases.[13	databases.[13	PROPN
ajst-32001	76	43	]	]	PUNCT
ajst-32001	76	44	pattern	pattern	NOUN
ajst-32001	76	45	recognizers	recognizer	NOUN
ajst-32001	76	46	for	for	ADP
ajst-32001	76	47	language	language	NOUN
ajst-32001	76	48	emotions	emotion	NOUN
ajst-32001	76	49	:	:	PUNCT
ajst-32001	76	50	100	100	NUM
ajst-32001	76	51	1	1	NUM
ajst-32001	76	52	.	.	PUNCT
ajst-32001	77	1	linear	linear	PROPN
ajst-32001	77	2	classifier	classifier	NOUN
ajst-32001	77	3	2	2	NUM
ajst-32001	77	4	.	.	PUNCT
ajst-32001	77	5	nonlinear	nonlinear	ADJ
ajst-32001	77	6	classifier[13	classifier[13	PROPN
ajst-32001	77	7	]	]	PUNCT
ajst-32001	77	8	in	in	ADP
ajst-32001	77	9	speech	speech	NOUN
ajst-32001	77	10	emotion	emotion	NOUN
ajst-32001	77	11	classification	classification	NOUN
ajst-32001	77	12	,	,	PUNCT
ajst-32001	77	13	the	the	DET
ajst-32001	77	14	selection	selection	NOUN
ajst-32001	77	15	logic	logic	NOUN
ajst-32001	77	16	between	between	ADP
ajst-32001	77	17	linear	linear	PROPN
ajst-32001	77	18	classifiers	classifier	NOUN
ajst-32001	77	19	and	and	CCONJ
ajst-32001	77	20	nonlinear	nonlinear	ADJ
ajst-32001	77	21	classifiers	classifier	NOUN
ajst-32001	77	22	is	be	AUX
ajst-32001	77	23	similar	similar	ADJ
ajst-32001	77	24	to	to	ADP
ajst-32001	77	25	that	that	PRON
ajst-32001	77	26	in	in	ADP
ajst-32001	77	27	speech	speech	NOUN
ajst-32001	77	28	emotion	emotion	NOUN
ajst-32001	77	29	recognition	recognition	NOUN
ajst-32001	77	30	(	(	PUNCT
ajst-32001	77	31	both	both	PRON
ajst-32001	77	32	depend	depend	VERB
ajst-32001	77	33	on	on	ADP
ajst-32001	77	34	feature	feature	NOUN
ajst-32001	77	35	correlation	correlation	NOUN
ajst-32001	77	36	,	,	PUNCT
ajst-32001	77	37	task	task	NOUN
ajst-32001	77	38	complexity	complexity	NOUN
ajst-32001	77	39	,	,	PUNCT
ajst-32001	77	40	etc	etc	X
ajst-32001	77	41	.	.	X
ajst-32001	77	42	)	)	PUNCT
ajst-32001	77	43	.	.	PUNCT
ajst-32001	78	1	however	however	ADV
ajst-32001	78	2	,	,	PUNCT
ajst-32001	78	3	compared	compare	VERB
ajst-32001	78	4	with	with	ADP
ajst-32001	78	5	"	"	PUNCT
ajst-32001	78	6	emotion	emotion	NOUN
ajst-32001	78	7	,	,	PUNCT
ajst-32001	78	8	"	"	PUNCT
ajst-32001	78	9	"	"	PUNCT
ajst-32001	78	10	affect	affect	NOUN
ajst-32001	78	11	"	"	PUNCT
ajst-32001	78	12	emphasizes	emphasize	VERB
ajst-32001	78	13	more	more	ADV
ajst-32001	78	14	delicate	delicate	ADJ
ajst-32001	78	15	,	,	PUNCT
ajst-32001	78	16	complex	complex	ADJ
ajst-32001	78	17	,	,	PUNCT
ajst-32001	78	18	and	and	CCONJ
ajst-32001	78	19	context	context	NOUN
ajst-32001	78	20	-	-	PUNCT
ajst-32001	78	21	dependent	dependent	ADJ
ajst-32001	78	22	psychological	psychological	ADJ
ajst-32001	78	23	states	state	NOUN
ajst-32001	78	24	such	such	ADJ
ajst-32001	78	25	as	as	ADP
ajst-32001	78	26	"	"	PUNCT
ajst-32001	78	27	relief	relief	NOUN
ajst-32001	78	28	,	,	PUNCT
ajst-32001	78	29	"	"	PUNCT
ajst-32001	78	30	"	"	PUNCT
ajst-32001	78	31	grievance	grievance	NOUN
ajst-32001	78	32	,	,	PUNCT
ajst-32001	78	33	"	"	PUNCT
ajst-32001	78	34	and	and	CCONJ
ajst-32001	78	35	"	"	PUNCT
ajst-32001	78	36	jealousy	jealousy	NOUN
ajst-32001	78	37	.	.	PUNCT
ajst-32001	78	38	"	"	PUNCT
ajst-32001	79	1	nonlinear	nonlinear	ADJ
ajst-32001	79	2	correlations	correlation	NOUN
ajst-32001	79	3	between	between	ADP
ajst-32001	79	4	features	feature	NOUN
ajst-32001	79	5	are	be	AUX
ajst-32001	79	6	more	more	ADV
ajst-32001	79	7	common	common	ADJ
ajst-32001	79	8	,	,	PUNCT
ajst-32001	79	9	so	so	SCONJ
ajst-32001	79	10	the	the	DET
ajst-32001	79	11	selection	selection	NOUN
ajst-32001	79	12	of	of	ADP
ajst-32001	79	13	classifiers	classifier	NOUN
ajst-32001	79	14	needs	need	VERB
ajst-32001	79	15	to	to	PART
ajst-32001	79	16	focus	focus	VERB
ajst-32001	79	17	more	more	ADV
ajst-32001	79	18	on	on	ADP
ajst-32001	79	19	capturing	capture	VERB
ajst-32001	79	20	"	"	PUNCT
ajst-32001	79	21	subtle	subtle	ADJ
ajst-32001	79	22	nonlinear	nonlinear	ADJ
ajst-32001	79	23	patterns	pattern	NOUN
ajst-32001	79	24	.	.	PUNCT
ajst-32001	79	25	"	"	PUNCT
ajst-32001	80	1	speech	speech	NOUN
ajst-32001	80	2	emotion	emotion	NOUN
ajst-32001	80	3	recognition	recognition	NOUN
ajst-32001	80	4	using	use	VERB
ajst-32001	80	5	deep	deep	ADJ
ajst-32001	80	6	learning	learning	NOUN
ajst-32001	80	7	technology	technology	NOUN
ajst-32001	80	8	:	:	PUNCT
ajst-32001	80	9	traditional	traditional	ADJ
ajst-32001	80	10	ser	ser	NOUN
ajst-32001	80	11	technologies	technology	NOUN
ajst-32001	80	12	:	:	PUNCT
ajst-32001	80	13	to	to	PART
ajst-32001	80	14	achieve	achieve	VERB
ajst-32001	80	15	speech	speech	NOUN
ajst-32001	80	16	emotion	emotion	NOUN
ajst-32001	80	17	recognition	recognition	NOUN
ajst-32001	80	18	,	,	PUNCT
ajst-32001	80	19	traditional	traditional	ADJ
ajst-32001	80	20	ser	ser	NOUN
ajst-32001	80	21	relies	rely	VERB
ajst-32001	80	22	on	on	ADP
ajst-32001	80	23	manually	manually	ADV
ajst-32001	80	24	designed	design	VERB
ajst-32001	80	25	features	feature	NOUN
ajst-32001	80	26	(	(	PUNCT
ajst-32001	80	27	such	such	ADJ
ajst-32001	80	28	as	as	ADP
ajst-32001	80	29	spectrum	spectrum	NOUN
ajst-32001	80	30	,	,	PUNCT
ajst-32001	80	31	prosody	prosody	NOUN
ajst-32001	80	32	)	)	PUNCT
ajst-32001	80	33	and	and	CCONJ
ajst-32001	80	34	statistical	statistical	ADJ
ajst-32001	80	35	model	model	NOUN
ajst-32001	80	36	classification	classification	NOUN
ajst-32001	80	37	[	[	X
ajst-32001	80	38	13	13	NUM
ajst-32001	80	39	,	,	PUNCT
ajst-32001	80	40	14	14	NUM
ajst-32001	80	41	]	]	PUNCT
ajst-32001	80	42	,	,	PUNCT
ajst-32001	80	43	and	and	CCONJ
ajst-32001	80	44	its	its	PRON
ajst-32001	80	45	process	process	NOUN
ajst-32001	80	46	includes	include	VERB
ajst-32001	80	47	:	:	PUNCT
ajst-32001	80	48	preprocessing	preprocesse	VERB
ajst-32001	80	49	→	→	SYM
ajst-32001	80	50	feature	feature	NOUN
ajst-32001	80	51	engineering	engineering	NOUN
ajst-32001	80	52	→	→	SYM
ajst-32001	80	53	model	model	NOUN
ajst-32001	80	54	inference	inference	NOUN
ajst-32001	80	55	.	.	PUNCT
ajst-32001	81	1	the	the	DET
ajst-32001	81	2	core	core	NOUN
ajst-32001	81	3	of	of	ADP
ajst-32001	81	4	this	this	DET
ajst-32001	81	5	process	process	NOUN
ajst-32001	81	6	lies	lie	VERB
ajst-32001	81	7	in	in	ADP
ajst-32001	81	8	its	its	PRON
ajst-32001	81	9	reliance	reliance	NOUN
ajst-32001	81	10	on	on	ADP
ajst-32001	81	11	manually	manually	ADV
ajst-32001	81	12	designed	design	VERB
ajst-32001	81	13	features	feature	NOUN
ajst-32001	81	14	.	.	PUNCT
ajst-32001	82	1	it	it	PRON
ajst-32001	82	2	requires	require	VERB
ajst-32001	82	3	manual	manual	ADJ
ajst-32001	82	4	design	design	NOUN
ajst-32001	82	5	of	of	ADP
ajst-32001	82	6	temporal	temporal	ADJ
ajst-32001	82	7	features	feature	NOUN
ajst-32001	82	8	,	,	PUNCT
ajst-32001	82	9	resulting	result	VERB
ajst-32001	82	10	in	in	ADP
ajst-32001	82	11	limited	limited	ADJ
ajst-32001	82	12	modeling	modeling	NOUN
ajst-32001	82	13	capabilities	capability	NOUN
ajst-32001	82	14	.	.	PUNCT
ajst-32001	83	1	moreover	moreover	ADV
ajst-32001	83	2	,	,	PUNCT
ajst-32001	83	3	it	it	PRON
ajst-32001	83	4	is	be	AUX
ajst-32001	83	5	more	more	ADV
ajst-32001	83	6	sensitive	sensitive	ADJ
ajst-32001	83	7	to	to	ADP
ajst-32001	83	8	noise	noise	NOUN
ajst-32001	83	9	and	and	CCONJ
ajst-32001	83	10	individual	individual	ADJ
ajst-32001	83	11	differences	difference	NOUN
ajst-32001	83	12	,	,	PUNCT
ajst-32001	83	13	with	with	ADP
ajst-32001	83	14	restricted	restrict	VERB
ajst-32001	83	15	generalization	generalization	NOUN
ajst-32001	83	16	performance.[13	performance.[13	NOUN
ajst-32001	83	17	]	]	PUNCT
ajst-32001	83	18	the	the	DET
ajst-32001	83	19	simplified	simplified	ADJ
ajst-32001	83	20	structure	structure	NOUN
ajst-32001	83	21	of	of	ADP
ajst-32001	83	22	traditional	traditional	ADJ
ajst-32001	83	23	speech	speech	NOUN
ajst-32001	83	24	emotion	emotion	NOUN
ajst-32001	83	25	recognition	recognition	NOUN
ajst-32001	83	26	is	be	AUX
ajst-32001	83	27	figure	figure	NOUN
ajst-32001	83	28	5	5	NUM
ajst-32001	83	29	[	[	SYM
ajst-32001	83	30	13	13	NUM
ajst-32001	83	31	]	]	PUNCT
ajst-32001	83	32	shown	show	VERB
ajst-32001	83	33	below	below	ADV
ajst-32001	83	34	.	.	PUNCT
ajst-32001	84	1	figure	figure	NOUN
ajst-32001	84	2	5	5	NUM
ajst-32001	84	3	.	.	PUNCT
ajst-32001	84	4	traditional	traditional	ADJ
ajst-32001	84	5	speech	speech	NOUN
ajst-32001	84	6	emotion	emotion	NOUN
ajst-32001	84	7	recognition	recognition	NOUN
ajst-32001	85	1	[	[	X
ajst-32001	85	2	13	13	NUM
ajst-32001	85	3	]	]	PUNCT
ajst-32001	85	4	however	however	ADV
ajst-32001	85	5	,	,	PUNCT
ajst-32001	85	6	the	the	DET
ajst-32001	85	7	feature	feature	NOUN
ajst-32001	85	8	extraction	extraction	NOUN
ajst-32001	85	9	of	of	ADP
ajst-32001	85	10	traditional	traditional	ADJ
ajst-32001	85	11	ser	ser	NOUN
ajst-32001	85	12	relies	rely	VERB
ajst-32001	85	13	on	on	ADP
ajst-32001	85	14	manual	manual	ADJ
ajst-32001	85	15	experience	experience	NOUN
ajst-32001	85	16	,	,	PUNCT
ajst-32001	85	17	and	and	CCONJ
ajst-32001	85	18	this	this	DET
ajst-32001	85	19	limitation	limitation	NOUN
ajst-32001	85	20	has	have	AUX
ajst-32001	85	21	promoted	promote	VERB
ajst-32001	85	22	the	the	DET
ajst-32001	85	23	application	application	NOUN
ajst-32001	85	24	of	of	ADP
ajst-32001	85	25	deep	deep	ADJ
ajst-32001	85	26	learning	learning	NOUN
ajst-32001	85	27	models	model	NOUN
ajst-32001	85	28	.	.	PUNCT
ajst-32001	86	1	deep	deep	ADJ
ajst-32001	86	2	learning	learning	NOUN
ajst-32001	86	3	technology	technology	NOUN
ajst-32001	86	4	utilizes	utilize	VERB
ajst-32001	86	5	several	several	ADJ
ajst-32001	86	6	key	key	ADJ
ajst-32001	86	7	functions	function	NOUN
ajst-32001	86	8	in	in	ADP
ajst-32001	86	9	the	the	DET
ajst-32001	86	10	field	field	NOUN
ajst-32001	86	11	of	of	ADP
ajst-32001	86	12	speech	speech	NOUN
ajst-32001	86	13	emotion	emotion	NOUN
ajst-32001	86	14	recognition	recognition	NOUN
ajst-32001	86	15	technology	technology	NOUN
ajst-32001	86	16	,	,	PUNCT
ajst-32001	86	17	as	as	SCONJ
ajst-32001	86	18	shown	show	VERB
ajst-32001	86	19	in	in	ADP
ajst-32001	86	20	figure	figure	NOUN
ajst-32001	86	21	6	6	NUM
ajst-32001	86	22	.	.	PUNCT
ajst-32001	86	23	analysis	analysis	NOUN
ajst-32001	86	24	and	and	CCONJ
ajst-32001	86	25	comparison	comparison	NOUN
ajst-32001	86	26	of	of	ADP
ajst-32001	86	27	the	the	DET
ajst-32001	86	28	respective	respective	ADJ
ajst-32001	86	29	advantages	advantage	NOUN
ajst-32001	86	30	of	of	ADP
ajst-32001	86	31	deep	deep	ADJ
ajst-32001	86	32	boltzmann	boltzmann	PROPN
ajst-32001	86	33	machine	machine	NOUN
ajst-32001	86	34	(	(	PUNCT
ajst-32001	86	35	dbm	dbm	PROPN
ajst-32001	86	36	)	)	PUNCT
ajst-32001	86	37	,	,	PUNCT
ajst-32001	86	38	recurrent	recurrent	ADJ
ajst-32001	86	39	neural	neural	ADJ
ajst-32001	86	40	network	network	NOUN
ajst-32001	86	41	(	(	PUNCT
ajst-32001	86	42	rnn	rnn	PROPN
ajst-32001	86	43	)	)	PUNCT
ajst-32001	86	44	,	,	PUNCT
ajst-32001	86	45	recursive	recursive	ADJ
ajst-32001	86	46	neural	neural	ADJ
ajst-32001	86	47	network	network	NOUN
ajst-32001	86	48	(	(	PUNCT
ajst-32001	86	49	rvnn	rvnn	NOUN
ajst-32001	86	50	)	)	PUNCT
ajst-32001	86	51	,	,	PUNCT
ajst-32001	86	52	deep	deep	ADJ
ajst-32001	86	53	belief	belief	NOUN
ajst-32001	86	54	network	network	NOUN
ajst-32001	86	55	(	(	PUNCT
ajst-32001	86	56	dbn	dbn	PROPN
ajst-32001	86	57	)	)	PUNCT
ajst-32001	86	58	,	,	PUNCT
ajst-32001	86	59	etc	etc	X
ajst-32001	86	60	.	.	X
ajst-32001	86	61	in	in	ADP
ajst-32001	86	62	emotion	emotion	NOUN
ajst-32001	86	63	recognition	recognition	NOUN
ajst-32001	86	64	.	.	PUNCT
ajst-32001	87	1	figure	figure	NOUN
ajst-32001	87	2	.	.	PUNCT
ajst-32001	88	1	6	6	NUM
ajst-32001	88	2	overview	overview	NOUN
ajst-32001	88	3	of	of	ADP
ajst-32001	88	4	deep	deep	ADJ
ajst-32001	88	5	learning	learning	NOUN
ajst-32001	88	6	techniques	technique	NOUN
ajst-32001	88	7	and	and	CCONJ
ajst-32001	88	8	their	their	PRON
ajst-32001	88	9	key	key	ADJ
ajst-32001	88	10	features	feature	NOUN
ajst-32001	88	11	adapted	adapt	VERB
ajst-32001	88	12	from	from	ADP
ajst-32001	88	13	khalil	khalil	PROPN
ajst-32001	88	14	et	et	PROPN
ajst-32001	88	15	al	al	PROPN
ajst-32001	88	16	.	.	PUNCT
ajst-32001	89	1	[	[	X
ajst-32001	89	2	14	14	NUM
ajst-32001	89	3	]	]	X
ajst-32001	89	4	deep	deep	ADJ
ajst-32001	89	5	boltzmann	boltzmann	PROPN
ajst-32001	89	6	machine	machine	NOUN
ajst-32001	89	7	(	(	PUNCT
ajst-32001	89	8	dbm	dbm	PROPN
ajst-32001	89	9	):	):	PUNCT
ajst-32001	89	10	dbm	dbm	PROPN
ajst-32001	89	11	can	can	AUX
ajst-32001	89	12	be	be	AUX
ajst-32001	89	13	combined	combine	VERB
ajst-32001	89	14	with	with	ADP
ajst-32001	89	15	other	other	ADJ
ajst-32001	89	16	models	model	NOUN
ajst-32001	89	17	,	,	PUNCT
ajst-32001	89	18	using	use	VERB
ajst-32001	89	19	unsupervised	unsupervised	ADJ
ajst-32001	89	20	training	training	NOUN
ajst-32001	89	21	to	to	PART
ajst-32001	89	22	identify	identify	VERB
ajst-32001	89	23	various	various	ADJ
ajst-32001	89	24	speech	speech	NOUN
ajst-32001	89	25	emotions	emotion	NOUN
ajst-32001	89	26	.	.	PUNCT
ajst-32001	90	1	the	the	DET
ajst-32001	90	2	main	main	ADJ
ajst-32001	90	3	advantage	advantage	NOUN
ajst-32001	90	4	of	of	ADP
ajst-32001	90	5	dbm	dbm	PROPN
ajst-32001	90	6	is	be	AUX
ajst-32001	90	7	that	that	SCONJ
ajst-32001	90	8	it	it	PRON
ajst-32001	90	9	it	it	PRON
ajst-32001	90	10	learns	learn	VERB
ajst-32001	90	11	things	thing	NOUN
ajst-32001	90	12	quickly	quickly	ADV
ajst-32001	90	13	and	and	CCONJ
ajst-32001	90	14	can	can	AUX
ajst-32001	90	15	also	also	ADV
ajst-32001	90	16	provide	provide	VERB
ajst-32001	90	17	useful	useful	ADJ
ajst-32001	90	18	information	information	NOUN
ajst-32001	90	19	representation	representation	NOUN
ajst-32001	90	20	.	.	PUNCT
ajst-32001	91	1	however	however	ADV
ajst-32001	91	2	,	,	PUNCT
ajst-32001	91	3	dbm	dbm	PROPN
ajst-32001	91	4	also	also	ADV
ajst-32001	91	5	has	have	VERB
ajst-32001	91	6	some	some	DET
ajst-32001	91	7	disadvantages	disadvantage	NOUN
ajst-32001	91	8	,	,	PUNCT
ajst-32001	91	9	such	such	ADJ
ajst-32001	91	10	as	as	ADP
ajst-32001	91	11	limited	limited	ADJ
ajst-32001	91	12	effectiveness	effectiveness	NOUN
ajst-32001	91	13	in	in	ADP
ajst-32001	91	14	certain	certain	ADJ
ajst-32001	91	15	cases	case	NOUN
ajst-32001	91	16	.	.	PUNCT
ajst-32001	92	1	101	101	NUM
ajst-32001	92	2	recurrent	recurrent	ADJ
ajst-32001	92	3	neural	neural	ADJ
ajst-32001	92	4	network	network	NOUN
ajst-32001	92	5	(	(	PUNCT
ajst-32001	92	6	rnn	rnn	PROPN
ajst-32001	92	7	):	):	PUNCT
ajst-32001	92	8	rnns	rnn	NOUN
ajst-32001	92	9	are	be	AUX
ajst-32001	92	10	suitable	suitable	ADJ
ajst-32001	92	11	for	for	ADP
ajst-32001	92	12	speech	speech	NOUN
ajst-32001	92	13	emotion	emotion	NOUN
ajst-32001	92	14	recognition	recognition	NOUN
ajst-32001	92	15	due	due	ADP
ajst-32001	92	16	to	to	ADP
ajst-32001	92	17	their	their	PRON
ajst-32001	92	18	short	short	ADJ
ajst-32001	92	19	-	-	PUNCT
ajst-32001	92	20	time	time	NOUN
ajst-32001	92	21	frame	frame	NOUN
ajst-32001	92	22	segmentation	segmentation	NOUN
ajst-32001	92	23	of	of	ADP
ajst-32001	92	24	acoustic	acoustic	ADJ
ajst-32001	92	25	features.[15	features.[15	NOUN
ajst-32001	92	26	]	]	PUNCT
ajst-32001	92	27	however	however	ADV
ajst-32001	92	28	,	,	PUNCT
ajst-32001	92	29	when	when	SCONJ
ajst-32001	92	30	rnns	rnns	NOUN
ajst-32001	92	31	process	process	NOUN
ajst-32001	92	32	long	long	ADJ
ajst-32001	92	33	sequence	sequence	NOUN
ajst-32001	92	34	data	datum	NOUN
ajst-32001	92	35	,	,	PUNCT
ajst-32001	92	36	the	the	DET
ajst-32001	92	37	gradients	gradient	NOUN
ajst-32001	92	38	gradually	gradually	ADV
ajst-32001	92	39	decay	decay	VERB
ajst-32001	92	40	as	as	SCONJ
ajst-32001	92	41	they	they	PRON
ajst-32001	92	42	propagate	propagate	VERB
ajst-32001	92	43	over	over	ADP
ajst-32001	92	44	time	time	NOUN
ajst-32001	92	45	,	,	PUNCT
ajst-32001	92	46	resulting	result	VERB
ajst-32001	92	47	in	in	ADP
ajst-32001	92	48	a	a	DET
ajst-32001	92	49	decreased	decrease	VERB
ajst-32001	92	50	ability	ability	NOUN
ajst-32001	92	51	to	to	PART
ajst-32001	92	52	capture	capture	VERB
ajst-32001	92	53	long	long	ADJ
ajst-32001	92	54	-	-	PUNCT
ajst-32001	92	55	distance	distance	NOUN
ajst-32001	92	56	information.[14	information.[14	PROPN
ajst-32001	92	57	]	]	PUNCT
ajst-32001	92	58	recursive	recursive	ADJ
ajst-32001	92	59	neural	neural	ADJ
ajst-32001	92	60	network	network	NOUN
ajst-32001	92	61	(	(	PUNCT
ajst-32001	92	62	rvnn	rvnn	NOUN
ajst-32001	92	63	):	):	PUNCT
ajst-32001	92	64	rvnn	rvnn	NOUN
ajst-32001	92	65	is	be	AUX
ajst-32001	92	66	mainly	mainly	ADV
ajst-32001	92	67	used	use	VERB
ajst-32001	92	68	in	in	ADP
ajst-32001	92	69	natural	natural	ADJ
ajst-32001	92	70	language	language	NOUN
ajst-32001	92	71	processing	processing	NOUN
ajst-32001	92	72	,	,	PUNCT
ajst-32001	92	73	and	and	CCONJ
ajst-32001	92	74	its	its	PRON
ajst-32001	92	75	structure	structure	NOUN
ajst-32001	92	76	can	can	AUX
ajst-32001	92	77	handle	handle	VERB
ajst-32001	92	78	different	different	ADJ
ajst-32001	92	79	situations	situation	NOUN
ajst-32001	92	80	.	.	PUNCT
ajst-32001	93	1	studies	study	NOUN
ajst-32001	93	2	have	have	AUX
ajst-32001	93	3	found	find	VERB
ajst-32001	93	4	that	that	SCONJ
ajst-32001	93	5	such	such	ADJ
ajst-32001	93	6	networks	network	NOUN
ajst-32001	93	7	can	can	AUX
ajst-32001	93	8	not	not	PART
ajst-32001	93	9	only	only	ADV
ajst-32001	93	10	classify	classify	VERB
ajst-32001	93	11	sentences	sentence	NOUN
ajst-32001	93	12	in	in	ADP
ajst-32001	93	13	natural	natural	ADJ
ajst-32001	93	14	language	language	NOUN
ajst-32001	93	15	but	but	CCONJ
ajst-32001	93	16	also	also	ADV
ajst-32001	93	17	be	be	AUX
ajst-32001	93	18	applied	apply	VERB
ajst-32001	93	19	to	to	ADP
ajst-32001	93	20	natural	natural	ADJ
ajst-32001	93	21	image	image	NOUN
ajst-32001	93	22	processing.[14	processing.[14	PROPN
ajst-32001	93	23	]	]	PUNCT
ajst-32001	93	24	deep	deep	ADJ
ajst-32001	93	25	belief	belief	NOUN
ajst-32001	93	26	network	network	NOUN
ajst-32001	93	27	(	(	PUNCT
ajst-32001	93	28	dbn	dbn	PROPN
ajst-32001	93	29	):	):	PUNCT
ajst-32001	93	30	dbn	dbn	PROPN
ajst-32001	93	31	does	do	AUX
ajst-32001	93	32	not	not	PART
ajst-32001	93	33	require	require	VERB
ajst-32001	93	34	manual	manual	ADJ
ajst-32001	93	35	feature	feature	NOUN
ajst-32001	93	36	design	design	NOUN
ajst-32001	93	37	and	and	CCONJ
ajst-32001	93	38	can	can	AUX
ajst-32001	93	39	automatically	automatically	ADV
ajst-32001	93	40	learn	learn	VERB
ajst-32001	93	41	multi	multi	ADJ
ajst-32001	93	42	-	-	ADJ
ajst-32001	93	43	level	level	ADJ
ajst-32001	93	44	abstract	abstract	ADJ
ajst-32001	93	45	features	feature	NOUN
ajst-32001	93	46	from	from	ADP
ajst-32001	93	47	raw	raw	ADJ
ajst-32001	93	48	data	datum	NOUN
ajst-32001	93	49	.	.	PUNCT
ajst-32001	94	1	for	for	ADP
ajst-32001	94	2	example	example	NOUN
ajst-32001	94	3	,	,	PUNCT
ajst-32001	94	4	it	it	PRON
ajst-32001	94	5	can	can	AUX
ajst-32001	94	6	extract	extract	VERB
ajst-32001	94	7	spectrum	spectrum	NOUN
ajst-32001	94	8	,	,	PUNCT
ajst-32001	94	9	prosody	prosody	NOUN
ajst-32001	94	10	,	,	PUNCT
ajst-32001	94	11	and	and	CCONJ
ajst-32001	94	12	emotion	emotion	NOUN
ajst-32001	94	13	-	-	PUNCT
ajst-32001	94	14	related	relate	VERB
ajst-32001	94	15	features	feature	NOUN
ajst-32001	94	16	layer	layer	NOUN
ajst-32001	94	17	by	by	ADP
ajst-32001	94	18	layer	layer	NOUN
ajst-32001	94	19	from	from	ADP
ajst-32001	94	20	speech	speech	NOUN
ajst-32001	94	21	signals	signal	NOUN
ajst-32001	94	22	.	.	PUNCT
ajst-32001	95	1	deep	deep	ADJ
ajst-32001	95	2	belief	belief	NOUN
ajst-32001	95	3	networks	network	NOUN
ajst-32001	95	4	(	(	PUNCT
ajst-32001	95	5	dbns	dbns	PROPN
ajst-32001	95	6	)	)	PUNCT
ajst-32001	95	7	have	have	VERB
ajst-32001	95	8	two	two	NUM
ajst-32001	95	9	main	main	ADJ
ajst-32001	95	10	advantages	advantage	NOUN
ajst-32001	95	11	.	.	PUNCT
ajst-32001	96	1	the	the	DET
ajst-32001	96	2	first	first	ADJ
ajst-32001	96	3	is	be	AUX
ajst-32001	96	4	that	that	SCONJ
ajst-32001	96	5	during	during	ADP
ajst-32001	96	6	pre	pre	ADJ
ajst-32001	96	7	-	-	NOUN
ajst-32001	96	8	training	training	NOUN
ajst-32001	96	9	,	,	PUNCT
ajst-32001	96	10	they	they	PRON
ajst-32001	96	11	can	can	AUX
ajst-32001	96	12	perform	perform	VERB
ajst-32001	96	13	unsupervised	unsupervised	ADJ
ajst-32001	96	14	learning	learning	NOUN
ajst-32001	96	15	on	on	ADP
ajst-32001	96	16	large	large	ADJ
ajst-32001	96	17	-	-	PUNCT
ajst-32001	96	18	scale	scale	NOUN
ajst-32001	96	19	unlabeled	unlabele	VERB
ajst-32001	96	20	databases	database	NOUN
ajst-32001	96	21	.	.	PUNCT
ajst-32001	97	1	and	and	CCONJ
ajst-32001	97	2	they	they	PRON
ajst-32001	97	3	do	do	AUX
ajst-32001	97	4	n’t	not	PART
ajst-32001	97	5	need	need	VERB
ajst-32001	97	6	the	the	DET
ajst-32001	97	7	manual	manual	ADJ
ajst-32001	97	8	data	datum	NOUN
ajst-32001	97	9	labeling	labeling	NOUN
ajst-32001	97	10	.	.	PUNCT
ajst-32001	98	1	the	the	DET
ajst-32001	98	2	second	second	ADJ
ajst-32001	98	3	is	be	AUX
ajst-32001	98	4	that	that	SCONJ
ajst-32001	98	5	they	they	PRON
ajst-32001	98	6	can	can	AUX
ajst-32001	98	7	approximate	approximate	VERB
ajst-32001	98	8	the	the	DET
ajst-32001	98	9	required	require	VERB
ajst-32001	98	10	output	output	NOUN
ajst-32001	98	11	weights	weight	NOUN
ajst-32001	98	12	of	of	ADP
ajst-32001	98	13	variables	variable	NOUN
ajst-32001	98	14	during	during	ADP
ajst-32001	98	15	the	the	DET
ajst-32001	98	16	inference	inference	NOUN
ajst-32001	98	17	process.[14	process.[14	PROPN
ajst-32001	98	18	]	]	PUNCT
ajst-32001	98	19	some	some	DET
ajst-32001	98	20	limitations	limitation	NOUN
ajst-32001	98	21	still	still	ADV
ajst-32001	98	22	exist	exist	VERB
ajst-32001	98	23	,	,	PUNCT
ajst-32001	98	24	such	such	ADJ
ajst-32001	98	25	as	as	ADP
ajst-32001	98	26	the	the	DET
ajst-32001	98	27	inference	inference	NOUN
ajst-32001	98	28	process	process	NOUN
ajst-32001	98	29	of	of	ADP
ajst-32001	98	30	deep	deep	ADJ
ajst-32001	98	31	belief	belief	NOUN
ajst-32001	98	32	network	network	NOUN
ajst-32001	98	33	(	(	PUNCT
ajst-32001	98	34	dbn	dbn	PROPN
ajst-32001	98	35	)	)	PUNCT
ajst-32001	98	36	is	be	AUX
ajst-32001	98	37	limited	limit	VERB
ajst-32001	98	38	to	to	ADP
ajst-32001	98	39	bottom	bottom	ADJ
ajst-32001	98	40	-	-	PUNCT
ajst-32001	98	41	up	up	ADP
ajst-32001	98	42	transmission	transmission	NOUN
ajst-32001	98	43	.	.	PUNCT
ajst-32001	99	1	5	5	X
ajst-32001	99	2	.	.	X
ajst-32001	99	3	conclusion	conclusion	NOUN
ajst-32001	99	4	although	although	SCONJ
ajst-32001	99	5	the	the	DET
ajst-32001	99	6	optimization	optimization	NOUN
ajst-32001	99	7	of	of	ADP
ajst-32001	99	8	these	these	DET
ajst-32001	99	9	three	three	NUM
ajst-32001	99	10	types	type	NOUN
ajst-32001	99	11	of	of	ADP
ajst-32001	99	12	emotion	emotion	NOUN
ajst-32001	99	13	recognition	recognition	NOUN
ajst-32001	99	14	technologies	technology	NOUN
ajst-32001	99	15	focuses	focus	VERB
ajst-32001	99	16	on	on	ADP
ajst-32001	99	17	different	different	ADJ
ajst-32001	99	18	data	datum	NOUN
ajst-32001	99	19	modalities	modality	NOUN
ajst-32001	99	20	,	,	PUNCT
ajst-32001	99	21	they	they	PRON
ajst-32001	99	22	share	share	VERB
ajst-32001	99	23	the	the	DET
ajst-32001	99	24	same	same	ADJ
ajst-32001	99	25	core	core	NOUN
ajst-32001	99	26	goal	goal	NOUN
ajst-32001	99	27	:	:	PUNCT
ajst-32001	99	28	enhancing	enhance	VERB
ajst-32001	99	29	the	the	DET
ajst-32001	99	30	precision	precision	NOUN
ajst-32001	99	31	and	and	CCONJ
ajst-32001	99	32	robustness	robustness	NOUN
ajst-32001	99	33	of	of	ADP
ajst-32001	99	34	emotion	emotion	NOUN
ajst-32001	99	35	perception	perception	NOUN
ajst-32001	99	36	.	.	PUNCT
ajst-32001	100	1	from	from	ADP
ajst-32001	100	2	a	a	DET
ajst-32001	100	3	technical	technical	ADJ
ajst-32001	100	4	perspective	perspective	NOUN
ajst-32001	100	5	,	,	PUNCT
ajst-32001	100	6	these	these	DET
ajst-32001	100	7	optimizations	optimization	NOUN
ajst-32001	100	8	have	have	AUX
ajst-32001	100	9	driven	drive	VERB
ajst-32001	100	10	innovations	innovation	NOUN
ajst-32001	100	11	in	in	ADP
ajst-32001	100	12	methods	method	NOUN
ajst-32001	100	13	such	such	ADJ
ajst-32001	100	14	as	as	ADP
ajst-32001	100	15	multimodal	multimodal	ADJ
ajst-32001	100	16	fusion	fusion	NOUN
ajst-32001	100	17	,	,	PUNCT
ajst-32001	100	18	causal	causal	ADJ
ajst-32001	100	19	reasoning	reasoning	NOUN
ajst-32001	100	20	,	,	PUNCT
ajst-32001	100	21	and	and	CCONJ
ajst-32001	100	22	lightweight	lightweight	ADJ
ajst-32001	100	23	models	model	NOUN
ajst-32001	100	24	.	.	PUNCT
ajst-32001	101	1	in	in	ADP
ajst-32001	101	2	terms	term	NOUN
ajst-32001	101	3	of	of	ADP
ajst-32001	101	4	application	application	NOUN
ajst-32001	101	5	value	value	NOUN
ajst-32001	101	6	,	,	PUNCT
ajst-32001	101	7	their	their	PRON
ajst-32001	101	8	outcomes	outcome	NOUN
ajst-32001	101	9	have	have	AUX
ajst-32001	101	10	supported	support	VERB
ajst-32001	101	11	the	the	DET
ajst-32001	101	12	transformation	transformation	NOUN
ajst-32001	101	13	of	of	ADP
ajst-32001	101	14	service	service	NOUN
ajst-32001	101	15	models	model	NOUN
ajst-32001	101	16	from	from	ADP
ajst-32001	101	17	"	"	PUNCT
ajst-32001	101	18	generalized	generalize	VERB
ajst-32001	101	19	"	"	PUNCT
ajst-32001	101	20	to	to	AUX
ajst-32001	101	21	"	"	PUNCT
ajst-32001	101	22	personalized	personalize	VERB
ajst-32001	101	23	"	"	PUNCT
ajst-32001	101	24	and	and	CCONJ
ajst-32001	101	25	"	"	PUNCT
ajst-32001	101	26	precision	precision	NOUN
ajst-32001	101	27	-	-	PUNCT
ajst-32001	101	28	oriented	orient	VERB
ajst-32001	101	29	.	.	PUNCT
ajst-32001	101	30	"	"	PUNCT
ajst-32001	102	1	although	although	SCONJ
ajst-32001	102	2	deep	deep	ADJ
ajst-32001	102	3	learning	learning	NOUN
ajst-32001	102	4	has	have	AUX
ajst-32001	102	5	improved	improve	VERB
ajst-32001	102	6	recognition	recognition	NOUN
ajst-32001	102	7	accuracy	accuracy	NOUN
ajst-32001	102	8	,	,	PUNCT
ajst-32001	102	9	current	current	ADJ
ajst-32001	102	10	technologies	technology	NOUN
ajst-32001	102	11	still	still	ADV
ajst-32001	102	12	face	face	VERB
ajst-32001	102	13	three	three	NUM
ajst-32001	102	14	major	major	ADJ
ajst-32001	102	15	challenges	challenge	NOUN
ajst-32001	102	16	:	:	PUNCT
ajst-32001	102	17	cross	cross	ADJ
ajst-32001	102	18	-	-	ADJ
ajst-32001	102	19	modal	modal	ADJ
ajst-32001	102	20	emotional	emotional	ADJ
ajst-32001	102	21	conflicts	conflict	NOUN
ajst-32001	102	22	,	,	PUNCT
ajst-32001	102	23	insufficient	insufficient	ADJ
ajst-32001	102	24	support	support	NOUN
ajst-32001	102	25	for	for	ADP
ajst-32001	102	26	low	low	ADJ
ajst-32001	102	27	-	-	PUNCT
ajst-32001	102	28	resource	resource	NOUN
ajst-32001	102	29	languages	language	NOUN
ajst-32001	102	30	,	,	PUNCT
ajst-32001	102	31	and	and	CCONJ
ajst-32001	102	32	energy	energy	NOUN
ajst-32001	102	33	consumption	consumption	NOUN
ajst-32001	102	34	limitations	limitation	NOUN
ajst-32001	102	35	of	of	ADP
ajst-32001	102	36	real	real	ADJ
ajst-32001	102	37	-	-	PUNCT
ajst-32001	102	38	time	time	NOUN
ajst-32001	102	39	systems	system	NOUN
ajst-32001	102	40	.	.	PUNCT
ajst-32001	103	1	we	we	PRON
ajst-32001	103	2	hope	hope	VERB
ajst-32001	103	3	that	that	SCONJ
ajst-32001	103	4	future	future	ADJ
ajst-32001	103	5	research	research	NOUN
ajst-32001	103	6	and	and	CCONJ
ajst-32001	103	7	development	development	NOUN
ajst-32001	103	8	can	can	AUX
ajst-32001	103	9	overcome	overcome	VERB
ajst-32001	103	10	these	these	DET
ajst-32001	103	11	challenges	challenge	NOUN
ajst-32001	103	12	.	.	PUNCT
ajst-32001	104	1	we	we	PRON
ajst-32001	104	2	also	also	ADV
ajst-32001	104	3	hope	hope	VERB
ajst-32001	104	4	that	that	SCONJ
ajst-32001	104	5	they	they	PRON
ajst-32001	104	6	will	will	AUX
ajst-32001	104	7	play	play	VERB
ajst-32001	104	8	a	a	DET
ajst-32001	104	9	more	more	ADV
ajst-32001	104	10	critical	critical	ADJ
ajst-32001	104	11	role	role	NOUN
ajst-32001	104	12	in	in	ADP
ajst-32001	104	13	fields	field	NOUN
ajst-32001	104	14	such	such	ADJ
ajst-32001	104	15	as	as	ADP
ajst-32001	104	16	education	education	NOUN
ajst-32001	104	17	,	,	PUNCT
ajst-32001	104	18	medical	medical	ADJ
ajst-32001	104	19	care	care	NOUN
ajst-32001	104	20	,	,	PUNCT
ajst-32001	104	21	retail	retail	NOUN
ajst-32001	104	22	,	,	PUNCT
ajst-32001	104	23	and	and	CCONJ
ajst-32001	104	24	public	public	ADJ
ajst-32001	104	25	security	security	NOUN
ajst-32001	104	26	in	in	ADP
ajst-32001	104	27	the	the	DET
ajst-32001	104	28	future	future	NOUN
ajst-32001	104	29	,	,	PUNCT
ajst-32001	104	30	and	and	CCONJ
ajst-32001	104	31	ultimately	ultimately	ADV
ajst-32001	104	32	achieve	achieve	VERB
ajst-32001	104	33	the	the	DET
ajst-32001	104	34	deep	deep	ADJ
ajst-32001	104	35	integration	integration	NOUN
ajst-32001	104	36	of	of	ADP
ajst-32001	104	37	technology	technology	NOUN
ajst-32001	104	38	and	and	CCONJ
ajst-32001	104	39	human	human	ADJ
ajst-32001	104	40	emotions	emotion	NOUN
ajst-32001	104	41	.	.	PUNCT
ajst-32001	105	1	references	reference	NOUN
ajst-32001	105	2	[	[	X
ajst-32001	105	3	1	1	NUM
ajst-32001	105	4	]	]	PUNCT
ajst-32001	105	5	batbaatar	batbaatar	NOUN
ajst-32001	105	6	,	,	PUNCT
ajst-32001	105	7	e.	e.	PROPN
ajst-32001	105	8	,	,	PUNCT
ajst-32001	105	9	li	li	PROPN
ajst-32001	105	10	,	,	PUNCT
ajst-32001	105	11	m.	m.	NOUN
ajst-32001	105	12	,	,	PUNCT
ajst-32001	105	13	&	&	CCONJ
ajst-32001	105	14	ryu	ryu	PROPN
ajst-32001	105	15	,	,	PUNCT
ajst-32001	105	16	k.	k.	PROPN
ajst-32001	105	17	h.	h.	PROPN
ajst-32001	105	18	(	(	PUNCT
ajst-32001	105	19	2019	2019	NUM
ajst-32001	105	20	)	)	PUNCT
ajst-32001	105	21	.	.	PUNCT
ajst-32001	106	1	semantic	semantic	ADJ
ajst-32001	106	2	-	-	PUNCT
ajst-32001	106	3	emotion	emotion	NOUN
ajst-32001	106	4	neural	neural	ADJ
ajst-32001	106	5	network	network	NOUN
ajst-32001	106	6	for	for	ADP
ajst-32001	106	7	emotion	emotion	NOUN
ajst-32001	106	8	recognition	recognition	NOUN
ajst-32001	106	9	from	from	ADP
ajst-32001	106	10	text	text	NOUN
ajst-32001	106	11	.	.	PUNCT
ajst-32001	107	1	ieee	ieee	NOUN
ajst-32001	107	2	access	access	NOUN
ajst-32001	107	3	,	,	PUNCT
ajst-32001	107	4	7	7	NUM
ajst-32001	107	5	,	,	PUNCT
ajst-32001	107	6	111866	111866	NUM
ajst-32001	107	7	-	-	SYM
ajst-32001	107	8	111878	111878	NUM
ajst-32001	107	9	.	.	PUNCT
ajst-32001	108	1	[	[	X
ajst-32001	108	2	2	2	NUM
ajst-32001	108	3	]	]	X
ajst-32001	108	4	hossain	hossain	PROPN
ajst-32001	108	5	,	,	PUNCT
ajst-32001	108	6	m.	m.	NOUN
ajst-32001	108	7	s.	s.	PROPN
ajst-32001	108	8	,	,	PUNCT
ajst-32001	108	9	&	&	CCONJ
ajst-32001	108	10	muhammad	muhammad	PROPN
ajst-32001	108	11	,	,	PUNCT
ajst-32001	108	12	g.	g.	PROPN
ajst-32001	108	13	(	(	PUNCT
ajst-32001	108	14	2019	2019	NUM
ajst-32001	108	15	)	)	PUNCT
ajst-32001	108	16	.	.	PUNCT
ajst-32001	109	1	emotion	emotion	NOUN
ajst-32001	109	2	recognition	recognition	NOUN
ajst-32001	109	3	using	use	VERB
ajst-32001	109	4	deep	deep	ADJ
ajst-32001	109	5	learning	learning	NOUN
ajst-32001	109	6	approach	approach	NOUN
ajst-32001	109	7	from	from	ADP
ajst-32001	109	8	audio	audio	ADJ
ajst-32001	109	9	–	–	PUNCT
ajst-32001	109	10	visual	visual	ADJ
ajst-32001	109	11	emotional	emotional	ADJ
ajst-32001	109	12	big	big	ADJ
ajst-32001	109	13	data	datum	NOUN
ajst-32001	109	14	.	.	PUNCT
ajst-32001	110	1	information	information	NOUN
ajst-32001	110	2	fusion	fusion	NOUN
ajst-32001	110	3	,	,	PUNCT
ajst-32001	110	4	49	49	NUM
ajst-32001	110	5	,	,	PUNCT
ajst-32001	110	6	69	69	NUM
ajst-32001	110	7	-	-	SYM
ajst-32001	110	8	78	78	NUM
ajst-32001	110	9	.	.	PUNCT
ajst-32001	111	1	[	[	X
ajst-32001	111	2	3	3	NUM
ajst-32001	111	3	]	]	X
ajst-32001	111	4	jiang	jiang	PROPN
ajst-32001	111	5	,	,	PUNCT
ajst-32001	111	6	y.	y.	PROPN
ajst-32001	111	7	,	,	PUNCT
ajst-32001	111	8	li	li	PROPN
ajst-32001	111	9	,	,	PUNCT
ajst-32001	111	10	w.	w.	PROPN
ajst-32001	111	11	,	,	PUNCT
ajst-32001	111	12	hossain	hossain	PROPN
ajst-32001	111	13	,	,	PUNCT
ajst-32001	111	14	m.	m.	PROPN
ajst-32001	111	15	s.	s.	PROPN
ajst-32001	111	16	,	,	PUNCT
ajst-32001	111	17	chen	chen	PROPN
ajst-32001	111	18	,	,	PUNCT
ajst-32001	111	19	m.	m.	NOUN
ajst-32001	111	20	,	,	PUNCT
ajst-32001	111	21	alelaiwi	alelaiwi	NOUN
ajst-32001	111	22	,	,	PUNCT
ajst-32001	111	23	a.	a.	NOUN
ajst-32001	111	24	,	,	PUNCT
ajst-32001	111	25	&	&	CCONJ
ajst-32001	111	26	al	al	PROPN
ajst-32001	111	27	-	-	PUNCT
ajst-32001	111	28	hammadi	hammadi	PROPN
ajst-32001	111	29	,	,	PUNCT
ajst-32001	111	30	m.	m.	NOUN
ajst-32001	111	31	(	(	PUNCT
ajst-32001	111	32	2020	2020	NUM
ajst-32001	111	33	)	)	PUNCT
ajst-32001	111	34	.	.	PUNCT
ajst-32001	112	1	a	a	DET
ajst-32001	112	2	snapshot	snapshot	ADJ
ajst-32001	112	3	research	research	NOUN
ajst-32001	112	4	and	and	CCONJ
ajst-32001	112	5	implementation	implementation	NOUN
ajst-32001	112	6	of	of	ADP
ajst-32001	112	7	multimodal	multimodal	ADJ
ajst-32001	112	8	information	information	NOUN
ajst-32001	112	9	fusion	fusion	NOUN
ajst-32001	112	10	for	for	ADP
ajst-32001	112	11	data	data	NOUN
ajst-32001	112	12	-	-	PUNCT
ajst-32001	112	13	driven	drive	VERB
ajst-32001	112	14	emotion	emotion	NOUN
ajst-32001	112	15	recognition	recognition	NOUN
ajst-32001	112	16	.	.	PUNCT
ajst-32001	113	1	information	information	NOUN
ajst-32001	113	2	fusion	fusion	NOUN
ajst-32001	113	3	,	,	PUNCT
ajst-32001	113	4	53	53	NUM
ajst-32001	113	5	,	,	PUNCT
ajst-32001	113	6	209	209	NUM
ajst-32001	113	7	-	-	SYM
ajst-32001	113	8	221	221	NUM
ajst-32001	113	9	.	.	PUNCT
ajst-32001	114	1	[	[	X
ajst-32001	114	2	4	4	NUM
ajst-32001	114	3	]	]	X
ajst-32001	114	4	koolagudi	koolagudi	PROPN
ajst-32001	114	5	,	,	PUNCT
ajst-32001	114	6	s.	s.	PROPN
ajst-32001	114	7	g.	g.	PROPN
ajst-32001	114	8	,	,	PUNCT
ajst-32001	114	9	&	&	CCONJ
ajst-32001	114	10	rao	rao	PROPN
ajst-32001	114	11	,	,	PUNCT
ajst-32001	114	12	k.	k.	PROPN
ajst-32001	114	13	s.	s.	PROPN
ajst-32001	114	14	(	(	PUNCT
ajst-32001	114	15	2012	2012	NUM
ajst-32001	114	16	)	)	PUNCT
ajst-32001	114	17	.	.	PUNCT
ajst-32001	115	1	emotion	emotion	NOUN
ajst-32001	115	2	recognition	recognition	NOUN
ajst-32001	115	3	from	from	ADP
ajst-32001	115	4	speech	speech	NOUN
ajst-32001	115	5	:	:	PUNCT
ajst-32001	115	6	a	a	DET
ajst-32001	115	7	review	review	NOUN
ajst-32001	115	8	.	.	PUNCT
ajst-32001	116	1	international	international	ADJ
ajst-32001	116	2	journal	journal	PROPN
ajst-32001	116	3	of	of	ADP
ajst-32001	116	4	speech	speech	NOUN
ajst-32001	116	5	technology	technology	NOUN
ajst-32001	116	6	,	,	PUNCT
ajst-32001	116	7	15	15	NUM
ajst-32001	116	8	,	,	PUNCT
ajst-32001	116	9	99	99	NUM
ajst-32001	116	10	-	-	SYM
ajst-32001	116	11	117	117	NUM
ajst-32001	116	12	.	.	PUNCT
ajst-32001	117	1	[	[	X
ajst-32001	117	2	5	5	NUM
ajst-32001	117	3	]	]	SYM
ajst-32001	117	4	yang	yang	PROPN
ajst-32001	117	5	,	,	PUNCT
ajst-32001	117	6	d.	d.	PROPN
ajst-32001	117	7	,	,	PUNCT
ajst-32001	117	8	chen	chen	PROPN
ajst-32001	117	9	,	,	PUNCT
ajst-32001	117	10	z.	z.	PROPN
ajst-32001	117	11	,	,	PUNCT
ajst-32001	117	12	wang	wang	PROPN
ajst-32001	117	13	,	,	PUNCT
ajst-32001	117	14	y.	y.	PROPN
ajst-32001	117	15	,	,	PUNCT
ajst-32001	117	16	wang	wang	PROPN
ajst-32001	117	17	,	,	PUNCT
ajst-32001	117	18	s.	s.	PROPN
ajst-32001	117	19	,	,	PUNCT
ajst-32001	117	20	li	li	PROPN
ajst-32001	117	21	,	,	PUNCT
ajst-32001	117	22	m.	m.	NOUN
ajst-32001	117	23	,	,	PUNCT
ajst-32001	117	24	liu	liu	PROPN
ajst-32001	117	25	,	,	PUNCT
ajst-32001	117	26	s.	s.	PROPN
ajst-32001	117	27	,	,	PUNCT
ajst-32001	117	28	zhao	zhao	PROPN
ajst-32001	117	29	,	,	PUNCT
ajst-32001	117	30	x.	x.	PROPN
ajst-32001	117	31	,	,	PUNCT
ajst-32001	117	32	huang	huang	PROPN
ajst-32001	117	33	,	,	PUNCT
ajst-32001	117	34	s.	s.	PROPN
ajst-32001	117	35	,	,	PUNCT
ajst-32001	117	36	dong	dong	PROPN
ajst-32001	117	37	,	,	PUNCT
ajst-32001	117	38	z.	z.	PROPN
ajst-32001	117	39	,	,	PUNCT
ajst-32001	117	40	zhai	zhai	PROPN
ajst-32001	117	41	,	,	PUNCT
ajst-32001	117	42	p.	p.	PROPN
ajst-32001	117	43	,	,	PUNCT
ajst-32001	117	44	&	&	CCONJ
ajst-32001	117	45	zhang	zhang	PROPN
ajst-32001	117	46	,	,	PUNCT
ajst-32001	117	47	l.	l.	PROPN
ajst-32001	117	48	context	context	PROPN
ajst-32001	117	49	de	de	PROPN
ajst-32001	117	50	-	-	ADJ
ajst-32001	117	51	confounded	confounded	ADJ
ajst-32001	117	52	emotion	emotion	NOUN
ajst-32001	117	53	recognition	recognition	NOUN
ajst-32001	117	54	.	.	PUNCT
ajst-32001	118	1	ieee	ieee	PROPN
ajst-32001	118	2	,	,	PUNCT
ajst-32001	118	3	19005	19005	NUM
ajst-32001	118	4	-	-	SYM
ajst-32001	118	5	19015	19015	NUM
ajst-32001	118	6	.	.	PUNCT
ajst-32001	119	1	[	[	X
ajst-32001	119	2	6	6	NUM
ajst-32001	119	3	]	]	PUNCT
ajst-32001	119	4	sati	sati	NOUN
ajst-32001	119	5	,	,	PUNCT
ajst-32001	119	6	v.	v.	PROPN
ajst-32001	119	7	,	,	PUNCT
ajst-32001	119	8	sánchez	sánchez	PROPN
ajst-32001	119	9	,	,	PUNCT
ajst-32001	119	10	s.	s.	PROPN
ajst-32001	119	11	m.	m.	PROPN
ajst-32001	119	12	,	,	PUNCT
ajst-32001	119	13	shoeibi	shoeibi	NOUN
ajst-32001	119	14	,	,	PUNCT
ajst-32001	119	15	n.	n.	NOUN
ajst-32001	119	16	,	,	PUNCT
ajst-32001	119	17	arora	arora	PROPN
ajst-32001	119	18	,	,	PUNCT
ajst-32001	119	19	a.	a.	PROPN
ajst-32001	119	20	,	,	PUNCT
ajst-32001	119	21	&	&	CCONJ
ajst-32001	119	22	corchado	corchado	PROPN
ajst-32001	119	23	,	,	PUNCT
ajst-32001	119	24	j.	j.	PROPN
ajst-32001	119	25	m.	m.	PROPN
ajst-32001	119	26	(	(	PUNCT
ajst-32001	119	27	2021	2021	NUM
ajst-32001	119	28	)	)	PUNCT
ajst-32001	119	29	.	.	PUNCT
ajst-32001	120	1	face	face	NOUN
ajst-32001	120	2	detection	detection	NOUN
ajst-32001	120	3	and	and	CCONJ
ajst-32001	120	4	recognition	recognition	NOUN
ajst-32001	120	5	,	,	PUNCT
ajst-32001	120	6	face	face	VERB
ajst-32001	120	7	emotion	emotion	NOUN
ajst-32001	120	8	recognition	recognition	NOUN
ajst-32001	120	9	through	through	ADP
ajst-32001	120	10	nvidia	nvidia	PROPN
ajst-32001	120	11	jetson	jetson	PROPN
ajst-32001	120	12	nano	nano	PROPN
ajst-32001	120	13	.	.	PUNCT
ajst-32001	121	1	in	in	ADP
ajst-32001	121	2	p.	p.	PROPN
ajst-32001	121	3	novais	novais	PROPN
ajst-32001	121	4	,	,	PUNCT
ajst-32001	121	5	g.	g.	PROPN
ajst-32001	121	6	vercelli	vercelli	PROPN
ajst-32001	121	7	,	,	PUNCT
ajst-32001	121	8	j.	j.	PROPN
ajst-32001	121	9	l.	l.	PROPN
ajst-32001	121	10	larriba	larriba	PROPN
ajst-32001	121	11	-	-	PUNCT
ajst-32001	121	12	pey	pey	NOUN
ajst-32001	121	13	,	,	PUNCT
ajst-32001	121	14	f.	f.	PROPN
ajst-32001	121	15	herrera	herrera	PROPN
ajst-32001	121	16	,	,	PUNCT
ajst-32001	121	17	&	&	CCONJ
ajst-32001	121	18	p.	p.	NOUN
ajst-32001	121	19	chamoso	chamoso	NOUN
ajst-32001	121	20	(	(	PUNCT
ajst-32001	121	21	eds	ed	NOUN
ajst-32001	121	22	.	.	PUNCT
ajst-32001	121	23	)	)	PUNCT
ajst-32001	121	24	,	,	PUNCT
ajst-32001	121	25	ambient	ambient	ADJ
ajst-32001	121	26	intelligence	intelligence	NOUN
ajst-32001	121	27	–	–	PUNCT
ajst-32001	121	28	software	software	NOUN
ajst-32001	121	29	and	and	CCONJ
ajst-32001	121	30	applications	application	NOUN
ajst-32001	121	31	(	(	PUNCT
ajst-32001	121	32	pp	pp	ADJ
ajst-32001	121	33	.	.	PUNCT
ajst-32001	121	34	1239	1239	NUM
ajst-32001	121	35	)	)	PUNCT
ajst-32001	121	36	.	.	PUNCT
ajst-32001	122	1	springer	springer	NOUN
ajst-32001	122	2	.	.	PUNCT
ajst-32001	123	1	[	[	X
ajst-32001	123	2	7	7	X
ajst-32001	123	3	]	]	X
ajst-32001	123	4	ko	ko	PROPN
ajst-32001	123	5	,	,	PUNCT
ajst-32001	123	6	b.	b.	PROPN
ajst-32001	123	7	c.	c.	PROPN
ajst-32001	123	8	(	(	PUNCT
ajst-32001	123	9	2018	2018	NUM
ajst-32001	123	10	)	)	PUNCT
ajst-32001	123	11	.	.	PUNCT
ajst-32001	124	1	a	a	DET
ajst-32001	124	2	brief	brief	ADJ
ajst-32001	124	3	review	review	NOUN
ajst-32001	124	4	of	of	ADP
ajst-32001	124	5	facial	facial	ADJ
ajst-32001	124	6	emotion	emotion	NOUN
ajst-32001	124	7	recognition	recognition	NOUN
ajst-32001	124	8	based	base	VERB
ajst-32001	124	9	on	on	ADP
ajst-32001	124	10	visual	visual	ADJ
ajst-32001	124	11	information	information	NOUN
ajst-32001	124	12	.	.	PUNCT
ajst-32001	125	1	sensors	sensor	NOUN
ajst-32001	125	2	,	,	PUNCT
ajst-32001	125	3	18(2	18(2	NUM
ajst-32001	125	4	)	)	PUNCT
ajst-32001	125	5	,	,	PUNCT
ajst-32001	125	6	401	401	NUM
ajst-32001	125	7	.	.	PUNCT
ajst-32001	126	1	[	[	X
ajst-32001	126	2	8	8	NUM
ajst-32001	126	3	]	]	X
ajst-32001	126	4	kahou	kahou	NOUN
ajst-32001	126	5	,	,	PUNCT
ajst-32001	126	6	s.	s.	PROPN
ajst-32001	126	7	e.	e.	PROPN
ajst-32001	126	8	,	,	PUNCT
ajst-32001	126	9	michalski	michalski	NOUN
ajst-32001	126	10	,	,	PUNCT
ajst-32001	126	11	v.	v.	PROPN
ajst-32001	126	12	,	,	PUNCT
ajst-32001	126	13	&	&	CCONJ
ajst-32001	126	14	konda	konda	PROPN
ajst-32001	126	15	,	,	PUNCT
ajst-32001	126	16	k.	k.	PROPN
ajst-32001	126	17	(	(	PUNCT
ajst-32001	126	18	2015	2015	NUM
ajst-32001	126	19	)	)	PUNCT
ajst-32001	126	20	.	.	PUNCT
ajst-32001	127	1	recurrent	recurrent	ADJ
ajst-32001	127	2	neural	neural	ADJ
ajst-32001	127	3	networks	network	NOUN
ajst-32001	127	4	for	for	ADP
ajst-32001	127	5	emotion	emotion	NOUN
ajst-32001	127	6	recognition	recognition	NOUN
ajst-32001	127	7	in	in	ADP
ajst-32001	127	8	video	video	NOUN
ajst-32001	127	9	.	.	PUNCT
ajst-32001	128	1	in	in	ADP
ajst-32001	128	2	proceedings	proceeding	NOUN
ajst-32001	128	3	of	of	ADP
ajst-32001	128	4	the	the	DET
ajst-32001	128	5	acm	acm	NOUN
ajst-32001	128	6	on	on	ADP
ajst-32001	128	7	international	international	ADJ
ajst-32001	128	8	conference	conference	NOUN
ajst-32001	128	9	on	on	ADP
ajst-32001	128	10	multimodal	multimodal	NOUN
ajst-32001	128	11	interaction	interaction	NOUN
ajst-32001	128	12	(	(	PUNCT
ajst-32001	128	13	pp	pp	PROPN
ajst-32001	128	14	.	.	PUNCT
ajst-32001	128	15	467	467	NUM
ajst-32001	128	16	-	-	SYM
ajst-32001	128	17	474	474	NUM
ajst-32001	128	18	)	)	PUNCT
ajst-32001	128	19	.	.	PUNCT
ajst-32001	129	1	seattle	seattle	PROPN
ajst-32001	129	2	,	,	PUNCT
ajst-32001	129	3	wa	wa	PROPN
ajst-32001	129	4	,	,	PUNCT
ajst-32001	129	5	usa	usa	PROPN
ajst-32001	129	6	:	:	PUNCT
ajst-32001	129	7	acm	acm	PROPN
ajst-32001	129	8	.	.	PUNCT
ajst-32001	130	1	102	102	NUM
ajst-32001	131	1	[	[	X
ajst-32001	131	2	9	9	NUM
ajst-32001	131	3	]	]	SYM
ajst-32001	131	4	zhang	zhang	PROPN
ajst-32001	131	5	,	,	PUNCT
ajst-32001	131	6	h.	h.	PROPN
ajst-32001	131	7	,	,	PUNCT
ajst-32001	131	8	jolfaei	jolfaei	PROPN
ajst-32001	131	9	,	,	PUNCT
ajst-32001	131	10	a.	a.	NOUN
ajst-32001	131	11	,	,	PUNCT
ajst-32001	131	12	&	&	CCONJ
ajst-32001	131	13	alazab	alazab	PROPN
ajst-32001	131	14	,	,	PUNCT
ajst-32001	131	15	m.	m.	NOUN
ajst-32001	131	16	(	(	PUNCT
ajst-32001	131	17	2019	2019	NUM
ajst-32001	131	18	)	)	PUNCT
ajst-32001	131	19	.	.	PUNCT
ajst-32001	132	1	a	a	DET
ajst-32001	132	2	face	face	NOUN
ajst-32001	132	3	emotion	emotion	NOUN
ajst-32001	132	4	recognition	recognition	NOUN
ajst-32001	132	5	method	method	NOUN
ajst-32001	132	6	using	use	VERB
ajst-32001	132	7	convolutional	convolutional	ADJ
ajst-32001	132	8	neural	neural	ADJ
ajst-32001	132	9	network	network	NOUN
ajst-32001	132	10	and	and	CCONJ
ajst-32001	132	11	image	image	NOUN
ajst-32001	132	12	edge	edge	NOUN
ajst-32001	132	13	computing	computing	NOUN
ajst-32001	132	14	.	.	PUNCT
ajst-32001	133	1	ieee	ieee	NOUN
ajst-32001	133	2	access	access	NOUN
ajst-32001	133	3	,	,	PUNCT
ajst-32001	133	4	7	7	NUM
ajst-32001	133	5	,	,	PUNCT
ajst-32001	133	6	159081	159081	NUM
ajst-32001	133	7	-	-	SYM
ajst-32001	133	8	159089	159089	NUM
ajst-32001	133	9	.	.	PUNCT
ajst-32001	134	1	[	[	X
ajst-32001	134	2	10	10	NUM
ajst-32001	134	3	]	]	X
ajst-32001	134	4	dwijayanti	dwijayanti	PROPN
ajst-32001	134	5	,	,	PUNCT
ajst-32001	134	6	s.	s.	PROPN
ajst-32001	134	7	,	,	PUNCT
ajst-32001	134	8	iqbal	iqbal	PROPN
ajst-32001	134	9	,	,	PUNCT
ajst-32001	134	10	m.	m.	NOUN
ajst-32001	134	11	,	,	PUNCT
ajst-32001	134	12	&	&	CCONJ
ajst-32001	134	13	suprapto	suprapto	PROPN
ajst-32001	134	14	,	,	PUNCT
ajst-32001	134	15	b.	b.	PROPN
ajst-32001	134	16	y.	y.	PROPN
ajst-32001	134	17	(	(	PUNCT
ajst-32001	134	18	2022	2022	NUM
ajst-32001	134	19	)	)	PUNCT
ajst-32001	134	20	.	.	PUNCT
ajst-32001	135	1	real	real	ADJ
ajst-32001	135	2	-	-	PUNCT
ajst-32001	135	3	time	time	NOUN
ajst-32001	135	4	implementation	implementation	NOUN
ajst-32001	135	5	of	of	ADP
ajst-32001	135	6	face	face	NOUN
ajst-32001	135	7	recognition	recognition	NOUN
ajst-32001	135	8	and	and	CCONJ
ajst-32001	135	9	emotion	emotion	NOUN
ajst-32001	135	10	recognition	recognition	NOUN
ajst-32001	135	11	in	in	ADP
ajst-32001	135	12	a	a	DET
ajst-32001	135	13	humanoid	humanoid	NOUN
ajst-32001	135	14	robot	robot	NOUN
ajst-32001	135	15	using	use	VERB
ajst-32001	135	16	a	a	DET
ajst-32001	135	17	convolutional	convolutional	ADJ
ajst-32001	135	18	neural	neural	ADJ
ajst-32001	135	19	network	network	NOUN
ajst-32001	135	20	.	.	PUNCT
ajst-32001	136	1	ieee	ieee	NOUN
ajst-32001	136	2	access	access	NOUN
ajst-32001	136	3	,	,	PUNCT
ajst-32001	136	4	10	10	NUM
ajst-32001	136	5	,	,	PUNCT
ajst-32001	136	6	89876	89876	NUM
ajst-32001	136	7	-	-	SYM
ajst-32001	136	8	89886	89886	NUM
ajst-32001	136	9	.	.	PUNCT
ajst-32001	137	1	[	[	X
ajst-32001	137	2	11	11	NUM
ajst-32001	137	3	]	]	PUNCT
ajst-32001	137	4	schuller	schuller	NOUN
ajst-32001	137	5	,	,	PUNCT
ajst-32001	137	6	b.	b.	PROPN
ajst-32001	137	7	w.	w.	PROPN
ajst-32001	137	8	(	(	PUNCT
ajst-32001	137	9	2018	2018	NUM
ajst-32001	137	10	)	)	PUNCT
ajst-32001	137	11	.	.	PUNCT
ajst-32001	138	1	speech	speech	NOUN
ajst-32001	138	2	emotion	emotion	NOUN
ajst-32001	138	3	recognition	recognition	NOUN
ajst-32001	138	4	:	:	PUNCT
ajst-32001	138	5	two	two	NUM
ajst-32001	138	6	decades	decade	NOUN
ajst-32001	138	7	in	in	ADP
ajst-32001	138	8	a	a	DET
ajst-32001	138	9	nutshell	nutshell	NOUN
ajst-32001	138	10	,	,	PUNCT
ajst-32001	138	11	benchmarks	benchmark	NOUN
ajst-32001	138	12	,	,	PUNCT
ajst-32001	138	13	and	and	CCONJ
ajst-32001	138	14	ongoing	ongoing	ADJ
ajst-32001	138	15	trends	trend	NOUN
ajst-32001	138	16	.	.	PUNCT
ajst-32001	139	1	communications	communication	NOUN
ajst-32001	139	2	of	of	ADP
ajst-32001	139	3	the	the	DET
ajst-32001	139	4	acm	acm	NOUN
ajst-32001	139	5	,	,	PUNCT
ajst-32001	139	6	61	61	NUM
ajst-32001	139	7	,	,	PUNCT
ajst-32001	139	8	90	90	NUM
ajst-32001	139	9	-	-	SYM
ajst-32001	139	10	99	99	NUM
ajst-32001	139	11	.	.	PUNCT
ajst-32001	140	1	[	[	X
ajst-32001	140	2	12	12	NUM
ajst-32001	140	3	]	]	X
ajst-32001	140	4	chen	chen	PROPN
ajst-32001	140	5	,	,	PUNCT
ajst-32001	140	6	m.	m.	NOUN
ajst-32001	140	7	,	,	PUNCT
ajst-32001	140	8	zhou	zhou	PROPN
ajst-32001	140	9	,	,	PUNCT
ajst-32001	140	10	p.	p.	NOUN
ajst-32001	140	11	,	,	PUNCT
ajst-32001	140	12	&	&	CCONJ
ajst-32001	140	13	fortino	fortino	PROPN
ajst-32001	140	14	,	,	PUNCT
ajst-32001	140	15	g.	g.	PROPN
ajst-32001	140	16	(	(	PUNCT
ajst-32001	140	17	2016	2016	NUM
ajst-32001	140	18	)	)	PUNCT
ajst-32001	140	19	.	.	PUNCT
ajst-32001	141	1	emotion	emotion	NOUN
ajst-32001	141	2	communication	communication	NOUN
ajst-32001	141	3	system	system	NOUN
ajst-32001	141	4	.	.	PUNCT
ajst-32001	142	1	ieee	ieee	NOUN
ajst-32001	142	2	access	access	NOUN
ajst-32001	142	3	,	,	PUNCT
ajst-32001	142	4	5	5	NUM
ajst-32001	142	5	,	,	PUNCT
ajst-32001	142	6	326	326	NUM
ajst-32001	142	7	-	-	SYM
ajst-32001	142	8	337	337	NUM
ajst-32001	142	9	.	.	PUNCT
ajst-32001	143	1	[	[	X
ajst-32001	143	2	13	13	NUM
ajst-32001	143	3	]	]	SYM
ajst-32001	143	4	koolagudi	koolagudi	PROPN
ajst-32001	143	5	,	,	PUNCT
ajst-32001	143	6	s.	s.	PROPN
ajst-32001	143	7	g.	g.	PROPN
ajst-32001	143	8	,	,	PUNCT
ajst-32001	143	9	&	&	CCONJ
ajst-32001	143	10	rao	rao	PROPN
ajst-32001	143	11	,	,	PUNCT
ajst-32001	143	12	k.	k.	PROPN
ajst-32001	143	13	s.	s.	PROPN
ajst-32001	143	14	(	(	PUNCT
ajst-32001	143	15	2012	2012	NUM
ajst-32001	143	16	)	)	PUNCT
ajst-32001	143	17	.	.	PUNCT
ajst-32001	144	1	emotion	emotion	NOUN
ajst-32001	144	2	recognition	recognition	NOUN
ajst-32001	144	3	from	from	ADP
ajst-32001	144	4	speech	speech	NOUN
ajst-32001	144	5	:	:	PUNCT
ajst-32001	144	6	a	a	DET
ajst-32001	144	7	review	review	NOUN
ajst-32001	144	8	.	.	PUNCT
ajst-32001	145	1	international	international	ADJ
ajst-32001	145	2	journal	journal	PROPN
ajst-32001	145	3	of	of	ADP
ajst-32001	145	4	speech	speech	NOUN
ajst-32001	145	5	technology	technology	NOUN
ajst-32001	145	6	,	,	PUNCT
ajst-32001	145	7	15	15	NUM
ajst-32001	145	8	,	,	PUNCT
ajst-32001	145	9	99	99	NUM
ajst-32001	145	10	-	-	SYM
ajst-32001	145	11	117	117	NUM
ajst-32001	145	12	.	.	PUNCT
ajst-32001	146	1	[	[	X
ajst-32001	146	2	14	14	NUM
ajst-32001	146	3	]	]	X
ajst-32001	146	4	khalil	khalil	PROPN
ajst-32001	146	5	,	,	PUNCT
ajst-32001	146	6	r.	r.	PROPN
ajst-32001	146	7	a.	a.	PROPN
ajst-32001	146	8	,	,	PUNCT
ajst-32001	146	9	jones	jones	PROPN
ajst-32001	146	10	,	,	PUNCT
ajst-32001	146	11	e.	e.	PROPN
ajst-32001	146	12	,	,	PUNCT
ajst-32001	146	13	babar	babar	PROPN
ajst-32001	146	14	,	,	PUNCT
ajst-32001	146	15	m.	m.	NOUN
ajst-32001	146	16	i.	i.	PROPN
ajst-32001	146	17	,	,	PUNCT
ajst-32001	146	18	jan	jan	PROPN
ajst-32001	146	19	,	,	PUNCT
ajst-32001	146	20	t.	t.	PROPN
ajst-32001	146	21	,	,	PUNCT
ajst-32001	146	22	zafar	zafar	PROPN
ajst-32001	146	23	,	,	PUNCT
ajst-32001	146	24	m.	m.	PROPN
ajst-32001	146	25	h.	h.	PROPN
ajst-32001	146	26	,	,	PUNCT
ajst-32001	146	27	&	&	CCONJ
ajst-32001	146	28	alhussain	alhussain	ADJ
ajst-32001	146	29	,	,	PUNCT
ajst-32001	146	30	t.	t.	NOUN
ajst-32001	146	31	(	(	PUNCT
ajst-32001	146	32	2019	2019	NUM
ajst-32001	146	33	)	)	PUNCT
ajst-32001	146	34	.	.	PUNCT
ajst-32001	147	1	speech	speech	NOUN
ajst-32001	147	2	emotion	emotion	NOUN
ajst-32001	147	3	recognition	recognition	NOUN
ajst-32001	147	4	using	use	VERB
ajst-32001	147	5	deep	deep	ADJ
ajst-32001	147	6	learning	learning	NOUN
ajst-32001	147	7	techniques	technique	NOUN
ajst-32001	147	8	:	:	PUNCT
ajst-32001	147	9	a	a	DET
ajst-32001	147	10	review	review	NOUN
ajst-32001	147	11	.	.	PUNCT
ajst-32001	148	1	ieee	ieee	NOUN
ajst-32001	148	2	access	access	NOUN
ajst-32001	148	3	,	,	PUNCT
ajst-32001	148	4	7	7	NUM
ajst-32001	148	5	,	,	PUNCT
ajst-32001	148	6	117327	117327	NUM
ajst-32001	148	7	-	-	SYM
ajst-32001	148	8	117345	117345	NUM
ajst-32001	148	9	.	.	PUNCT
ajst-32001	149	1	[	[	X
ajst-32001	149	2	15	15	NUM
ajst-32001	149	3	]	]	X
ajst-32001	149	4	chernykh	chernykh	NOUN
ajst-32001	149	5	,	,	PUNCT
ajst-32001	149	6	v.	v.	PROPN
ajst-32001	149	7	,	,	PUNCT
ajst-32001	149	8	&	&	CCONJ
ajst-32001	149	9	prihodko	prihodko	PROPN
ajst-32001	149	10	,	,	PUNCT
ajst-32001	149	11	p.	p.	NOUN
ajst-32001	149	12	(	(	PUNCT
ajst-32001	149	13	2017	2017	NUM
ajst-32001	149	14	)	)	PUNCT
ajst-32001	149	15	.	.	PUNCT
ajst-32001	150	1	emotion	emotion	NOUN
ajst-32001	150	2	recognition	recognition	NOUN
ajst-32001	150	3	from	from	ADP
ajst-32001	150	4	speech	speech	NOUN
ajst-32001	150	5	with	with	ADP
ajst-32001	150	6	recurrent	recurrent	ADJ
ajst-32001	150	7	neural	neural	ADJ
ajst-32001	150	8	networks	network	NOUN
ajst-32001	151	1	[	[	X
ajst-32001	151	2	preprint	preprint	NOUN
ajst-32001	151	3	]	]	X
ajst-32001	151	4	.	.	PUNCT
ajst-32001	152	1	arxiv:1701.08071	arxiv:1701.08071	PROPN
ajst-32001	152	2	https://arxiv.org/abs/1701.08071	https://arxiv.org/abs/1701.08071	PROPN