id	sid	tid	token	lemma	pos
easat-5713	1	1	edelweiss	edelweiss	PROPN
easat-5713	1	2	applied	apply	VERB
easat-5713	1	3	science	science	NOUN
easat-5713	1	4	and	and	CCONJ
easat-5713	1	5	technology	technology	NOUN
easat-5713	1	6	issn	issn	PROPN
easat-5713	1	7	:	:	PUNCT
easat-5713	1	8	2576	2576	NUM
easat-5713	1	9	-	-	SYM
easat-5713	1	10	8484	8484	NUM
easat-5713	1	11	vol	vol	NOUN
easat-5713	1	12	.	.	PROPN
easat-5713	2	1	9	9	NUM
easat-5713	2	2	,	,	PUNCT
easat-5713	2	3	no	no	INTJ
easat-5713	2	4	.	.	NOUN
easat-5713	2	5	3	3	NUM
easat-5713	2	6	,	,	PUNCT
easat-5713	2	7	1855	1855	NUM
easat-5713	2	8	-	-	SYM
easat-5713	2	9	1869	1869	NUM
easat-5713	2	10	2025	2025	NUM
easat-5713	2	11	publisher	publisher	NOUN
easat-5713	2	12	:	:	PUNCT
easat-5713	2	13	learning	learn	VERB
easat-5713	2	14	gate	gate	NOUN
easat-5713	2	15	doi	doi	PROPN
easat-5713	2	16	:	:	PUNCT
easat-5713	2	17	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	2	18	©	©	ADP
easat-5713	2	19	2025	2025	NUM
easat-5713	2	20	by	by	ADP
easat-5713	2	21	the	the	DET
easat-5713	2	22	authors	author	NOUN
easat-5713	2	23	;	;	PUNCT
easat-5713	2	24	licensee	licensee	PROPN
easat-5713	2	25	learning	learning	NOUN
easat-5713	2	26	gate	gate	NOUN
easat-5713	3	1	©	©	PROPN
easat-5713	3	2	2025	2025	NUM
easat-5713	3	3	by	by	ADP
easat-5713	3	4	the	the	DET
easat-5713	3	5	authors	author	NOUN
easat-5713	3	6	;	;	PUNCT
easat-5713	3	7	licensee	licensee	PROPN
easat-5713	3	8	learning	learning	NOUN
easat-5713	3	9	gate	gate	NOUN
easat-5713	3	10	history	history	NOUN
easat-5713	3	11	:	:	PUNCT
easat-5713	3	12	received	receive	VERB
easat-5713	3	13	:	:	PUNCT
easat-5713	3	14	3	3	NUM
easat-5713	3	15	january	january	NOUN
easat-5713	3	16	2025	2025	NUM
easat-5713	3	17	;	;	PUNCT
easat-5713	3	18	revised	revise	VERB
easat-5713	3	19	:	:	PUNCT
easat-5713	3	20	10	10	NUM
easat-5713	3	21	march	march	NOUN
easat-5713	3	22	2025	2025	NUM
easat-5713	3	23	;	;	PUNCT
easat-5713	3	24	accepted	accept	VERB
easat-5713	3	25	:	:	PUNCT
easat-5713	3	26	12	12	NUM
easat-5713	3	27	march	march	NOUN
easat-5713	3	28	2025	2025	NUM
easat-5713	3	29	;	;	PUNCT
easat-5713	3	30	published	publish	VERB
easat-5713	3	31	:	:	PUNCT
easat-5713	3	32	24	24	NUM
easat-5713	3	33	march	march	NOUN
easat-5713	3	34	2025	2025	NUM
easat-5713	3	35	*	*	PUNCT
easat-5713	3	36	correspondence	correspondence	NOUN
easat-5713	3	37	:	:	PUNCT
easat-5713	3	38	a.kartbayev@gmail.com	a.kartbayev@gmail.com	NOUN
easat-5713	3	39	evaluating	evaluate	VERB
easat-5713	3	40	the	the	DET
easat-5713	3	41	impact	impact	NOUN
easat-5713	3	42	of	of	ADP
easat-5713	3	43	ai	ai	VERB
easat-5713	3	44	-	-	PUNCT
easat-5713	3	45	generated	generate	VERB
easat-5713	3	46	synthetic	synthetic	ADJ
easat-5713	3	47	noisy	noisy	ADJ
easat-5713	3	48	data	datum	NOUN
easat-5713	3	49	on	on	ADP
easat-5713	3	50	human	human	ADJ
easat-5713	3	51	emotion	emotion	NOUN
easat-5713	3	52	recognition	recognition	NOUN
easat-5713	3	53	models	model	NOUN
easat-5713	3	54	dastan	dastan	ADJ
easat-5713	3	55	sultanov1	sultanov1	NOUN
easat-5713	3	56	,	,	PUNCT
easat-5713	3	57	sanzhar	sanzhar	PROPN
easat-5713	3	58	seitbekov2	seitbekov2	PROPN
easat-5713	3	59	,	,	PUNCT
easat-5713	3	60	shynara	shynara	PROPN
easat-5713	3	61	ayanbek3	ayanbek3	PROPN
easat-5713	3	62	,	,	PUNCT
easat-5713	3	63	miras	miras	PROPN
easat-5713	3	64	assubay4	assubay4	PROPN
easat-5713	3	65	,	,	PUNCT
easat-5713	3	66	zhaisang	zhaisang	PROPN
easat-5713	3	67	sarsengaliyev5	sarsengaliyev5	PROPN
easat-5713	3	68	,	,	PUNCT
easat-5713	3	69	amandyk	amandyk	NOUN
easat-5713	3	70	kartbayev6	kartbayev6	PROPN
easat-5713	3	71	*	*	SYM
easat-5713	3	72	1,2,3,4,5,6school	1,2,3,4,5,6school	NOUN
easat-5713	3	73	of	of	ADP
easat-5713	3	74	it	it	PRON
easat-5713	3	75	and	and	CCONJ
easat-5713	3	76	engineering	engineering	NOUN
easat-5713	3	77	,	,	PUNCT
easat-5713	3	78	kazakh	kazakh	PROPN
easat-5713	3	79	-	-	PUNCT
easat-5713	3	80	british	british	PROPN
easat-5713	3	81	technical	technical	PROPN
easat-5713	3	82	university	university	PROPN
easat-5713	3	83	,	,	PUNCT
easat-5713	3	84	almaty	almaty	PROPN
easat-5713	3	85	,	,	PUNCT
easat-5713	3	86	kazakhstan	kazakhstan	PROPN
easat-5713	3	87	;	;	PUNCT
easat-5713	3	88	a.kartbayev@gmail.com	a.kartbayev@gmail.com	PROPN
easat-5713	3	89	(	(	PUNCT
easat-5713	3	90	a.k	a.k	PROPN
easat-5713	3	91	.	.	PUNCT
easat-5713	3	92	)	)	PUNCT
easat-5713	4	1	abstract	abstract	NOUN
easat-5713	4	2	:	:	PUNCT
easat-5713	4	3	human	human	ADJ
easat-5713	4	4	emotion	emotion	NOUN
easat-5713	4	5	recognition	recognition	NOUN
easat-5713	4	6	remains	remain	VERB
easat-5713	4	7	a	a	DET
easat-5713	4	8	challenging	challenging	ADJ
easat-5713	4	9	task	task	NOUN
easat-5713	4	10	due	due	ADP
easat-5713	4	11	to	to	ADP
easat-5713	4	12	the	the	DET
easat-5713	4	13	complexity	complexity	NOUN
easat-5713	4	14	and	and	CCONJ
easat-5713	4	15	variability	variability	NOUN
easat-5713	4	16	of	of	ADP
easat-5713	4	17	human	human	ADJ
easat-5713	4	18	emotions	emotion	NOUN
easat-5713	4	19	in	in	ADP
easat-5713	4	20	real	real	ADJ
easat-5713	4	21	-	-	PUNCT
easat-5713	4	22	world	world	NOUN
easat-5713	4	23	scenarios	scenario	NOUN
easat-5713	4	24	.	.	PUNCT
easat-5713	5	1	this	this	DET
easat-5713	5	2	study	study	NOUN
easat-5713	5	3	investigates	investigate	VERB
easat-5713	5	4	the	the	DET
easat-5713	5	5	impact	impact	NOUN
easat-5713	5	6	of	of	ADP
easat-5713	5	7	ai	ai	VERB
easat-5713	5	8	-	-	PUNCT
easat-5713	5	9	generated	generate	VERB
easat-5713	5	10	synthetic	synthetic	ADJ
easat-5713	5	11	data	datum	NOUN
easat-5713	5	12	on	on	ADP
easat-5713	5	13	enhancing	enhance	VERB
easat-5713	5	14	facial	facial	ADJ
easat-5713	5	15	expression	expression	NOUN
easat-5713	5	16	recognition	recognition	NOUN
easat-5713	5	17	(	(	PUNCT
easat-5713	5	18	fer	fer	PROPN
easat-5713	5	19	)	)	PUNCT
easat-5713	5	20	model	model	NOUN
easat-5713	5	21	performance	performance	NOUN
easat-5713	5	22	.	.	PUNCT
easat-5713	6	1	using	use	VERB
easat-5713	6	2	the	the	DET
easat-5713	6	3	juggernaut	juggernaut	PROPN
easat-5713	6	4	xl	xl	PROPN
easat-5713	6	5	model	model	PROPN
easat-5713	6	6	,	,	PUNCT
easat-5713	6	7	we	we	PRON
easat-5713	6	8	generated	generate	VERB
easat-5713	6	9	300	300	NUM
easat-5713	6	10	synthetic	synthetic	ADJ
easat-5713	6	11	images	image	NOUN
easat-5713	6	12	per	per	ADP
easat-5713	6	13	emotion	emotion	NOUN
easat-5713	6	14	category	category	NOUN
easat-5713	6	15	from	from	ADP
easat-5713	6	16	the	the	DET
easat-5713	6	17	fer2013	fer2013	ADJ
easat-5713	6	18	dataset	dataset	NOUN
easat-5713	6	19	and	and	CCONJ
easat-5713	6	20	incorporated	incorporate	VERB
easat-5713	6	21	them	they	PRON
easat-5713	6	22	into	into	ADP
easat-5713	6	23	the	the	DET
easat-5713	6	24	training	training	NOUN
easat-5713	6	25	process	process	NOUN
easat-5713	6	26	of	of	ADP
easat-5713	6	27	a	a	DET
easat-5713	6	28	vgg-19	vgg-19	ADV
easat-5713	6	29	-	-	PUNCT
easat-5713	6	30	based	base	VERB
easat-5713	6	31	fer	fer	PROPN
easat-5713	6	32	model	model	PROPN
easat-5713	6	33	.	.	PUNCT
easat-5713	7	1	experimental	experimental	ADJ
easat-5713	7	2	results	result	NOUN
easat-5713	7	3	revealed	reveal	VERB
easat-5713	7	4	that	that	SCONJ
easat-5713	7	5	the	the	DET
easat-5713	7	6	synthetic	synthetic	ADJ
easat-5713	7	7	data	datum	NOUN
easat-5713	7	8	did	do	AUX
easat-5713	7	9	not	not	PART
easat-5713	7	10	improve	improve	VERB
easat-5713	7	11	key	key	ADJ
easat-5713	7	12	performance	performance	NOUN
easat-5713	7	13	metrics	metric	NOUN
easat-5713	7	14	,	,	PUNCT
easat-5713	7	15	with	with	ADP
easat-5713	7	16	the	the	DET
easat-5713	7	17	originally	originally	ADV
easat-5713	7	18	trained	train	VERB
easat-5713	7	19	model	model	NOUN
easat-5713	7	20	achieving	achieve	VERB
easat-5713	7	21	an	an	DET
easat-5713	7	22	accuracy	accuracy	NOUN
easat-5713	7	23	of	of	ADP
easat-5713	7	24	65	65	NUM
easat-5713	7	25	%	%	NOUN
easat-5713	7	26	,	,	PUNCT
easat-5713	7	27	compared	compare	VERB
easat-5713	7	28	to	to	ADP
easat-5713	7	29	63	63	NUM
easat-5713	7	30	%	%	NOUN
easat-5713	7	31	for	for	ADP
easat-5713	7	32	the	the	DET
easat-5713	7	33	augmented	augment	VERB
easat-5713	7	34	dataset	dataset	NOUN
easat-5713	7	35	.	.	PUNCT
easat-5713	8	1	precision	precision	NOUN
easat-5713	8	2	,	,	PUNCT
easat-5713	8	3	recall	recall	NOUN
easat-5713	8	4	,	,	PUNCT
easat-5713	8	5	and	and	CCONJ
easat-5713	8	6	f1	f1	NOUN
easat-5713	8	7	-	-	PUNCT
easat-5713	8	8	score	score	NOUN
easat-5713	8	9	also	also	ADV
easat-5713	8	10	exhibited	exhibit	VERB
easat-5713	8	11	fluctuations	fluctuation	NOUN
easat-5713	8	12	across	across	ADP
easat-5713	8	13	different	different	ADJ
easat-5713	8	14	emotion	emotion	NOUN
easat-5713	8	15	categories	category	NOUN
easat-5713	8	16	,	,	PUNCT
easat-5713	8	17	as	as	SCONJ
easat-5713	8	18	illustrated	illustrate	VERB
easat-5713	8	19	by	by	ADP
easat-5713	8	20	confusion	confusion	NOUN
easat-5713	8	21	matrices	matrix	NOUN
easat-5713	8	22	.	.	PUNCT
easat-5713	9	1	the	the	DET
easat-5713	9	2	findings	finding	NOUN
easat-5713	9	3	suggest	suggest	VERB
easat-5713	9	4	that	that	SCONJ
easat-5713	9	5	the	the	DET
easat-5713	9	6	quality	quality	NOUN
easat-5713	9	7	of	of	ADP
easat-5713	9	8	synthetic	synthetic	ADJ
easat-5713	9	9	images	image	NOUN
easat-5713	9	10	plays	play	VERB
easat-5713	9	11	a	a	DET
easat-5713	9	12	crucial	crucial	ADJ
easat-5713	9	13	role	role	NOUN
easat-5713	9	14	in	in	ADP
easat-5713	9	15	model	model	NOUN
easat-5713	9	16	effectiveness	effectiveness	NOUN
easat-5713	9	17	,	,	PUNCT
easat-5713	9	18	as	as	SCONJ
easat-5713	9	19	insufficient	insufficient	ADJ
easat-5713	9	20	diversity	diversity	NOUN
easat-5713	9	21	may	may	AUX
easat-5713	9	22	introduce	introduce	VERB
easat-5713	9	23	noise	noise	NOUN
easat-5713	9	24	rather	rather	ADV
easat-5713	9	25	than	than	ADP
easat-5713	9	26	beneficial	beneficial	ADJ
easat-5713	9	27	augmentation	augmentation	NOUN
easat-5713	9	28	.	.	PUNCT
easat-5713	10	1	factors	factor	NOUN
easat-5713	10	2	such	such	ADJ
easat-5713	10	3	as	as	ADP
easat-5713	10	4	limited	limited	ADJ
easat-5713	10	5	training	training	NOUN
easat-5713	10	6	epochs	epoch	NOUN
easat-5713	10	7	and	and	CCONJ
easat-5713	10	8	potential	potential	ADJ
easat-5713	10	9	dataset	dataset	NOUN
easat-5713	10	10	biases	bias	NOUN
easat-5713	10	11	may	may	AUX
easat-5713	10	12	have	have	AUX
easat-5713	10	13	also	also	ADV
easat-5713	10	14	influenced	influence	VERB
easat-5713	10	15	the	the	DET
easat-5713	10	16	outcomes	outcome	NOUN
easat-5713	10	17	.	.	PUNCT
easat-5713	11	1	this	this	DET
easat-5713	11	2	study	study	NOUN
easat-5713	11	3	highlights	highlight	VERB
easat-5713	11	4	the	the	DET
easat-5713	11	5	importance	importance	NOUN
easat-5713	11	6	of	of	ADP
easat-5713	11	7	optimizing	optimize	VERB
easat-5713	11	8	synthetic	synthetic	ADJ
easat-5713	11	9	image	image	NOUN
easat-5713	11	10	realism	realism	NOUN
easat-5713	11	11	to	to	PART
easat-5713	11	12	improve	improve	VERB
easat-5713	11	13	fer	fer	PROPN
easat-5713	11	14	models	model	NOUN
easat-5713	11	15	and	and	CCONJ
easat-5713	11	16	offers	offer	VERB
easat-5713	11	17	practical	practical	ADJ
easat-5713	11	18	insights	insight	NOUN
easat-5713	11	19	for	for	ADP
easat-5713	11	20	future	future	ADJ
easat-5713	11	21	ai	ai	ADJ
easat-5713	11	22	-	-	PUNCT
easat-5713	11	23	driven	drive	VERB
easat-5713	11	24	applications	application	NOUN
easat-5713	11	25	using	use	VERB
easat-5713	11	26	data	datum	NOUN
easat-5713	11	27	augmentation	augmentation	NOUN
easat-5713	11	28	.	.	PUNCT
easat-5713	12	1	keywords	keyword	NOUN
easat-5713	12	2	:	:	PUNCT
easat-5713	12	3	data	datum	NOUN
easat-5713	12	4	augmentation	augmentation	NOUN
easat-5713	12	5	emotion	emotion	NOUN
easat-5713	12	6	recognition	recognition	NOUN
easat-5713	12	7	,	,	PUNCT
easat-5713	12	8	facial	facial	ADJ
easat-5713	12	9	expression	expression	NOUN
easat-5713	12	10	recognition	recognition	NOUN
easat-5713	12	11	,	,	PUNCT
easat-5713	12	12	medical	medical	ADJ
easat-5713	12	13	images	image	NOUN
easat-5713	12	14	,	,	PUNCT
easat-5713	12	15	prediction	prediction	NOUN
easat-5713	12	16	models	model	NOUN
easat-5713	12	17	,	,	PUNCT
easat-5713	12	18	synthetic	synthetic	ADJ
easat-5713	12	19	data	datum	NOUN
easat-5713	12	20	,	,	PUNCT
easat-5713	12	21	system	system	NOUN
easat-5713	12	22	design	design	NOUN
easat-5713	12	23	.	.	PUNCT
easat-5713	13	1	1	1	X
easat-5713	13	2	.	.	X
easat-5713	13	3	introduction	introduction	NOUN
easat-5713	13	4	imagine	imagine	VERB
easat-5713	13	5	a	a	DET
easat-5713	13	6	world	world	NOUN
easat-5713	13	7	where	where	SCONJ
easat-5713	13	8	entertainment	entertainment	NOUN
easat-5713	13	9	systems	system	NOUN
easat-5713	13	10	adapt	adapt	VERB
easat-5713	13	11	to	to	ADP
easat-5713	13	12	your	your	PRON
easat-5713	13	13	emotional	emotional	ADJ
easat-5713	13	14	responses	response	NOUN
easat-5713	13	15	in	in	ADP
easat-5713	13	16	real	real	ADJ
easat-5713	13	17	-	-	PUNCT
easat-5713	13	18	time	time	NOUN
easat-5713	13	19	,	,	PUNCT
easat-5713	13	20	adjusting	adjust	VERB
easat-5713	13	21	movie	movie	NOUN
easat-5713	13	22	recommendations	recommendation	NOUN
easat-5713	13	23	or	or	CCONJ
easat-5713	13	24	modifying	modify	VERB
easat-5713	13	25	gameplay	gameplay	NOUN
easat-5713	13	26	difficulty	difficulty	NOUN
easat-5713	13	27	based	base	VERB
easat-5713	13	28	on	on	ADP
easat-5713	13	29	facial	facial	ADJ
easat-5713	13	30	expressions	expression	NOUN
easat-5713	13	31	.	.	PUNCT
easat-5713	14	1	consider	consider	VERB
easat-5713	14	2	a	a	DET
easat-5713	14	3	vehicle	vehicle	NOUN
easat-5713	14	4	that	that	PRON
easat-5713	14	5	monitors	monitor	VERB
easat-5713	14	6	driver	driver	NOUN
easat-5713	14	7	fatigue	fatigue	NOUN
easat-5713	14	8	and	and	CCONJ
easat-5713	14	9	distraction	distraction	NOUN
easat-5713	14	10	through	through	ADP
easat-5713	14	11	facial	facial	ADJ
easat-5713	14	12	emotion	emotion	NOUN
easat-5713	14	13	recognition	recognition	NOUN
easat-5713	14	14	(	(	PUNCT
easat-5713	14	15	fer	fer	PROPN
easat-5713	14	16	)	)	PUNCT
easat-5713	14	17	to	to	PART
easat-5713	14	18	enhance	enhance	VERB
easat-5713	14	19	road	road	NOUN
easat-5713	14	20	safety	safety	NOUN
easat-5713	14	21	by	by	ADP
easat-5713	14	22	providing	provide	VERB
easat-5713	14	23	timely	timely	ADJ
easat-5713	14	24	alerts	alert	NOUN
easat-5713	14	25	and	and	CCONJ
easat-5713	14	26	interventions	intervention	NOUN
easat-5713	14	27	.	.	PUNCT
easat-5713	15	1	these	these	DET
easat-5713	15	2	applications	application	NOUN
easat-5713	15	3	highlight	highlight	VERB
easat-5713	15	4	the	the	DET
easat-5713	15	5	transformative	transformative	ADJ
easat-5713	15	6	potential	potential	NOUN
easat-5713	15	7	of	of	ADP
easat-5713	15	8	fer	fer	PROPN
easat-5713	15	9	in	in	ADP
easat-5713	15	10	artificial	artificial	ADJ
easat-5713	15	11	intelligence	intelligence	NOUN
easat-5713	15	12	,	,	PUNCT
easat-5713	15	13	with	with	ADP
easat-5713	15	14	applications	application	NOUN
easat-5713	15	15	spanning	span	VERB
easat-5713	15	16	humancomputer	humancomputer	NOUN
easat-5713	15	17	interaction	interaction	NOUN
easat-5713	15	18	,	,	PUNCT
easat-5713	15	19	healthcare	healthcare	PROPN
easat-5713	15	20	,	,	PUNCT
easat-5713	15	21	security	security	NOUN
easat-5713	15	22	,	,	PUNCT
easat-5713	15	23	and	and	CCONJ
easat-5713	15	24	automotive	automotive	ADJ
easat-5713	15	25	safety	safety	NOUN
easat-5713	15	26	.	.	PUNCT
easat-5713	16	1	however	however	ADV
easat-5713	16	2	,	,	PUNCT
easat-5713	16	3	despite	despite	SCONJ
easat-5713	16	4	its	its	PRON
easat-5713	16	5	growing	grow	VERB
easat-5713	16	6	importance	importance	NOUN
easat-5713	16	7	,	,	PUNCT
easat-5713	16	8	fer	fer	PROPN
easat-5713	16	9	systems	system	NOUN
easat-5713	16	10	face	face	VERB
easat-5713	16	11	significant	significant	ADJ
easat-5713	16	12	challenges	challenge	NOUN
easat-5713	16	13	,	,	PUNCT
easat-5713	16	14	particularly	particularly	ADV
easat-5713	16	15	in	in	ADP
easat-5713	16	16	ensuring	ensure	VERB
easat-5713	16	17	high	high	ADJ
easat-5713	16	18	accuracy	accuracy	NOUN
easat-5713	16	19	and	and	CCONJ
easat-5713	16	20	robustness	robustness	NOUN
easat-5713	16	21	across	across	ADP
easat-5713	16	22	diverse	diverse	ADJ
easat-5713	16	23	,	,	PUNCT
easat-5713	16	24	real	real	ADJ
easat-5713	16	25	-	-	PUNCT
easat-5713	16	26	world	world	NOUN
easat-5713	16	27	conditions	condition	NOUN
easat-5713	16	28	.	.	PUNCT
easat-5713	17	1	one	one	NUM
easat-5713	17	2	of	of	ADP
easat-5713	17	3	the	the	DET
easat-5713	17	4	key	key	ADJ
easat-5713	17	5	barriers	barrier	NOUN
easat-5713	17	6	to	to	ADP
easat-5713	17	7	effective	effective	ADJ
easat-5713	17	8	deployment	deployment	NOUN
easat-5713	17	9	is	be	AUX
easat-5713	17	10	the	the	DET
easat-5713	17	11	inherent	inherent	ADJ
easat-5713	17	12	limitations	limitation	NOUN
easat-5713	17	13	of	of	ADP
easat-5713	17	14	existing	exist	VERB
easat-5713	17	15	datasets	dataset	NOUN
easat-5713	17	16	.	.	PUNCT
easat-5713	18	1	traditional	traditional	ADJ
easat-5713	18	2	datasets	dataset	NOUN
easat-5713	18	3	like	like	ADP
easat-5713	18	4	fer2013	fer2013	ADJ
easat-5713	18	5	,	,	PUNCT
easat-5713	18	6	despite	despite	SCONJ
easat-5713	18	7	being	be	AUX
easat-5713	18	8	widely	widely	ADV
easat-5713	18	9	used	use	VERB
easat-5713	18	10	in	in	ADP
easat-5713	18	11	fer	fer	PROPN
easat-5713	18	12	research	research	PROPN
easat-5713	18	13	,	,	PUNCT
easat-5713	18	14	suffer	suffer	VERB
easat-5713	18	15	from	from	ADP
easat-5713	18	16	issues	issue	NOUN
easat-5713	18	17	such	such	ADJ
easat-5713	18	18	as	as	ADP
easat-5713	18	19	insufficient	insufficient	ADJ
easat-5713	18	20	diversity	diversity	NOUN
easat-5713	18	21	in	in	ADP
easat-5713	18	22	facial	facial	ADJ
easat-5713	18	23	expressions	expression	NOUN
easat-5713	18	24	,	,	PUNCT
easat-5713	18	25	demographic	demographic	ADJ
easat-5713	18	26	representation	representation	NOUN
easat-5713	18	27	biases	bias	NOUN
easat-5713	18	28	,	,	PUNCT
easat-5713	18	29	and	and	CCONJ
easat-5713	18	30	ethical	ethical	ADJ
easat-5713	18	31	concerns	concern	NOUN
easat-5713	18	32	related	relate	VERB
easat-5713	18	33	to	to	ADP
easat-5713	18	34	privacy	privacy	NOUN
easat-5713	18	35	[	[	X
easat-5713	18	36	1	1	NUM
easat-5713	18	37	]	]	PUNCT
easat-5713	18	38	.	.	PUNCT
easat-5713	19	1	more	more	ADV
easat-5713	19	2	critically	critically	ADV
easat-5713	19	3	,	,	PUNCT
easat-5713	19	4	these	these	DET
easat-5713	19	5	datasets	dataset	NOUN
easat-5713	19	6	contain	contain	VERB
easat-5713	19	7	noisy	noisy	ADJ
easat-5713	19	8	images	image	NOUN
easat-5713	19	9	—	—	PUNCT
easat-5713	19	10	low	low	ADJ
easat-5713	19	11	-	-	PUNCT
easat-5713	19	12	resolution	resolution	NOUN
easat-5713	19	13	,	,	PUNCT
easat-5713	19	14	blurred	blurred	ADJ
easat-5713	19	15	,	,	PUNCT
easat-5713	19	16	or	or	CCONJ
easat-5713	19	17	occluded	occlude	VERB
easat-5713	19	18	facial	facial	ADJ
easat-5713	19	19	expressions	expression	NOUN
easat-5713	19	20	that	that	PRON
easat-5713	19	21	degrade	degrade	VERB
easat-5713	19	22	model	model	NOUN
easat-5713	19	23	performance	performance	NOUN
easat-5713	19	24	.	.	PUNCT
easat-5713	20	1	the	the	DET
easat-5713	20	2	presence	presence	NOUN
easat-5713	20	3	of	of	ADP
easat-5713	20	4	noise	noise	NOUN
easat-5713	20	5	makes	make	VERB
easat-5713	20	6	it	it	PRON
easat-5713	20	7	difficult	difficult	ADJ
easat-5713	20	8	for	for	ADP
easat-5713	20	9	deep	deep	ADJ
easat-5713	20	10	learning	learning	NOUN
easat-5713	20	11	models	model	NOUN
easat-5713	20	12	to	to	PART
easat-5713	20	13	learn	learn	VERB
easat-5713	20	14	meaningful	meaningful	ADJ
easat-5713	20	15	facial	facial	ADJ
easat-5713	20	16	features	feature	NOUN
easat-5713	20	17	,	,	PUNCT
easat-5713	20	18	leading	lead	VERB
easat-5713	20	19	to	to	ADP
easat-5713	20	20	misclassifications	misclassification	NOUN
easat-5713	20	21	and	and	CCONJ
easat-5713	20	22	reduced	reduced	ADJ
easat-5713	20	23	accuracy	accuracy	NOUN
easat-5713	20	24	[	[	X
easat-5713	20	25	2	2	NUM
easat-5713	20	26	,	,	PUNCT
easat-5713	20	27	3	3	NUM
easat-5713	20	28	]	]	PUNCT
easat-5713	20	29	.	.	PUNCT
easat-5713	21	1	addressing	address	VERB
easat-5713	21	2	these	these	DET
easat-5713	21	3	challenges	challenge	NOUN
easat-5713	21	4	is	be	AUX
easat-5713	21	5	crucial	crucial	ADJ
easat-5713	21	6	for	for	ADP
easat-5713	21	7	improving	improve	VERB
easat-5713	21	8	fer	fer	PROPN
easat-5713	21	9	systems	system	NOUN
easat-5713	21	10	'	'	PART
easat-5713	21	11	ability	ability	NOUN
easat-5713	21	12	to	to	PART
easat-5713	21	13	generalize	generalize	VERB
easat-5713	21	14	across	across	ADP
easat-5713	21	15	different	different	ADJ
easat-5713	21	16	environments	environment	NOUN
easat-5713	21	17	,	,	PUNCT
easat-5713	21	18	lighting	lighting	NOUN
easat-5713	21	19	conditions	condition	NOUN
easat-5713	21	20	,	,	PUNCT
easat-5713	21	21	and	and	CCONJ
easat-5713	21	22	cultural	cultural	ADJ
easat-5713	21	23	variations	variation	NOUN
easat-5713	21	24	.	.	PUNCT
easat-5713	22	1	1856	1856	NUM
easat-5713	22	2	edelweiss	edelweiss	PROPN
easat-5713	22	3	applied	apply	VERB
easat-5713	22	4	science	science	NOUN
easat-5713	22	5	and	and	CCONJ
easat-5713	22	6	technology	technology	NOUN
easat-5713	22	7	issn	issn	PROPN
easat-5713	22	8	:	:	PUNCT
easat-5713	22	9	2576	2576	NUM
easat-5713	22	10	-	-	SYM
easat-5713	22	11	8484	8484	NUM
easat-5713	22	12	vol	vol	NOUN
easat-5713	22	13	.	.	PROPN
easat-5713	23	1	9	9	NUM
easat-5713	23	2	,	,	PUNCT
easat-5713	23	3	no	no	INTJ
easat-5713	23	4	.	.	NOUN
easat-5713	24	1	3	3	NUM
easat-5713	24	2	:	:	PUNCT
easat-5713	24	3	1855	1855	NUM
easat-5713	24	4	-	-	SYM
easat-5713	24	5	1869	1869	NUM
easat-5713	24	6	,	,	PUNCT
easat-5713	24	7	2025	2025	NUM
easat-5713	24	8	doi	doi	NOUN
easat-5713	24	9	:	:	PUNCT
easat-5713	24	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	24	11	©	©	ADP
easat-5713	24	12	2025	2025	NUM
easat-5713	24	13	by	by	ADP
easat-5713	24	14	the	the	DET
easat-5713	24	15	authors	author	NOUN
easat-5713	24	16	;	;	PUNCT
easat-5713	24	17	licensee	licensee	PROPN
easat-5713	24	18	learning	learning	NOUN
easat-5713	24	19	gate	gate	NOUN
easat-5713	24	20	to	to	PART
easat-5713	24	21	mitigate	mitigate	VERB
easat-5713	24	22	these	these	DET
easat-5713	24	23	issues	issue	NOUN
easat-5713	24	24	,	,	PUNCT
easat-5713	24	25	researchers	researcher	NOUN
easat-5713	24	26	have	have	AUX
easat-5713	24	27	explored	explore	VERB
easat-5713	24	28	various	various	ADJ
easat-5713	24	29	methods	method	NOUN
easat-5713	24	30	,	,	PUNCT
easat-5713	24	31	including	include	VERB
easat-5713	24	32	data	data	NOUN
easat-5713	24	33	augmentation	augmentation	NOUN
easat-5713	24	34	,	,	PUNCT
easat-5713	24	35	transfer	transfer	NOUN
easat-5713	24	36	learning	learning	NOUN
easat-5713	24	37	,	,	PUNCT
easat-5713	24	38	and	and	CCONJ
easat-5713	24	39	the	the	DET
easat-5713	24	40	integration	integration	NOUN
easat-5713	24	41	of	of	ADP
easat-5713	24	42	synthetic	synthetic	ADJ
easat-5713	24	43	data	datum	NOUN
easat-5713	24	44	[	[	X
easat-5713	24	45	4	4	NUM
easat-5713	24	46	,	,	PUNCT
easat-5713	24	47	5	5	NUM
easat-5713	24	48	]	]	PUNCT
easat-5713	24	49	.	.	PUNCT
easat-5713	25	1	the	the	DET
easat-5713	25	2	emergence	emergence	NOUN
easat-5713	25	3	of	of	ADP
easat-5713	25	4	ai	ai	NOUN
easat-5713	25	5	-	-	PUNCT
easat-5713	25	6	generated	generate	VERB
easat-5713	25	7	synthetic	synthetic	ADJ
easat-5713	25	8	images	image	NOUN
easat-5713	25	9	has	have	AUX
easat-5713	25	10	opened	open	VERB
easat-5713	25	11	new	new	ADJ
easat-5713	25	12	avenues	avenue	NOUN
easat-5713	25	13	for	for	ADP
easat-5713	25	14	enhancing	enhance	VERB
easat-5713	25	15	fer	fer	PROPN
easat-5713	25	16	datasets	dataset	NOUN
easat-5713	25	17	,	,	PUNCT
easat-5713	25	18	allowing	allow	VERB
easat-5713	25	19	for	for	ADP
easat-5713	25	20	increased	increase	VERB
easat-5713	25	21	diversity	diversity	NOUN
easat-5713	25	22	and	and	CCONJ
easat-5713	25	23	improved	improved	ADJ
easat-5713	25	24	model	model	NOUN
easat-5713	25	25	generalization	generalization	NOUN
easat-5713	25	26	.	.	PUNCT
easat-5713	26	1	advanced	advanced	ADJ
easat-5713	26	2	generative	generative	ADJ
easat-5713	26	3	models	model	NOUN
easat-5713	26	4	,	,	PUNCT
easat-5713	26	5	such	such	ADJ
easat-5713	26	6	as	as	ADP
easat-5713	26	7	generative	generative	ADJ
easat-5713	26	8	adversarial	adversarial	ADJ
easat-5713	26	9	networks	network	NOUN
easat-5713	26	10	(	(	PUNCT
easat-5713	26	11	gans	gan	NOUN
easat-5713	26	12	)	)	PUNCT
easat-5713	26	13	and	and	CCONJ
easat-5713	26	14	variational	variational	ADJ
easat-5713	26	15	autoencoders	autoencoder	NOUN
easat-5713	26	16	(	(	PUNCT
easat-5713	26	17	vaes	vaes	ADJ
easat-5713	26	18	)	)	PUNCT
easat-5713	26	19	,	,	PUNCT
easat-5713	26	20	have	have	AUX
easat-5713	26	21	demonstrated	demonstrate	VERB
easat-5713	26	22	their	their	PRON
easat-5713	26	23	ability	ability	NOUN
easat-5713	26	24	to	to	PART
easat-5713	26	25	create	create	VERB
easat-5713	26	26	high	high	ADJ
easat-5713	26	27	-	-	PUNCT
easat-5713	26	28	fidelity	fidelity	NOUN
easat-5713	26	29	images	image	NOUN
easat-5713	26	30	that	that	PRON
easat-5713	26	31	supplement	supplement	VERB
easat-5713	26	32	real	real	ADJ
easat-5713	26	33	datasets	dataset	NOUN
easat-5713	26	34	without	without	ADP
easat-5713	26	35	ethical	ethical	ADJ
easat-5713	26	36	or	or	CCONJ
easat-5713	26	37	logistical	logistical	ADJ
easat-5713	26	38	constraints	constraint	NOUN
easat-5713	26	39	[	[	X
easat-5713	26	40	6	6	NUM
easat-5713	26	41	]	]	PUNCT
easat-5713	26	42	.	.	PUNCT
easat-5713	27	1	these	these	DET
easat-5713	27	2	synthetic	synthetic	ADJ
easat-5713	27	3	images	image	NOUN
easat-5713	27	4	offer	offer	VERB
easat-5713	27	5	the	the	DET
easat-5713	27	6	potential	potential	NOUN
easat-5713	27	7	to	to	PART
easat-5713	27	8	address	address	VERB
easat-5713	27	9	class	class	NOUN
easat-5713	27	10	imbalances	imbalance	NOUN
easat-5713	27	11	and	and	CCONJ
easat-5713	27	12	enrich	enrich	VERB
easat-5713	27	13	training	training	NOUN
easat-5713	27	14	data	datum	NOUN
easat-5713	27	15	,	,	PUNCT
easat-5713	27	16	leading	lead	VERB
easat-5713	27	17	to	to	ADP
easat-5713	27	18	more	more	ADV
easat-5713	27	19	robust	robust	ADJ
easat-5713	27	20	fer	fer	PROPN
easat-5713	27	21	models	model	NOUN
easat-5713	27	22	.	.	PUNCT
easat-5713	28	1	the	the	DET
easat-5713	28	2	landscape	landscape	NOUN
easat-5713	28	3	of	of	ADP
easat-5713	28	4	the	the	DET
easat-5713	28	5	research	research	NOUN
easat-5713	28	6	has	have	AUX
easat-5713	28	7	evolved	evolve	VERB
easat-5713	28	8	significantly	significantly	ADV
easat-5713	28	9	,	,	PUNCT
easat-5713	28	10	with	with	ADP
easat-5713	28	11	ai	ai	ADJ
easat-5713	28	12	applications	application	NOUN
easat-5713	28	13	playing	play	VERB
easat-5713	28	14	a	a	DET
easat-5713	28	15	pivotal	pivotal	ADJ
easat-5713	28	16	role	role	NOUN
easat-5713	28	17	in	in	ADP
easat-5713	28	18	overcoming	overcome	VERB
easat-5713	28	19	dataset	dataset	NOUN
easat-5713	28	20	limitations	limitation	NOUN
easat-5713	28	21	.	.	PUNCT
easat-5713	29	1	the	the	DET
easat-5713	29	2	introduction	introduction	NOUN
easat-5713	29	3	of	of	ADP
easat-5713	29	4	gans	gan	NOUN
easat-5713	29	5	revolutionized	revolutionize	VERB
easat-5713	29	6	image	image	NOUN
easat-5713	29	7	synthesis	synthesis	NOUN
easat-5713	29	8	,	,	PUNCT
easat-5713	29	9	enabling	enable	VERB
easat-5713	29	10	models	model	NOUN
easat-5713	29	11	to	to	PART
easat-5713	29	12	generate	generate	VERB
easat-5713	29	13	realistic	realistic	ADJ
easat-5713	29	14	images	image	NOUN
easat-5713	29	15	by	by	ADP
easat-5713	29	16	learning	learn	VERB
easat-5713	29	17	the	the	DET
easat-5713	29	18	underlying	underlie	VERB
easat-5713	29	19	distribution	distribution	NOUN
easat-5713	29	20	of	of	ADP
easat-5713	29	21	real	real	ADJ
easat-5713	29	22	-	-	PUNCT
easat-5713	29	23	world	world	NOUN
easat-5713	29	24	facial	facial	ADJ
easat-5713	29	25	expressions	expression	NOUN
easat-5713	29	26	[	[	X
easat-5713	29	27	7	7	NUM
easat-5713	29	28	,	,	PUNCT
easat-5713	29	29	8	8	NUM
easat-5713	29	30	]	]	PUNCT
easat-5713	29	31	.	.	PUNCT
easat-5713	30	1	subsequent	subsequent	ADJ
easat-5713	30	2	improvements	improvement	NOUN
easat-5713	30	3	in	in	ADP
easat-5713	30	4	generative	generative	ADJ
easat-5713	30	5	models	model	NOUN
easat-5713	30	6	have	have	AUX
easat-5713	30	7	made	make	VERB
easat-5713	30	8	synthetic	synthetic	ADJ
easat-5713	30	9	images	image	NOUN
easat-5713	30	10	nearly	nearly	ADV
easat-5713	30	11	indistinguishable	indistinguishable	ADJ
easat-5713	30	12	from	from	ADP
easat-5713	30	13	real	real	ADJ
easat-5713	30	14	ones	one	NOUN
easat-5713	30	15	,	,	PUNCT
easat-5713	30	16	leading	lead	VERB
easat-5713	30	17	to	to	ADP
easat-5713	30	18	promising	promising	ADJ
easat-5713	30	19	applications	application	NOUN
easat-5713	30	20	in	in	ADP
easat-5713	30	21	fer	fer	PROPN
easat-5713	30	22	dataset	dataset	NOUN
easat-5713	30	23	augmentation	augmentation	NOUN
easat-5713	30	24	[	[	X
easat-5713	30	25	9	9	NUM
easat-5713	30	26	]	]	PUNCT
easat-5713	30	27	.	.	PUNCT
easat-5713	31	1	several	several	ADJ
easat-5713	31	2	studies	study	NOUN
easat-5713	31	3	have	have	AUX
easat-5713	31	4	explored	explore	VERB
easat-5713	31	5	the	the	DET
easat-5713	31	6	impact	impact	NOUN
easat-5713	31	7	of	of	ADP
easat-5713	31	8	synthetic	synthetic	ADJ
easat-5713	31	9	data	datum	NOUN
easat-5713	31	10	on	on	ADP
easat-5713	31	11	emotion	emotion	NOUN
easat-5713	31	12	recognition	recognition	NOUN
easat-5713	31	13	performance	performance	NOUN
easat-5713	31	14	.	.	PUNCT
easat-5713	32	1	the	the	DET
easat-5713	32	2	research	research	NOUN
easat-5713	32	3	by	by	ADP
easat-5713	32	4	ma	ma	PROPN
easat-5713	32	5	and	and	CCONJ
easat-5713	32	6	nguyen	nguyen	NOUN
easat-5713	33	1	[	[	X
easat-5713	33	2	10	10	NUM
easat-5713	33	3	]	]	PUNCT
easat-5713	33	4	and	and	CCONJ
easat-5713	33	5	he	he	PRON
easat-5713	33	6	,	,	PUNCT
easat-5713	33	7	et	et	PROPN
easat-5713	33	8	al	al	PROPN
easat-5713	33	9	.	.	PUNCT
easat-5713	34	1	[	[	X
easat-5713	34	2	11	11	NUM
easat-5713	34	3	]	]	PUNCT
easat-5713	34	4	demonstrated	demonstrate	VERB
easat-5713	34	5	that	that	SCONJ
easat-5713	34	6	incorporating	incorporate	VERB
easat-5713	34	7	gan	gan	NOUN
easat-5713	34	8	-	-	PUNCT
easat-5713	34	9	generated	generate	VERB
easat-5713	34	10	images	image	NOUN
easat-5713	34	11	into	into	ADP
easat-5713	34	12	training	training	NOUN
easat-5713	34	13	datasets	dataset	NOUN
easat-5713	34	14	could	could	AUX
easat-5713	34	15	enhance	enhance	VERB
easat-5713	34	16	model	model	NOUN
easat-5713	34	17	accuracy	accuracy	NOUN
easat-5713	34	18	and	and	CCONJ
easat-5713	34	19	improve	improve	VERB
easat-5713	34	20	recognition	recognition	NOUN
easat-5713	34	21	of	of	ADP
easat-5713	34	22	underrepresented	underrepresented	ADJ
easat-5713	34	23	emotions	emotion	NOUN
easat-5713	34	24	.	.	PUNCT
easat-5713	35	1	similarly	similarly	ADV
easat-5713	35	2	,	,	PUNCT
easat-5713	35	3	meng	meng	PROPN
easat-5713	35	4	,	,	PUNCT
easat-5713	35	5	et	et	PROPN
easat-5713	35	6	al	al	PROPN
easat-5713	35	7	.	.	PUNCT
easat-5713	36	1	[	[	X
easat-5713	36	2	12	12	NUM
easat-5713	36	3	]	]	PUNCT
easat-5713	36	4	and	and	CCONJ
easat-5713	36	5	kartbayev	kartbayev	X
easat-5713	37	1	[	[	X
easat-5713	37	2	13	13	NUM
easat-5713	37	3	]	]	PUNCT
easat-5713	37	4	showed	show	VERB
easat-5713	37	5	that	that	SCONJ
easat-5713	37	6	synthetic	synthetic	ADJ
easat-5713	37	7	data	datum	NOUN
easat-5713	37	8	could	could	AUX
easat-5713	37	9	increase	increase	VERB
easat-5713	37	10	the	the	DET
easat-5713	37	11	diversity	diversity	NOUN
easat-5713	37	12	of	of	ADP
easat-5713	37	13	training	training	NOUN
easat-5713	37	14	sets	set	NOUN
easat-5713	37	15	,	,	PUNCT
easat-5713	37	16	improving	improve	VERB
easat-5713	37	17	generalization	generalization	NOUN
easat-5713	37	18	to	to	ADP
easat-5713	37	19	unseen	unseen	ADJ
easat-5713	37	20	expressions	expression	NOUN
easat-5713	37	21	.	.	PUNCT
easat-5713	38	1	these	these	DET
easat-5713	38	2	findings	finding	NOUN
easat-5713	38	3	suggest	suggest	VERB
easat-5713	38	4	that	that	SCONJ
easat-5713	38	5	ai	ai	ADJ
easat-5713	38	6	-	-	PUNCT
easat-5713	38	7	generated	generate	VERB
easat-5713	38	8	images	image	NOUN
easat-5713	38	9	can	can	AUX
easat-5713	38	10	complement	complement	VERB
easat-5713	38	11	traditional	traditional	ADJ
easat-5713	38	12	datasets	dataset	NOUN
easat-5713	38	13	,	,	PUNCT
easat-5713	38	14	addressing	address	VERB
easat-5713	38	15	some	some	PRON
easat-5713	38	16	of	of	ADP
easat-5713	38	17	their	their	PRON
easat-5713	38	18	inherent	inherent	ADJ
easat-5713	38	19	weaknesses	weakness	NOUN
easat-5713	38	20	.	.	PUNCT
easat-5713	39	1	however	however	ADV
easat-5713	39	2	,	,	PUNCT
easat-5713	39	3	challenges	challenge	NOUN
easat-5713	39	4	remain	remain	VERB
easat-5713	39	5	in	in	ADP
easat-5713	39	6	effectively	effectively	ADV
easat-5713	39	7	integrating	integrate	VERB
easat-5713	39	8	synthetic	synthetic	ADJ
easat-5713	39	9	data	datum	NOUN
easat-5713	39	10	into	into	ADP
easat-5713	39	11	fer	fer	PROPN
easat-5713	39	12	models	model	NOUN
easat-5713	39	13	.	.	PUNCT
easat-5713	40	1	despite	despite	SCONJ
easat-5713	40	2	the	the	DET
easat-5713	40	3	success	success	NOUN
easat-5713	40	4	of	of	ADP
easat-5713	40	5	gans	gan	NOUN
easat-5713	40	6	in	in	ADP
easat-5713	40	7	generating	generate	VERB
easat-5713	40	8	high	high	ADJ
easat-5713	40	9	-	-	PUNCT
easat-5713	40	10	quality	quality	NOUN
easat-5713	40	11	images	image	NOUN
easat-5713	40	12	,	,	PUNCT
easat-5713	40	13	the	the	DET
easat-5713	40	14	realism	realism	NOUN
easat-5713	40	15	and	and	CCONJ
easat-5713	40	16	variability	variability	NOUN
easat-5713	40	17	of	of	ADP
easat-5713	40	18	synthetic	synthetic	ADJ
easat-5713	40	19	expressions	expression	NOUN
easat-5713	40	20	are	be	AUX
easat-5713	40	21	critical	critical	ADJ
easat-5713	40	22	factors	factor	NOUN
easat-5713	40	23	influencing	influence	VERB
easat-5713	40	24	their	their	PRON
easat-5713	40	25	effectiveness	effectiveness	NOUN
easat-5713	40	26	[	[	X
easat-5713	40	27	14	14	NUM
easat-5713	40	28	]	]	PUNCT
easat-5713	40	29	.	.	PUNCT
easat-5713	41	1	for	for	ADP
easat-5713	41	2	instance	instance	NOUN
easat-5713	41	3	,	,	PUNCT
easat-5713	41	4	studies	study	NOUN
easat-5713	41	5	by	by	ADP
easat-5713	41	6	kim	kim	PROPN
easat-5713	41	7	and	and	CCONJ
easat-5713	41	8	park	park	NOUN
easat-5713	41	9	[	[	X
easat-5713	41	10	15	15	NUM
easat-5713	41	11	]	]	PUNCT
easat-5713	41	12	and	and	CCONJ
easat-5713	41	13	yalçin	yalçin	PROPN
easat-5713	41	14	and	and	CCONJ
easat-5713	41	15	alisawi	alisawi	VERB
easat-5713	42	1	[	[	X
easat-5713	42	2	16	16	NUM
easat-5713	42	3	]	]	PUNCT
easat-5713	42	4	highlighted	highlight	VERB
easat-5713	42	5	that	that	SCONJ
easat-5713	42	6	many	many	ADJ
easat-5713	42	7	synthetic	synthetic	ADJ
easat-5713	42	8	datasets	dataset	NOUN
easat-5713	42	9	fail	fail	VERB
easat-5713	42	10	to	to	PART
easat-5713	42	11	capture	capture	VERB
easat-5713	42	12	the	the	DET
easat-5713	42	13	subtle	subtle	ADJ
easat-5713	42	14	variations	variation	NOUN
easat-5713	42	15	and	and	CCONJ
easat-5713	42	16	context	context	NOUN
easat-5713	42	17	-	-	PUNCT
easat-5713	42	18	specific	specific	ADJ
easat-5713	42	19	nuances	nuance	NOUN
easat-5713	42	20	of	of	ADP
easat-5713	42	21	human	human	ADJ
easat-5713	42	22	facial	facial	ADJ
easat-5713	42	23	expressions	expression	NOUN
easat-5713	42	24	.	.	PUNCT
easat-5713	43	1	poor	poor	ADJ
easat-5713	43	2	-	-	PUNCT
easat-5713	43	3	quality	quality	NOUN
easat-5713	43	4	synthetic	synthetic	ADJ
easat-5713	43	5	images	image	NOUN
easat-5713	43	6	may	may	AUX
easat-5713	43	7	introduce	introduce	VERB
easat-5713	43	8	noise	noise	NOUN
easat-5713	43	9	rather	rather	ADV
easat-5713	43	10	than	than	ADP
easat-5713	43	11	valuable	valuable	ADJ
easat-5713	43	12	learning	learning	NOUN
easat-5713	43	13	signals	signal	NOUN
easat-5713	43	14	,	,	PUNCT
easat-5713	43	15	potentially	potentially	ADV
easat-5713	43	16	degrading	degrading	ADJ
easat-5713	43	17	model	model	NOUN
easat-5713	43	18	performance	performance	NOUN
easat-5713	43	19	rather	rather	ADV
easat-5713	43	20	than	than	ADP
easat-5713	43	21	improving	improve	VERB
easat-5713	43	22	it	it	PRON
easat-5713	43	23	.	.	PUNCT
easat-5713	44	1	another	another	DET
easat-5713	44	2	key	key	ADJ
easat-5713	44	3	challenge	challenge	NOUN
easat-5713	44	4	is	be	AUX
easat-5713	44	5	the	the	DET
easat-5713	44	6	lack	lack	NOUN
easat-5713	44	7	of	of	ADP
easat-5713	44	8	systematic	systematic	ADJ
easat-5713	44	9	evaluations	evaluation	NOUN
easat-5713	44	10	of	of	ADP
easat-5713	44	11	synthetic	synthetic	ADJ
easat-5713	44	12	data	datum	NOUN
easat-5713	44	13	integration	integration	NOUN
easat-5713	44	14	in	in	ADP
easat-5713	44	15	fer	fer	PROPN
easat-5713	44	16	models	model	NOUN
easat-5713	44	17	.	.	PUNCT
easat-5713	45	1	while	while	SCONJ
easat-5713	45	2	numerous	numerous	ADJ
easat-5713	45	3	studies	study	NOUN
easat-5713	45	4	have	have	AUX
easat-5713	45	5	demonstrated	demonstrate	VERB
easat-5713	45	6	improvements	improvement	NOUN
easat-5713	45	7	in	in	ADP
easat-5713	45	8	fer	fer	PROPN
easat-5713	45	9	performance	performance	NOUN
easat-5713	45	10	using	use	VERB
easat-5713	45	11	gans	gan	NOUN
easat-5713	45	12	,	,	PUNCT
easat-5713	45	13	few	few	ADJ
easat-5713	45	14	have	have	AUX
easat-5713	45	15	analyzed	analyze	VERB
easat-5713	45	16	the	the	DET
easat-5713	45	17	specific	specific	ADJ
easat-5713	45	18	effects	effect	NOUN
easat-5713	45	19	of	of	ADP
easat-5713	45	20	synthetic	synthetic	ADJ
easat-5713	45	21	images	image	NOUN
easat-5713	45	22	on	on	ADP
easat-5713	45	23	different	different	ADJ
easat-5713	45	24	emotion	emotion	NOUN
easat-5713	45	25	categories	category	NOUN
easat-5713	45	26	or	or	CCONJ
easat-5713	45	27	explored	explore	VERB
easat-5713	45	28	their	their	PRON
easat-5713	45	29	impact	impact	NOUN
easat-5713	45	30	on	on	ADP
easat-5713	45	31	key	key	ADJ
easat-5713	45	32	metrics	metric	NOUN
easat-5713	45	33	such	such	ADJ
easat-5713	45	34	as	as	ADP
easat-5713	45	35	precision	precision	NOUN
easat-5713	45	36	,	,	PUNCT
easat-5713	45	37	recall	recall	NOUN
easat-5713	45	38	,	,	PUNCT
easat-5713	45	39	and	and	CCONJ
easat-5713	45	40	f1	f1	NOUN
easat-5713	45	41	-	-	PUNCT
easat-5713	45	42	score	score	NOUN
easat-5713	45	43	[	[	X
easat-5713	45	44	17	17	NUM
easat-5713	45	45	]	]	PUNCT
easat-5713	45	46	.	.	PUNCT
easat-5713	46	1	a	a	DET
easat-5713	46	2	comprehensive	comprehensive	ADJ
easat-5713	46	3	evaluation	evaluation	NOUN
easat-5713	46	4	of	of	ADP
easat-5713	46	5	synthetic	synthetic	ADJ
easat-5713	46	6	data	datum	NOUN
easat-5713	46	7	’s	’s	PART
easat-5713	46	8	role	role	NOUN
easat-5713	46	9	in	in	ADP
easat-5713	46	10	mitigating	mitigate	VERB
easat-5713	46	11	dataset	dataset	NOUN
easat-5713	46	12	limitations	limitation	NOUN
easat-5713	46	13	,	,	PUNCT
easat-5713	46	14	especially	especially	ADV
easat-5713	46	15	noisy	noisy	ADJ
easat-5713	46	16	images	image	NOUN
easat-5713	46	17	,	,	PUNCT
easat-5713	46	18	is	be	AUX
easat-5713	46	19	essential	essential	ADJ
easat-5713	46	20	for	for	ADP
easat-5713	46	21	advancing	advance	VERB
easat-5713	46	22	this	this	DET
easat-5713	46	23	research	research	NOUN
easat-5713	46	24	.	.	PUNCT
easat-5713	47	1	this	this	DET
easat-5713	47	2	study	study	NOUN
easat-5713	47	3	aims	aim	VERB
easat-5713	47	4	to	to	PART
easat-5713	47	5	address	address	VERB
easat-5713	47	6	these	these	DET
easat-5713	47	7	research	research	NOUN
easat-5713	47	8	gaps	gap	NOUN
easat-5713	47	9	by	by	ADP
easat-5713	47	10	systematically	systematically	ADV
easat-5713	47	11	investigating	investigate	VERB
easat-5713	47	12	the	the	DET
easat-5713	47	13	impact	impact	NOUN
easat-5713	47	14	of	of	ADP
easat-5713	47	15	aigenerated	aigenerate	VERB
easat-5713	47	16	synthetic	synthetic	ADJ
easat-5713	47	17	data	datum	NOUN
easat-5713	47	18	on	on	ADP
easat-5713	47	19	fer	fer	PROPN
easat-5713	47	20	model	model	NOUN
easat-5713	47	21	performance	performance	NOUN
easat-5713	47	22	.	.	PUNCT
easat-5713	48	1	specifically	specifically	ADV
easat-5713	48	2	,	,	PUNCT
easat-5713	48	3	we	we	PRON
easat-5713	48	4	seek	seek	VERB
easat-5713	48	5	to	to	PART
easat-5713	48	6	answer	answer	VERB
easat-5713	48	7	the	the	DET
easat-5713	48	8	following	follow	VERB
easat-5713	48	9	research	research	NOUN
easat-5713	48	10	questions	question	NOUN
easat-5713	48	11	:	:	PUNCT
easat-5713	49	1	1	1	X
easat-5713	49	2	.	.	X
easat-5713	49	3	how	how	SCONJ
easat-5713	49	4	does	do	AUX
easat-5713	49	5	the	the	DET
easat-5713	49	6	integration	integration	NOUN
easat-5713	49	7	of	of	ADP
easat-5713	49	8	ai	ai	VERB
easat-5713	49	9	-	-	PUNCT
easat-5713	49	10	generated	generate	VERB
easat-5713	49	11	synthetic	synthetic	ADJ
easat-5713	49	12	images	image	NOUN
easat-5713	49	13	affect	affect	VERB
easat-5713	49	14	the	the	DET
easat-5713	49	15	accuracy	accuracy	NOUN
easat-5713	49	16	of	of	ADP
easat-5713	49	17	fer	fer	PROPN
easat-5713	49	18	models	model	NOUN
easat-5713	49	19	,	,	PUNCT
easat-5713	49	20	particularly	particularly	ADV
easat-5713	49	21	in	in	ADP
easat-5713	49	22	handling	handle	VERB
easat-5713	49	23	noisy	noisy	ADJ
easat-5713	49	24	data	datum	NOUN
easat-5713	49	25	?	?	PUNCT
easat-5713	50	1	2	2	X
easat-5713	50	2	.	.	X
easat-5713	50	3	do	do	AUX
easat-5713	50	4	synthetic	synthetic	ADJ
easat-5713	50	5	images	image	NOUN
easat-5713	50	6	improve	improve	VERB
easat-5713	50	7	the	the	DET
easat-5713	50	8	recognition	recognition	NOUN
easat-5713	50	9	of	of	ADP
easat-5713	50	10	specific	specific	ADJ
easat-5713	50	11	emotion	emotion	NOUN
easat-5713	50	12	categories	category	NOUN
easat-5713	50	13	,	,	PUNCT
easat-5713	50	14	or	or	CCONJ
easat-5713	50	15	do	do	AUX
easat-5713	50	16	they	they	PRON
easat-5713	50	17	introduce	introduce	VERB
easat-5713	50	18	additional	additional	ADJ
easat-5713	50	19	noise	noise	NOUN
easat-5713	50	20	into	into	ADP
easat-5713	50	21	the	the	DET
easat-5713	50	22	dataset	dataset	NOUN
easat-5713	50	23	?	?	PUNCT
easat-5713	51	1	3	3	X
easat-5713	51	2	.	.	X
easat-5713	51	3	what	what	PRON
easat-5713	51	4	are	be	AUX
easat-5713	51	5	the	the	DET
easat-5713	51	6	key	key	ADJ
easat-5713	51	7	factors	factor	NOUN
easat-5713	51	8	influencing	influence	VERB
easat-5713	51	9	the	the	DET
easat-5713	51	10	effectiveness	effectiveness	NOUN
easat-5713	51	11	of	of	ADP
easat-5713	51	12	synthetic	synthetic	ADJ
easat-5713	51	13	data	datum	NOUN
easat-5713	51	14	in	in	ADP
easat-5713	51	15	fer	fer	PROPN
easat-5713	51	16	,	,	PUNCT
easat-5713	51	17	including	include	VERB
easat-5713	51	18	realism	realism	NOUN
easat-5713	51	19	,	,	PUNCT
easat-5713	51	20	diversity	diversity	NOUN
easat-5713	51	21	,	,	PUNCT
easat-5713	51	22	and	and	CCONJ
easat-5713	51	23	integration	integration	NOUN
easat-5713	51	24	strategies	strategy	NOUN
easat-5713	51	25	?	?	PUNCT
easat-5713	52	1	based	base	VERB
easat-5713	52	2	on	on	ADP
easat-5713	52	3	prior	prior	ADJ
easat-5713	52	4	research	research	NOUN
easat-5713	52	5	and	and	CCONJ
easat-5713	52	6	existing	exist	VERB
easat-5713	52	7	challenges	challenge	NOUN
easat-5713	52	8	,	,	PUNCT
easat-5713	52	9	we	we	PRON
easat-5713	52	10	formulate	formulate	VERB
easat-5713	52	11	the	the	DET
easat-5713	52	12	following	following	ADJ
easat-5713	52	13	hypotheses	hypothesis	NOUN
easat-5713	52	14	:	:	PUNCT
easat-5713	52	15	h1	h1	NOUN
easat-5713	52	16	:	:	PUNCT
easat-5713	52	17	the	the	DET
easat-5713	52	18	inclusion	inclusion	NOUN
easat-5713	52	19	of	of	ADP
easat-5713	52	20	ai	ai	ADJ
easat-5713	52	21	-	-	PUNCT
easat-5713	52	22	generated	generate	VERB
easat-5713	52	23	synthetic	synthetic	ADJ
easat-5713	52	24	images	image	NOUN
easat-5713	52	25	in	in	ADP
easat-5713	52	26	fer	fer	PROPN
easat-5713	52	27	model	model	NOUN
easat-5713	52	28	training	training	NOUN
easat-5713	52	29	will	will	AUX
easat-5713	52	30	improve	improve	VERB
easat-5713	52	31	classification	classification	NOUN
easat-5713	52	32	accuracy	accuracy	NOUN
easat-5713	52	33	,	,	PUNCT
easat-5713	52	34	particularly	particularly	ADV
easat-5713	52	35	for	for	ADP
easat-5713	52	36	underrepresented	underrepresented	ADJ
easat-5713	52	37	emotion	emotion	NOUN
easat-5713	52	38	categories	category	NOUN
easat-5713	52	39	.	.	PUNCT
easat-5713	53	1	h2	h2	NOUN
easat-5713	53	2	:	:	PUNCT
easat-5713	53	3	synthetic	synthetic	ADJ
easat-5713	53	4	images	image	NOUN
easat-5713	53	5	will	will	AUX
easat-5713	53	6	help	help	VERB
easat-5713	53	7	reduce	reduce	VERB
easat-5713	53	8	the	the	DET
easat-5713	53	9	impact	impact	NOUN
easat-5713	53	10	of	of	ADP
easat-5713	53	11	noisy	noisy	ADJ
easat-5713	53	12	images	image	NOUN
easat-5713	53	13	by	by	ADP
easat-5713	53	14	providing	provide	VERB
easat-5713	53	15	additional	additional	ADJ
easat-5713	53	16	training	training	NOUN
easat-5713	53	17	examples	example	NOUN
easat-5713	53	18	that	that	PRON
easat-5713	53	19	enhance	enhance	VERB
easat-5713	53	20	model	model	NOUN
easat-5713	53	21	generalization	generalization	NOUN
easat-5713	53	22	.	.	PUNCT
easat-5713	54	1	h3	h3	NOUN
easat-5713	54	2	:	:	PUNCT
easat-5713	54	3	the	the	DET
easat-5713	54	4	effectiveness	effectiveness	NOUN
easat-5713	54	5	of	of	ADP
easat-5713	54	6	synthetic	synthetic	ADJ
easat-5713	54	7	data	datum	NOUN
easat-5713	54	8	augmentation	augmentation	NOUN
easat-5713	54	9	is	be	AUX
easat-5713	54	10	highly	highly	ADV
easat-5713	54	11	dependent	dependent	ADJ
easat-5713	54	12	on	on	ADP
easat-5713	54	13	the	the	DET
easat-5713	54	14	quality	quality	NOUN
easat-5713	54	15	and	and	CCONJ
easat-5713	54	16	realism	realism	NOUN
easat-5713	54	17	of	of	ADP
easat-5713	54	18	generated	generate	VERB
easat-5713	54	19	images	image	NOUN
easat-5713	54	20	;	;	PUNCT
easat-5713	54	21	lower	low	ADJ
easat-5713	54	22	-	-	PUNCT
easat-5713	54	23	quality	quality	NOUN
easat-5713	54	24	images	image	NOUN
easat-5713	54	25	may	may	AUX
easat-5713	54	26	introduce	introduce	VERB
easat-5713	54	27	noise	noise	NOUN
easat-5713	54	28	and	and	CCONJ
easat-5713	54	29	reduce	reduce	VERB
easat-5713	54	30	model	model	NOUN
easat-5713	54	31	performance	performance	NOUN
easat-5713	54	32	.	.	PUNCT
easat-5713	55	1	1857	1857	NUM
easat-5713	55	2	edelweiss	edelweiss	PROPN
easat-5713	55	3	applied	apply	VERB
easat-5713	55	4	science	science	NOUN
easat-5713	55	5	and	and	CCONJ
easat-5713	55	6	technology	technology	NOUN
easat-5713	55	7	issn	issn	PROPN
easat-5713	55	8	:	:	PUNCT
easat-5713	55	9	2576	2576	NUM
easat-5713	55	10	-	-	SYM
easat-5713	55	11	8484	8484	NUM
easat-5713	55	12	vol	vol	NOUN
easat-5713	55	13	.	.	PROPN
easat-5713	56	1	9	9	NUM
easat-5713	56	2	,	,	PUNCT
easat-5713	56	3	no	no	INTJ
easat-5713	56	4	.	.	NOUN
easat-5713	57	1	3	3	NUM
easat-5713	57	2	:	:	PUNCT
easat-5713	57	3	1855	1855	NUM
easat-5713	57	4	-	-	SYM
easat-5713	57	5	1869	1869	NUM
easat-5713	57	6	,	,	PUNCT
easat-5713	57	7	2025	2025	NUM
easat-5713	57	8	doi	doi	NOUN
easat-5713	57	9	:	:	PUNCT
easat-5713	57	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	57	11	©	©	ADP
easat-5713	57	12	2025	2025	NUM
easat-5713	57	13	by	by	ADP
easat-5713	57	14	the	the	DET
easat-5713	57	15	authors	author	NOUN
easat-5713	57	16	;	;	PUNCT
easat-5713	57	17	licensee	licensee	PROPN
easat-5713	57	18	learning	learn	VERB
easat-5713	57	19	gate	gate	NOUN
easat-5713	57	20	to	to	PART
easat-5713	57	21	test	test	VERB
easat-5713	57	22	these	these	DET
easat-5713	57	23	hypotheses	hypothesis	NOUN
easat-5713	57	24	,	,	PUNCT
easat-5713	57	25	we	we	PRON
easat-5713	57	26	utilize	utilize	VERB
easat-5713	57	27	the	the	DET
easat-5713	57	28	juggernaut	juggernaut	PROPN
easat-5713	57	29	xl	xl	PROPN
easat-5713	57	30	model	model	PROPN
easat-5713	57	31	to	to	PART
easat-5713	57	32	generate	generate	VERB
easat-5713	57	33	300	300	NUM
easat-5713	57	34	synthetic	synthetic	ADJ
easat-5713	57	35	images	image	NOUN
easat-5713	57	36	for	for	ADP
easat-5713	57	37	each	each	DET
easat-5713	57	38	emotion	emotion	NOUN
easat-5713	57	39	category	category	NOUN
easat-5713	57	40	in	in	ADP
easat-5713	57	41	the	the	DET
easat-5713	57	42	fer2013	fer2013	ADJ
easat-5713	57	43	dataset	dataset	NOUN
easat-5713	57	44	[	[	X
easat-5713	57	45	18	18	NUM
easat-5713	57	46	]	]	PUNCT
easat-5713	57	47	.	.	PUNCT
easat-5713	58	1	these	these	DET
easat-5713	58	2	images	image	NOUN
easat-5713	58	3	are	be	AUX
easat-5713	58	4	then	then	ADV
easat-5713	58	5	integrated	integrate	VERB
easat-5713	58	6	into	into	ADP
easat-5713	58	7	the	the	DET
easat-5713	58	8	training	training	NOUN
easat-5713	58	9	process	process	NOUN
easat-5713	58	10	of	of	ADP
easat-5713	58	11	a	a	DET
easat-5713	58	12	vgg-19	vgg-19	ADV
easat-5713	58	13	-	-	PUNCT
easat-5713	58	14	based	base	VERB
easat-5713	58	15	model	model	NOUN
easat-5713	58	16	,	,	PUNCT
easat-5713	58	17	allowing	allow	VERB
easat-5713	58	18	for	for	ADP
easat-5713	58	19	a	a	DET
easat-5713	58	20	direct	direct	ADJ
easat-5713	58	21	comparison	comparison	NOUN
easat-5713	58	22	of	of	ADP
easat-5713	58	23	model	model	NOUN
easat-5713	58	24	performance	performance	NOUN
easat-5713	58	25	with	with	ADP
easat-5713	58	26	and	and	CCONJ
easat-5713	58	27	without	without	ADP
easat-5713	58	28	synthetic	synthetic	ADJ
easat-5713	58	29	data	datum	NOUN
easat-5713	58	30	augmentation	augmentation	NOUN
easat-5713	58	31	.	.	PUNCT
easat-5713	59	1	key	key	ADJ
easat-5713	59	2	evaluation	evaluation	NOUN
easat-5713	59	3	metrics	metric	NOUN
easat-5713	59	4	include	include	VERB
easat-5713	59	5	accuracy	accuracy	NOUN
easat-5713	59	6	,	,	PUNCT
easat-5713	59	7	precision	precision	NOUN
easat-5713	59	8	,	,	PUNCT
easat-5713	59	9	recall	recall	NOUN
easat-5713	59	10	,	,	PUNCT
easat-5713	59	11	and	and	CCONJ
easat-5713	59	12	f1score	f1score	NOUN
easat-5713	59	13	,	,	PUNCT
easat-5713	59	14	analyzed	analyze	VERB
easat-5713	59	15	across	across	ADP
easat-5713	59	16	different	different	ADJ
easat-5713	59	17	emotion	emotion	NOUN
easat-5713	59	18	categories	category	NOUN
easat-5713	59	19	to	to	PART
easat-5713	59	20	assess	assess	VERB
easat-5713	59	21	the	the	DET
easat-5713	59	22	impact	impact	NOUN
easat-5713	59	23	of	of	ADP
easat-5713	59	24	synthetic	synthetic	ADJ
easat-5713	59	25	images	image	NOUN
easat-5713	59	26	comprehensively	comprehensively	ADV
easat-5713	59	27	.	.	PUNCT
easat-5713	60	1	a	a	DET
easat-5713	60	2	critical	critical	ADJ
easat-5713	60	3	aspect	aspect	NOUN
easat-5713	60	4	of	of	ADP
easat-5713	60	5	our	our	PRON
easat-5713	60	6	approach	approach	NOUN
easat-5713	60	7	is	be	AUX
easat-5713	60	8	evaluating	evaluate	VERB
easat-5713	60	9	how	how	SCONJ
easat-5713	60	10	synthetic	synthetic	ADJ
easat-5713	60	11	images	image	NOUN
easat-5713	60	12	influence	influence	NOUN
easat-5713	60	13	model	model	NOUN
easat-5713	60	14	performance	performance	NOUN
easat-5713	60	15	in	in	ADP
easat-5713	60	16	the	the	DET
easat-5713	60	17	presence	presence	NOUN
easat-5713	60	18	of	of	ADP
easat-5713	60	19	noisy	noisy	ADJ
easat-5713	60	20	data	datum	NOUN
easat-5713	60	21	.	.	PUNCT
easat-5713	61	1	the	the	DET
easat-5713	61	2	fer2013	fer2013	ADJ
easat-5713	61	3	dataset	dataset	NOUN
easat-5713	61	4	is	be	AUX
easat-5713	61	5	known	know	VERB
easat-5713	61	6	to	to	PART
easat-5713	61	7	contain	contain	VERB
easat-5713	61	8	a	a	DET
easat-5713	61	9	significant	significant	ADJ
easat-5713	61	10	number	number	NOUN
easat-5713	61	11	of	of	ADP
easat-5713	61	12	lowquality	lowquality	NOUN
easat-5713	61	13	images	image	NOUN
easat-5713	61	14	,	,	PUNCT
easat-5713	61	15	including	include	VERB
easat-5713	61	16	blurred	blurred	ADJ
easat-5713	61	17	,	,	PUNCT
easat-5713	61	18	occluded	occluded	ADJ
easat-5713	61	19	,	,	PUNCT
easat-5713	61	20	and	and	CCONJ
easat-5713	61	21	poorly	poorly	ADV
easat-5713	61	22	lit	light	VERB
easat-5713	61	23	expressions	expression	NOUN
easat-5713	61	24	.	.	PUNCT
easat-5713	62	1	by	by	ADP
easat-5713	62	2	examining	examine	VERB
easat-5713	62	3	classification	classification	NOUN
easat-5713	62	4	results	result	NOUN
easat-5713	62	5	using	use	VERB
easat-5713	62	6	confusion	confusion	NOUN
easat-5713	62	7	matrices	matrix	NOUN
easat-5713	62	8	and	and	CCONJ
easat-5713	62	9	classification	classification	NOUN
easat-5713	62	10	reports	report	NOUN
easat-5713	62	11	,	,	PUNCT
easat-5713	62	12	we	we	PRON
easat-5713	62	13	aim	aim	VERB
easat-5713	62	14	to	to	PART
easat-5713	62	15	determine	determine	VERB
easat-5713	62	16	whether	whether	SCONJ
easat-5713	62	17	synthetic	synthetic	ADJ
easat-5713	62	18	data	datum	NOUN
easat-5713	62	19	helps	help	VERB
easat-5713	62	20	mitigate	mitigate	VERB
easat-5713	62	21	these	these	DET
easat-5713	62	22	challenges	challenge	NOUN
easat-5713	62	23	or	or	CCONJ
easat-5713	62	24	exacerbates	exacerbate	VERB
easat-5713	62	25	them	they	PRON
easat-5713	62	26	.	.	PUNCT
easat-5713	63	1	this	this	DET
easat-5713	63	2	research	research	NOUN
easat-5713	63	3	contributes	contribute	VERB
easat-5713	63	4	to	to	ADP
easat-5713	63	5	the	the	DET
easat-5713	63	6	field	field	NOUN
easat-5713	63	7	of	of	ADP
easat-5713	63	8	fer	fer	PROPN
easat-5713	63	9	in	in	ADP
easat-5713	63	10	several	several	ADJ
easat-5713	63	11	key	key	ADJ
easat-5713	63	12	ways	way	NOUN
easat-5713	63	13	:	:	PUNCT
easat-5713	63	14	•	•	NUM
easat-5713	63	15	evaluation	evaluation	NOUN
easat-5713	63	16	of	of	ADP
easat-5713	63	17	synthetic	synthetic	ADJ
easat-5713	63	18	data	datum	NOUN
easat-5713	63	19	:	:	PUNCT
easat-5713	63	20	unlike	unlike	ADP
easat-5713	63	21	prior	prior	ADJ
easat-5713	63	22	studies	study	NOUN
easat-5713	63	23	that	that	PRON
easat-5713	63	24	focus	focus	VERB
easat-5713	63	25	solely	solely	ADV
easat-5713	63	26	on	on	ADP
easat-5713	63	27	generating	generate	VERB
easat-5713	63	28	synthetic	synthetic	ADJ
easat-5713	63	29	images	image	NOUN
easat-5713	63	30	,	,	PUNCT
easat-5713	63	31	we	we	PRON
easat-5713	63	32	systematically	systematically	ADV
easat-5713	63	33	assess	assess	VERB
easat-5713	63	34	their	their	PRON
easat-5713	63	35	impact	impact	NOUN
easat-5713	63	36	on	on	ADP
easat-5713	63	37	fer	fer	PROPN
easat-5713	63	38	model	model	NOUN
easat-5713	63	39	performance	performance	NOUN
easat-5713	63	40	across	across	ADP
easat-5713	63	41	multiple	multiple	ADJ
easat-5713	63	42	metrics	metric	NOUN
easat-5713	63	43	.	.	PUNCT
easat-5713	64	1	•	•	NUM
easat-5713	64	2	addressing	address	VERB
easat-5713	64	3	noisy	noisy	ADJ
easat-5713	64	4	data	datum	NOUN
easat-5713	64	5	challenges	challenge	NOUN
easat-5713	64	6	:	:	PUNCT
easat-5713	65	1	by	by	ADP
easat-5713	65	2	analyzing	analyze	VERB
easat-5713	65	3	how	how	SCONJ
easat-5713	65	4	synthetic	synthetic	ADJ
easat-5713	65	5	data	data	NOUN
easat-5713	65	6	interacts	interact	VERB
easat-5713	65	7	with	with	ADP
easat-5713	65	8	noisy	noisy	ADJ
easat-5713	65	9	images	image	NOUN
easat-5713	65	10	,	,	PUNCT
easat-5713	65	11	this	this	DET
easat-5713	65	12	study	study	NOUN
easat-5713	65	13	provides	provide	VERB
easat-5713	65	14	insights	insight	NOUN
easat-5713	65	15	into	into	ADP
easat-5713	65	16	the	the	DET
easat-5713	65	17	effectiveness	effectiveness	NOUN
easat-5713	65	18	of	of	ADP
easat-5713	65	19	ai	ai	ADJ
easat-5713	65	20	-	-	PUNCT
easat-5713	65	21	generated	generate	VERB
easat-5713	65	22	augmentation	augmentation	NOUN
easat-5713	65	23	in	in	ADP
easat-5713	65	24	real	real	ADJ
easat-5713	65	25	-	-	PUNCT
easat-5713	65	26	world	world	NOUN
easat-5713	65	27	fer	fer	PROPN
easat-5713	65	28	applications	application	NOUN
easat-5713	65	29	.	.	PUNCT
easat-5713	66	1	•	•	NUM
easat-5713	66	2	optimizing	optimize	VERB
easat-5713	66	3	synthetic	synthetic	ADJ
easat-5713	66	4	image	image	NOUN
easat-5713	66	5	generation	generation	NOUN
easat-5713	66	6	:	:	PUNCT
easat-5713	66	7	the	the	DET
easat-5713	66	8	findings	finding	NOUN
easat-5713	66	9	will	will	AUX
easat-5713	66	10	inform	inform	VERB
easat-5713	66	11	future	future	ADJ
easat-5713	66	12	work	work	NOUN
easat-5713	66	13	on	on	ADP
easat-5713	66	14	optimizing	optimize	VERB
easat-5713	66	15	synthetic	synthetic	ADJ
easat-5713	66	16	image	image	NOUN
easat-5713	66	17	generation	generation	NOUN
easat-5713	66	18	and	and	CCONJ
easat-5713	66	19	integration	integration	NOUN
easat-5713	66	20	strategies	strategy	NOUN
easat-5713	66	21	,	,	PUNCT
easat-5713	66	22	paving	pave	VERB
easat-5713	66	23	the	the	DET
easat-5713	66	24	way	way	NOUN
easat-5713	66	25	for	for	ADP
easat-5713	66	26	more	more	ADV
easat-5713	66	27	effective	effective	ADJ
easat-5713	66	28	fer	fer	PROPN
easat-5713	66	29	models	model	NOUN
easat-5713	66	30	.	.	PUNCT
easat-5713	67	1	despite	despite	SCONJ
easat-5713	67	2	the	the	DET
easat-5713	67	3	growing	grow	VERB
easat-5713	67	4	interest	interest	NOUN
easat-5713	67	5	in	in	ADP
easat-5713	67	6	ai	ai	ADJ
easat-5713	67	7	-	-	PUNCT
easat-5713	67	8	generated	generate	VERB
easat-5713	67	9	data	datum	NOUN
easat-5713	67	10	augmentation	augmentation	NOUN
easat-5713	67	11	,	,	PUNCT
easat-5713	67	12	many	many	ADJ
easat-5713	67	13	questions	question	NOUN
easat-5713	67	14	remain	remain	VERB
easat-5713	67	15	regarding	regard	VERB
easat-5713	67	16	its	its	PRON
easat-5713	67	17	effectiveness	effectiveness	NOUN
easat-5713	67	18	and	and	CCONJ
easat-5713	67	19	practical	practical	ADJ
easat-5713	67	20	implementation	implementation	NOUN
easat-5713	67	21	.	.	PUNCT
easat-5713	68	1	by	by	ADP
easat-5713	68	2	investigating	investigate	VERB
easat-5713	68	3	the	the	DET
easat-5713	68	4	potential	potential	ADJ
easat-5713	68	5	benefits	benefit	NOUN
easat-5713	68	6	and	and	CCONJ
easat-5713	68	7	limitations	limitation	NOUN
easat-5713	68	8	of	of	ADP
easat-5713	68	9	synthetic	synthetic	ADJ
easat-5713	68	10	images	image	NOUN
easat-5713	68	11	in	in	ADP
easat-5713	68	12	fer	fer	NOUN
easat-5713	68	13	,	,	PUNCT
easat-5713	68	14	this	this	DET
easat-5713	68	15	study	study	NOUN
easat-5713	68	16	seeks	seek	VERB
easat-5713	68	17	to	to	PART
easat-5713	68	18	advance	advance	VERB
easat-5713	68	19	the	the	DET
easat-5713	68	20	field	field	NOUN
easat-5713	68	21	and	and	CCONJ
easat-5713	68	22	contribute	contribute	VERB
easat-5713	68	23	to	to	ADP
easat-5713	68	24	the	the	DET
easat-5713	68	25	development	development	NOUN
easat-5713	68	26	of	of	ADP
easat-5713	68	27	more	more	ADV
easat-5713	68	28	accurate	accurate	ADJ
easat-5713	68	29	and	and	CCONJ
easat-5713	68	30	robust	robust	ADJ
easat-5713	68	31	emotion	emotion	NOUN
easat-5713	68	32	recognition	recognition	NOUN
easat-5713	68	33	systems	system	NOUN
easat-5713	68	34	.	.	PUNCT
easat-5713	69	1	2	2	X
easat-5713	69	2	.	.	X
easat-5713	69	3	methods	method	NOUN
easat-5713	69	4	and	and	CCONJ
easat-5713	69	5	methodology	methodology	NOUN
easat-5713	69	6	the	the	DET
easat-5713	69	7	development	development	NOUN
easat-5713	69	8	of	of	ADP
easat-5713	69	9	a	a	DET
easat-5713	69	10	robust	robust	ADJ
easat-5713	69	11	model	model	NOUN
easat-5713	69	12	using	use	VERB
easat-5713	69	13	ai	ai	NOUN
easat-5713	69	14	-	-	PUNCT
easat-5713	69	15	generated	generate	VERB
easat-5713	69	16	synthetic	synthetic	ADJ
easat-5713	69	17	data	datum	NOUN
easat-5713	69	18	involves	involve	VERB
easat-5713	69	19	a	a	DET
easat-5713	69	20	series	series	NOUN
easat-5713	69	21	of	of	ADP
easat-5713	69	22	methodical	methodical	ADJ
easat-5713	69	23	steps	step	NOUN
easat-5713	69	24	,	,	PUNCT
easat-5713	69	25	including	include	VERB
easat-5713	69	26	data	datum	NOUN
easat-5713	69	27	preprocessing	preprocessing	NOUN
easat-5713	69	28	,	,	PUNCT
easat-5713	69	29	synthetic	synthetic	ADJ
easat-5713	69	30	image	image	NOUN
easat-5713	69	31	generation	generation	NOUN
easat-5713	69	32	,	,	PUNCT
easat-5713	69	33	model	model	NOUN
easat-5713	69	34	training	training	NOUN
easat-5713	69	35	,	,	PUNCT
easat-5713	69	36	and	and	CCONJ
easat-5713	69	37	performance	performance	NOUN
easat-5713	69	38	evaluation	evaluation	NOUN
easat-5713	69	39	.	.	PUNCT
easat-5713	70	1	this	this	DET
easat-5713	70	2	research	research	NOUN
easat-5713	70	3	focuses	focus	VERB
easat-5713	70	4	on	on	ADP
easat-5713	70	5	the	the	DET
easat-5713	70	6	integration	integration	NOUN
easat-5713	70	7	of	of	ADP
easat-5713	70	8	synthetic	synthetic	ADJ
easat-5713	70	9	images	image	NOUN
easat-5713	70	10	generated	generate	VERB
easat-5713	70	11	by	by	ADP
easat-5713	70	12	the	the	DET
easat-5713	70	13	juggernaut	juggernaut	PROPN
easat-5713	70	14	xl	xl	PROPN
easat-5713	70	15	model	model	PROPN
easat-5713	70	16	and	and	CCONJ
easat-5713	70	17	the	the	DET
easat-5713	70	18	training	training	NOUN
easat-5713	70	19	of	of	ADP
easat-5713	70	20	a	a	DET
easat-5713	70	21	vgg-19	vgg-19	ADV
easat-5713	70	22	-	-	PUNCT
easat-5713	70	23	based	base	VERB
easat-5713	70	24	fer	fer	PROPN
easat-5713	70	25	model	model	NOUN
easat-5713	70	26	to	to	PART
easat-5713	70	27	analyze	analyze	VERB
easat-5713	70	28	the	the	DET
easat-5713	70	29	impact	impact	NOUN
easat-5713	70	30	of	of	ADP
easat-5713	70	31	synthetic	synthetic	ADJ
easat-5713	70	32	data	datum	NOUN
easat-5713	70	33	on	on	ADP
easat-5713	70	34	overall	overall	ADJ
easat-5713	70	35	model	model	NOUN
easat-5713	70	36	performance	performance	NOUN
easat-5713	70	37	,	,	PUNCT
easat-5713	70	38	particularly	particularly	ADV
easat-5713	70	39	in	in	ADP
easat-5713	70	40	reducing	reduce	VERB
easat-5713	70	41	the	the	DET
easat-5713	70	42	effects	effect	NOUN
easat-5713	70	43	of	of	ADP
easat-5713	70	44	noisy	noisy	ADJ
easat-5713	70	45	images	image	NOUN
easat-5713	70	46	.	.	PUNCT
easat-5713	71	1	a	a	DET
easat-5713	71	2	core	core	NOUN
easat-5713	71	3	objective	objective	NOUN
easat-5713	71	4	of	of	ADP
easat-5713	71	5	this	this	DET
easat-5713	71	6	study	study	NOUN
easat-5713	71	7	is	be	AUX
easat-5713	71	8	to	to	PART
easat-5713	71	9	assess	assess	VERB
easat-5713	71	10	whether	whether	SCONJ
easat-5713	71	11	synthetic	synthetic	ADJ
easat-5713	71	12	images	image	NOUN
easat-5713	71	13	help	help	VERB
easat-5713	71	14	mitigate	mitigate	VERB
easat-5713	71	15	issues	issue	NOUN
easat-5713	71	16	related	relate	VERB
easat-5713	71	17	to	to	ADP
easat-5713	71	18	noisy	noisy	ADJ
easat-5713	71	19	data	datum	NOUN
easat-5713	71	20	,	,	PUNCT
easat-5713	71	21	a	a	DET
easat-5713	71	22	persistent	persistent	ADJ
easat-5713	71	23	challenge	challenge	NOUN
easat-5713	71	24	in	in	ADP
easat-5713	71	25	fer	fer	PROPN
easat-5713	71	26	models	model	NOUN
easat-5713	71	27	.	.	PUNCT
easat-5713	72	1	2.1	2.1	NUM
easat-5713	72	2	.	.	PUNCT
easat-5713	72	3	data	datum	NOUN
easat-5713	72	4	preprocessing	preprocesse	VERB
easat-5713	72	5	the	the	DET
easat-5713	72	6	fer2013	fer2013	ADJ
easat-5713	72	7	dataset	dataset	NOUN
easat-5713	72	8	is	be	AUX
easat-5713	72	9	the	the	DET
easat-5713	72	10	foundation	foundation	NOUN
easat-5713	72	11	for	for	ADP
easat-5713	72	12	this	this	DET
easat-5713	72	13	study	study	NOUN
easat-5713	72	14	and	and	CCONJ
easat-5713	72	15	requires	require	VERB
easat-5713	72	16	thorough	thorough	ADJ
easat-5713	72	17	preprocessing	preprocessing	NOUN
easat-5713	72	18	to	to	PART
easat-5713	72	19	enhance	enhance	VERB
easat-5713	72	20	data	datum	NOUN
easat-5713	72	21	quality	quality	NOUN
easat-5713	72	22	and	and	CCONJ
easat-5713	72	23	consistency	consistency	NOUN
easat-5713	72	24	.	.	PUNCT
easat-5713	73	1	the	the	DET
easat-5713	73	2	dataset	dataset	NOUN
easat-5713	73	3	comprises	comprise	VERB
easat-5713	73	4	grayscale	grayscale	NOUN
easat-5713	73	5	facial	facial	ADJ
easat-5713	73	6	images	image	NOUN
easat-5713	73	7	classified	classify	VERB
easat-5713	73	8	into	into	ADP
easat-5713	73	9	seven	seven	NUM
easat-5713	73	10	emotion	emotion	NOUN
easat-5713	73	11	categories	category	NOUN
easat-5713	73	12	:	:	PUNCT
easat-5713	73	13	anger	anger	NOUN
easat-5713	73	14	,	,	PUNCT
easat-5713	73	15	disgust	disgust	NOUN
easat-5713	73	16	,	,	PUNCT
easat-5713	73	17	fear	fear	NOUN
easat-5713	73	18	,	,	PUNCT
easat-5713	73	19	happiness	happiness	NOUN
easat-5713	73	20	,	,	PUNCT
easat-5713	73	21	sadness	sadness	NOUN
easat-5713	73	22	,	,	PUNCT
easat-5713	73	23	surprise	surprise	NOUN
easat-5713	73	24	,	,	PUNCT
easat-5713	73	25	and	and	CCONJ
easat-5713	73	26	neutral	neutral	ADJ
easat-5713	73	27	.	.	PUNCT
easat-5713	74	1	to	to	PART
easat-5713	74	2	improve	improve	VERB
easat-5713	74	3	data	datum	NOUN
easat-5713	74	4	quality	quality	NOUN
easat-5713	74	5	and	and	CCONJ
easat-5713	74	6	standardize	standardize	VERB
easat-5713	74	7	inputs	input	NOUN
easat-5713	74	8	,	,	PUNCT
easat-5713	74	9	we	we	PRON
easat-5713	74	10	applied	apply	VERB
easat-5713	74	11	the	the	DET
easat-5713	74	12	following	follow	VERB
easat-5713	74	13	preprocessing	preprocesse	VERB
easat-5713	74	14	steps	step	NOUN
easat-5713	74	15	.	.	PUNCT
easat-5713	75	1	each	each	DET
easat-5713	75	2	image	image	NOUN
easat-5713	75	3	is	be	AUX
easat-5713	75	4	resized	resize	VERB
easat-5713	75	5	to	to	ADP
easat-5713	75	6	48x48	48x48	NUM
easat-5713	75	7	pixels	pixel	NOUN
easat-5713	75	8	,	,	PUNCT
easat-5713	75	9	which	which	PRON
easat-5713	75	10	is	be	AUX
easat-5713	75	11	the	the	DET
easat-5713	75	12	standard	standard	ADJ
easat-5713	75	13	input	input	NOUN
easat-5713	75	14	size	size	NOUN
easat-5713	75	15	for	for	ADP
easat-5713	75	16	most	most	ADJ
easat-5713	75	17	deep	deep	ADJ
easat-5713	75	18	learning	learning	NOUN
easat-5713	75	19	-	-	PUNCT
easat-5713	75	20	based	base	VERB
easat-5713	75	21	fer	fer	PROPN
easat-5713	75	22	models	model	NOUN
easat-5713	75	23	[	[	X
easat-5713	75	24	19	19	NUM
easat-5713	75	25	]	]	PUNCT
easat-5713	75	26	.	.	PUNCT
easat-5713	76	1	normalization	normalization	NOUN
easat-5713	76	2	is	be	AUX
easat-5713	76	3	applied	apply	VERB
easat-5713	76	4	to	to	PART
easat-5713	76	5	scale	scale	VERB
easat-5713	76	6	pixel	pixel	ADJ
easat-5713	76	7	values	value	NOUN
easat-5713	76	8	to	to	ADP
easat-5713	76	9	the	the	DET
easat-5713	76	10	range	range	NOUN
easat-5713	76	11	[	[	X
easat-5713	76	12	0	0	NUM
easat-5713	76	13	,	,	PUNCT
easat-5713	76	14	1	1	NUM
easat-5713	76	15	]	]	PUNCT
easat-5713	76	16	,	,	PUNCT
easat-5713	76	17	improving	improve	VERB
easat-5713	76	18	the	the	DET
easat-5713	76	19	stability	stability	NOUN
easat-5713	76	20	of	of	ADP
easat-5713	76	21	the	the	DET
easat-5713	76	22	learning	learning	NOUN
easat-5713	76	23	process	process	NOUN
easat-5713	76	24	and	and	CCONJ
easat-5713	76	25	preventing	prevent	VERB
easat-5713	76	26	numerical	numerical	ADJ
easat-5713	76	27	instabilities	instability	NOUN
easat-5713	76	28	.	.	PUNCT
easat-5713	77	1	histogram	histogram	PROPN
easat-5713	77	2	equalization	equalization	NOUN
easat-5713	77	3	was	be	AUX
easat-5713	77	4	applied	apply	VERB
easat-5713	77	5	to	to	PART
easat-5713	77	6	improve	improve	VERB
easat-5713	77	7	contrast	contrast	NOUN
easat-5713	77	8	in	in	ADP
easat-5713	77	9	low	low	ADJ
easat-5713	77	10	-	-	PUNCT
easat-5713	77	11	light	light	NOUN
easat-5713	77	12	images	image	NOUN
easat-5713	77	13	.	.	PUNCT
easat-5713	78	1	one	one	NUM
easat-5713	78	2	of	of	ADP
easat-5713	78	3	the	the	DET
easat-5713	78	4	primary	primary	ADJ
easat-5713	78	5	concerns	concern	NOUN
easat-5713	78	6	in	in	ADP
easat-5713	78	7	fer	fer	PROPN
easat-5713	78	8	datasets	dataset	NOUN
easat-5713	78	9	is	be	AUX
easat-5713	78	10	the	the	DET
easat-5713	78	11	presence	presence	NOUN
easat-5713	78	12	of	of	ADP
easat-5713	78	13	noise	noise	NOUN
easat-5713	78	14	,	,	PUNCT
easat-5713	78	15	which	which	PRON
easat-5713	78	16	can	can	AUX
easat-5713	78	17	arise	arise	VERB
easat-5713	78	18	from	from	ADP
easat-5713	78	19	low	low	ADJ
easat-5713	78	20	-	-	PUNCT
easat-5713	78	21	resolution	resolution	NOUN
easat-5713	78	22	images	image	NOUN
easat-5713	78	23	,	,	PUNCT
easat-5713	78	24	occlusions	occlusion	NOUN
easat-5713	78	25	,	,	PUNCT
easat-5713	78	26	varying	vary	VERB
easat-5713	78	27	lighting	lighting	NOUN
easat-5713	78	28	conditions	condition	NOUN
easat-5713	78	29	,	,	PUNCT
easat-5713	78	30	and	and	CCONJ
easat-5713	78	31	blurred	blur	VERB
easat-5713	78	32	facial	facial	ADJ
easat-5713	78	33	features	feature	NOUN
easat-5713	78	34	.	.	PUNCT
easat-5713	79	1	to	to	PART
easat-5713	79	2	mitigate	mitigate	VERB
easat-5713	79	3	this	this	DET
easat-5713	79	4	issue	issue	NOUN
easat-5713	79	5	,	,	PUNCT
easat-5713	79	6	we	we	PRON
easat-5713	79	7	apply	apply	VERB
easat-5713	79	8	data	data	NOUN
easat-5713	79	9	augmentation	augmentation	NOUN
easat-5713	79	10	techniques	technique	NOUN
easat-5713	79	11	such	such	ADJ
easat-5713	79	12	as	as	ADP
easat-5713	79	13	random	random	ADJ
easat-5713	79	14	rotation	rotation	NOUN
easat-5713	79	15	,	,	PUNCT
easat-5713	79	16	horizontal	horizontal	ADJ
easat-5713	79	17	flipping	flipping	NOUN
easat-5713	79	18	,	,	PUNCT
easat-5713	79	19	and	and	CCONJ
easat-5713	79	20	gaussian	gaussian	ADJ
easat-5713	79	21	blur	blur	NOUN
easat-5713	79	22	reduction	reduction	NOUN
easat-5713	79	23	was	be	AUX
easat-5713	79	24	used	use	VERB
easat-5713	79	25	to	to	PART
easat-5713	79	26	sharpen	sharpen	VERB
easat-5713	79	27	blurred	blurred	ADJ
easat-5713	79	28	images	image	NOUN
easat-5713	79	29	.	.	PUNCT
easat-5713	80	1	this	this	PRON
easat-5713	80	2	ensures	ensure	VERB
easat-5713	80	3	that	that	SCONJ
easat-5713	80	4	the	the	DET
easat-5713	80	5	model	model	NOUN
easat-5713	80	6	learns	learn	VERB
easat-5713	80	7	from	from	ADP
easat-5713	80	8	a	a	DET
easat-5713	80	9	diverse	diverse	ADJ
easat-5713	80	10	set	set	NOUN
easat-5713	80	11	of	of	ADP
easat-5713	80	12	expressions	expression	NOUN
easat-5713	80	13	while	while	SCONJ
easat-5713	80	14	being	be	AUX
easat-5713	80	15	less	less	ADV
easat-5713	80	16	sensitive	sensitive	ADJ
easat-5713	80	17	to	to	ADP
easat-5713	80	18	minor	minor	ADJ
easat-5713	80	19	distortions	distortion	NOUN
easat-5713	80	20	that	that	PRON
easat-5713	80	21	naturally	naturally	ADV
easat-5713	80	22	occur	occur	VERB
easat-5713	80	23	in	in	ADP
easat-5713	80	24	real	real	ADJ
easat-5713	80	25	-	-	PUNCT
easat-5713	80	26	world	world	NOUN
easat-5713	80	27	images	image	NOUN
easat-5713	80	28	.	.	PUNCT
easat-5713	81	1	rotation	rotation	NOUN
easat-5713	81	2	(	(	PUNCT
easat-5713	81	3	±15	±15	NOUN
easat-5713	81	4	°	°	NOUN
easat-5713	81	5	)	)	PUNCT
easat-5713	81	6	,	,	PUNCT
easat-5713	81	7	flipping	flip	VERB
easat-5713	81	8	,	,	PUNCT
easat-5713	81	9	zooming	zooming	NOUN
easat-5713	81	10	,	,	PUNCT
easat-5713	81	11	and	and	CCONJ
easat-5713	81	12	random	random	ADJ
easat-5713	81	13	cropping	cropping	NOUN
easat-5713	81	14	were	be	AUX
easat-5713	81	15	applied	apply	VERB
easat-5713	81	16	to	to	PART
easat-5713	81	17	artificially	artificially	ADV
easat-5713	81	18	expand	expand	VERB
easat-5713	81	19	dataset	dataset	ADJ
easat-5713	81	20	diversity	diversity	NOUN
easat-5713	81	21	.	.	PUNCT
easat-5713	82	1	using	use	VERB
easat-5713	82	2	entropy	entropy	NOUN
easat-5713	82	3	-	-	PUNCT
easat-5713	82	4	based	base	VERB
easat-5713	82	5	filtering	filtering	NOUN
easat-5713	82	6	,	,	PUNCT
easat-5713	82	7	excessively	excessively	ADV
easat-5713	82	8	noisy	noisy	ADJ
easat-5713	82	9	or	or	CCONJ
easat-5713	82	10	unrecognizable	unrecognizable	ADJ
easat-5713	82	11	images	image	NOUN
easat-5713	82	12	were	be	AUX
easat-5713	82	13	flagged	flag	VERB
easat-5713	82	14	and	and	CCONJ
easat-5713	82	15	excluded	exclude	VERB
easat-5713	82	16	.	.	PUNCT
easat-5713	83	1	the	the	DET
easat-5713	83	2	goal	goal	NOUN
easat-5713	83	3	of	of	ADP
easat-5713	83	4	this	this	DET
easat-5713	83	5	step	step	NOUN
easat-5713	83	6	is	be	AUX
easat-5713	83	7	to	to	ADP
easat-5713	83	8	1858	1858	NUM
easat-5713	83	9	edelweiss	edelweiss	PROPN
easat-5713	83	10	applied	apply	VERB
easat-5713	83	11	science	science	NOUN
easat-5713	83	12	and	and	CCONJ
easat-5713	83	13	technology	technology	NOUN
easat-5713	83	14	issn	issn	PROPN
easat-5713	83	15	:	:	PUNCT
easat-5713	83	16	2576	2576	NUM
easat-5713	83	17	-	-	SYM
easat-5713	83	18	8484	8484	NUM
easat-5713	83	19	vol	vol	NOUN
easat-5713	83	20	.	.	PROPN
easat-5713	84	1	9	9	NUM
easat-5713	84	2	,	,	PUNCT
easat-5713	84	3	no	no	INTJ
easat-5713	84	4	.	.	NOUN
easat-5713	85	1	3	3	NUM
easat-5713	85	2	:	:	PUNCT
easat-5713	85	3	1855	1855	NUM
easat-5713	85	4	-	-	SYM
easat-5713	85	5	1869	1869	NUM
easat-5713	85	6	,	,	PUNCT
easat-5713	85	7	2025	2025	NUM
easat-5713	85	8	doi	doi	NOUN
easat-5713	85	9	:	:	PUNCT
easat-5713	85	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	85	11	©	©	ADP
easat-5713	85	12	2025	2025	NUM
easat-5713	85	13	by	by	ADP
easat-5713	85	14	the	the	DET
easat-5713	85	15	authors	author	NOUN
easat-5713	85	16	;	;	PUNCT
easat-5713	85	17	licensee	licensee	PROPN
easat-5713	85	18	learning	learn	VERB
easat-5713	85	19	gate	gate	NOUN
easat-5713	85	20	enhance	enhance	VERB
easat-5713	85	21	the	the	DET
easat-5713	85	22	model	model	NOUN
easat-5713	85	23	’s	’s	PART
easat-5713	85	24	generalization	generalization	NOUN
easat-5713	85	25	ability	ability	NOUN
easat-5713	85	26	,	,	PUNCT
easat-5713	85	27	reducing	reduce	VERB
easat-5713	85	28	the	the	DET
easat-5713	85	29	likelihood	likelihood	NOUN
easat-5713	85	30	of	of	ADP
easat-5713	85	31	overfitting	overfitte	VERB
easat-5713	85	32	and	and	CCONJ
easat-5713	85	33	making	make	VERB
easat-5713	85	34	it	it	PRON
easat-5713	85	35	more	more	ADV
easat-5713	85	36	resilient	resilient	ADJ
easat-5713	85	37	to	to	ADP
easat-5713	85	38	noisy	noisy	ADJ
easat-5713	85	39	images	image	NOUN
easat-5713	85	40	.	.	PUNCT
easat-5713	86	1	examples	example	NOUN
easat-5713	86	2	of	of	ADP
easat-5713	86	3	the	the	DET
easat-5713	86	4	fer2013	fer2013	ADJ
easat-5713	86	5	dataset	dataset	NOUN
easat-5713	86	6	’s	’s	PART
easat-5713	86	7	inherent	inherent	ADJ
easat-5713	86	8	variability	variability	NOUN
easat-5713	86	9	can	can	AUX
easat-5713	86	10	be	be	AUX
easat-5713	86	11	found	find	VERB
easat-5713	86	12	in	in	ADP
easat-5713	86	13	figure	figure	NOUN
easat-5713	86	14	1	1	NUM
easat-5713	86	15	.	.	PUNCT
easat-5713	86	16	figure	figure	NOUN
easat-5713	86	17	1	1	NUM
easat-5713	86	18	.	.	PUNCT
easat-5713	87	1	sample	sample	NOUN
easat-5713	87	2	images	image	NOUN
easat-5713	87	3	from	from	ADP
easat-5713	87	4	fer2013	fer2013	ADJ
easat-5713	87	5	dataset	dataset	NOUN
easat-5713	87	6	.	.	PUNCT
easat-5713	88	1	2.2	2.2	NUM
easat-5713	88	2	.	.	PUNCT
easat-5713	88	3	model	model	NOUN
easat-5713	88	4	training	training	NOUN
easat-5713	88	5	training	training	NOUN
easat-5713	88	6	the	the	DET
easat-5713	88	7	model	model	NOUN
easat-5713	88	8	is	be	AUX
easat-5713	88	9	performed	perform	VERB
easat-5713	88	10	using	use	VERB
easat-5713	88	11	the	the	DET
easat-5713	88	12	augmented	augment	VERB
easat-5713	88	13	dataset	dataset	NOUN
easat-5713	88	14	to	to	PART
easat-5713	88	15	assess	assess	VERB
easat-5713	88	16	the	the	DET
easat-5713	88	17	impact	impact	NOUN
easat-5713	88	18	of	of	ADP
easat-5713	88	19	synthetic	synthetic	ADJ
easat-5713	88	20	data	datum	NOUN
easat-5713	88	21	on	on	ADP
easat-5713	88	22	model	model	NOUN
easat-5713	88	23	performance	performance	NOUN
easat-5713	88	24	.	.	PUNCT
easat-5713	89	1	we	we	PRON
easat-5713	89	2	employ	employ	VERB
easat-5713	89	3	the	the	DET
easat-5713	89	4	segmentation	segmentation	NOUN
easat-5713	89	5	vgg-19	vgg-19	PROPN
easat-5713	89	6	model	model	NOUN
easat-5713	89	7	,	,	PUNCT
easat-5713	89	8	a	a	DET
easat-5713	89	9	deep	deep	ADJ
easat-5713	89	10	convolutional	convolutional	ADJ
easat-5713	89	11	neural	neural	ADJ
easat-5713	89	12	network	network	NOUN
easat-5713	89	13	known	know	VERB
easat-5713	89	14	for	for	ADP
easat-5713	89	15	its	its	PRON
easat-5713	89	16	strong	strong	ADJ
easat-5713	89	17	feature	feature	NOUN
easat-5713	89	18	extraction	extraction	NOUN
easat-5713	89	19	capabilities	capability	NOUN
easat-5713	89	20	.	.	PUNCT
easat-5713	90	1	the	the	DET
easat-5713	90	2	vgg-19	vgg-19	PROPN
easat-5713	90	3	model	model	NOUN
easat-5713	90	4	consists	consist	VERB
easat-5713	90	5	of	of	ADP
easat-5713	90	6	16	16	NUM
easat-5713	90	7	convolutional	convolutional	ADJ
easat-5713	90	8	layers	layer	NOUN
easat-5713	90	9	followed	follow	VERB
easat-5713	90	10	by	by	ADP
easat-5713	90	11	3	3	NUM
easat-5713	90	12	fully	fully	ADV
easat-5713	90	13	connected	connected	ADJ
easat-5713	90	14	layers	layer	NOUN
easat-5713	90	15	,	,	PUNCT
easat-5713	90	16	utilizing	utilize	VERB
easat-5713	90	17	relu	relu	NOUN
easat-5713	90	18	activations	activation	NOUN
easat-5713	90	19	and	and	CCONJ
easat-5713	90	20	max	max	PROPN
easat-5713	90	21	-	-	PUNCT
easat-5713	90	22	pooling	pool	VERB
easat-5713	90	23	layers	layer	NOUN
easat-5713	90	24	to	to	PART
easat-5713	90	25	capture	capture	VERB
easat-5713	90	26	hierarchical	hierarchical	ADJ
easat-5713	90	27	facial	facial	ADJ
easat-5713	90	28	features	feature	NOUN
easat-5713	90	29	[	[	X
easat-5713	90	30	20	20	NUM
easat-5713	90	31	]	]	PUNCT
easat-5713	90	32	.	.	PUNCT
easat-5713	91	1	given	give	VERB
easat-5713	91	2	the	the	DET
easat-5713	91	3	complexity	complexity	NOUN
easat-5713	91	4	of	of	ADP
easat-5713	91	5	facial	facial	ADJ
easat-5713	91	6	expressions	expression	NOUN
easat-5713	91	7	,	,	PUNCT
easat-5713	91	8	deep	deep	ADJ
easat-5713	91	9	architectures	architecture	NOUN
easat-5713	91	10	like	like	ADP
easat-5713	91	11	vgg-19	vgg-19	NUM
easat-5713	91	12	are	be	AUX
easat-5713	91	13	well	well	ADV
easat-5713	91	14	-	-	PUNCT
easat-5713	91	15	suited	suit	VERB
easat-5713	91	16	for	for	ADP
easat-5713	91	17	learning	learn	VERB
easat-5713	91	18	fine	fine	ADV
easat-5713	91	19	-	-	PUNCT
easat-5713	91	20	grained	grain	VERB
easat-5713	91	21	differences	difference	NOUN
easat-5713	91	22	between	between	ADP
easat-5713	91	23	emotions	emotion	NOUN
easat-5713	91	24	.	.	PUNCT
easat-5713	92	1	the	the	DET
easat-5713	92	2	overall	overall	ADJ
easat-5713	92	3	architecture	architecture	NOUN
easat-5713	92	4	of	of	ADP
easat-5713	92	5	the	the	DET
easat-5713	92	6	vgg-19	vgg-19	PROPN
easat-5713	92	7	model	model	NOUN
easat-5713	92	8	used	use	VERB
easat-5713	92	9	in	in	ADP
easat-5713	92	10	this	this	DET
easat-5713	92	11	study	study	NOUN
easat-5713	92	12	is	be	AUX
easat-5713	92	13	shown	show	VERB
easat-5713	92	14	in	in	ADP
easat-5713	92	15	figure	figure	NOUN
easat-5713	92	16	2	2	NUM
easat-5713	92	17	.	.	SYM
easat-5713	92	18	1859	1859	NUM
easat-5713	92	19	edelweiss	edelweiss	PROPN
easat-5713	92	20	applied	apply	VERB
easat-5713	92	21	science	science	NOUN
easat-5713	92	22	and	and	CCONJ
easat-5713	92	23	technology	technology	NOUN
easat-5713	92	24	issn	issn	PROPN
easat-5713	92	25	:	:	PUNCT
easat-5713	92	26	2576	2576	NUM
easat-5713	92	27	-	-	SYM
easat-5713	92	28	8484	8484	NUM
easat-5713	92	29	vol	vol	NOUN
easat-5713	92	30	.	.	PROPN
easat-5713	93	1	9	9	NUM
easat-5713	93	2	,	,	PUNCT
easat-5713	93	3	no	no	INTJ
easat-5713	93	4	.	.	NOUN
easat-5713	94	1	3	3	NUM
easat-5713	94	2	:	:	PUNCT
easat-5713	94	3	1855	1855	NUM
easat-5713	94	4	-	-	SYM
easat-5713	94	5	1869	1869	NUM
easat-5713	94	6	,	,	PUNCT
easat-5713	94	7	2025	2025	NUM
easat-5713	94	8	doi	doi	NOUN
easat-5713	94	9	:	:	PUNCT
easat-5713	94	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	94	11	©	©	ADP
easat-5713	94	12	2025	2025	NUM
easat-5713	94	13	by	by	ADP
easat-5713	94	14	the	the	DET
easat-5713	94	15	authors	author	NOUN
easat-5713	94	16	;	;	PUNCT
easat-5713	94	17	licensee	licensee	PROPN
easat-5713	94	18	learning	learn	VERB
easat-5713	94	19	gate	gate	NOUN
easat-5713	94	20	figure	figure	NOUN
easat-5713	94	21	2	2	NUM
easat-5713	94	22	.	.	PUNCT
easat-5713	95	1	the	the	DET
easat-5713	95	2	architecture	architecture	NOUN
easat-5713	95	3	of	of	ADP
easat-5713	95	4	the	the	DET
easat-5713	95	5	vgg-19	vgg-19	PROPN
easat-5713	95	6	segmentation	segmentation	NOUN
easat-5713	95	7	model	model	NOUN
easat-5713	95	8	.	.	PUNCT
easat-5713	96	1	the	the	DET
easat-5713	96	2	model	model	NOUN
easat-5713	96	3	is	be	AUX
easat-5713	96	4	trained	train	VERB
easat-5713	96	5	using	use	VERB
easat-5713	96	6	the	the	DET
easat-5713	96	7	adam	adam	PROPN
easat-5713	96	8	optimizer	optimizer	NOUN
easat-5713	96	9	,	,	PUNCT
easat-5713	96	10	which	which	PRON
easat-5713	96	11	is	be	AUX
easat-5713	96	12	well	well	ADV
easat-5713	96	13	-	-	PUNCT
easat-5713	96	14	suited	suit	VERB
easat-5713	96	15	for	for	ADP
easat-5713	96	16	handling	handle	VERB
easat-5713	96	17	noisy	noisy	ADJ
easat-5713	96	18	gradients	gradient	NOUN
easat-5713	96	19	and	and	CCONJ
easat-5713	96	20	adaptive	adaptive	ADJ
easat-5713	96	21	learning	learning	NOUN
easat-5713	96	22	rates	rate	NOUN
easat-5713	96	23	.	.	PUNCT
easat-5713	97	1	the	the	DET
easat-5713	97	2	hyperparameters	hyperparameter	NOUN
easat-5713	97	3	used	use	VERB
easat-5713	97	4	include	include	VERB
easat-5713	97	5	a	a	DET
easat-5713	97	6	learning	learning	NOUN
easat-5713	97	7	rate	rate	NOUN
easat-5713	97	8	of	of	ADP
easat-5713	97	9	0.0001	0.0001	NUM
easat-5713	97	10	,	,	PUNCT
easat-5713	97	11	a	a	DET
easat-5713	97	12	batch	batch	NOUN
easat-5713	97	13	size	size	NOUN
easat-5713	97	14	of	of	ADP
easat-5713	97	15	8	8	NUM
easat-5713	97	16	,	,	PUNCT
easat-5713	97	17	and	and	CCONJ
easat-5713	97	18	a	a	DET
easat-5713	97	19	total	total	NOUN
easat-5713	97	20	of	of	ADP
easat-5713	97	21	25	25	NUM
easat-5713	97	22	epochs	epoch	NOUN
easat-5713	97	23	.	.	PUNCT
easat-5713	98	1	due	due	ADP
easat-5713	98	2	to	to	ADP
easat-5713	98	3	hardware	hardware	NOUN
easat-5713	98	4	constraints	constraint	NOUN
easat-5713	98	5	,	,	PUNCT
easat-5713	98	6	the	the	DET
easat-5713	98	7	training	training	NOUN
easat-5713	98	8	process	process	NOUN
easat-5713	98	9	is	be	AUX
easat-5713	98	10	limited	limit	VERB
easat-5713	98	11	to	to	ADP
easat-5713	98	12	this	this	DET
easat-5713	98	13	number	number	NOUN
easat-5713	98	14	of	of	ADP
easat-5713	98	15	epochs	epoch	NOUN
easat-5713	98	16	,	,	PUNCT
easat-5713	98	17	though	though	SCONJ
easat-5713	98	18	prior	prior	ADJ
easat-5713	98	19	studies	study	NOUN
easat-5713	98	20	have	have	AUX
easat-5713	98	21	indicated	indicate	VERB
easat-5713	98	22	that	that	SCONJ
easat-5713	98	23	optimal	optimal	ADJ
easat-5713	98	24	performance	performance	NOUN
easat-5713	98	25	for	for	ADP
easat-5713	98	26	vgg-19	vgg-19	PROPN
easat-5713	98	27	is	be	AUX
easat-5713	98	28	achieved	achieve	VERB
easat-5713	98	29	at	at	ADP
easat-5713	98	30	around	around	ADV
easat-5713	98	31	67	67	NUM
easat-5713	98	32	epochs	epoch	NOUN
easat-5713	99	1	[	[	X
easat-5713	99	2	21	21	NUM
easat-5713	99	3	]	]	PUNCT
easat-5713	99	4	.	.	PUNCT
easat-5713	100	1	the	the	DET
easat-5713	100	2	categorical	categorical	ADJ
easat-5713	100	3	cross	cross	ADJ
easat-5713	100	4	-	-	ADJ
easat-5713	100	5	entropy	entropy	ADJ
easat-5713	100	6	loss	loss	NOUN
easat-5713	100	7	function	function	NOUN
easat-5713	100	8	is	be	AUX
easat-5713	100	9	used	use	VERB
easat-5713	100	10	to	to	PART
easat-5713	100	11	optimize	optimize	VERB
easat-5713	100	12	the	the	DET
easat-5713	100	13	model	model	NOUN
easat-5713	100	14	,	,	PUNCT
easat-5713	100	15	given	give	VERB
easat-5713	100	16	that	that	PRON
easat-5713	100	17	fer	fer	PROPN
easat-5713	100	18	is	be	AUX
easat-5713	100	19	a	a	DET
easat-5713	100	20	multi	multi	ADJ
easat-5713	100	21	-	-	ADJ
easat-5713	100	22	class	class	ADJ
easat-5713	100	23	classification	classification	NOUN
easat-5713	100	24	problem	problem	NOUN
easat-5713	100	25	.	.	PUNCT
easat-5713	101	1	to	to	PART
easat-5713	101	2	further	far	ADV
easat-5713	101	3	address	address	VERB
easat-5713	101	4	the	the	DET
easat-5713	101	5	impact	impact	NOUN
easat-5713	101	6	of	of	ADP
easat-5713	101	7	noisy	noisy	ADJ
easat-5713	101	8	images	image	NOUN
easat-5713	101	9	on	on	ADP
easat-5713	101	10	training	training	NOUN
easat-5713	101	11	,	,	PUNCT
easat-5713	101	12	additional	additional	ADJ
easat-5713	101	13	preprocessing	preprocessing	NOUN
easat-5713	101	14	steps	step	NOUN
easat-5713	101	15	such	such	ADJ
easat-5713	101	16	as	as	ADP
easat-5713	101	17	adaptive	adaptive	ADJ
easat-5713	101	18	histogram	histogram	NOUN
easat-5713	101	19	equalization	equalization	NOUN
easat-5713	101	20	(	(	PUNCT
easat-5713	101	21	ahe	ahe	NOUN
easat-5713	101	22	)	)	PUNCT
easat-5713	101	23	and	and	CCONJ
easat-5713	101	24	gaussian	gaussian	ADJ
easat-5713	101	25	blurring	blurring	NOUN
easat-5713	101	26	are	be	AUX
easat-5713	101	27	explored	explore	VERB
easat-5713	101	28	.	.	PUNCT
easat-5713	102	1	these	these	DET
easat-5713	102	2	methods	method	NOUN
easat-5713	102	3	help	help	VERB
easat-5713	102	4	in	in	ADP
easat-5713	102	5	normalizing	normalize	VERB
easat-5713	102	6	image	image	NOUN
easat-5713	102	7	contrast	contrast	NOUN
easat-5713	102	8	and	and	CCONJ
easat-5713	102	9	reducing	reduce	VERB
easat-5713	102	10	noise	noise	NOUN
easat-5713	102	11	artifacts	artifact	NOUN
easat-5713	102	12	that	that	PRON
easat-5713	102	13	could	could	AUX
easat-5713	102	14	mislead	mislead	VERB
easat-5713	102	15	the	the	DET
easat-5713	102	16	model	model	NOUN
easat-5713	102	17	.	.	PUNCT
easat-5713	103	1	2.3	2.3	NUM
easat-5713	103	2	.	.	PUNCT
easat-5713	104	1	performance	performance	NOUN
easat-5713	104	2	evaluation	evaluation	NOUN
easat-5713	104	3	the	the	DET
easat-5713	104	4	effectiveness	effectiveness	NOUN
easat-5713	104	5	of	of	ADP
easat-5713	104	6	the	the	DET
easat-5713	104	7	trained	train	VERB
easat-5713	104	8	model	model	NOUN
easat-5713	104	9	is	be	AUX
easat-5713	104	10	evaluated	evaluate	VERB
easat-5713	104	11	using	use	VERB
easat-5713	104	12	multiple	multiple	ADJ
easat-5713	104	13	performance	performance	NOUN
easat-5713	104	14	metrics	metric	NOUN
easat-5713	104	15	,	,	PUNCT
easat-5713	104	16	including	include	VERB
easat-5713	104	17	accuracy	accuracy	NOUN
easat-5713	104	18	,	,	PUNCT
easat-5713	104	19	precision	precision	NOUN
easat-5713	104	20	,	,	PUNCT
easat-5713	104	21	recall	recall	NOUN
easat-5713	104	22	,	,	PUNCT
easat-5713	104	23	and	and	CCONJ
easat-5713	104	24	f1	f1	NOUN
easat-5713	104	25	-	-	PUNCT
easat-5713	104	26	score	score	NOUN
easat-5713	104	27	.	.	PUNCT
easat-5713	105	1	these	these	DET
easat-5713	105	2	metrics	metric	NOUN
easat-5713	105	3	provide	provide	VERB
easat-5713	105	4	a	a	DET
easat-5713	105	5	comprehensive	comprehensive	ADJ
easat-5713	105	6	understanding	understanding	NOUN
easat-5713	105	7	of	of	ADP
easat-5713	105	8	the	the	DET
easat-5713	105	9	model	model	NOUN
easat-5713	105	10	's	's	PART
easat-5713	105	11	ability	ability	NOUN
easat-5713	105	12	to	to	PART
easat-5713	105	13	correctly	correctly	ADV
easat-5713	105	14	classify	classify	VERB
easat-5713	105	15	facial	facial	ADJ
easat-5713	105	16	expressions	expression	NOUN
easat-5713	105	17	across	across	ADP
easat-5713	105	18	different	different	ADJ
easat-5713	105	19	emotion	emotion	NOUN
easat-5713	105	20	categories	category	NOUN
easat-5713	105	21	.	.	PUNCT
easat-5713	106	1	given	give	VERB
easat-5713	106	2	the	the	DET
easat-5713	106	3	known	know	VERB
easat-5713	106	4	challenges	challenge	NOUN
easat-5713	106	5	of	of	ADP
easat-5713	106	6	datasets	dataset	NOUN
easat-5713	106	7	,	,	PUNCT
easat-5713	106	8	particularly	particularly	ADV
easat-5713	106	9	those	those	PRON
easat-5713	106	10	related	relate	VERB
easat-5713	106	11	to	to	ADP
easat-5713	106	12	noise	noise	NOUN
easat-5713	106	13	,	,	PUNCT
easat-5713	106	14	additional	additional	ADJ
easat-5713	106	15	evaluation	evaluation	NOUN
easat-5713	106	16	methods	method	NOUN
easat-5713	106	17	are	be	AUX
easat-5713	106	18	incorporated	incorporate	VERB
easat-5713	106	19	to	to	PART
easat-5713	106	20	analyze	analyze	VERB
easat-5713	106	21	the	the	DET
easat-5713	106	22	impact	impact	NOUN
easat-5713	106	23	of	of	ADP
easat-5713	106	24	synthetic	synthetic	ADJ
easat-5713	106	25	data	datum	NOUN
easat-5713	106	26	on	on	ADP
easat-5713	106	27	model	model	NOUN
easat-5713	106	28	robustness	robustness	NOUN
easat-5713	106	29	.	.	PUNCT
easat-5713	107	1	another	another	DET
easat-5713	107	2	critical	critical	ADJ
easat-5713	107	3	aspect	aspect	NOUN
easat-5713	107	4	of	of	ADP
easat-5713	107	5	the	the	DET
easat-5713	107	6	evaluation	evaluation	NOUN
easat-5713	107	7	is	be	AUX
easat-5713	107	8	comparing	compare	VERB
easat-5713	107	9	the	the	DET
easat-5713	107	10	model	model	NOUN
easat-5713	107	11	's	's	PART
easat-5713	107	12	performance	performance	NOUN
easat-5713	107	13	on	on	ADP
easat-5713	107	14	specific	specific	ADJ
easat-5713	107	15	noisy	noisy	ADJ
easat-5713	107	16	images	image	NOUN
easat-5713	107	17	.	.	PUNCT
easat-5713	108	1	using	use	VERB
easat-5713	108	2	the	the	DET
easat-5713	108	3	classification	classification	NOUN
easat-5713	108	4	reports	report	NOUN
easat-5713	108	5	in	in	ADP
easat-5713	108	6	tables	table	NOUN
easat-5713	108	7	3	3	NUM
easat-5713	108	8	and	and	CCONJ
easat-5713	108	9	4	4	NUM
easat-5713	108	10	,	,	PUNCT
easat-5713	108	11	we	we	PRON
easat-5713	108	12	analyze	analyze	VERB
easat-5713	108	13	precision	precision	NOUN
easat-5713	108	14	and	and	CCONJ
easat-5713	108	15	recall	recall	NOUN
easat-5713	108	16	values	value	NOUN
easat-5713	108	17	for	for	ADP
easat-5713	108	18	each	each	DET
easat-5713	108	19	emotion	emotion	NOUN
easat-5713	108	20	class	class	NOUN
easat-5713	108	21	.	.	PUNCT
easat-5713	109	1	a	a	DET
easat-5713	109	2	confusion	confusion	NOUN
easat-5713	109	3	matrix	matrix	NOUN
easat-5713	109	4	is	be	AUX
easat-5713	109	5	constructed	construct	VERB
easat-5713	109	6	to	to	PART
easat-5713	109	7	visualize	visualize	VERB
easat-5713	109	8	the	the	DET
easat-5713	109	9	distribution	distribution	NOUN
easat-5713	109	10	of	of	ADP
easat-5713	109	11	true	true	ADJ
easat-5713	109	12	positives	positive	NOUN
easat-5713	109	13	,	,	PUNCT
easat-5713	109	14	false	false	ADJ
easat-5713	109	15	positives	positive	NOUN
easat-5713	109	16	,	,	PUNCT
easat-5713	109	17	and	and	CCONJ
easat-5713	109	18	false	false	ADJ
easat-5713	109	19	negatives	negative	NOUN
easat-5713	109	20	across	across	ADP
easat-5713	109	21	emotion	emotion	NOUN
easat-5713	109	22	categories	category	NOUN
easat-5713	109	23	.	.	PUNCT
easat-5713	110	1	this	this	PRON
easat-5713	110	2	helps	help	VERB
easat-5713	110	3	in	in	ADP
easat-5713	110	4	identifying	identify	VERB
easat-5713	110	5	which	which	PRON
easat-5713	110	6	emotions	emotion	NOUN
easat-5713	110	7	are	be	AUX
easat-5713	110	8	most	most	ADV
easat-5713	110	9	affected	affect	VERB
easat-5713	110	10	by	by	ADP
easat-5713	110	11	noisy	noisy	ADJ
easat-5713	110	12	data	datum	NOUN
easat-5713	110	13	and	and	CCONJ
easat-5713	110	14	whether	whether	SCONJ
easat-5713	110	15	the	the	DET
easat-5713	110	16	addition	addition	NOUN
easat-5713	110	17	of	of	ADP
easat-5713	110	18	synthetic	synthetic	ADJ
easat-5713	110	19	images	image	NOUN
easat-5713	110	20	improves	improve	VERB
easat-5713	110	21	classification	classification	NOUN
easat-5713	110	22	performance	performance	NOUN
easat-5713	110	23	in	in	ADP
easat-5713	110	24	these	these	DET
easat-5713	110	25	cases	case	NOUN
easat-5713	110	26	.	.	PUNCT
easat-5713	111	1	a	a	DET
easat-5713	111	2	key	key	ADJ
easat-5713	111	3	focus	focus	NOUN
easat-5713	111	4	is	be	AUX
easat-5713	111	5	on	on	ADP
easat-5713	111	6	whether	whether	SCONJ
easat-5713	111	7	synthetic	synthetic	ADJ
easat-5713	111	8	data	datum	NOUN
easat-5713	111	9	helps	helps	AUX
easat-5713	111	10	mitigate	mitigate	VERB
easat-5713	111	11	misclassification	misclassification	NOUN
easat-5713	111	12	in	in	ADP
easat-5713	111	13	expressions	expression	NOUN
easat-5713	111	14	commonly	commonly	ADV
easat-5713	111	15	affected	affect	VERB
easat-5713	111	16	by	by	ADP
easat-5713	111	17	noise	noise	NOUN
easat-5713	111	18	,	,	PUNCT
easat-5713	111	19	such	such	ADJ
easat-5713	111	20	as	as	ADP
easat-5713	111	21	"	"	PUNCT
easat-5713	111	22	disgust	disgust	NOUN
easat-5713	111	23	"	"	PUNCT
easat-5713	111	24	and	and	CCONJ
easat-5713	111	25	"	"	PUNCT
easat-5713	111	26	fear	fear	NOUN
easat-5713	111	27	,	,	PUNCT
easat-5713	111	28	"	"	PUNCT
easat-5713	111	29	which	which	PRON
easat-5713	111	30	have	have	AUX
easat-5713	111	31	historically	historically	ADV
easat-5713	111	32	exhibited	exhibit	VERB
easat-5713	111	33	lower	low	ADJ
easat-5713	111	34	classification	classification	NOUN
easat-5713	111	35	accuracy	accuracy	NOUN
easat-5713	111	36	due	due	ADP
easat-5713	111	37	to	to	ADP
easat-5713	111	38	subtle	subtle	ADJ
easat-5713	111	39	variations	variation	NOUN
easat-5713	111	40	in	in	ADP
easat-5713	111	41	facial	facial	ADJ
easat-5713	111	42	features	feature	NOUN
easat-5713	111	43	.	.	PUNCT
easat-5713	112	1	by	by	ADP
easat-5713	112	2	systematically	systematically	ADV
easat-5713	112	3	evaluating	evaluate	VERB
easat-5713	112	4	the	the	DET
easat-5713	112	5	model	model	NOUN
easat-5713	112	6	's	's	PART
easat-5713	112	7	ability	ability	NOUN
easat-5713	112	8	to	to	PART
easat-5713	112	9	handle	handle	VERB
easat-5713	112	10	noise	noise	NOUN
easat-5713	112	11	,	,	PUNCT
easat-5713	112	12	this	this	DET
easat-5713	112	13	study	study	NOUN
easat-5713	112	14	aims	aim	VERB
easat-5713	112	15	to	to	PART
easat-5713	112	16	determine	determine	VERB
easat-5713	112	17	whether	whether	SCONJ
easat-5713	112	18	ai	ai	AUX
easat-5713	112	19	-	-	PUNCT
easat-5713	112	20	generated	generate	VERB
easat-5713	112	21	synthetic	synthetic	ADJ
easat-5713	112	22	images	image	NOUN
easat-5713	112	23	provide	provide	VERB
easat-5713	112	24	meaningful	meaningful	ADJ
easat-5713	112	25	improvements	improvement	NOUN
easat-5713	112	26	in	in	ADP
easat-5713	112	27	accuracy	accuracy	NOUN
easat-5713	112	28	.	.	PUNCT
easat-5713	113	1	the	the	DET
easat-5713	113	2	findings	finding	NOUN
easat-5713	113	3	will	will	AUX
easat-5713	113	4	contribute	contribute	VERB
easat-5713	113	5	to	to	ADP
easat-5713	113	6	ongoing	ongoing	ADJ
easat-5713	113	7	discussions	discussion	NOUN
easat-5713	113	8	about	about	ADP
easat-5713	113	9	the	the	DET
easat-5713	113	10	viability	viability	NOUN
easat-5713	113	11	of	of	ADP
easat-5713	113	12	synthetic	synthetic	ADJ
easat-5713	113	13	data	datum	NOUN
easat-5713	113	14	in	in	ADP
easat-5713	113	15	deep	deep	ADJ
easat-5713	113	16	learning	learning	NOUN
easat-5713	113	17	applications	application	NOUN
easat-5713	113	18	and	and	CCONJ
easat-5713	113	19	addresses	address	VERB
easat-5713	113	20	key	key	ADJ
easat-5713	113	21	limitations	limitation	NOUN
easat-5713	113	22	in	in	ADP
easat-5713	113	23	datasets	dataset	NOUN
easat-5713	113	24	,	,	PUNCT
easat-5713	113	25	including	include	VERB
easat-5713	113	26	noise	noise	NOUN
easat-5713	113	27	,	,	PUNCT
easat-5713	113	28	class	class	NOUN
easat-5713	113	29	imbalances	imbalance	NOUN
easat-5713	113	30	,	,	PUNCT
easat-5713	113	31	and	and	CCONJ
easat-5713	113	32	lack	lack	NOUN
easat-5713	113	33	of	of	ADP
easat-5713	113	34	diversity	diversity	NOUN
easat-5713	113	35	.	.	PUNCT
easat-5713	114	1	by	by	ADP
easat-5713	114	2	leveraging	leverage	VERB
easat-5713	114	3	ai	ai	ADJ
easat-5713	114	4	-	-	PUNCT
easat-5713	114	5	generated	generate	VERB
easat-5713	114	6	images	image	NOUN
easat-5713	114	7	,	,	PUNCT
easat-5713	114	8	we	we	PRON
easat-5713	114	9	seek	seek	VERB
easat-5713	114	10	to	to	PART
easat-5713	114	11	enhance	enhance	VERB
easat-5713	114	12	the	the	DET
easat-5713	114	13	model	model	NOUN
easat-5713	114	14	robustness	robustness	NOUN
easat-5713	114	15	while	while	SCONJ
easat-5713	114	16	providing	provide	VERB
easat-5713	114	17	a	a	DET
easat-5713	114	18	structured	structured	ADJ
easat-5713	114	19	evaluation	evaluation	NOUN
easat-5713	114	20	framework	framework	NOUN
easat-5713	114	21	to	to	PART
easat-5713	114	22	assess	assess	VERB
easat-5713	114	23	the	the	DET
easat-5713	114	24	real	real	ADJ
easat-5713	114	25	impact	impact	NOUN
easat-5713	114	26	of	of	ADP
easat-5713	114	27	synthetic	synthetic	ADJ
easat-5713	114	28	data	datum	NOUN
easat-5713	114	29	augmentation	augmentation	NOUN
easat-5713	114	30	.	.	PUNCT
easat-5713	115	1	the	the	DET
easat-5713	115	2	overall	overall	ADJ
easat-5713	115	3	approach	approach	NOUN
easat-5713	115	4	is	be	AUX
easat-5713	115	5	summarized	summarize	VERB
easat-5713	115	6	in	in	ADP
easat-5713	115	7	figure	figure	NOUN
easat-5713	115	8	3	3	NUM
easat-5713	115	9	.	.	NOUN
easat-5713	115	10	1860	1860	NUM
easat-5713	115	11	edelweiss	edelweiss	PROPN
easat-5713	115	12	applied	apply	VERB
easat-5713	115	13	science	science	NOUN
easat-5713	115	14	and	and	CCONJ
easat-5713	115	15	technology	technology	NOUN
easat-5713	115	16	issn	issn	PROPN
easat-5713	115	17	:	:	PUNCT
easat-5713	115	18	2576	2576	NUM
easat-5713	115	19	-	-	SYM
easat-5713	115	20	8484	8484	NUM
easat-5713	115	21	vol	vol	NOUN
easat-5713	115	22	.	.	PROPN
easat-5713	116	1	9	9	NUM
easat-5713	116	2	,	,	PUNCT
easat-5713	116	3	no	no	INTJ
easat-5713	116	4	.	.	NOUN
easat-5713	117	1	3	3	NUM
easat-5713	117	2	:	:	PUNCT
easat-5713	117	3	1855	1855	NUM
easat-5713	117	4	-	-	SYM
easat-5713	117	5	1869	1869	NUM
easat-5713	117	6	,	,	PUNCT
easat-5713	117	7	2025	2025	NUM
easat-5713	117	8	doi	doi	NOUN
easat-5713	117	9	:	:	PUNCT
easat-5713	117	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	117	11	©	©	ADP
easat-5713	117	12	2025	2025	NUM
easat-5713	117	13	by	by	ADP
easat-5713	117	14	the	the	DET
easat-5713	117	15	authors	author	NOUN
easat-5713	117	16	;	;	PUNCT
easat-5713	117	17	licensee	licensee	PROPN
easat-5713	117	18	learning	learn	VERB
easat-5713	117	19	gate	gate	PROPN
easat-5713	117	20	figure	figure	NOUN
easat-5713	117	21	3	3	NUM
easat-5713	117	22	.	.	PUNCT
easat-5713	118	1	the	the	DET
easat-5713	118	2	methodological	methodological	ADJ
easat-5713	118	3	approach	approach	NOUN
easat-5713	118	4	of	of	ADP
easat-5713	118	5	the	the	DET
easat-5713	118	6	study	study	NOUN
easat-5713	118	7	.	.	PUNCT
easat-5713	119	1	3	3	X
easat-5713	119	2	.	.	X
easat-5713	119	3	results	result	VERB
easat-5713	119	4	the	the	DET
easat-5713	119	5	results	result	NOUN
easat-5713	119	6	of	of	ADP
easat-5713	119	7	this	this	DET
easat-5713	119	8	study	study	NOUN
easat-5713	119	9	reveal	reveal	VERB
easat-5713	119	10	a	a	DET
easat-5713	119	11	significant	significant	ADJ
easat-5713	119	12	impact	impact	NOUN
easat-5713	119	13	of	of	ADP
easat-5713	119	14	ai	ai	VERB
easat-5713	119	15	-	-	PUNCT
easat-5713	119	16	generated	generate	VERB
easat-5713	119	17	synthetic	synthetic	ADJ
easat-5713	119	18	data	datum	NOUN
easat-5713	119	19	on	on	ADP
easat-5713	119	20	the	the	DET
easat-5713	119	21	performance	performance	NOUN
easat-5713	119	22	of	of	ADP
easat-5713	119	23	a	a	DET
easat-5713	119	24	facial	facial	ADJ
easat-5713	119	25	emotion	emotion	NOUN
easat-5713	119	26	recognition	recognition	NOUN
easat-5713	119	27	model	model	NOUN
easat-5713	119	28	.	.	PUNCT
easat-5713	120	1	given	give	VERB
easat-5713	120	2	the	the	DET
easat-5713	120	3	inherent	inherent	ADJ
easat-5713	120	4	challenges	challenge	NOUN
easat-5713	120	5	in	in	ADP
easat-5713	120	6	real	real	ADJ
easat-5713	120	7	-	-	PUNCT
easat-5713	120	8	world	world	NOUN
easat-5713	120	9	datasets	dataset	NOUN
easat-5713	120	10	,	,	PUNCT
easat-5713	120	11	including	include	VERB
easat-5713	120	12	class	class	NOUN
easat-5713	120	13	imbalances	imbalance	NOUN
easat-5713	120	14	,	,	PUNCT
easat-5713	120	15	limited	limited	ADJ
easat-5713	120	16	diversity	diversity	NOUN
easat-5713	120	17	,	,	PUNCT
easat-5713	120	18	and	and	CCONJ
easat-5713	120	19	noise	noise	NOUN
easat-5713	120	20	such	such	ADJ
easat-5713	120	21	as	as	ADP
easat-5713	120	22	blurriness	blurriness	NOUN
easat-5713	120	23	,	,	PUNCT
easat-5713	120	24	occlusions	occlusion	NOUN
easat-5713	120	25	,	,	PUNCT
easat-5713	120	26	and	and	CCONJ
easat-5713	120	27	varying	vary	VERB
easat-5713	120	28	lighting	lighting	NOUN
easat-5713	120	29	conditions	condition	NOUN
easat-5713	120	30	,	,	PUNCT
easat-5713	120	31	synthetic	synthetic	ADJ
easat-5713	120	32	data	datum	NOUN
easat-5713	120	33	has	have	AUX
easat-5713	120	34	been	be	AUX
easat-5713	120	35	explored	explore	VERB
easat-5713	120	36	as	as	ADP
easat-5713	120	37	a	a	DET
easat-5713	120	38	potential	potential	ADJ
easat-5713	120	39	enhancement	enhancement	NOUN
easat-5713	120	40	.	.	PUNCT
easat-5713	121	1	the	the	DET
easat-5713	121	2	juggernaut	juggernaut	PROPN
easat-5713	121	3	xl	xl	PROPN
easat-5713	121	4	model	model	PROPN
easat-5713	121	5	,	,	PUNCT
easat-5713	121	6	a	a	DET
easat-5713	121	7	state	state	NOUN
easat-5713	121	8	-	-	PUNCT
easat-5713	121	9	of	of	ADP
easat-5713	121	10	-	-	PUNCT
easat-5713	121	11	the	the	DET
easat-5713	121	12	-	-	PUNCT
easat-5713	121	13	art	art	NOUN
easat-5713	121	14	text	text	NOUN
easat-5713	121	15	-	-	PUNCT
easat-5713	121	16	to	to	ADP
easat-5713	121	17	-	-	PUNCT
easat-5713	121	18	image	image	NOUN
easat-5713	121	19	generative	generative	ADJ
easat-5713	121	20	model	model	NOUN
easat-5713	121	21	,	,	PUNCT
easat-5713	121	22	was	be	AUX
easat-5713	121	23	used	use	VERB
easat-5713	121	24	to	to	PART
easat-5713	121	25	create	create	VERB
easat-5713	121	26	additional	additional	ADJ
easat-5713	121	27	facial	facial	ADJ
easat-5713	121	28	images	image	NOUN
easat-5713	121	29	that	that	PRON
easat-5713	121	30	aimed	aim	VERB
easat-5713	121	31	to	to	PART
easat-5713	121	32	supplement	supplement	VERB
easat-5713	121	33	the	the	DET
easat-5713	121	34	existing	exist	VERB
easat-5713	121	35	fer2013	fer2013	ADJ
easat-5713	121	36	dataset	dataset	NOUN
easat-5713	121	37	.	.	PUNCT
easat-5713	122	1	unlike	unlike	ADP
easat-5713	122	2	traditional	traditional	ADJ
easat-5713	122	3	data	datum	NOUN
easat-5713	122	4	augmentation	augmentation	NOUN
easat-5713	122	5	techniques	technique	NOUN
easat-5713	122	6	,	,	PUNCT
easat-5713	122	7	which	which	PRON
easat-5713	122	8	modify	modify	VERB
easat-5713	122	9	existing	exist	VERB
easat-5713	122	10	images	image	NOUN
easat-5713	122	11	through	through	ADP
easat-5713	122	12	transformations	transformation	NOUN
easat-5713	122	13	like	like	ADP
easat-5713	122	14	rotation	rotation	NOUN
easat-5713	122	15	and	and	CCONJ
easat-5713	122	16	flipping	flip	VERB
easat-5713	122	17	,	,	PUNCT
easat-5713	122	18	aigenerated	aigenerate	VERB
easat-5713	122	19	synthetic	synthetic	ADJ
easat-5713	122	20	data	datum	NOUN
easat-5713	122	21	introduces	introduce	NOUN
easat-5713	122	22	entirely	entirely	ADV
easat-5713	122	23	new	new	ADJ
easat-5713	122	24	images	image	NOUN
easat-5713	122	25	.	.	PUNCT
easat-5713	123	1	this	this	DET
easat-5713	123	2	approach	approach	NOUN
easat-5713	123	3	offers	offer	VERB
easat-5713	123	4	greater	great	ADJ
easat-5713	123	5	diversity	diversity	NOUN
easat-5713	123	6	in	in	ADP
easat-5713	123	7	facial	facial	ADJ
easat-5713	123	8	expressions	expression	NOUN
easat-5713	123	9	,	,	PUNCT
easat-5713	123	10	environmental	environmental	ADJ
easat-5713	123	11	conditions	condition	NOUN
easat-5713	123	12	,	,	PUNCT
easat-5713	123	13	and	and	CCONJ
easat-5713	123	14	demographic	demographic	ADJ
easat-5713	123	15	representation	representation	NOUN
easat-5713	123	16	,	,	PUNCT
easat-5713	123	17	which	which	PRON
easat-5713	123	18	are	be	AUX
easat-5713	123	19	crucial	crucial	ADJ
easat-5713	123	20	for	for	ADP
easat-5713	123	21	improving	improve	VERB
easat-5713	123	22	model	model	NOUN
easat-5713	123	23	generalization	generalization	NOUN
easat-5713	123	24	.	.	PUNCT
easat-5713	124	1	to	to	PART
easat-5713	124	2	ensure	ensure	VERB
easat-5713	124	3	realistic	realistic	ADJ
easat-5713	124	4	and	and	CCONJ
easat-5713	124	5	diverse	diverse	ADJ
easat-5713	124	6	synthetic	synthetic	ADJ
easat-5713	124	7	images	image	NOUN
easat-5713	124	8	,	,	PUNCT
easat-5713	124	9	we	we	PRON
easat-5713	124	10	constructed	construct	VERB
easat-5713	124	11	detailed	detailed	ADJ
easat-5713	124	12	text	text	NOUN
easat-5713	124	13	prompts	prompt	NOUN
easat-5713	124	14	capturing	capture	VERB
easat-5713	124	15	natural	natural	ADJ
easat-5713	124	16	variations	variation	NOUN
easat-5713	124	17	observed	observe	VERB
easat-5713	124	18	in	in	ADP
easat-5713	124	19	real	real	ADJ
easat-5713	124	20	-	-	PUNCT
easat-5713	124	21	world	world	NOUN
easat-5713	124	22	images	image	NOUN
easat-5713	124	23	.	.	PUNCT
easat-5713	125	1	these	these	DET
easat-5713	125	2	variations	variation	NOUN
easat-5713	125	3	included	include	VERB
easat-5713	125	4	age	age	NOUN
easat-5713	125	5	,	,	PUNCT
easat-5713	125	6	gender	gender	NOUN
easat-5713	125	7	,	,	PUNCT
easat-5713	125	8	ethnicity	ethnicity	NOUN
easat-5713	125	9	,	,	PUNCT
easat-5713	125	10	1861	1861	NUM
easat-5713	125	11	edelweiss	edelweiss	PROPN
easat-5713	125	12	applied	apply	VERB
easat-5713	125	13	science	science	NOUN
easat-5713	125	14	and	and	CCONJ
easat-5713	125	15	technology	technology	NOUN
easat-5713	125	16	issn	issn	PROPN
easat-5713	125	17	:	:	PUNCT
easat-5713	125	18	2576	2576	NUM
easat-5713	125	19	-	-	SYM
easat-5713	125	20	8484	8484	NUM
easat-5713	125	21	vol	vol	NOUN
easat-5713	125	22	.	.	PROPN
easat-5713	126	1	9	9	NUM
easat-5713	126	2	,	,	PUNCT
easat-5713	126	3	no	no	INTJ
easat-5713	126	4	.	.	NOUN
easat-5713	127	1	3	3	NUM
easat-5713	127	2	:	:	PUNCT
easat-5713	127	3	1855	1855	NUM
easat-5713	127	4	-	-	SYM
easat-5713	127	5	1869	1869	NUM
easat-5713	127	6	,	,	PUNCT
easat-5713	127	7	2025	2025	NUM
easat-5713	127	8	doi	doi	NOUN
easat-5713	127	9	:	:	PUNCT
easat-5713	127	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	127	11	©	©	ADP
easat-5713	127	12	2025	2025	NUM
easat-5713	127	13	by	by	ADP
easat-5713	127	14	the	the	DET
easat-5713	127	15	authors	author	NOUN
easat-5713	127	16	;	;	PUNCT
easat-5713	127	17	licensee	licensee	PROPN
easat-5713	127	18	learning	learn	VERB
easat-5713	127	19	gate	gate	NOUN
easat-5713	127	20	lighting	lighting	NOUN
easat-5713	127	21	conditions	condition	NOUN
easat-5713	127	22	,	,	PUNCT
easat-5713	127	23	camera	camera	NOUN
easat-5713	127	24	angles	angle	NOUN
easat-5713	127	25	,	,	PUNCT
easat-5713	127	26	obstructions	obstruction	NOUN
easat-5713	127	27	such	such	ADJ
easat-5713	127	28	as	as	ADP
easat-5713	127	29	hands	hand	NOUN
easat-5713	127	30	,	,	PUNCT
easat-5713	127	31	hair	hair	NOUN
easat-5713	127	32	,	,	PUNCT
easat-5713	127	33	or	or	CCONJ
easat-5713	127	34	glasses	glass	NOUN
easat-5713	127	35	,	,	PUNCT
easat-5713	127	36	and	and	CCONJ
easat-5713	127	37	intentional	intentional	ADJ
easat-5713	127	38	blur	blur	NOUN
easat-5713	127	39	,	,	PUNCT
easat-5713	127	40	all	all	PRON
easat-5713	127	41	of	of	ADP
easat-5713	127	42	which	which	PRON
easat-5713	127	43	can	can	AUX
easat-5713	127	44	significantly	significantly	ADV
easat-5713	127	45	impact	impact	VERB
easat-5713	127	46	the	the	DET
easat-5713	127	47	accuracy	accuracy	NOUN
easat-5713	127	48	of	of	ADP
easat-5713	127	49	fer	fer	PROPN
easat-5713	127	50	models	model	NOUN
easat-5713	127	51	.	.	PUNCT
easat-5713	128	1	table	table	NOUN
easat-5713	128	2	1	1	NUM
easat-5713	128	3	illustrates	illustrate	VERB
easat-5713	128	4	the	the	DET
easat-5713	128	5	different	different	ADJ
easat-5713	128	6	variations	variation	NOUN
easat-5713	128	7	considered	consider	VERB
easat-5713	128	8	in	in	ADP
easat-5713	128	9	the	the	DET
easat-5713	128	10	prompt	prompt	ADJ
easat-5713	128	11	design	design	NOUN
easat-5713	128	12	process	process	NOUN
easat-5713	128	13	.	.	PUNCT
easat-5713	129	1	an	an	DET
easat-5713	129	2	example	example	NOUN
easat-5713	129	3	of	of	ADP
easat-5713	129	4	a	a	DET
easat-5713	129	5	generated	generate	VERB
easat-5713	129	6	prompt	prompt	NOUN
easat-5713	129	7	is	be	AUX
easat-5713	129	8	:	:	PUNCT
easat-5713	129	9	“	"	PUNCT
easat-5713	129	10	an	an	DET
easat-5713	129	11	image	image	NOUN
easat-5713	129	12	of	of	ADP
easat-5713	129	13	a	a	DET
easat-5713	129	14	child	child	NOUN
easat-5713	129	15	,	,	PUNCT
easat-5713	129	16	middle	middle	ADJ
easat-5713	129	17	eastern	eastern	ADJ
easat-5713	129	18	male	male	NOUN
easat-5713	129	19	with	with	ADP
easat-5713	129	20	a	a	DET
easat-5713	129	21	scowling	scowling	ADJ
easat-5713	129	22	expression	expression	NOUN
easat-5713	129	23	and	and	CCONJ
easat-5713	129	24	tight	tight	ADJ
easat-5713	129	25	lips	lip	NOUN
easat-5713	129	26	,	,	PUNCT
easat-5713	129	27	captured	capture	VERB
easat-5713	129	28	in	in	ADP
easat-5713	129	29	shadowy	shadowy	ADJ
easat-5713	129	30	light	light	NOUN
easat-5713	129	31	,	,	PUNCT
easat-5713	129	32	from	from	ADP
easat-5713	129	33	a	a	DET
easat-5713	129	34	side	side	NOUN
easat-5713	129	35	angle	angle	NOUN
easat-5713	129	36	,	,	PUNCT
easat-5713	129	37	covered	cover	VERB
easat-5713	129	38	with	with	ADP
easat-5713	129	39	hair	hair	NOUN
easat-5713	129	40	,	,	PUNCT
easat-5713	129	41	with	with	ADP
easat-5713	129	42	slight	slight	ADJ
easat-5713	129	43	blur	blur	PROPN
easat-5713	129	44	,	,	PUNCT
easat-5713	129	45	expressing	express	VERB
easat-5713	129	46	anger	anger	NOUN
easat-5713	129	47	.	.	PUNCT
easat-5713	129	48	”	"	PUNCT
easat-5713	130	1	this	this	DET
easat-5713	130	2	method	method	NOUN
easat-5713	130	3	ensures	ensure	VERB
easat-5713	130	4	that	that	SCONJ
easat-5713	130	5	the	the	DET
easat-5713	130	6	synthetic	synthetic	ADJ
easat-5713	130	7	images	image	NOUN
easat-5713	130	8	closely	closely	ADV
easat-5713	130	9	resemble	resemble	VERB
easat-5713	130	10	real	real	ADJ
easat-5713	130	11	-	-	PUNCT
easat-5713	130	12	world	world	NOUN
easat-5713	130	13	conditions	condition	NOUN
easat-5713	130	14	while	while	SCONJ
easat-5713	130	15	maintaining	maintain	VERB
easat-5713	130	16	a	a	DET
easat-5713	130	17	high	high	ADJ
easat-5713	130	18	level	level	NOUN
easat-5713	130	19	of	of	ADP
easat-5713	130	20	diversity	diversity	NOUN
easat-5713	130	21	and	and	CCONJ
easat-5713	130	22	fidelity	fidelity	NOUN
easat-5713	130	23	in	in	ADP
easat-5713	130	24	facial	facial	ADJ
easat-5713	130	25	expressions	expression	NOUN
easat-5713	130	26	.	.	PUNCT
easat-5713	131	1	in	in	ADP
easat-5713	131	2	terms	term	NOUN
easat-5713	131	3	of	of	ADP
easat-5713	131	4	model	model	NOUN
easat-5713	131	5	accuracy	accuracy	NOUN
easat-5713	131	6	,	,	PUNCT
easat-5713	131	7	which	which	PRON
easat-5713	131	8	serves	serve	VERB
easat-5713	131	9	as	as	ADP
easat-5713	131	10	a	a	DET
easat-5713	131	11	fundamental	fundamental	ADJ
easat-5713	131	12	metric	metric	NOUN
easat-5713	131	13	for	for	ADP
easat-5713	131	14	classification	classification	NOUN
easat-5713	131	15	performance	performance	NOUN
easat-5713	131	16	,	,	PUNCT
easat-5713	131	17	the	the	DET
easat-5713	131	18	vgg-19	vgg-19	NUM
easat-5713	131	19	model	model	NOUN
easat-5713	131	20	trained	train	VERB
easat-5713	131	21	on	on	ADP
easat-5713	131	22	the	the	DET
easat-5713	131	23	original	original	ADJ
easat-5713	131	24	fer2013	fer2013	ADJ
easat-5713	131	25	dataset	dataset	NOUN
easat-5713	131	26	achieved	achieve	VERB
easat-5713	131	27	an	an	DET
easat-5713	131	28	accuracy	accuracy	NOUN
easat-5713	131	29	of	of	ADP
easat-5713	131	30	65	65	NUM
easat-5713	131	31	%	%	NOUN
easat-5713	131	32	,	,	PUNCT
easat-5713	131	33	whereas	whereas	SCONJ
easat-5713	131	34	the	the	DET
easat-5713	131	35	model	model	NOUN
easat-5713	131	36	trained	train	VERB
easat-5713	131	37	on	on	ADP
easat-5713	131	38	the	the	DET
easat-5713	131	39	augmented	augment	VERB
easat-5713	131	40	dataset	dataset	NOUN
easat-5713	131	41	containing	contain	VERB
easat-5713	131	42	synthetic	synthetic	ADJ
easat-5713	131	43	images	image	NOUN
easat-5713	131	44	obtained	obtain	VERB
easat-5713	131	45	a	a	DET
easat-5713	131	46	slightly	slightly	ADV
easat-5713	131	47	lower	low	ADJ
easat-5713	131	48	accuracy	accuracy	NOUN
easat-5713	131	49	of	of	ADP
easat-5713	131	50	63	63	NUM
easat-5713	131	51	%	%	NOUN
easat-5713	131	52	.	.	PUNCT
easat-5713	132	1	this	this	PRON
easat-5713	132	2	suggests	suggest	VERB
easat-5713	132	3	that	that	SCONJ
easat-5713	132	4	the	the	DET
easat-5713	132	5	inclusion	inclusion	NOUN
easat-5713	132	6	of	of	ADP
easat-5713	132	7	synthetic	synthetic	ADJ
easat-5713	132	8	images	image	NOUN
easat-5713	132	9	did	do	AUX
easat-5713	132	10	not	not	PART
easat-5713	132	11	lead	lead	VERB
easat-5713	132	12	to	to	ADP
easat-5713	132	13	the	the	DET
easat-5713	132	14	anticipated	anticipate	VERB
easat-5713	132	15	improvement	improvement	NOUN
easat-5713	132	16	in	in	ADP
easat-5713	132	17	model	model	NOUN
easat-5713	132	18	performance	performance	NOUN
easat-5713	132	19	.	.	PUNCT
easat-5713	133	1	table	table	NOUN
easat-5713	133	2	1	1	NUM
easat-5713	133	3	.	.	PUNCT
easat-5713	133	4	natural	natural	ADJ
easat-5713	133	5	variations	variation	NOUN
easat-5713	133	6	of	of	ADP
easat-5713	133	7	textures	texture	NOUN
easat-5713	133	8	that	that	PRON
easat-5713	133	9	occur	occur	VERB
easat-5713	133	10	in	in	ADP
easat-5713	133	11	the	the	DET
easat-5713	133	12	images	image	NOUN
easat-5713	133	13	.	.	PUNCT
easat-5713	134	1	age	age	NOUN
easat-5713	134	2	group	group	NOUN
easat-5713	134	3	gender	gender	NOUN
easat-5713	134	4	ethnicity	ethnicity	NOUN
easat-5713	134	5	lighting	lighting	NOUN
easat-5713	134	6	angle	angle	NOUN
easat-5713	134	7	obstructions	obstruction	NOUN
easat-5713	134	8	blurriness	blurriness	NOUN
easat-5713	134	9	child	child	NOUN
easat-5713	134	10	male	male	PROPN
easat-5713	134	11	caucasian	caucasian	ADJ
easat-5713	134	12	bright	bright	ADJ
easat-5713	134	13	frontal	frontal	ADJ
easat-5713	134	14	hands	hand	NOUN
easat-5713	134	15	slight	slight	ADJ
easat-5713	134	16	teenager	teenager	NOUN
easat-5713	134	17	female	female	ADJ
easat-5713	134	18	african	african	ADJ
easat-5713	134	19	low	low	ADJ
easat-5713	134	20	side	side	NOUN
easat-5713	134	21	hair	hair	NOUN
easat-5713	134	22	moderate	moderate	ADJ
easat-5713	134	23	young	young	ADJ
easat-5713	134	24	adult	adult	NOUN
easat-5713	134	25	asian	asian	ADJ
easat-5713	134	26	shadowy	shadowy	ADJ
easat-5713	134	27	top	top	ADJ
easat-5713	134	28	-	-	PUNCT
easat-5713	134	29	down	down	ADP
easat-5713	134	30	glasses	glass	NOUN
easat-5713	134	31	middle	middle	ADJ
easat-5713	134	32	-	-	PUNCT
easat-5713	134	33	aged	aged	ADJ
easat-5713	134	34	hispanic	hispanic	ADJ
easat-5713	134	35	bottom	bottom	NOUN
easat-5713	134	36	-	-	PUNCT
easat-5713	134	37	up	up	ADP
easat-5713	134	38	mask	mask	NOUN
easat-5713	134	39	elderly	elderly	ADJ
easat-5713	134	40	middle	middle	ADJ
easat-5713	134	41	eastern	eastern	ADJ
easat-5713	134	42	a	a	DET
easat-5713	134	43	total	total	NOUN
easat-5713	134	44	of	of	ADP
easat-5713	134	45	300	300	NUM
easat-5713	134	46	synthetic	synthetic	ADJ
easat-5713	134	47	images	image	NOUN
easat-5713	134	48	were	be	AUX
easat-5713	134	49	generated	generate	VERB
easat-5713	134	50	for	for	ADP
easat-5713	134	51	each	each	DET
easat-5713	134	52	emotion	emotion	NOUN
easat-5713	134	53	category	category	NOUN
easat-5713	134	54	,	,	PUNCT
easat-5713	134	55	which	which	PRON
easat-5713	134	56	were	be	AUX
easat-5713	134	57	then	then	ADV
easat-5713	134	58	appended	append	VERB
easat-5713	134	59	to	to	ADP
easat-5713	134	60	the	the	DET
easat-5713	134	61	fer2013	fer2013	ADJ
easat-5713	134	62	dataset	dataset	NOUN
easat-5713	134	63	to	to	PART
easat-5713	134	64	form	form	VERB
easat-5713	134	65	an	an	DET
easat-5713	134	66	augmented	augment	VERB
easat-5713	134	67	dataset	dataset	NOUN
easat-5713	134	68	.	.	PUNCT
easat-5713	135	1	the	the	DET
easat-5713	135	2	composition	composition	NOUN
easat-5713	135	3	of	of	ADP
easat-5713	135	4	the	the	DET
easat-5713	135	5	augmented	augment	VERB
easat-5713	135	6	dataset	dataset	NOUN
easat-5713	135	7	,	,	PUNCT
easat-5713	135	8	as	as	SCONJ
easat-5713	135	9	shown	show	VERB
easat-5713	135	10	in	in	ADP
easat-5713	135	11	table	table	NOUN
easat-5713	135	12	2	2	NUM
easat-5713	135	13	,	,	PUNCT
easat-5713	135	14	resulted	result	VERB
easat-5713	135	15	in	in	ADP
easat-5713	135	16	a	a	DET
easat-5713	135	17	balanced	balanced	ADJ
easat-5713	135	18	distribution	distribution	NOUN
easat-5713	135	19	of	of	ADP
easat-5713	135	20	real	real	ADJ
easat-5713	135	21	and	and	CCONJ
easat-5713	135	22	synthetic	synthetic	ADJ
easat-5713	135	23	images	image	NOUN
easat-5713	135	24	across	across	ADP
easat-5713	135	25	all	all	DET
easat-5713	135	26	emotion	emotion	NOUN
easat-5713	135	27	categories	category	NOUN
easat-5713	135	28	.	.	PUNCT
easat-5713	136	1	sample	sample	NOUN
easat-5713	136	2	synthetic	synthetic	NOUN
easat-5713	136	3	images	image	NOUN
easat-5713	136	4	generated	generate	VERB
easat-5713	136	5	by	by	ADP
easat-5713	136	6	the	the	DET
easat-5713	136	7	juggernaut	juggernaut	PROPN
easat-5713	136	8	xl	xl	PROPN
easat-5713	136	9	model	model	PROPN
easat-5713	136	10	were	be	AUX
easat-5713	136	11	analyzed	analyze	VERB
easat-5713	136	12	for	for	ADP
easat-5713	136	13	quality	quality	NOUN
easat-5713	136	14	and	and	CCONJ
easat-5713	136	15	consistency	consistency	NOUN
easat-5713	136	16	before	before	ADP
easat-5713	136	17	being	be	AUX
easat-5713	136	18	incorporated	incorporate	VERB
easat-5713	136	19	into	into	ADP
easat-5713	136	20	the	the	DET
easat-5713	136	21	training	training	NOUN
easat-5713	136	22	pipeline	pipeline	NOUN
easat-5713	136	23	.	.	PUNCT
easat-5713	137	1	we	we	PRON
easat-5713	137	2	trained	train	VERB
easat-5713	137	3	the	the	DET
easat-5713	137	4	vgg-19	vgg-19	PROPN
easat-5713	137	5	model	model	NOUN
easat-5713	137	6	on	on	ADP
easat-5713	137	7	both	both	CCONJ
easat-5713	137	8	the	the	DET
easat-5713	137	9	original	original	ADJ
easat-5713	137	10	fer2013	fer2013	ADJ
easat-5713	137	11	dataset	dataset	NOUN
easat-5713	137	12	and	and	CCONJ
easat-5713	137	13	the	the	DET
easat-5713	137	14	augmented	augment	VERB
easat-5713	137	15	version	version	NOUN
easat-5713	137	16	,	,	PUNCT
easat-5713	137	17	which	which	PRON
easat-5713	137	18	included	include	VERB
easat-5713	137	19	synthetic	synthetic	ADJ
easat-5713	137	20	images	image	NOUN
easat-5713	137	21	.	.	PUNCT
easat-5713	138	1	the	the	DET
easat-5713	138	2	vgg-19	vgg-19	PROPN
easat-5713	138	3	model	model	NOUN
easat-5713	138	4	,	,	PUNCT
easat-5713	138	5	a	a	DET
easat-5713	138	6	widely	widely	ADV
easat-5713	138	7	used	use	VERB
easat-5713	138	8	convolutional	convolutional	ADJ
easat-5713	138	9	neural	neural	ADJ
easat-5713	138	10	network	network	NOUN
easat-5713	138	11	for	for	ADP
easat-5713	138	12	image	image	NOUN
easat-5713	138	13	classification	classification	NOUN
easat-5713	138	14	,	,	PUNCT
easat-5713	138	15	was	be	AUX
easat-5713	138	16	trained	train	VERB
easat-5713	138	17	using	use	VERB
easat-5713	138	18	the	the	DET
easat-5713	138	19	adam	adam	PROPN
easat-5713	138	20	optimizer	optimizer	NOUN
easat-5713	138	21	with	with	ADP
easat-5713	138	22	a	a	DET
easat-5713	138	23	learning	learn	VERB
easat-5713	138	24	rate	rate	NOUN
easat-5713	138	25	of	of	ADP
easat-5713	138	26	0.0001	0.0001	NUM
easat-5713	138	27	,	,	PUNCT
easat-5713	138	28	a	a	DET
easat-5713	138	29	batch	batch	NOUN
easat-5713	138	30	size	size	NOUN
easat-5713	138	31	of	of	ADP
easat-5713	138	32	8	8	NUM
easat-5713	138	33	,	,	PUNCT
easat-5713	138	34	and	and	CCONJ
easat-5713	138	35	for	for	ADP
easat-5713	138	36	25	25	NUM
easat-5713	138	37	epochs	epoch	NOUN
easat-5713	138	38	.	.	PUNCT
easat-5713	139	1	table	table	NOUN
easat-5713	139	2	2	2	NUM
easat-5713	139	3	.	.	PUNCT
easat-5713	139	4	fer2013	fer2013	PROPN
easat-5713	139	5	dataset	dataset	PROPN
easat-5713	139	6	augmented	augment	VERB
easat-5713	139	7	with	with	ADP
easat-5713	139	8	ai	ai	ADP
easat-5713	139	9	generated	generate	VERB
easat-5713	139	10	images	image	NOUN
easat-5713	139	11	.	.	PUNCT
easat-5713	140	1	emotion	emotion	NOUN
easat-5713	140	2	(	(	PUNCT
easat-5713	140	3	class	class	NOUN
easat-5713	140	4	)	)	PUNCT
easat-5713	140	5	training	training	NOUN
easat-5713	140	6	validation	validation	NOUN
easat-5713	140	7	testing	testing	NOUN
easat-5713	140	8	class	class	NOUN
easat-5713	140	9	total	total	NOUN
easat-5713	140	10	angry	angry	ADJ
easat-5713	140	11	4231	4231	NUM
easat-5713	140	12	497	497	NUM
easat-5713	140	13	525	525	NUM
easat-5713	140	14	5253	5253	NUM
easat-5713	140	15	disgust	disgust	NOUN
easat-5713	140	16	666	666	NUM
easat-5713	140	17	87	87	NUM
easat-5713	140	18	94	94	NUM
easat-5713	140	19	847	847	NUM
easat-5713	140	20	fear	fear	VERB
easat-5713	140	21	4344	4344	NUM
easat-5713	140	22	524	524	NUM
easat-5713	140	23	553	553	NUM
easat-5713	140	24	5421	5421	NUM
easat-5713	140	25	happy	happy	ADJ
easat-5713	140	26	7461	7461	NUM
easat-5713	140	27	922	922	NUM
easat-5713	140	28	906	906	NUM
easat-5713	140	29	9289	9289	NUM
easat-5713	140	30	sad	sad	ADJ
easat-5713	140	31	5065	5065	NUM
easat-5713	140	32	682	682	NUM
easat-5713	140	33	630	630	NUM
easat-5713	140	34	6377	6377	NUM
easat-5713	140	35	surprise	surprise	NOUN
easat-5713	140	36	3410	3410	NUM
easat-5713	140	37	450	450	NUM
easat-5713	140	38	442	442	NUM
easat-5713	140	39	4302	4302	NUM
easat-5713	140	40	neutral	neutral	ADJ
easat-5713	140	41	5211	5211	NUM
easat-5713	140	42	637	637	NUM
easat-5713	140	43	650	650	NUM
easat-5713	140	44	6498	6498	NUM
easat-5713	140	45	total	total	ADJ
easat-5713	140	46	30388	30388	NUM
easat-5713	140	47	3799	3799	NUM
easat-5713	140	48	3800	3800	NUM
easat-5713	140	49	37987	37987	NUM
easat-5713	140	50	additionally	additionally	ADV
easat-5713	140	51	,	,	PUNCT
easat-5713	140	52	precision	precision	NOUN
easat-5713	140	53	,	,	PUNCT
easat-5713	140	54	recall	recall	NOUN
easat-5713	140	55	,	,	PUNCT
easat-5713	140	56	and	and	CCONJ
easat-5713	140	57	f1	f1	NOUN
easat-5713	140	58	-	-	PUNCT
easat-5713	140	59	score	score	NOUN
easat-5713	140	60	results	result	NOUN
easat-5713	140	61	exhibited	exhibit	VERB
easat-5713	140	62	variability	variability	NOUN
easat-5713	140	63	across	across	ADP
easat-5713	140	64	different	different	ADJ
easat-5713	140	65	emotion	emotion	NOUN
easat-5713	140	66	categories	category	NOUN
easat-5713	140	67	when	when	SCONJ
easat-5713	140	68	training	train	VERB
easat-5713	140	69	on	on	ADP
easat-5713	140	70	the	the	DET
easat-5713	140	71	augmented	augment	VERB
easat-5713	140	72	dataset	dataset	NOUN
easat-5713	140	73	.	.	PUNCT
easat-5713	141	1	the	the	DET
easat-5713	141	2	classification	classification	NOUN
easat-5713	141	3	reports	report	NOUN
easat-5713	141	4	for	for	ADP
easat-5713	141	5	both	both	DET
easat-5713	141	6	datasets	dataset	NOUN
easat-5713	141	7	,	,	PUNCT
easat-5713	141	8	provided	provide	VERB
easat-5713	141	9	in	in	ADP
easat-5713	141	10	tables	table	NOUN
easat-5713	141	11	3	3	NUM
easat-5713	141	12	and	and	CCONJ
easat-5713	141	13	4	4	NUM
easat-5713	141	14	,	,	PUNCT
easat-5713	141	15	highlight	highlight	VERB
easat-5713	141	16	these	these	DET
easat-5713	141	17	differences	difference	NOUN
easat-5713	141	18	in	in	ADP
easat-5713	141	19	performance	performance	NOUN
easat-5713	141	20	.	.	PUNCT
easat-5713	142	1	while	while	SCONJ
easat-5713	142	2	some	some	DET
easat-5713	142	3	emotions	emotion	NOUN
easat-5713	142	4	benefited	benefit	VERB
easat-5713	142	5	slightly	slightly	ADV
easat-5713	142	6	from	from	ADP
easat-5713	142	7	the	the	DET
easat-5713	142	8	additional	additional	ADJ
easat-5713	142	9	data	datum	NOUN
easat-5713	142	10	,	,	PUNCT
easat-5713	142	11	others	other	NOUN
easat-5713	142	12	experienced	experience	VERB
easat-5713	142	13	increased	increase	VERB
easat-5713	142	14	misclassification	misclassification	NOUN
easat-5713	142	15	rates	rate	NOUN
easat-5713	142	16	,	,	PUNCT
easat-5713	142	17	particularly	particularly	ADV
easat-5713	142	18	those	those	DET
easat-5713	142	19	expressions	expression	NOUN
easat-5713	142	20	that	that	PRON
easat-5713	142	21	require	require	VERB
easat-5713	142	22	nuanced	nuanced	ADJ
easat-5713	142	23	feature	feature	NOUN
easat-5713	142	24	recognition	recognition	NOUN
easat-5713	142	25	,	,	PUNCT
easat-5713	142	26	such	such	ADJ
easat-5713	142	27	as	as	ADP
easat-5713	142	28	"	"	PUNCT
easat-5713	142	29	disgust	disgust	NOUN
easat-5713	142	30	"	"	PUNCT
easat-5713	142	31	and	and	CCONJ
easat-5713	142	32	"	"	PUNCT
easat-5713	142	33	happy	happy	ADJ
easat-5713	142	34	.	.	PUNCT
easat-5713	142	35	"	"	PUNCT
easat-5713	143	1	one	one	NUM
easat-5713	143	2	of	of	ADP
easat-5713	143	3	the	the	DET
easat-5713	143	4	primary	primary	ADJ
easat-5713	143	5	motivations	motivation	NOUN
easat-5713	143	6	for	for	ADP
easat-5713	143	7	integrating	integrate	VERB
easat-5713	143	8	synthetic	synthetic	ADJ
easat-5713	143	9	data	datum	NOUN
easat-5713	143	10	was	be	AUX
easat-5713	143	11	to	to	PART
easat-5713	143	12	assess	assess	VERB
easat-5713	143	13	whether	whether	SCONJ
easat-5713	143	14	it	it	PRON
easat-5713	143	15	could	could	AUX
easat-5713	143	16	reduce	reduce	VERB
easat-5713	143	17	the	the	DET
easat-5713	143	18	impact	impact	NOUN
easat-5713	143	19	of	of	ADP
easat-5713	143	20	noisy	noisy	ADJ
easat-5713	143	21	images	image	NOUN
easat-5713	143	22	in	in	ADP
easat-5713	143	23	the	the	DET
easat-5713	143	24	training	training	NOUN
easat-5713	143	25	process	process	NOUN
easat-5713	143	26	.	.	PUNCT
easat-5713	144	1	the	the	DET
easat-5713	144	2	fer2013	fer2013	ADJ
easat-5713	144	3	dataset	dataset	NOUN
easat-5713	144	4	contains	contain	VERB
easat-5713	144	5	a	a	DET
easat-5713	144	6	significant	significant	ADJ
easat-5713	144	7	number	number	NOUN
easat-5713	144	8	of	of	ADP
easat-5713	144	9	low	low	ADJ
easat-5713	144	10	-	-	PUNCT
easat-5713	144	11	quality	quality	NOUN
easat-5713	144	12	images	image	NOUN
easat-5713	144	13	,	,	PUNCT
easat-5713	144	14	including	include	VERB
easat-5713	144	15	blurred	blurred	ADJ
easat-5713	144	16	expressions	expression	NOUN
easat-5713	144	17	,	,	PUNCT
easat-5713	144	18	occluded	occlude	VERB
easat-5713	144	19	faces	face	NOUN
easat-5713	144	20	,	,	PUNCT
easat-5713	144	21	and	and	CCONJ
easat-5713	144	22	variations	variation	NOUN
easat-5713	144	23	in	in	ADP
easat-5713	144	24	lighting	lighting	NOUN
easat-5713	144	25	that	that	PRON
easat-5713	144	26	make	make	VERB
easat-5713	144	27	it	it	PRON
easat-5713	144	28	difficult	difficult	ADJ
easat-5713	144	29	for	for	SCONJ
easat-5713	144	30	machine	machine	NOUN
easat-5713	144	31	learning	learning	NOUN
easat-5713	144	32	models	model	NOUN
easat-5713	144	33	to	to	PART
easat-5713	144	34	extract	extract	VERB
easat-5713	144	35	relevant	relevant	ADJ
easat-5713	144	36	features	feature	NOUN
easat-5713	144	37	.	.	PUNCT
easat-5713	145	1	by	by	ADP
easat-5713	145	2	generating	generate	VERB
easat-5713	145	3	synthetic	synthetic	ADJ
easat-5713	145	4	1862	1862	NUM
easat-5713	145	5	edelweiss	edelweiss	PROPN
easat-5713	145	6	applied	apply	VERB
easat-5713	145	7	science	science	NOUN
easat-5713	145	8	and	and	CCONJ
easat-5713	145	9	technology	technology	NOUN
easat-5713	145	10	issn	issn	PROPN
easat-5713	145	11	:	:	PUNCT
easat-5713	145	12	2576	2576	NUM
easat-5713	145	13	-	-	SYM
easat-5713	145	14	8484	8484	NUM
easat-5713	145	15	vol	vol	NOUN
easat-5713	145	16	.	.	PROPN
easat-5713	146	1	9	9	NUM
easat-5713	146	2	,	,	PUNCT
easat-5713	146	3	no	no	INTJ
easat-5713	146	4	.	.	NOUN
easat-5713	147	1	3	3	NUM
easat-5713	147	2	:	:	PUNCT
easat-5713	147	3	1855	1855	NUM
easat-5713	147	4	-	-	SYM
easat-5713	147	5	1869	1869	NUM
easat-5713	147	6	,	,	PUNCT
easat-5713	147	7	2025	2025	NUM
easat-5713	147	8	doi	doi	NOUN
easat-5713	147	9	:	:	PUNCT
easat-5713	147	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	147	11	©	©	ADP
easat-5713	147	12	2025	2025	NUM
easat-5713	147	13	by	by	ADP
easat-5713	147	14	the	the	DET
easat-5713	147	15	authors	author	NOUN
easat-5713	147	16	;	;	PUNCT
easat-5713	147	17	licensee	licensee	PROPN
easat-5713	147	18	learning	learn	VERB
easat-5713	147	19	gate	gate	NOUN
easat-5713	147	20	images	image	NOUN
easat-5713	147	21	with	with	ADP
easat-5713	147	22	controlled	control	VERB
easat-5713	147	23	variations	variation	NOUN
easat-5713	147	24	in	in	ADP
easat-5713	147	25	lighting	lighting	NOUN
easat-5713	147	26	,	,	PUNCT
easat-5713	147	27	angles	angle	NOUN
easat-5713	147	28	,	,	PUNCT
easat-5713	147	29	and	and	CCONJ
easat-5713	147	30	occlusions	occlusion	NOUN
easat-5713	147	31	,	,	PUNCT
easat-5713	147	32	we	we	PRON
easat-5713	147	33	hypothesized	hypothesize	VERB
easat-5713	147	34	that	that	SCONJ
easat-5713	147	35	the	the	DET
easat-5713	147	36	model	model	NOUN
easat-5713	147	37	would	would	AUX
easat-5713	147	38	learn	learn	VERB
easat-5713	147	39	more	more	ADV
easat-5713	147	40	robust	robust	ADJ
easat-5713	147	41	features	feature	NOUN
easat-5713	147	42	and	and	CCONJ
easat-5713	147	43	perform	perform	VERB
easat-5713	147	44	better	well	ADV
easat-5713	147	45	in	in	ADP
easat-5713	147	46	real	real	ADJ
easat-5713	147	47	-	-	PUNCT
easat-5713	147	48	world	world	NOUN
easat-5713	147	49	settings	setting	NOUN
easat-5713	147	50	[	[	X
easat-5713	147	51	22	22	NUM
easat-5713	147	52	]	]	PUNCT
easat-5713	147	53	.	.	PUNCT
easat-5713	148	1	table	table	NOUN
easat-5713	148	2	3	3	NUM
easat-5713	148	3	.	.	PUNCT
easat-5713	149	1	the	the	DET
easat-5713	149	2	report	report	NOUN
easat-5713	149	3	on	on	ADP
easat-5713	149	4	classification	classification	NOUN
easat-5713	149	5	of	of	ADP
easat-5713	149	6	the	the	DET
easat-5713	149	7	original	original	ADJ
easat-5713	149	8	fer2013	fer2013	ADJ
easat-5713	149	9	dataset	dataset	NOUN
easat-5713	149	10	.	.	PUNCT
easat-5713	150	1	class	class	NOUN
easat-5713	150	2	precision	precision	NOUN
easat-5713	150	3	recall	recall	NOUN
easat-5713	150	4	f1	f1	NOUN
easat-5713	150	5	-	-	PUNCT
easat-5713	150	6	score	score	NOUN
easat-5713	150	7	neutral	neutral	ADJ
easat-5713	150	8	0.58	0.58	NUM
easat-5713	150	9	0.55	0.55	NUM
easat-5713	150	10	0.56	0.56	NUM
easat-5713	150	11	angry	angry	ADJ
easat-5713	150	12	0.50	0.50	NUM
easat-5713	150	13	0.28	0.28	NUM
easat-5713	150	14	0.36	0.36	NUM
easat-5713	150	15	disgust	disgust	NOUN
easat-5713	150	16	0.59	0.59	NUM
easat-5713	150	17	0.32	0.32	NUM
easat-5713	150	18	0.42	0.42	NUM
easat-5713	150	19	fear	fear	NOUN
easat-5713	150	20	0.85	0.85	NUM
easat-5713	150	21	0.87	0.87	NUM
easat-5713	150	22	0.86	0.86	NUM
easat-5713	150	23	happy	happy	ADJ
easat-5713	150	24	0.50	0.50	NUM
easat-5713	150	25	0.63	0.63	NUM
easat-5713	150	26	0.56	0.56	NUM
easat-5713	150	27	sad	sad	ADJ
easat-5713	150	28	0.75	0.75	NUM
easat-5713	150	29	0.82	0.82	NUM
easat-5713	150	30	0.78	0.78	NUM
easat-5713	150	31	surprise	surprise	NOUN
easat-5713	150	32	0.58	0.58	NUM
easat-5713	150	33	0.56	0.56	NUM
easat-5713	150	34	0.57	0.57	NUM
easat-5713	150	35	table	table	NOUN
easat-5713	150	36	4	4	NUM
easat-5713	150	37	.	.	PUNCT
easat-5713	151	1	the	the	DET
easat-5713	151	2	report	report	NOUN
easat-5713	151	3	on	on	ADP
easat-5713	151	4	classification	classification	NOUN
easat-5713	151	5	of	of	ADP
easat-5713	151	6	the	the	DET
easat-5713	151	7	augmented	augment	VERB
easat-5713	151	8	fer2013	fer2013	ADJ
easat-5713	151	9	dataset	dataset	NOUN
easat-5713	151	10	.	.	PUNCT
easat-5713	152	1	class	class	NOUN
easat-5713	152	2	precision	precision	NOUN
easat-5713	152	3	recall	recall	NOUN
easat-5713	152	4	f1	f1	NOUN
easat-5713	152	5	-	-	PUNCT
easat-5713	152	6	score	score	NOUN
easat-5713	152	7	neutral	neutral	ADJ
easat-5713	152	8	0.58	0.58	NUM
easat-5713	152	9	0.61	0.61	NUM
easat-5713	152	10	0.59	0.59	NUM
easat-5713	152	11	angry	angry	ADJ
easat-5713	152	12	0.59	0.59	NUM
easat-5713	152	13	0.48	0.48	NUM
easat-5713	152	14	0.53	0.53	NUM
easat-5713	152	15	disgust	disgust	NOUN
easat-5713	152	16	0.40	0.40	NUM
easat-5713	152	17	0.55	0.55	NUM
easat-5713	152	18	0.47	0.47	NUM
easat-5713	152	19	fear	fear	VERB
easat-5713	152	20	0.90	0.90	NUM
easat-5713	152	21	0.83	0.83	NUM
easat-5713	152	22	0.86	0.86	NUM
easat-5713	152	23	happy	happy	ADJ
easat-5713	152	24	0.55	0.55	NUM
easat-5713	152	25	0.47	0.47	NUM
easat-5713	152	26	0.51	0.51	NUM
easat-5713	152	27	sad	sad	ADJ
easat-5713	152	28	0.74	0.74	NUM
easat-5713	152	29	0.80	0.80	NUM
easat-5713	152	30	0.77	0.77	NUM
easat-5713	152	31	surprise	surprise	NOUN
easat-5713	152	32	0.57	0.57	NUM
easat-5713	152	33	0.58	0.58	NUM
easat-5713	152	34	0.58	0.58	NUM
easat-5713	152	35	the	the	DET
easat-5713	152	36	results	result	NOUN
easat-5713	152	37	suggest	suggest	VERB
easat-5713	152	38	that	that	SCONJ
easat-5713	152	39	the	the	DET
easat-5713	152	40	synthetic	synthetic	ADJ
easat-5713	152	41	data	datum	NOUN
easat-5713	152	42	generated	generate	VERB
easat-5713	152	43	by	by	ADP
easat-5713	152	44	the	the	DET
easat-5713	152	45	juggernaut	juggernaut	PROPN
easat-5713	152	46	xl	xl	PROPN
easat-5713	152	47	model	model	PROPN
easat-5713	152	48	did	do	AUX
easat-5713	152	49	not	not	PART
easat-5713	152	50	significantly	significantly	ADV
easat-5713	152	51	enhance	enhance	VERB
easat-5713	152	52	the	the	DET
easat-5713	152	53	model	model	NOUN
easat-5713	152	54	performance	performance	NOUN
easat-5713	152	55	.	.	PUNCT
easat-5713	153	1	one	one	NUM
easat-5713	153	2	plausible	plausible	ADJ
easat-5713	153	3	explanation	explanation	NOUN
easat-5713	153	4	is	be	AUX
easat-5713	153	5	the	the	DET
easat-5713	153	6	quality	quality	NOUN
easat-5713	153	7	and	and	CCONJ
easat-5713	153	8	realism	realism	NOUN
easat-5713	153	9	of	of	ADP
easat-5713	153	10	the	the	DET
easat-5713	153	11	synthetic	synthetic	ADJ
easat-5713	153	12	images	image	NOUN
easat-5713	153	13	.	.	PUNCT
easat-5713	154	1	despite	despite	SCONJ
easat-5713	154	2	carefully	carefully	ADV
easat-5713	154	3	crafted	craft	VERB
easat-5713	154	4	prompts	prompt	NOUN
easat-5713	154	5	,	,	PUNCT
easat-5713	154	6	the	the	DET
easat-5713	154	7	synthetic	synthetic	ADJ
easat-5713	154	8	images	image	NOUN
easat-5713	154	9	may	may	AUX
easat-5713	154	10	not	not	PART
easat-5713	154	11	have	have	AUX
easat-5713	154	12	captured	capture	VERB
easat-5713	154	13	the	the	DET
easat-5713	154	14	fine	fine	ADV
easat-5713	154	15	-	-	PUNCT
easat-5713	154	16	grained	grain	VERB
easat-5713	154	17	variations	variation	NOUN
easat-5713	154	18	in	in	ADP
easat-5713	154	19	real	real	ADJ
easat-5713	154	20	-	-	PUNCT
easat-5713	154	21	world	world	NOUN
easat-5713	154	22	facial	facial	ADJ
easat-5713	154	23	expressions	expression	NOUN
easat-5713	154	24	.	.	PUNCT
easat-5713	155	1	the	the	DET
easat-5713	155	2	presence	presence	NOUN
easat-5713	155	3	of	of	ADP
easat-5713	155	4	artifacts	artifact	NOUN
easat-5713	155	5	,	,	PUNCT
easat-5713	155	6	exaggerated	exaggerated	ADJ
easat-5713	155	7	features	feature	NOUN
easat-5713	155	8	,	,	PUNCT
easat-5713	155	9	or	or	CCONJ
easat-5713	155	10	inconsistencies	inconsistency	NOUN
easat-5713	155	11	in	in	ADP
easat-5713	155	12	shading	shading	NOUN
easat-5713	155	13	and	and	CCONJ
easat-5713	155	14	texture	texture	NOUN
easat-5713	155	15	could	could	AUX
easat-5713	155	16	have	have	AUX
easat-5713	155	17	introduced	introduce	VERB
easat-5713	155	18	additional	additional	ADJ
easat-5713	155	19	noise	noise	NOUN
easat-5713	155	20	rather	rather	ADV
easat-5713	155	21	than	than	ADP
easat-5713	155	22	meaningful	meaningful	ADJ
easat-5713	155	23	training	training	NOUN
easat-5713	155	24	data	datum	NOUN
easat-5713	155	25	.	.	PUNCT
easat-5713	156	1	1863	1863	NUM
easat-5713	156	2	edelweiss	edelweiss	PROPN
easat-5713	156	3	applied	apply	VERB
easat-5713	156	4	science	science	NOUN
easat-5713	156	5	and	and	CCONJ
easat-5713	156	6	technology	technology	NOUN
easat-5713	156	7	issn	issn	PROPN
easat-5713	156	8	:	:	PUNCT
easat-5713	156	9	2576	2576	NUM
easat-5713	156	10	-	-	SYM
easat-5713	156	11	8484	8484	NUM
easat-5713	156	12	vol	vol	NOUN
easat-5713	156	13	.	.	PROPN
easat-5713	157	1	9	9	NUM
easat-5713	157	2	,	,	PUNCT
easat-5713	157	3	no	no	INTJ
easat-5713	157	4	.	.	NOUN
easat-5713	158	1	3	3	NUM
easat-5713	158	2	:	:	PUNCT
easat-5713	158	3	1855	1855	NUM
easat-5713	158	4	-	-	SYM
easat-5713	158	5	1869	1869	NUM
easat-5713	158	6	,	,	PUNCT
easat-5713	158	7	2025	2025	NUM
easat-5713	158	8	doi	doi	NOUN
easat-5713	158	9	:	:	PUNCT
easat-5713	158	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	158	11	©	©	ADP
easat-5713	158	12	2025	2025	NUM
easat-5713	158	13	by	by	ADP
easat-5713	158	14	the	the	DET
easat-5713	158	15	authors	author	NOUN
easat-5713	158	16	;	;	PUNCT
easat-5713	158	17	licensee	licensee	PROPN
easat-5713	158	18	learning	learn	VERB
easat-5713	158	19	gate	gate	NOUN
easat-5713	158	20	figure	figure	NOUN
easat-5713	158	21	4	4	NUM
easat-5713	158	22	.	.	PUNCT
easat-5713	158	23	confusion	confusion	NOUN
easat-5713	158	24	matrix	matrix	NOUN
easat-5713	158	25	of	of	ADP
easat-5713	158	26	vgg-19	vgg-19	PROPN
easat-5713	158	27	model	model	NOUN
easat-5713	158	28	on	on	ADP
easat-5713	158	29	the	the	DET
easat-5713	158	30	original	original	ADJ
easat-5713	158	31	fer-2013	fer-2013	NOUN
easat-5713	158	32	dataset	dataset	NOUN
easat-5713	158	33	.	.	PUNCT
easat-5713	159	1	to	to	PART
easat-5713	159	2	analyze	analyze	VERB
easat-5713	159	3	the	the	DET
easat-5713	159	4	effect	effect	NOUN
easat-5713	159	5	of	of	ADP
easat-5713	159	6	synthetic	synthetic	ADJ
easat-5713	159	7	data	datum	NOUN
easat-5713	159	8	on	on	ADP
easat-5713	159	9	noise	noise	NOUN
easat-5713	159	10	reduction	reduction	NOUN
easat-5713	159	11	,	,	PUNCT
easat-5713	159	12	confusion	confusion	NOUN
easat-5713	159	13	matrices	matrix	NOUN
easat-5713	159	14	,	,	PUNCT
easat-5713	159	15	as	as	SCONJ
easat-5713	159	16	shown	show	VERB
easat-5713	159	17	in	in	ADP
easat-5713	159	18	figures	figure	NOUN
easat-5713	159	19	4	4	NUM
easat-5713	159	20	and	and	CCONJ
easat-5713	159	21	5	5	NUM
easat-5713	159	22	,	,	PUNCT
easat-5713	159	23	were	be	AUX
easat-5713	159	24	generated	generate	VERB
easat-5713	159	25	for	for	ADP
easat-5713	159	26	both	both	CCONJ
easat-5713	159	27	the	the	DET
easat-5713	159	28	original	original	ADJ
easat-5713	159	29	and	and	CCONJ
easat-5713	159	30	augmented	augment	VERB
easat-5713	159	31	datasets	dataset	NOUN
easat-5713	159	32	.	.	PUNCT
easat-5713	160	1	these	these	DET
easat-5713	160	2	matrices	matrix	NOUN
easat-5713	160	3	provide	provide	VERB
easat-5713	160	4	a	a	DET
easat-5713	160	5	detailed	detailed	ADJ
easat-5713	160	6	breakdown	breakdown	NOUN
easat-5713	160	7	of	of	ADP
easat-5713	160	8	true	true	ADJ
easat-5713	160	9	positives	positive	NOUN
easat-5713	160	10	,	,	PUNCT
easat-5713	160	11	false	false	ADJ
easat-5713	160	12	positives	positive	NOUN
easat-5713	160	13	,	,	PUNCT
easat-5713	160	14	and	and	CCONJ
easat-5713	160	15	false	false	ADJ
easat-5713	160	16	negatives	negative	NOUN
easat-5713	160	17	for	for	ADP
easat-5713	160	18	each	each	DET
easat-5713	160	19	emotion	emotion	NOUN
easat-5713	160	20	category	category	NOUN
easat-5713	160	21	.	.	PUNCT
easat-5713	161	1	the	the	DET
easat-5713	161	2	model	model	NOUN
easat-5713	161	3	trained	train	VERB
easat-5713	161	4	on	on	ADP
easat-5713	161	5	the	the	DET
easat-5713	161	6	original	original	ADJ
easat-5713	161	7	fer2013	fer2013	ADJ
easat-5713	161	8	dataset	dataset	NOUN
easat-5713	161	9	performed	perform	VERB
easat-5713	161	10	well	well	ADV
easat-5713	161	11	in	in	ADP
easat-5713	161	12	classifying	classify	VERB
easat-5713	161	13	“	"	PUNCT
easat-5713	161	14	fear	fear	NOUN
easat-5713	161	15	”	"	PUNCT
easat-5713	161	16	with	with	ADP
easat-5713	161	17	777	777	NUM
easat-5713	161	18	true	true	ADJ
easat-5713	161	19	positives	positive	NOUN
easat-5713	161	20	but	but	CCONJ
easat-5713	161	21	struggled	struggle	VERB
easat-5713	161	22	with	with	ADP
easat-5713	161	23	“	"	PUNCT
easat-5713	161	24	disgust	disgust	NOUN
easat-5713	161	25	,	,	PUNCT
easat-5713	161	26	”	"	PUNCT
easat-5713	161	27	which	which	PRON
easat-5713	161	28	had	have	VERB
easat-5713	161	29	a	a	DET
easat-5713	161	30	high	high	ADJ
easat-5713	161	31	number	number	NOUN
easat-5713	161	32	of	of	ADP
easat-5713	161	33	misclassifications	misclassification	NOUN
easat-5713	161	34	.	.	PUNCT
easat-5713	162	1	the	the	DET
easat-5713	162	2	augmented	augment	VERB
easat-5713	162	3	dataset	dataset	NOUN
easat-5713	162	4	continued	continue	VERB
easat-5713	162	5	to	to	PART
easat-5713	162	6	classify	classify	VERB
easat-5713	162	7	“	"	PUNCT
easat-5713	162	8	fear	fear	NOUN
easat-5713	162	9	”	"	PUNCT
easat-5713	162	10	effectively	effectively	ADV
easat-5713	162	11	,	,	PUNCT
easat-5713	162	12	with	with	ADP
easat-5713	162	13	763	763	NUM
easat-5713	162	14	true	true	ADJ
easat-5713	162	15	positives	positive	NOUN
easat-5713	162	16	,	,	PUNCT
easat-5713	162	17	but	but	CCONJ
easat-5713	162	18	performance	performance	NOUN
easat-5713	162	19	variations	variation	NOUN
easat-5713	162	20	were	be	AUX
easat-5713	162	21	observed	observe	VERB
easat-5713	162	22	for	for	ADP
easat-5713	162	23	other	other	ADJ
easat-5713	162	24	emotions	emotion	NOUN
easat-5713	162	25	,	,	PUNCT
easat-5713	162	26	particularly	particularly	ADV
easat-5713	162	27	“	"	PUNCT
easat-5713	162	28	happy	happy	ADJ
easat-5713	162	29	”	"	PUNCT
easat-5713	162	30	and	and	CCONJ
easat-5713	162	31	“	"	PUNCT
easat-5713	162	32	disgust	disgust	NOUN
easat-5713	162	33	.	.	PUNCT
easat-5713	162	34	”	"	PUNCT
easat-5713	163	1	this	this	PRON
easat-5713	163	2	indicates	indicate	VERB
easat-5713	163	3	that	that	SCONJ
easat-5713	163	4	while	while	SCONJ
easat-5713	163	5	synthetic	synthetic	ADJ
easat-5713	163	6	images	image	NOUN
easat-5713	163	7	-	-	PUNCT
easat-5713	163	8	maintained	maintain	VERB
easat-5713	163	9	performance	performance	NOUN
easat-5713	163	10	consistency	consistency	NOUN
easat-5713	163	11	for	for	ADP
easat-5713	163	12	certain	certain	ADJ
easat-5713	163	13	emotion	emotion	NOUN
easat-5713	163	14	classes	class	NOUN
easat-5713	163	15	,	,	PUNCT
easat-5713	163	16	they	they	PRON
easat-5713	163	17	did	do	AUX
easat-5713	163	18	not	not	PART
easat-5713	163	19	universally	universally	ADV
easat-5713	163	20	improve	improve	VERB
easat-5713	163	21	recognition	recognition	NOUN
easat-5713	163	22	for	for	ADP
easat-5713	163	23	complex	complex	ADJ
easat-5713	163	24	facial	facial	ADJ
easat-5713	163	25	expressions	expression	NOUN
easat-5713	163	26	that	that	PRON
easat-5713	163	27	require	require	VERB
easat-5713	163	28	subtle	subtle	ADJ
easat-5713	163	29	feature	feature	NOUN
easat-5713	163	30	differentiation	differentiation	NOUN
easat-5713	163	31	.	.	PUNCT
easat-5713	164	1	another	another	DET
easat-5713	164	2	contributing	contribute	VERB
easat-5713	164	3	factor	factor	NOUN
easat-5713	164	4	could	could	AUX
easat-5713	164	5	be	be	AUX
easat-5713	164	6	the	the	DET
easat-5713	164	7	limited	limited	ADJ
easat-5713	164	8	diversity	diversity	NOUN
easat-5713	164	9	in	in	ADP
easat-5713	164	10	synthetic	synthetic	ADJ
easat-5713	164	11	data	datum	NOUN
easat-5713	164	12	despite	despite	SCONJ
easat-5713	164	13	controlled	control	VERB
easat-5713	164	14	variations	variation	NOUN
easat-5713	164	15	in	in	ADP
easat-5713	164	16	demographics	demographic	NOUN
easat-5713	164	17	and	and	CCONJ
easat-5713	164	18	environmental	environmental	ADJ
easat-5713	164	19	conditions	condition	NOUN
easat-5713	164	20	.	.	PUNCT
easat-5713	165	1	while	while	SCONJ
easat-5713	165	2	synthetic	synthetic	ADJ
easat-5713	165	3	data	data	NOUN
easat-5713	165	4	generation	generation	NOUN
easat-5713	165	5	aimed	aim	VERB
easat-5713	165	6	to	to	PART
easat-5713	165	7	enhance	enhance	VERB
easat-5713	165	8	class	class	NOUN
easat-5713	165	9	balance	balance	NOUN
easat-5713	165	10	and	and	CCONJ
easat-5713	165	11	representation	representation	NOUN
easat-5713	165	12	,	,	PUNCT
easat-5713	165	13	it	it	PRON
easat-5713	165	14	may	may	AUX
easat-5713	165	15	not	not	PART
easat-5713	165	16	have	have	AUX
easat-5713	165	17	sufficiently	sufficiently	ADV
easat-5713	165	18	accounted	account	VERB
easat-5713	165	19	for	for	ADP
easat-5713	165	20	expression	expression	NOUN
easat-5713	165	21	dynamics	dynamic	NOUN
easat-5713	165	22	,	,	PUNCT
easat-5713	165	23	microexpressions	microexpression	NOUN
easat-5713	165	24	,	,	PUNCT
easat-5713	165	25	and	and	CCONJ
easat-5713	165	26	natural	natural	ADJ
easat-5713	165	27	distortions	distortion	NOUN
easat-5713	165	28	that	that	PRON
easat-5713	165	29	occur	occur	VERB
easat-5713	165	30	in	in	ADP
easat-5713	165	31	real	real	ADJ
easat-5713	165	32	facial	facial	ADJ
easat-5713	165	33	movements	movement	NOUN
easat-5713	165	34	.	.	PUNCT
easat-5713	166	1	this	this	DET
easat-5713	166	2	limitation	limitation	NOUN
easat-5713	166	3	could	could	AUX
easat-5713	166	4	explain	explain	VERB
easat-5713	166	5	the	the	DET
easat-5713	166	6	marginal	marginal	ADJ
easat-5713	166	7	decrease	decrease	NOUN
easat-5713	166	8	in	in	ADP
easat-5713	166	9	accuracy	accuracy	NOUN
easat-5713	166	10	when	when	SCONJ
easat-5713	166	11	incorporating	incorporate	VERB
easat-5713	166	12	synthetic	synthetic	ADJ
easat-5713	166	13	images	image	NOUN
easat-5713	166	14	[	[	X
easat-5713	166	15	23	23	NUM
easat-5713	166	16	]	]	PUNCT
easat-5713	166	17	.	.	PUNCT
easat-5713	167	1	1864	1864	NUM
easat-5713	167	2	edelweiss	edelweiss	PROPN
easat-5713	167	3	applied	apply	VERB
easat-5713	167	4	science	science	NOUN
easat-5713	167	5	and	and	CCONJ
easat-5713	167	6	technology	technology	NOUN
easat-5713	167	7	issn	issn	PROPN
easat-5713	167	8	:	:	PUNCT
easat-5713	167	9	2576	2576	NUM
easat-5713	167	10	-	-	SYM
easat-5713	167	11	8484	8484	NUM
easat-5713	167	12	vol	vol	NOUN
easat-5713	167	13	.	.	PROPN
easat-5713	168	1	9	9	NUM
easat-5713	168	2	,	,	PUNCT
easat-5713	168	3	no	no	INTJ
easat-5713	168	4	.	.	NOUN
easat-5713	169	1	3	3	NUM
easat-5713	169	2	:	:	PUNCT
easat-5713	169	3	1855	1855	NUM
easat-5713	169	4	-	-	SYM
easat-5713	169	5	1869	1869	NUM
easat-5713	169	6	,	,	PUNCT
easat-5713	169	7	2025	2025	NUM
easat-5713	169	8	doi	doi	NOUN
easat-5713	169	9	:	:	PUNCT
easat-5713	169	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	169	11	©	©	ADP
easat-5713	169	12	2025	2025	NUM
easat-5713	169	13	by	by	ADP
easat-5713	169	14	the	the	DET
easat-5713	169	15	authors	author	NOUN
easat-5713	169	16	;	;	PUNCT
easat-5713	169	17	licensee	licensee	PROPN
easat-5713	169	18	learning	learn	VERB
easat-5713	169	19	gate	gate	PROPN
easat-5713	169	20	figure	figure	NOUN
easat-5713	169	21	5	5	NUM
easat-5713	169	22	.	.	PUNCT
easat-5713	169	23	confusion	confusion	NOUN
easat-5713	169	24	matrix	matrix	NOUN
easat-5713	169	25	of	of	ADP
easat-5713	169	26	vgg-19	vgg-19	PROPN
easat-5713	169	27	model	model	NOUN
easat-5713	169	28	on	on	ADP
easat-5713	169	29	our	our	PRON
easat-5713	169	30	ai	ai	ADJ
easat-5713	169	31	augmented	augment	VERB
easat-5713	169	32	fer-2013	fer-2013	NOUN
easat-5713	169	33	dataset	dataset	NOUN
easat-5713	169	34	.	.	PUNCT
easat-5713	170	1	this	this	DET
easat-5713	170	2	outcome	outcome	NOUN
easat-5713	170	3	suggests	suggest	VERB
easat-5713	170	4	that	that	SCONJ
easat-5713	170	5	the	the	DET
easat-5713	170	6	effectiveness	effectiveness	NOUN
easat-5713	170	7	of	of	ADP
easat-5713	170	8	data	datum	NOUN
easat-5713	170	9	augmentation	augmentation	NOUN
easat-5713	170	10	using	use	VERB
easat-5713	170	11	synthetic	synthetic	ADJ
easat-5713	170	12	images	image	NOUN
easat-5713	170	13	is	be	AUX
easat-5713	170	14	highly	highly	ADV
easat-5713	170	15	dependent	dependent	ADJ
easat-5713	170	16	on	on	ADP
easat-5713	170	17	the	the	DET
easat-5713	170	18	quality	quality	NOUN
easat-5713	170	19	and	and	CCONJ
easat-5713	170	20	realism	realism	NOUN
easat-5713	170	21	of	of	ADP
easat-5713	170	22	the	the	DET
easat-5713	170	23	generated	generate	VERB
easat-5713	170	24	images	image	NOUN
easat-5713	170	25	.	.	PUNCT
easat-5713	171	1	while	while	SCONJ
easat-5713	171	2	the	the	DET
easat-5713	171	3	synthetic	synthetic	ADJ
easat-5713	171	4	images	image	NOUN
easat-5713	171	5	were	be	AUX
easat-5713	171	6	carefully	carefully	ADV
easat-5713	171	7	designed	design	VERB
easat-5713	171	8	using	use	VERB
easat-5713	171	9	prompts	prompt	NOUN
easat-5713	171	10	to	to	PART
easat-5713	171	11	capture	capture	VERB
easat-5713	171	12	diverse	diverse	ADJ
easat-5713	171	13	demographic	demographic	ADJ
easat-5713	171	14	and	and	CCONJ
easat-5713	171	15	environmental	environmental	ADJ
easat-5713	171	16	conditions	condition	NOUN
easat-5713	171	17	,	,	PUNCT
easat-5713	171	18	they	they	PRON
easat-5713	171	19	may	may	AUX
easat-5713	171	20	not	not	PART
easat-5713	171	21	have	have	AUX
easat-5713	171	22	fully	fully	ADV
easat-5713	171	23	replicated	replicate	VERB
easat-5713	171	24	the	the	DET
easat-5713	171	25	subtle	subtle	ADJ
easat-5713	171	26	variations	variation	NOUN
easat-5713	171	27	in	in	ADP
easat-5713	171	28	real	real	ADJ
easat-5713	171	29	-	-	PUNCT
easat-5713	171	30	world	world	NOUN
easat-5713	171	31	facial	facial	ADJ
easat-5713	171	32	expressions	expression	NOUN
easat-5713	171	33	.	.	PUNCT
easat-5713	172	1	instead	instead	ADV
easat-5713	172	2	of	of	ADP
easat-5713	172	3	providing	provide	VERB
easat-5713	172	4	additional	additional	ADJ
easat-5713	172	5	useful	useful	ADJ
easat-5713	172	6	training	training	NOUN
easat-5713	172	7	data	datum	NOUN
easat-5713	172	8	,	,	PUNCT
easat-5713	172	9	these	these	DET
easat-5713	172	10	images	image	NOUN
easat-5713	172	11	might	might	AUX
easat-5713	172	12	have	have	AUX
easat-5713	172	13	introduced	introduce	VERB
easat-5713	172	14	noise	noise	NOUN
easat-5713	172	15	,	,	PUNCT
easat-5713	172	16	making	make	VERB
easat-5713	172	17	it	it	PRON
easat-5713	172	18	more	more	ADV
easat-5713	172	19	difficult	difficult	ADJ
easat-5713	172	20	for	for	SCONJ
easat-5713	172	21	the	the	DET
easat-5713	172	22	model	model	NOUN
easat-5713	172	23	to	to	PART
easat-5713	172	24	extract	extract	VERB
easat-5713	172	25	meaningful	meaningful	ADJ
easat-5713	172	26	features	feature	NOUN
easat-5713	172	27	.	.	PUNCT
easat-5713	173	1	prior	prior	ADJ
easat-5713	173	2	research	research	NOUN
easat-5713	173	3	has	have	AUX
easat-5713	173	4	demonstrated	demonstrate	VERB
easat-5713	173	5	that	that	SCONJ
easat-5713	173	6	gan	gan	NOUN
easat-5713	173	7	-	-	PUNCT
easat-5713	173	8	generated	generate	VERB
easat-5713	173	9	images	image	NOUN
easat-5713	173	10	can	can	AUX
easat-5713	173	11	enhance	enhance	VERB
easat-5713	173	12	the	the	DET
easat-5713	173	13	model	model	NOUN
easat-5713	173	14	robustness	robustness	NOUN
easat-5713	173	15	and	and	CCONJ
easat-5713	173	16	accuracy	accuracy	NOUN
easat-5713	173	17	,	,	PUNCT
easat-5713	173	18	particularly	particularly	ADV
easat-5713	173	19	in	in	ADP
easat-5713	173	20	addressing	address	VERB
easat-5713	173	21	class	class	NOUN
easat-5713	173	22	imbalances	imbalance	NOUN
easat-5713	173	23	.	.	PUNCT
easat-5713	174	1	however	however	ADV
easat-5713	174	2	,	,	PUNCT
easat-5713	174	3	these	these	DET
easat-5713	174	4	studies	study	NOUN
easat-5713	174	5	often	often	ADV
easat-5713	174	6	utilize	utilize	VERB
easat-5713	174	7	different	different	ADJ
easat-5713	174	8	generative	generative	ADJ
easat-5713	174	9	architectures	architecture	NOUN
easat-5713	174	10	and	and	CCONJ
easat-5713	174	11	datasets	dataset	NOUN
easat-5713	174	12	,	,	PUNCT
easat-5713	174	13	which	which	PRON
easat-5713	174	14	could	could	AUX
easat-5713	174	15	explain	explain	VERB
easat-5713	174	16	the	the	DET
easat-5713	174	17	discrepancy	discrepancy	NOUN
easat-5713	174	18	in	in	ADP
easat-5713	174	19	results	result	NOUN
easat-5713	174	20	.	.	PUNCT
easat-5713	175	1	gan	gin	VERB
easat-5713	175	2	-	-	PUNCT
easat-5713	175	3	based	base	VERB
easat-5713	175	4	approaches	approach	NOUN
easat-5713	175	5	have	have	AUX
easat-5713	175	6	been	be	AUX
easat-5713	175	7	shown	show	VERB
easat-5713	175	8	to	to	PART
easat-5713	175	9	produce	produce	VERB
easat-5713	175	10	highly	highly	ADV
easat-5713	175	11	realistic	realistic	ADJ
easat-5713	175	12	images	image	NOUN
easat-5713	175	13	,	,	PUNCT
easat-5713	175	14	whereas	whereas	SCONJ
easat-5713	175	15	the	the	DET
easat-5713	175	16	juggernaut	juggernaut	NOUN
easat-5713	175	17	xl	xl	PROPN
easat-5713	175	18	model	model	PROPN
easat-5713	175	19	,	,	PUNCT
easat-5713	175	20	while	while	SCONJ
easat-5713	175	21	advanced	advanced	ADJ
easat-5713	175	22	,	,	PUNCT
easat-5713	175	23	may	may	AUX
easat-5713	175	24	require	require	VERB
easat-5713	175	25	further	further	ADJ
easat-5713	175	26	fine	fine	ADV
easat-5713	175	27	-	-	PUNCT
easat-5713	175	28	tuning	tuning	NOUN
easat-5713	175	29	to	to	PART
easat-5713	175	30	generate	generate	VERB
easat-5713	175	31	images	image	NOUN
easat-5713	175	32	that	that	PRON
easat-5713	175	33	effectively	effectively	ADV
easat-5713	175	34	complement	complement	VERB
easat-5713	175	35	realworld	realworld	PROPN
easat-5713	175	36	datasets	datasets	PROPN
easat-5713	175	37	.	.	PUNCT
easat-5713	176	1	figure	figure	NOUN
easat-5713	176	2	6	6	NUM
easat-5713	176	3	illustrates	illustrate	VERB
easat-5713	176	4	several	several	ADJ
easat-5713	176	5	examples	example	NOUN
easat-5713	176	6	of	of	ADP
easat-5713	176	7	high	high	ADJ
easat-5713	176	8	-	-	PUNCT
easat-5713	176	9	quality	quality	NOUN
easat-5713	176	10	synthetic	synthetic	ADJ
easat-5713	176	11	images	image	NOUN
easat-5713	176	12	that	that	PRON
easat-5713	176	13	may	may	AUX
easat-5713	176	14	have	have	AUX
easat-5713	176	15	contributed	contribute	VERB
easat-5713	176	16	to	to	ADP
easat-5713	176	17	the	the	DET
easat-5713	176	18	observed	observe	VERB
easat-5713	176	19	performance	performance	NOUN
easat-5713	176	20	improvement	improvement	NOUN
easat-5713	176	21	.	.	PUNCT
easat-5713	177	1	1865	1865	NUM
easat-5713	177	2	edelweiss	edelweiss	PROPN
easat-5713	177	3	applied	apply	VERB
easat-5713	177	4	science	science	NOUN
easat-5713	177	5	and	and	CCONJ
easat-5713	177	6	technology	technology	NOUN
easat-5713	177	7	issn	issn	PROPN
easat-5713	177	8	:	:	PUNCT
easat-5713	177	9	2576	2576	NUM
easat-5713	177	10	-	-	SYM
easat-5713	177	11	8484	8484	NUM
easat-5713	177	12	vol	vol	NOUN
easat-5713	177	13	.	.	PROPN
easat-5713	178	1	9	9	NUM
easat-5713	178	2	,	,	PUNCT
easat-5713	178	3	no	no	INTJ
easat-5713	178	4	.	.	NOUN
easat-5713	179	1	3	3	NUM
easat-5713	179	2	:	:	PUNCT
easat-5713	179	3	1855	1855	NUM
easat-5713	179	4	-	-	SYM
easat-5713	179	5	1869	1869	NUM
easat-5713	179	6	,	,	PUNCT
easat-5713	179	7	2025	2025	NUM
easat-5713	179	8	doi	doi	NOUN
easat-5713	179	9	:	:	PUNCT
easat-5713	179	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	179	11	©	©	ADP
easat-5713	179	12	2025	2025	NUM
easat-5713	179	13	by	by	ADP
easat-5713	179	14	the	the	DET
easat-5713	179	15	authors	author	NOUN
easat-5713	179	16	;	;	PUNCT
easat-5713	180	1	licensee	licensee	PROPN
easat-5713	180	2	learning	learn	VERB
easat-5713	180	3	gate	gate	PROPN
easat-5713	180	4	figure	figure	NOUN
easat-5713	180	5	6	6	NUM
easat-5713	180	6	.	.	PUNCT
easat-5713	180	7	improved	improve	VERB
easat-5713	180	8	ai	ai	PROPN
easat-5713	180	9	-	-	PUNCT
easat-5713	180	10	generated	generate	VERB
easat-5713	180	11	images	image	NOUN
easat-5713	180	12	using	use	VERB
easat-5713	180	13	juggernaut	juggernaut	PROPN
easat-5713	180	14	xl	xl	PROPN
easat-5713	180	15	model	model	PROPN
easat-5713	180	16	.	.	PUNCT
easat-5713	181	1	figure	figure	VERB
easat-5713	181	2	7	7	NUM
easat-5713	181	3	presents	present	VERB
easat-5713	181	4	the	the	DET
easat-5713	181	5	application	application	NOUN
easat-5713	181	6	used	use	VERB
easat-5713	181	7	for	for	ADP
easat-5713	181	8	testing	test	VERB
easat-5713	181	9	the	the	DET
easat-5713	181	10	models	model	NOUN
easat-5713	181	11	,	,	PUNCT
easat-5713	181	12	allowing	allow	VERB
easat-5713	181	13	real	real	ADJ
easat-5713	181	14	-	-	PUNCT
easat-5713	181	15	time	time	NOUN
easat-5713	181	16	evaluation	evaluation	NOUN
easat-5713	181	17	of	of	ADP
easat-5713	181	18	music	music	NOUN
easat-5713	181	19	emotion	emotion	NOUN
easat-5713	181	20	classification	classification	NOUN
easat-5713	181	21	and	and	CCONJ
easat-5713	181	22	facial	facial	ADJ
easat-5713	181	23	emotion	emotion	NOUN
easat-5713	181	24	recognition	recognition	NOUN
easat-5713	181	25	.	.	PUNCT
easat-5713	182	1	the	the	DET
easat-5713	182	2	interface	interface	NOUN
easat-5713	182	3	facilitates	facilitate	VERB
easat-5713	182	4	emotion	emotion	NOUN
easat-5713	182	5	mapping	mapping	NOUN
easat-5713	182	6	and	and	CCONJ
easat-5713	182	7	dynamic	dynamic	ADJ
easat-5713	182	8	playlist	playlist	NOUN
easat-5713	182	9	generation	generation	NOUN
easat-5713	182	10	based	base	VERB
easat-5713	182	11	on	on	ADP
easat-5713	182	12	detected	detect	VERB
easat-5713	182	13	emotional	emotional	ADJ
easat-5713	182	14	states	state	NOUN
easat-5713	182	15	.	.	PUNCT
easat-5713	183	1	the	the	DET
easat-5713	183	2	adaptive	adaptive	ADJ
easat-5713	183	3	music	music	NOUN
easat-5713	183	4	player	player	NOUN
easat-5713	183	5	further	far	ADV
easat-5713	183	6	contributed	contribute	VERB
easat-5713	183	7	to	to	ADP
easat-5713	183	8	user	user	NOUN
easat-5713	183	9	experience	experience	NOUN
easat-5713	183	10	by	by	ADP
easat-5713	183	11	dynamically	dynamically	ADV
easat-5713	183	12	updating	update	VERB
easat-5713	183	13	playlists	playlist	NOUN
easat-5713	183	14	in	in	ADP
easat-5713	183	15	real	real	ADJ
easat-5713	183	16	-	-	PUNCT
easat-5713	183	17	time	time	NOUN
easat-5713	183	18	,	,	PUNCT
easat-5713	183	19	learning	learn	VERB
easat-5713	183	20	from	from	ADP
easat-5713	183	21	user	user	NOUN
easat-5713	183	22	interactions	interaction	NOUN
easat-5713	183	23	,	,	PUNCT
easat-5713	183	24	and	and	CCONJ
easat-5713	183	25	refining	refine	VERB
easat-5713	183	26	its	its	PRON
easat-5713	183	27	recommendations	recommendation	NOUN
easat-5713	183	28	.	.	PUNCT
easat-5713	184	1	the	the	DET
easat-5713	184	2	system	system	NOUN
easat-5713	184	3	’s	’s	PART
easat-5713	184	4	ability	ability	NOUN
easat-5713	184	5	to	to	PART
easat-5713	184	6	respond	respond	VERB
easat-5713	184	7	to	to	ADP
easat-5713	184	8	real	real	ADJ
easat-5713	184	9	-	-	PUNCT
easat-5713	184	10	time	time	NOUN
easat-5713	184	11	emotional	emotional	ADJ
easat-5713	184	12	changes	change	NOUN
easat-5713	184	13	with	with	ADP
easat-5713	184	14	a	a	DET
easat-5713	184	15	response	response	NOUN
easat-5713	184	16	time	time	NOUN
easat-5713	184	17	of	of	ADP
easat-5713	184	18	under	under	ADP
easat-5713	184	19	2	2	NUM
easat-5713	184	20	seconds	second	NOUN
easat-5713	184	21	underscores	underscore	VERB
easat-5713	184	22	its	its	PRON
easat-5713	184	23	practicality	practicality	NOUN
easat-5713	184	24	for	for	ADP
easat-5713	184	25	real	real	ADJ
easat-5713	184	26	-	-	PUNCT
easat-5713	184	27	world	world	NOUN
easat-5713	184	28	applications	application	NOUN
easat-5713	184	29	.	.	PUNCT
easat-5713	185	1	optimizing	optimize	VERB
easat-5713	185	2	training	training	NOUN
easat-5713	185	3	strategies	strategy	NOUN
easat-5713	185	4	,	,	PUNCT
easat-5713	185	5	including	include	VERB
easat-5713	185	6	longer	long	ADJ
easat-5713	185	7	training	training	NOUN
easat-5713	185	8	periods	period	NOUN
easat-5713	185	9	,	,	PUNCT
easat-5713	185	10	fine	fine	ADV
easat-5713	185	11	-	-	PUNCT
easat-5713	185	12	tuning	tune	VERB
easat-5713	185	13	approaches	approach	NOUN
easat-5713	185	14	,	,	PUNCT
easat-5713	185	15	and	and	CCONJ
easat-5713	185	16	selective	selective	ADJ
easat-5713	185	17	filtering	filtering	NOUN
easat-5713	185	18	of	of	ADP
easat-5713	185	19	synthetic	synthetic	ADJ
easat-5713	185	20	images	image	NOUN
easat-5713	185	21	based	base	VERB
easat-5713	185	22	on	on	ADP
easat-5713	185	23	quality	quality	NOUN
easat-5713	185	24	assessment	assessment	NOUN
easat-5713	185	25	,	,	PUNCT
easat-5713	185	26	could	could	AUX
easat-5713	185	27	improve	improve	VERB
easat-5713	185	28	the	the	DET
easat-5713	185	29	system	system	NOUN
easat-5713	185	30	’s	’s	PART
easat-5713	185	31	design	design	NOUN
easat-5713	185	32	.	.	PUNCT
easat-5713	186	1	1866	1866	PROPN
easat-5713	187	1	edelweiss	edelweiss	PROPN
easat-5713	187	2	applied	apply	VERB
easat-5713	187	3	science	science	NOUN
easat-5713	187	4	and	and	CCONJ
easat-5713	187	5	technology	technology	NOUN
easat-5713	187	6	issn	issn	PROPN
easat-5713	187	7	:	:	PUNCT
easat-5713	187	8	2576	2576	NUM
easat-5713	187	9	-	-	SYM
easat-5713	187	10	8484	8484	NUM
easat-5713	187	11	vol	vol	NOUN
easat-5713	187	12	.	.	PROPN
easat-5713	188	1	9	9	NUM
easat-5713	188	2	,	,	PUNCT
easat-5713	188	3	no	no	INTJ
easat-5713	188	4	.	.	NOUN
easat-5713	189	1	3	3	NUM
easat-5713	189	2	:	:	PUNCT
easat-5713	189	3	1855	1855	NUM
easat-5713	189	4	-	-	SYM
easat-5713	189	5	1869	1869	NUM
easat-5713	189	6	,	,	PUNCT
easat-5713	189	7	2025	2025	NUM
easat-5713	189	8	doi	doi	NOUN
easat-5713	189	9	:	:	PUNCT
easat-5713	189	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	189	11	©	©	ADP
easat-5713	189	12	2025	2025	NUM
easat-5713	189	13	by	by	ADP
easat-5713	189	14	the	the	DET
easat-5713	189	15	authors	author	NOUN
easat-5713	189	16	;	;	PUNCT
easat-5713	189	17	licensee	licensee	PROPN
easat-5713	189	18	learning	learn	VERB
easat-5713	189	19	gate	gate	NOUN
easat-5713	189	20	figure	figure	NOUN
easat-5713	189	21	7	7	NUM
easat-5713	189	22	.	.	PUNCT
easat-5713	190	1	the	the	DET
easat-5713	190	2	application	application	NOUN
easat-5713	190	3	developed	develop	VERB
easat-5713	190	4	for	for	ADP
easat-5713	190	5	testing	test	VERB
easat-5713	190	6	the	the	DET
easat-5713	190	7	models	model	NOUN
easat-5713	190	8	.	.	PUNCT
easat-5713	191	1	4	4	X
easat-5713	191	2	.	.	X
easat-5713	191	3	discussion	discussion	NOUN
easat-5713	191	4	this	this	DET
easat-5713	191	5	study	study	NOUN
easat-5713	191	6	examined	examine	VERB
easat-5713	191	7	the	the	DET
easat-5713	191	8	impact	impact	NOUN
easat-5713	191	9	of	of	ADP
easat-5713	191	10	ai	ai	VERB
easat-5713	191	11	-	-	PUNCT
easat-5713	191	12	generated	generate	VERB
easat-5713	191	13	synthetic	synthetic	ADJ
easat-5713	191	14	images	image	NOUN
easat-5713	191	15	on	on	ADP
easat-5713	191	16	fer	fer	PROPN
easat-5713	191	17	model	model	NOUN
easat-5713	191	18	performance	performance	NOUN
easat-5713	191	19	,	,	PUNCT
easat-5713	191	20	revealing	reveal	VERB
easat-5713	191	21	that	that	SCONJ
easat-5713	191	22	while	while	SCONJ
easat-5713	191	23	synthetic	synthetic	ADJ
easat-5713	191	24	data	data	NOUN
easat-5713	191	25	-	-	PUNCT
easat-5713	191	26	maintained	maintain	VERB
easat-5713	191	27	performance	performance	NOUN
easat-5713	191	28	stability	stability	NOUN
easat-5713	191	29	for	for	ADP
easat-5713	191	30	certain	certain	ADJ
easat-5713	191	31	emotion	emotion	NOUN
easat-5713	191	32	categories	category	NOUN
easat-5713	191	33	,	,	PUNCT
easat-5713	191	34	it	it	PRON
easat-5713	191	35	did	do	AUX
easat-5713	191	36	not	not	PART
easat-5713	191	37	significantly	significantly	ADV
easat-5713	191	38	enhance	enhance	VERB
easat-5713	191	39	accuracy	accuracy	NOUN
easat-5713	191	40	.	.	PUNCT
easat-5713	192	1	the	the	DET
easat-5713	192	2	findings	finding	NOUN
easat-5713	192	3	emphasize	emphasize	VERB
easat-5713	192	4	the	the	DET
easat-5713	192	5	importance	importance	NOUN
easat-5713	192	6	of	of	ADP
easat-5713	192	7	synthetic	synthetic	ADJ
easat-5713	192	8	data	datum	NOUN
easat-5713	192	9	realism	realism	NOUN
easat-5713	192	10	and	and	CCONJ
easat-5713	192	11	suggest	suggest	VERB
easat-5713	192	12	that	that	SCONJ
easat-5713	192	13	further	further	ADJ
easat-5713	192	14	refinements	refinement	NOUN
easat-5713	192	15	in	in	ADP
easat-5713	192	16	generation	generation	NOUN
easat-5713	192	17	techniques	technique	NOUN
easat-5713	192	18	and	and	CCONJ
easat-5713	192	19	training	training	NOUN
easat-5713	192	20	methodologies	methodology	NOUN
easat-5713	192	21	are	be	AUX
easat-5713	192	22	necessary	necessary	ADJ
easat-5713	192	23	for	for	ADP
easat-5713	192	24	effective	effective	ADJ
easat-5713	192	25	integration	integration	NOUN
easat-5713	192	26	.	.	PUNCT
easat-5713	193	1	future	future	ADJ
easat-5713	193	2	research	research	NOUN
easat-5713	193	3	should	should	AUX
easat-5713	193	4	focus	focus	VERB
easat-5713	193	5	on	on	ADP
easat-5713	193	6	improving	improve	VERB
easat-5713	193	7	synthetic	synthetic	ADJ
easat-5713	193	8	image	image	NOUN
easat-5713	193	9	quality	quality	NOUN
easat-5713	193	10	,	,	PUNCT
easat-5713	193	11	increasing	increase	VERB
easat-5713	193	12	training	training	NOUN
easat-5713	193	13	epochs	epoch	NOUN
easat-5713	193	14	,	,	PUNCT
easat-5713	193	15	and	and	CCONJ
easat-5713	193	16	exploring	explore	VERB
easat-5713	193	17	alternative	alternative	ADJ
easat-5713	193	18	generative	generative	ADJ
easat-5713	193	19	models	model	NOUN
easat-5713	193	20	to	to	PART
easat-5713	193	21	maximize	maximize	VERB
easat-5713	193	22	the	the	DET
easat-5713	193	23	benefits	benefit	NOUN
easat-5713	193	24	of	of	ADP
easat-5713	193	25	synthetic	synthetic	ADJ
easat-5713	193	26	data	datum	NOUN
easat-5713	193	27	augmentation	augmentation	NOUN
easat-5713	193	28	in	in	ADP
easat-5713	193	29	fer	fer	PROPN
easat-5713	193	30	.	.	PUNCT
easat-5713	194	1	this	this	DET
easat-5713	194	2	research	research	NOUN
easat-5713	194	3	aimed	aim	VERB
easat-5713	194	4	to	to	PART
easat-5713	194	5	improve	improve	VERB
easat-5713	194	6	the	the	DET
easat-5713	194	7	models	model	NOUN
easat-5713	194	8	by	by	ADP
easat-5713	194	9	incorporating	incorporate	VERB
easat-5713	194	10	ai	ai	NOUN
easat-5713	194	11	-	-	PUNCT
easat-5713	194	12	generated	generate	VERB
easat-5713	194	13	synthetic	synthetic	ADJ
easat-5713	194	14	data	datum	NOUN
easat-5713	194	15	into	into	ADP
easat-5713	194	16	the	the	DET
easat-5713	194	17	training	training	NOUN
easat-5713	194	18	process	process	NOUN
easat-5713	194	19	.	.	PUNCT
easat-5713	195	1	using	use	VERB
easat-5713	195	2	the	the	DET
easat-5713	195	3	juggernaut	juggernaut	PROPN
easat-5713	195	4	xl	xl	PROPN
easat-5713	195	5	model	model	PROPN
easat-5713	195	6	,	,	PUNCT
easat-5713	195	7	we	we	PRON
easat-5713	195	8	generated	generate	VERB
easat-5713	195	9	300	300	NUM
easat-5713	195	10	synthetic	synthetic	ADJ
easat-5713	195	11	images	image	NOUN
easat-5713	195	12	for	for	ADP
easat-5713	195	13	each	each	DET
easat-5713	195	14	emotion	emotion	NOUN
easat-5713	195	15	category	category	NOUN
easat-5713	195	16	in	in	ADP
easat-5713	195	17	the	the	DET
easat-5713	195	18	fer2013	fer2013	ADJ
easat-5713	195	19	dataset	dataset	NOUN
easat-5713	195	20	and	and	CCONJ
easat-5713	195	21	integrated	integrate	VERB
easat-5713	195	22	them	they	PRON
easat-5713	195	23	into	into	ADP
easat-5713	195	24	the	the	DET
easat-5713	195	25	training	training	NOUN
easat-5713	195	26	pipeline	pipeline	NOUN
easat-5713	195	27	of	of	ADP
easat-5713	195	28	a	a	DET
easat-5713	195	29	vgg-19	vgg-19	ADV
easat-5713	195	30	-	-	PUNCT
easat-5713	195	31	based	base	VERB
easat-5713	195	32	fer	fer	PROPN
easat-5713	195	33	model	model	PROPN
easat-5713	195	34	.	.	PUNCT
easat-5713	196	1	the	the	DET
easat-5713	196	2	primary	primary	ADJ
easat-5713	196	3	objective	objective	NOUN
easat-5713	196	4	was	be	AUX
easat-5713	196	5	to	to	PART
easat-5713	196	6	determine	determine	VERB
easat-5713	196	7	whether	whether	SCONJ
easat-5713	196	8	introducing	introduce	VERB
easat-5713	196	9	synthetic	synthetic	ADJ
easat-5713	196	10	images	image	NOUN
easat-5713	196	11	could	could	AUX
easat-5713	196	12	enhance	enhance	VERB
easat-5713	196	13	key	key	ADJ
easat-5713	196	14	performance	performance	NOUN
easat-5713	196	15	metrics	metric	NOUN
easat-5713	196	16	such	such	ADJ
easat-5713	196	17	as	as	ADP
easat-5713	196	18	accuracy	accuracy	NOUN
easat-5713	196	19	,	,	PUNCT
easat-5713	196	20	precision	precision	NOUN
easat-5713	196	21	,	,	PUNCT
easat-5713	196	22	recall	recall	NOUN
easat-5713	196	23	,	,	PUNCT
easat-5713	196	24	and	and	CCONJ
easat-5713	196	25	f1	f1	NOUN
easat-5713	196	26	-	-	PUNCT
easat-5713	196	27	score	score	NOUN
easat-5713	196	28	.	.	PUNCT
easat-5713	197	1	given	give	VERB
easat-5713	197	2	the	the	DET
easat-5713	197	3	challenges	challenge	NOUN
easat-5713	197	4	posed	pose	VERB
easat-5713	197	5	by	by	ADP
easat-5713	197	6	noisy	noisy	ADJ
easat-5713	197	7	images	image	NOUN
easat-5713	197	8	,	,	PUNCT
easat-5713	197	9	class	class	NOUN
easat-5713	197	10	imbalances	imbalance	NOUN
easat-5713	197	11	,	,	PUNCT
easat-5713	197	12	and	and	CCONJ
easat-5713	197	13	limited	limited	ADJ
easat-5713	197	14	diversity	diversity	NOUN
easat-5713	197	15	in	in	ADP
easat-5713	197	16	existing	exist	VERB
easat-5713	197	17	fer	fer	PROPN
easat-5713	197	18	datasets	dataset	NOUN
easat-5713	197	19	,	,	PUNCT
easat-5713	197	20	synthetic	synthetic	ADJ
easat-5713	197	21	data	datum	NOUN
easat-5713	197	22	was	be	AUX
easat-5713	197	23	expected	expect	VERB
easat-5713	197	24	to	to	PART
easat-5713	197	25	contribute	contribute	VERB
easat-5713	197	26	to	to	ADP
easat-5713	197	27	generalizable	generalizable	ADJ
easat-5713	197	28	model	model	NOUN
easat-5713	197	29	performance	performance	NOUN
easat-5713	197	30	.	.	PUNCT
easat-5713	198	1	however	however	ADV
easat-5713	198	2	,	,	PUNCT
easat-5713	198	3	our	our	PRON
easat-5713	198	4	results	result	NOUN
easat-5713	198	5	indicated	indicate	VERB
easat-5713	198	6	that	that	SCONJ
easat-5713	198	7	the	the	DET
easat-5713	198	8	inclusion	inclusion	NOUN
easat-5713	198	9	of	of	ADP
easat-5713	198	10	synthetic	synthetic	ADJ
easat-5713	198	11	data	datum	NOUN
easat-5713	198	12	did	do	AUX
easat-5713	198	13	not	not	PART
easat-5713	198	14	significantly	significantly	ADV
easat-5713	198	15	improve	improve	VERB
easat-5713	198	16	the	the	DET
easat-5713	198	17	model	model	NOUN
easat-5713	198	18	's	's	PART
easat-5713	198	19	performance	performance	NOUN
easat-5713	198	20	.	.	PUNCT
easat-5713	199	1	in	in	ADP
easat-5713	199	2	fact	fact	NOUN
easat-5713	199	3	,	,	PUNCT
easat-5713	199	4	the	the	DET
easat-5713	199	5	model	model	NOUN
easat-5713	199	6	trained	train	VERB
easat-5713	199	7	on	on	ADP
easat-5713	199	8	the	the	DET
easat-5713	199	9	augmented	augment	VERB
easat-5713	199	10	dataset	dataset	NOUN
easat-5713	199	11	with	with	ADP
easat-5713	199	12	synthetic	synthetic	ADJ
easat-5713	199	13	images	image	NOUN
easat-5713	199	14	achieved	achieve	VERB
easat-5713	199	15	slightly	slightly	ADV
easat-5713	199	16	lower	low	ADJ
easat-5713	199	17	accuracy	accuracy	NOUN
easat-5713	199	18	than	than	ADP
easat-5713	199	19	the	the	DET
easat-5713	199	20	model	model	NOUN
easat-5713	199	21	trained	train	VERB
easat-5713	199	22	on	on	ADP
easat-5713	199	23	the	the	DET
easat-5713	199	24	original	original	ADJ
easat-5713	199	25	fer2013	fer2013	ADJ
easat-5713	199	26	dataset	dataset	NOUN
easat-5713	199	27	.	.	PUNCT
easat-5713	200	1	one	one	NUM
easat-5713	200	2	possible	possible	ADJ
easat-5713	200	3	explanation	explanation	NOUN
easat-5713	200	4	for	for	ADP
easat-5713	200	5	the	the	DET
easat-5713	200	6	lack	lack	NOUN
easat-5713	200	7	of	of	ADP
easat-5713	200	8	performance	performance	NOUN
easat-5713	200	9	improvement	improvement	NOUN
easat-5713	200	10	is	be	AUX
easat-5713	200	11	the	the	DET
easat-5713	200	12	limitations	limitation	NOUN
easat-5713	200	13	of	of	ADP
easat-5713	200	14	the	the	DET
easat-5713	200	15	synthetic	synthetic	ADJ
easat-5713	200	16	images	image	NOUN
easat-5713	200	17	.	.	PUNCT
easat-5713	201	1	although	although	SCONJ
easat-5713	201	2	ai	ai	ADJ
easat-5713	201	3	-	-	PUNCT
easat-5713	201	4	generated	generate	VERB
easat-5713	201	5	data	datum	NOUN
easat-5713	201	6	can	can	AUX
easat-5713	201	7	expand	expand	VERB
easat-5713	201	8	the	the	DET
easat-5713	201	9	diversity	diversity	NOUN
easat-5713	201	10	of	of	ADP
easat-5713	201	11	training	training	NOUN
easat-5713	201	12	datasets	dataset	NOUN
easat-5713	201	13	,	,	PUNCT
easat-5713	201	14	it	it	PRON
easat-5713	201	15	does	do	AUX
easat-5713	201	16	not	not	PART
easat-5713	201	17	always	always	ADV
easat-5713	201	18	guarantee	guarantee	VERB
easat-5713	201	19	a	a	DET
easat-5713	201	20	perfect	perfect	ADJ
easat-5713	201	21	representation	representation	NOUN
easat-5713	201	22	of	of	ADP
easat-5713	201	23	real	real	ADJ
easat-5713	201	24	-	-	PUNCT
easat-5713	201	25	world	world	NOUN
easat-5713	201	26	expressions	expression	NOUN
easat-5713	201	27	.	.	PUNCT
easat-5713	202	1	the	the	DET
easat-5713	202	2	juggernaut	juggernaut	PROPN
easat-5713	202	3	xl	xl	PROPN
easat-5713	202	4	model	model	PROPN
easat-5713	202	5	,	,	PUNCT
easat-5713	202	6	despite	despite	SCONJ
easat-5713	202	7	being	be	AUX
easat-5713	202	8	an	an	DET
easat-5713	202	9	advanced	advanced	ADJ
easat-5713	202	10	generative	generative	ADJ
easat-5713	202	11	model	model	NOUN
easat-5713	202	12	,	,	PUNCT
easat-5713	202	13	may	may	AUX
easat-5713	202	14	not	not	PART
easat-5713	202	15	have	have	AUX
easat-5713	202	16	produced	produce	VERB
easat-5713	202	17	images	image	NOUN
easat-5713	202	18	with	with	ADP
easat-5713	202	19	sufficient	sufficient	ADJ
easat-5713	202	20	realism	realism	NOUN
easat-5713	202	21	to	to	PART
easat-5713	202	22	enhance	enhance	VERB
easat-5713	202	23	fer	fer	PROPN
easat-5713	202	24	performance	performance	NOUN
easat-5713	202	25	[	[	X
easat-5713	202	26	24	24	NUM
easat-5713	202	27	]	]	PUNCT
easat-5713	202	28	.	.	PUNCT
easat-5713	203	1	previous	previous	ADJ
easat-5713	203	2	studies	study	NOUN
easat-5713	203	3	have	have	AUX
easat-5713	203	4	demonstrated	demonstrate	VERB
easat-5713	203	5	that	that	SCONJ
easat-5713	203	6	synthetic	synthetic	ADJ
easat-5713	203	7	images	image	NOUN
easat-5713	203	8	generated	generate	VERB
easat-5713	203	9	by	by	ADP
easat-5713	203	10	gans	gan	NOUN
easat-5713	203	11	or	or	CCONJ
easat-5713	203	12	1867	1867	NUM
easat-5713	203	13	edelweiss	edelweiss	PROPN
easat-5713	203	14	applied	apply	VERB
easat-5713	203	15	science	science	NOUN
easat-5713	203	16	and	and	CCONJ
easat-5713	203	17	technology	technology	NOUN
easat-5713	203	18	issn	issn	PROPN
easat-5713	203	19	:	:	PUNCT
easat-5713	203	20	2576	2576	NUM
easat-5713	203	21	-	-	SYM
easat-5713	203	22	8484	8484	NUM
easat-5713	203	23	vol	vol	NOUN
easat-5713	203	24	.	.	PROPN
easat-5713	204	1	9	9	NUM
easat-5713	204	2	,	,	PUNCT
easat-5713	204	3	no	no	INTJ
easat-5713	204	4	.	.	NOUN
easat-5713	205	1	3	3	NUM
easat-5713	205	2	:	:	PUNCT
easat-5713	205	3	1855	1855	NUM
easat-5713	205	4	-	-	SYM
easat-5713	205	5	1869	1869	NUM
easat-5713	205	6	,	,	PUNCT
easat-5713	205	7	2025	2025	NUM
easat-5713	205	8	doi	doi	NOUN
easat-5713	205	9	:	:	PUNCT
easat-5713	205	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	205	11	©	©	ADP
easat-5713	205	12	2025	2025	NUM
easat-5713	205	13	by	by	ADP
easat-5713	205	14	the	the	DET
easat-5713	205	15	authors	author	NOUN
easat-5713	205	16	;	;	PUNCT
easat-5713	205	17	licensee	licensee	PROPN
easat-5713	205	18	learning	learn	VERB
easat-5713	205	19	gate	gate	NOUN
easat-5713	205	20	diffusion	diffusion	NOUN
easat-5713	205	21	models	model	NOUN
easat-5713	205	22	can	can	AUX
easat-5713	205	23	improve	improve	VERB
easat-5713	205	24	models	model	NOUN
easat-5713	205	25	,	,	PUNCT
easat-5713	205	26	particularly	particularly	ADV
easat-5713	205	27	by	by	ADP
easat-5713	205	28	addressing	address	VERB
easat-5713	205	29	class	class	NOUN
easat-5713	205	30	imbalances	imbalance	NOUN
easat-5713	205	31	.	.	PUNCT
easat-5713	206	1	however	however	ADV
easat-5713	206	2	,	,	PUNCT
easat-5713	206	3	these	these	DET
easat-5713	206	4	studies	study	NOUN
easat-5713	206	5	often	often	ADV
easat-5713	206	6	rely	rely	VERB
easat-5713	206	7	on	on	ADP
easat-5713	206	8	highly	highly	ADV
easat-5713	206	9	refined	refined	ADJ
easat-5713	206	10	generative	generative	ADJ
easat-5713	206	11	techniques	technique	NOUN
easat-5713	206	12	and	and	CCONJ
easat-5713	206	13	extensive	extensive	ADJ
easat-5713	206	14	post	post	ADJ
easat-5713	206	15	-	-	ADJ
easat-5713	206	16	processing	processing	NOUN
easat-5713	206	17	to	to	PART
easat-5713	206	18	ensure	ensure	VERB
easat-5713	206	19	the	the	DET
easat-5713	206	20	quality	quality	NOUN
easat-5713	206	21	of	of	ADP
easat-5713	206	22	synthetic	synthetic	ADJ
easat-5713	206	23	images	image	NOUN
easat-5713	206	24	.	.	PUNCT
easat-5713	207	1	the	the	DET
easat-5713	207	2	discrepancies	discrepancy	NOUN
easat-5713	207	3	between	between	ADP
easat-5713	207	4	our	our	PRON
easat-5713	207	5	findings	finding	NOUN
easat-5713	207	6	and	and	CCONJ
easat-5713	207	7	prior	prior	ADJ
easat-5713	207	8	research	research	NOUN
easat-5713	207	9	suggest	suggest	VERB
easat-5713	207	10	that	that	SCONJ
easat-5713	207	11	generative	generative	ADJ
easat-5713	207	12	models	model	NOUN
easat-5713	207	13	must	must	AUX
easat-5713	207	14	be	be	AUX
easat-5713	207	15	optimized	optimize	VERB
easat-5713	207	16	further	far	ADV
easat-5713	207	17	to	to	PART
easat-5713	207	18	produce	produce	VERB
easat-5713	207	19	high	high	ADJ
easat-5713	207	20	-	-	PUNCT
easat-5713	207	21	quality	quality	NOUN
easat-5713	207	22	images	image	NOUN
easat-5713	207	23	that	that	PRON
easat-5713	207	24	closely	closely	ADV
easat-5713	207	25	mimic	mimic	ADJ
easat-5713	207	26	natural	natural	ADJ
easat-5713	207	27	facial	facial	ADJ
easat-5713	207	28	expressions	expression	NOUN
easat-5713	207	29	[	[	X
easat-5713	207	30	25	25	NUM
easat-5713	207	31	]	]	PUNCT
easat-5713	207	32	.	.	PUNCT
easat-5713	208	1	several	several	ADJ
easat-5713	208	2	experimental	experimental	ADJ
easat-5713	208	3	limitations	limitation	NOUN
easat-5713	208	4	may	may	AUX
easat-5713	208	5	have	have	AUX
easat-5713	208	6	also	also	ADV
easat-5713	208	7	contributed	contribute	VERB
easat-5713	208	8	to	to	ADP
easat-5713	208	9	the	the	DET
easat-5713	208	10	lack	lack	NOUN
easat-5713	208	11	of	of	ADP
easat-5713	208	12	performance	performance	NOUN
easat-5713	208	13	gains	gain	NOUN
easat-5713	208	14	.	.	PUNCT
easat-5713	209	1	one	one	NUM
easat-5713	209	2	of	of	ADP
easat-5713	209	3	the	the	DET
easat-5713	209	4	main	main	ADJ
easat-5713	209	5	constraints	constraint	NOUN
easat-5713	209	6	was	be	AUX
easat-5713	209	7	the	the	DET
easat-5713	209	8	training	training	NOUN
easat-5713	209	9	duration	duration	NOUN
easat-5713	209	10	of	of	ADP
easat-5713	209	11	the	the	DET
easat-5713	209	12	vgg-19	vgg-19	PROPN
easat-5713	209	13	model	model	NOUN
easat-5713	209	14	.	.	PUNCT
easat-5713	210	1	due	due	ADP
easat-5713	210	2	to	to	ADP
easat-5713	210	3	hardware	hardware	NOUN
easat-5713	210	4	limitations	limitation	NOUN
easat-5713	210	5	,	,	PUNCT
easat-5713	210	6	the	the	DET
easat-5713	210	7	model	model	NOUN
easat-5713	210	8	was	be	AUX
easat-5713	210	9	trained	train	VERB
easat-5713	210	10	for	for	ADP
easat-5713	210	11	only	only	ADV
easat-5713	210	12	25	25	NUM
easat-5713	210	13	epochs	epoch	NOUN
easat-5713	210	14	,	,	PUNCT
easat-5713	210	15	whereas	whereas	SCONJ
easat-5713	210	16	the	the	DET
easat-5713	210	17	original	original	ADJ
easat-5713	210	18	segmentation	segmentation	NOUN
easat-5713	210	19	vgg-19	vgg-19	NUM
easat-5713	210	20	model	model	NOUN
easat-5713	210	21	achieved	achieve	VERB
easat-5713	210	22	optimal	optimal	ADJ
easat-5713	210	23	performance	performance	NOUN
easat-5713	210	24	after	after	ADP
easat-5713	210	25	67	67	NUM
easat-5713	210	26	epochs	epoch	NOUN
easat-5713	210	27	.	.	PUNCT
easat-5713	211	1	given	give	VERB
easat-5713	211	2	the	the	DET
easat-5713	211	3	complexity	complexity	NOUN
easat-5713	211	4	of	of	ADP
easat-5713	211	5	fer	fer	PROPN
easat-5713	211	6	tasks	task	NOUN
easat-5713	211	7	,	,	PUNCT
easat-5713	211	8	a	a	DET
easat-5713	211	9	longer	long	ADV
easat-5713	211	10	training	training	NOUN
easat-5713	211	11	duration	duration	NOUN
easat-5713	211	12	may	may	AUX
easat-5713	211	13	have	have	AUX
easat-5713	211	14	allowed	allow	VERB
easat-5713	211	15	the	the	DET
easat-5713	211	16	model	model	NOUN
easat-5713	211	17	to	to	PART
easat-5713	211	18	better	well	ADV
easat-5713	211	19	adapt	adapt	VERB
easat-5713	211	20	to	to	ADP
easat-5713	211	21	the	the	DET
easat-5713	211	22	synthetic	synthetic	ADJ
easat-5713	211	23	data	datum	NOUN
easat-5713	211	24	and	and	CCONJ
easat-5713	211	25	extract	extract	VERB
easat-5713	211	26	more	more	ADV
easat-5713	211	27	meaningful	meaningful	ADJ
easat-5713	211	28	features	feature	NOUN
easat-5713	211	29	from	from	ADP
easat-5713	211	30	it	it	PRON
easat-5713	211	31	.	.	PUNCT
easat-5713	212	1	while	while	SCONJ
easat-5713	212	2	generating	generate	VERB
easat-5713	212	3	300	300	NUM
easat-5713	212	4	synthetic	synthetic	ADJ
easat-5713	212	5	images	image	NOUN
easat-5713	212	6	per	per	ADP
easat-5713	212	7	emotion	emotion	NOUN
easat-5713	212	8	category	category	NOUN
easat-5713	212	9	provided	provide	VERB
easat-5713	212	10	a	a	DET
easat-5713	212	11	substantial	substantial	ADJ
easat-5713	212	12	augmentation	augmentation	NOUN
easat-5713	212	13	,	,	PUNCT
easat-5713	212	14	this	this	DET
easat-5713	212	15	number	number	NOUN
easat-5713	212	16	may	may	AUX
easat-5713	212	17	still	still	ADV
easat-5713	212	18	be	be	AUX
easat-5713	212	19	insufficient	insufficient	ADJ
easat-5713	212	20	to	to	PART
easat-5713	212	21	significantly	significantly	ADV
easat-5713	212	22	impact	impact	VERB
easat-5713	212	23	the	the	DET
easat-5713	212	24	training	training	NOUN
easat-5713	212	25	process	process	NOUN
easat-5713	212	26	.	.	PUNCT
easat-5713	213	1	a	a	DET
easat-5713	213	2	larger	large	ADJ
easat-5713	213	3	dataset	dataset	NOUN
easat-5713	213	4	with	with	ADP
easat-5713	213	5	more	more	ADJ
easat-5713	213	6	variations	variation	NOUN
easat-5713	213	7	in	in	ADP
easat-5713	213	8	facial	facial	ADJ
easat-5713	213	9	expressions	expression	NOUN
easat-5713	213	10	and	and	CCONJ
easat-5713	213	11	environmental	environmental	ADJ
easat-5713	213	12	conditions	condition	NOUN
easat-5713	213	13	could	could	AUX
easat-5713	213	14	lead	lead	VERB
easat-5713	213	15	to	to	ADP
easat-5713	213	16	better	well	ADJ
easat-5713	213	17	generalization	generalization	NOUN
easat-5713	213	18	and	and	CCONJ
easat-5713	213	19	improved	improved	ADJ
easat-5713	213	20	recognition	recognition	NOUN
easat-5713	213	21	performance	performance	NOUN
easat-5713	213	22	.	.	PUNCT
easat-5713	214	1	future	future	ADJ
easat-5713	214	2	research	research	NOUN
easat-5713	214	3	should	should	AUX
easat-5713	214	4	explore	explore	VERB
easat-5713	214	5	multiple	multiple	ADJ
easat-5713	214	6	strategies	strategy	NOUN
easat-5713	214	7	to	to	PART
easat-5713	214	8	maximize	maximize	VERB
easat-5713	214	9	the	the	DET
easat-5713	214	10	potential	potential	NOUN
easat-5713	214	11	of	of	ADP
easat-5713	214	12	synthetic	synthetic	ADJ
easat-5713	214	13	data	datum	NOUN
easat-5713	214	14	in	in	ADP
easat-5713	214	15	fer	fer	PROPN
easat-5713	214	16	.	.	PUNCT
easat-5713	215	1	improving	improve	VERB
easat-5713	215	2	the	the	DET
easat-5713	215	3	quality	quality	NOUN
easat-5713	215	4	and	and	CCONJ
easat-5713	215	5	realism	realism	NOUN
easat-5713	215	6	of	of	ADP
easat-5713	215	7	synthetic	synthetic	ADJ
easat-5713	215	8	images	image	NOUN
easat-5713	215	9	through	through	ADP
easat-5713	215	10	advanced	advanced	ADJ
easat-5713	215	11	generative	generative	ADJ
easat-5713	215	12	models	model	NOUN
easat-5713	215	13	is	be	AUX
easat-5713	215	14	a	a	DET
easat-5713	215	15	crucial	crucial	ADJ
easat-5713	215	16	first	first	ADJ
easat-5713	215	17	step	step	NOUN
easat-5713	215	18	.	.	PUNCT
easat-5713	216	1	while	while	SCONJ
easat-5713	216	2	the	the	DET
easat-5713	216	3	juggernaut	juggernaut	NOUN
easat-5713	216	4	xl	xl	PROPN
easat-5713	216	5	model	model	PROPN
easat-5713	216	6	provided	provide	VERB
easat-5713	216	7	a	a	DET
easat-5713	216	8	useful	useful	ADJ
easat-5713	216	9	baseline	baseline	NOUN
easat-5713	216	10	,	,	PUNCT
easat-5713	216	11	exploring	explore	VERB
easat-5713	216	12	alternative	alternative	ADJ
easat-5713	216	13	architectures	architecture	NOUN
easat-5713	216	14	such	such	ADJ
easat-5713	216	15	as	as	ADP
easat-5713	216	16	diffusion	diffusion	NOUN
easat-5713	216	17	models	model	NOUN
easat-5713	216	18	or	or	CCONJ
easat-5713	216	19	stylegan	stylegan	NOUN
easat-5713	216	20	could	could	AUX
easat-5713	216	21	yield	yield	VERB
easat-5713	216	22	higher	high	ADJ
easat-5713	216	23	-	-	PUNCT
easat-5713	216	24	fidelity	fidelity	NOUN
easat-5713	216	25	images	image	NOUN
easat-5713	216	26	that	that	PRON
easat-5713	216	27	better	well	ADV
easat-5713	216	28	capture	capture	VERB
easat-5713	216	29	the	the	DET
easat-5713	216	30	complexities	complexity	NOUN
easat-5713	216	31	of	of	ADP
easat-5713	216	32	human	human	ADJ
easat-5713	216	33	facial	facial	ADJ
easat-5713	216	34	expressions	expression	NOUN
easat-5713	216	35	.	.	PUNCT
easat-5713	217	1	high	high	ADJ
easat-5713	217	2	-	-	PUNCT
easat-5713	217	3	quality	quality	NOUN
easat-5713	217	4	synthetic	synthetic	ADJ
easat-5713	217	5	images	image	NOUN
easat-5713	217	6	that	that	PRON
easat-5713	217	7	accurately	accurately	ADV
easat-5713	217	8	mimic	mimic	VERB
easat-5713	217	9	real	real	ADJ
easat-5713	217	10	-	-	PUNCT
easat-5713	217	11	world	world	NOUN
easat-5713	217	12	variations	variation	NOUN
easat-5713	217	13	can	can	AUX
easat-5713	217	14	enhance	enhance	VERB
easat-5713	217	15	model	model	NOUN
easat-5713	217	16	robustness	robustness	NOUN
easat-5713	217	17	,	,	PUNCT
easat-5713	217	18	making	make	VERB
easat-5713	217	19	fer	fer	PROPN
easat-5713	217	20	systems	system	NOUN
easat-5713	217	21	more	more	ADV
easat-5713	217	22	reliable	reliable	ADJ
easat-5713	217	23	in	in	ADP
easat-5713	217	24	practical	practical	ADJ
easat-5713	217	25	applications	application	NOUN
easat-5713	217	26	[	[	X
easat-5713	217	27	26	26	NUM
easat-5713	217	28	]	]	PUNCT
easat-5713	217	29	.	.	PUNCT
easat-5713	218	1	one	one	NUM
easat-5713	218	2	potential	potential	ADJ
easat-5713	218	3	approach	approach	NOUN
easat-5713	218	4	is	be	AUX
easat-5713	218	5	using	use	VERB
easat-5713	218	6	an	an	DET
easat-5713	218	7	automated	automate	VERB
easat-5713	218	8	filtering	filtering	NOUN
easat-5713	218	9	mechanism	mechanism	NOUN
easat-5713	218	10	to	to	PART
easat-5713	218	11	assess	assess	VERB
easat-5713	218	12	the	the	DET
easat-5713	218	13	realism	realism	NOUN
easat-5713	218	14	of	of	ADP
easat-5713	218	15	generated	generate	VERB
easat-5713	218	16	images	image	NOUN
easat-5713	218	17	before	before	ADP
easat-5713	218	18	integrating	integrate	VERB
easat-5713	218	19	them	they	PRON
easat-5713	218	20	into	into	ADP
easat-5713	218	21	the	the	DET
easat-5713	218	22	dataset	dataset	NOUN
easat-5713	218	23	.	.	PUNCT
easat-5713	219	1	this	this	PRON
easat-5713	219	2	could	could	AUX
easat-5713	219	3	be	be	AUX
easat-5713	219	4	achieved	achieve	VERB
easat-5713	219	5	through	through	ADP
easat-5713	219	6	discriminator	discriminator	NOUN
easat-5713	219	7	networks	network	NOUN
easat-5713	219	8	that	that	PRON
easat-5713	219	9	score	score	VERB
easat-5713	219	10	the	the	DET
easat-5713	219	11	quality	quality	NOUN
easat-5713	219	12	of	of	ADP
easat-5713	219	13	synthetic	synthetic	ADJ
easat-5713	219	14	images	image	NOUN
easat-5713	219	15	or	or	CCONJ
easat-5713	219	16	through	through	ADP
easat-5713	219	17	human	human	NOUN
easat-5713	219	18	-	-	PUNCT
easat-5713	219	19	in	in	ADP
easat-5713	219	20	-	-	PUNCT
easat-5713	219	21	theloop	theloop	NOUN
easat-5713	219	22	verification	verification	NOUN
easat-5713	219	23	processes	process	NOUN
easat-5713	219	24	[	[	X
easat-5713	219	25	27	27	NUM
easat-5713	219	26	]	]	PUNCT
easat-5713	219	27	.	.	PUNCT
easat-5713	220	1	another	another	DET
easat-5713	220	2	promising	promising	ADJ
easat-5713	220	3	direction	direction	NOUN
easat-5713	220	4	for	for	ADP
easat-5713	220	5	future	future	ADJ
easat-5713	220	6	research	research	NOUN
easat-5713	220	7	is	be	AUX
easat-5713	220	8	leveraging	leverage	VERB
easat-5713	220	9	more	more	ADV
easat-5713	220	10	advanced	advanced	ADJ
easat-5713	220	11	fine	fine	ADJ
easat-5713	220	12	-	-	PUNCT
easat-5713	220	13	tuning	tune	VERB
easat-5713	220	14	techniques	technique	NOUN
easat-5713	220	15	to	to	PART
easat-5713	220	16	train	train	VERB
easat-5713	220	17	models	model	NOUN
easat-5713	220	18	entirely	entirely	ADV
easat-5713	220	19	on	on	ADP
easat-5713	220	20	synthetic	synthetic	ADJ
easat-5713	220	21	datasets	dataset	NOUN
easat-5713	220	22	.	.	PUNCT
easat-5713	221	1	one	one	NUM
easat-5713	221	2	such	such	ADJ
easat-5713	221	3	technique	technique	NOUN
easat-5713	221	4	is	be	AUX
easat-5713	221	5	dreambooth	dreambooth	NOUN
easat-5713	221	6	,	,	PUNCT
easat-5713	221	7	which	which	DET
easat-5713	221	8	fine	fine	ADJ
easat-5713	221	9	-	-	PUNCT
easat-5713	221	10	tunes	tune	NOUN
easat-5713	221	11	a	a	DET
easat-5713	221	12	diffusion	diffusion	NOUN
easat-5713	221	13	model	model	NOUN
easat-5713	221	14	using	use	VERB
easat-5713	221	15	only	only	ADV
easat-5713	221	16	a	a	DET
easat-5713	221	17	small	small	ADJ
easat-5713	221	18	number	number	NOUN
easat-5713	221	19	of	of	ADP
easat-5713	221	20	real	real	ADJ
easat-5713	221	21	images	image	NOUN
easat-5713	221	22	to	to	PART
easat-5713	221	23	generate	generate	VERB
easat-5713	221	24	highly	highly	ADV
easat-5713	221	25	realistic	realistic	ADJ
easat-5713	221	26	synthetic	synthetic	ADJ
easat-5713	221	27	data	datum	NOUN
easat-5713	221	28	.	.	PUNCT
easat-5713	222	1	dreambooth	dreambooth	NOUN
easat-5713	222	2	allows	allow	VERB
easat-5713	222	3	models	model	NOUN
easat-5713	222	4	to	to	PART
easat-5713	222	5	learn	learn	VERB
easat-5713	222	6	details	detail	NOUN
easat-5713	222	7	from	from	ADP
easat-5713	222	8	a	a	DET
easat-5713	222	9	small	small	ADJ
easat-5713	222	10	dataset	dataset	NOUN
easat-5713	222	11	while	while	SCONJ
easat-5713	222	12	maintaining	maintain	VERB
easat-5713	222	13	high	high	ADJ
easat-5713	222	14	consistency	consistency	NOUN
easat-5713	222	15	in	in	ADP
easat-5713	222	16	generated	generate	VERB
easat-5713	222	17	images	image	NOUN
easat-5713	222	18	[	[	X
easat-5713	222	19	28	28	NUM
easat-5713	222	20	]	]	PUNCT
easat-5713	222	21	.	.	PUNCT
easat-5713	223	1	a	a	DET
easat-5713	223	2	key	key	ADJ
easat-5713	223	3	challenge	challenge	NOUN
easat-5713	223	4	in	in	ADP
easat-5713	223	5	integrating	integrate	VERB
easat-5713	223	6	synthetic	synthetic	ADJ
easat-5713	223	7	data	datum	NOUN
easat-5713	223	8	is	be	AUX
easat-5713	223	9	ensuring	ensure	VERB
easat-5713	223	10	that	that	SCONJ
easat-5713	223	11	it	it	PRON
easat-5713	223	12	complements	complement	VERB
easat-5713	223	13	real	real	ADJ
easat-5713	223	14	-	-	PUNCT
easat-5713	223	15	world	world	NOUN
easat-5713	223	16	images	image	NOUN
easat-5713	223	17	rather	rather	ADV
easat-5713	223	18	than	than	ADP
easat-5713	223	19	distorting	distort	VERB
easat-5713	223	20	the	the	DET
easat-5713	223	21	training	training	NOUN
easat-5713	223	22	distribution	distribution	NOUN
easat-5713	223	23	.	.	PUNCT
easat-5713	224	1	increasing	increase	VERB
easat-5713	224	2	the	the	DET
easat-5713	224	3	volume	volume	NOUN
easat-5713	224	4	and	and	CCONJ
easat-5713	224	5	diversity	diversity	NOUN
easat-5713	224	6	of	of	ADP
easat-5713	224	7	synthetic	synthetic	ADJ
easat-5713	224	8	images	image	NOUN
easat-5713	224	9	is	be	AUX
easat-5713	224	10	another	another	DET
easat-5713	224	11	promising	promising	ADJ
easat-5713	224	12	direction	direction	NOUN
easat-5713	224	13	.	.	PUNCT
easat-5713	225	1	while	while	SCONJ
easat-5713	225	2	this	this	DET
easat-5713	225	3	study	study	NOUN
easat-5713	225	4	generated	generate	VERB
easat-5713	225	5	a	a	DET
easat-5713	225	6	fixed	fix	VERB
easat-5713	225	7	number	number	NOUN
easat-5713	225	8	of	of	ADP
easat-5713	225	9	synthetic	synthetic	ADJ
easat-5713	225	10	images	image	NOUN
easat-5713	225	11	per	per	ADP
easat-5713	225	12	emotion	emotion	NOUN
easat-5713	225	13	category	category	NOUN
easat-5713	225	14	,	,	PUNCT
easat-5713	225	15	future	future	ADJ
easat-5713	225	16	research	research	NOUN
easat-5713	225	17	could	could	AUX
easat-5713	225	18	investigate	investigate	VERB
easat-5713	225	19	the	the	DET
easat-5713	225	20	impact	impact	NOUN
easat-5713	225	21	of	of	ADP
easat-5713	225	22	varying	vary	VERB
easat-5713	225	23	the	the	DET
easat-5713	225	24	number	number	NOUN
easat-5713	225	25	of	of	ADP
easat-5713	225	26	synthetic	synthetic	ADJ
easat-5713	225	27	images	image	NOUN
easat-5713	225	28	to	to	PART
easat-5713	225	29	determine	determine	VERB
easat-5713	225	30	the	the	DET
easat-5713	225	31	optimal	optimal	ADJ
easat-5713	225	32	ratio	ratio	NOUN
easat-5713	225	33	of	of	ADP
easat-5713	225	34	real	real	ADJ
easat-5713	225	35	to	to	ADP
easat-5713	225	36	synthetic	synthetic	ADJ
easat-5713	225	37	data	datum	NOUN
easat-5713	225	38	.	.	PUNCT
easat-5713	226	1	fine	fine	ADJ
easat-5713	226	2	-	-	PUNCT
easat-5713	226	3	tuning	tune	VERB
easat-5713	226	4	strategies	strategy	NOUN
easat-5713	226	5	such	such	ADJ
easat-5713	226	6	as	as	ADP
easat-5713	226	7	transfer	transfer	NOUN
easat-5713	226	8	learning	learning	NOUN
easat-5713	226	9	,	,	PUNCT
easat-5713	226	10	domain	domain	NOUN
easat-5713	226	11	adaptation	adaptation	NOUN
easat-5713	226	12	,	,	PUNCT
easat-5713	226	13	and	and	CCONJ
easat-5713	226	14	feature	feature	NOUN
easat-5713	226	15	alignment	alignment	NOUN
easat-5713	226	16	techniques	technique	NOUN
easat-5713	226	17	could	could	AUX
easat-5713	226	18	also	also	ADV
easat-5713	226	19	be	be	AUX
easat-5713	226	20	explored	explore	VERB
easat-5713	226	21	to	to	PART
easat-5713	226	22	ensure	ensure	VERB
easat-5713	226	23	that	that	SCONJ
easat-5713	226	24	synthetic	synthetic	ADJ
easat-5713	226	25	images	image	NOUN
easat-5713	226	26	contribute	contribute	VERB
easat-5713	226	27	positively	positively	ADV
easat-5713	226	28	to	to	ADP
easat-5713	226	29	model	model	NOUN
easat-5713	226	30	performance	performance	NOUN
easat-5713	226	31	rather	rather	ADV
easat-5713	226	32	than	than	ADP
easat-5713	226	33	introducing	introduce	VERB
easat-5713	226	34	additional	additional	ADJ
easat-5713	226	35	noise	noise	NOUN
easat-5713	226	36	[	[	X
easat-5713	226	37	29	29	NUM
easat-5713	226	38	]	]	PUNCT
easat-5713	226	39	.	.	PUNCT
easat-5713	227	1	also	also	ADV
easat-5713	227	2	exploring	explore	VERB
easat-5713	227	3	adversarial	adversarial	ADJ
easat-5713	227	4	training	training	NOUN
easat-5713	227	5	techniques	technique	NOUN
easat-5713	227	6	,	,	PUNCT
easat-5713	227	7	where	where	SCONJ
easat-5713	227	8	a	a	DET
easat-5713	227	9	fer	fer	PROPN
easat-5713	227	10	model	model	NOUN
easat-5713	227	11	is	be	AUX
easat-5713	227	12	trained	train	VERB
easat-5713	227	13	alongside	alongside	ADP
easat-5713	227	14	a	a	DET
easat-5713	227	15	generative	generative	ADJ
easat-5713	227	16	model	model	NOUN
easat-5713	227	17	to	to	PART
easat-5713	227	18	improve	improve	VERB
easat-5713	227	19	its	its	PRON
easat-5713	227	20	ability	ability	NOUN
easat-5713	227	21	to	to	PART
easat-5713	227	22	distinguish	distinguish	VERB
easat-5713	227	23	between	between	ADP
easat-5713	227	24	real	real	ADJ
easat-5713	227	25	and	and	CCONJ
easat-5713	227	26	synthetic	synthetic	ADJ
easat-5713	227	27	expressions	expression	NOUN
easat-5713	227	28	,	,	PUNCT
easat-5713	227	29	could	could	AUX
easat-5713	227	30	lead	lead	VERB
easat-5713	227	31	to	to	ADP
easat-5713	227	32	more	more	ADV
easat-5713	227	33	robust	robust	ADJ
easat-5713	227	34	results	result	NOUN
easat-5713	227	35	.	.	PUNCT
easat-5713	228	1	5	5	X
easat-5713	228	2	.	.	X
easat-5713	228	3	conclusion	conclusion	NOUN
easat-5713	228	4	this	this	DET
easat-5713	228	5	study	study	NOUN
easat-5713	228	6	investigated	investigate	VERB
easat-5713	228	7	the	the	DET
easat-5713	228	8	impact	impact	NOUN
easat-5713	228	9	of	of	ADP
easat-5713	228	10	ai	ai	VERB
easat-5713	228	11	-	-	PUNCT
easat-5713	228	12	generated	generate	VERB
easat-5713	228	13	synthetic	synthetic	ADJ
easat-5713	228	14	data	datum	NOUN
easat-5713	228	15	on	on	ADP
easat-5713	228	16	facial	facial	ADJ
easat-5713	228	17	emotion	emotion	NOUN
easat-5713	228	18	recognition	recognition	NOUN
easat-5713	228	19	model	model	NOUN
easat-5713	228	20	performance	performance	NOUN
easat-5713	228	21	by	by	ADP
easat-5713	228	22	integrating	integrate	VERB
easat-5713	228	23	synthetic	synthetic	ADJ
easat-5713	228	24	images	image	NOUN
easat-5713	228	25	generated	generate	VERB
easat-5713	228	26	using	use	VERB
easat-5713	228	27	the	the	DET
easat-5713	228	28	juggernaut	juggernaut	NOUN
easat-5713	228	29	xl	xl	PROPN
easat-5713	228	30	model	model	PROPN
easat-5713	228	31	into	into	ADP
easat-5713	228	32	the	the	DET
easat-5713	228	33	fer2013	fer2013	ADJ
easat-5713	228	34	dataset	dataset	NOUN
easat-5713	228	35	.	.	PUNCT
easat-5713	229	1	the	the	DET
easat-5713	229	2	primary	primary	ADJ
easat-5713	229	3	goal	goal	NOUN
easat-5713	229	4	was	be	AUX
easat-5713	229	5	to	to	PART
easat-5713	229	6	determine	determine	VERB
easat-5713	229	7	whether	whether	SCONJ
easat-5713	229	8	synthetic	synthetic	ADJ
easat-5713	229	9	data	datum	NOUN
easat-5713	229	10	could	could	AUX
easat-5713	229	11	enhance	enhance	VERB
easat-5713	229	12	model	model	NOUN
easat-5713	229	13	accuracy	accuracy	NOUN
easat-5713	229	14	,	,	PUNCT
easat-5713	229	15	precision	precision	NOUN
easat-5713	229	16	,	,	PUNCT
easat-5713	229	17	recall	recall	NOUN
easat-5713	229	18	,	,	PUNCT
easat-5713	229	19	and	and	CCONJ
easat-5713	229	20	f1	f1	NOUN
easat-5713	229	21	-	-	PUNCT
easat-5713	229	22	score	score	NOUN
easat-5713	229	23	by	by	ADP
easat-5713	229	24	mitigating	mitigate	VERB
easat-5713	229	25	the	the	DET
easat-5713	229	26	limitations	limitation	NOUN
easat-5713	229	27	of	of	ADP
easat-5713	229	28	real	real	ADJ
easat-5713	229	29	-	-	PUNCT
easat-5713	229	30	world	world	NOUN
easat-5713	229	31	datasets	dataset	NOUN
easat-5713	229	32	,	,	PUNCT
easat-5713	229	33	such	such	ADJ
easat-5713	229	34	as	as	ADP
easat-5713	229	35	class	class	NOUN
easat-5713	229	36	imbalances	imbalance	NOUN
easat-5713	229	37	,	,	PUNCT
easat-5713	229	38	noise	noise	NOUN
easat-5713	229	39	,	,	PUNCT
easat-5713	229	40	and	and	CCONJ
easat-5713	229	41	limited	limited	ADJ
easat-5713	229	42	diversity	diversity	NOUN
easat-5713	229	43	.	.	PUNCT
easat-5713	230	1	however	however	ADV
easat-5713	230	2	,	,	PUNCT
easat-5713	230	3	the	the	DET
easat-5713	230	4	results	result	NOUN
easat-5713	230	5	indicate	indicate	VERB
easat-5713	230	6	that	that	SCONJ
easat-5713	230	7	the	the	DET
easat-5713	230	8	inclusion	inclusion	NOUN
easat-5713	230	9	of	of	ADP
easat-5713	230	10	synthetic	synthetic	ADJ
easat-5713	230	11	images	image	NOUN
easat-5713	230	12	did	do	AUX
easat-5713	230	13	not	not	PART
easat-5713	230	14	significantly	significantly	ADV
easat-5713	230	15	improve	improve	VERB
easat-5713	230	16	the	the	DET
easat-5713	230	17	model	model	NOUN
easat-5713	230	18	’s	’s	PART
easat-5713	230	19	performance	performance	NOUN
easat-5713	230	20	.	.	PUNCT
easat-5713	231	1	instead	instead	ADV
easat-5713	231	2	,	,	PUNCT
easat-5713	231	3	the	the	DET
easat-5713	231	4	model	model	NOUN
easat-5713	231	5	trained	train	VERB
easat-5713	231	6	on	on	ADP
easat-5713	231	7	the	the	DET
easat-5713	231	8	augmented	augment	VERB
easat-5713	231	9	dataset	dataset	NOUN
easat-5713	231	10	achieved	achieve	VERB
easat-5713	231	11	slightly	slightly	ADV
easat-5713	231	12	lower	low	ADJ
easat-5713	231	13	accuracy	accuracy	NOUN
easat-5713	231	14	compared	compare	VERB
easat-5713	231	15	to	to	ADP
easat-5713	231	16	the	the	DET
easat-5713	231	17	one	one	NOUN
easat-5713	231	18	trained	train	VERB
easat-5713	231	19	on	on	ADP
easat-5713	231	20	the	the	DET
easat-5713	231	21	original	original	ADJ
easat-5713	231	22	fer2013	fer2013	ADJ
easat-5713	231	23	dataset	dataset	NOUN
easat-5713	231	24	.	.	PUNCT
easat-5713	232	1	one	one	NUM
easat-5713	232	2	of	of	ADP
easat-5713	232	3	the	the	DET
easat-5713	232	4	key	key	ADJ
easat-5713	232	5	findings	finding	NOUN
easat-5713	232	6	is	be	AUX
easat-5713	232	7	that	that	SCONJ
easat-5713	232	8	the	the	DET
easat-5713	232	9	effectiveness	effectiveness	NOUN
easat-5713	232	10	of	of	ADP
easat-5713	232	11	synthetic	synthetic	ADJ
easat-5713	232	12	data	datum	NOUN
easat-5713	232	13	for	for	ADP
easat-5713	232	14	fer	fer	PROPN
easat-5713	232	15	is	be	AUX
easat-5713	232	16	highly	highly	ADV
easat-5713	232	17	dependent	dependent	ADJ
easat-5713	232	18	on	on	ADP
easat-5713	232	19	the	the	DET
easat-5713	232	20	quality	quality	NOUN
easat-5713	232	21	and	and	CCONJ
easat-5713	232	22	realism	realism	NOUN
easat-5713	232	23	of	of	ADP
easat-5713	232	24	the	the	DET
easat-5713	232	25	generated	generate	VERB
easat-5713	232	26	images	image	NOUN
easat-5713	232	27	.	.	PUNCT
easat-5713	233	1	despite	despite	SCONJ
easat-5713	233	2	carefully	carefully	ADV
easat-5713	233	3	designed	design	VERB
easat-5713	233	4	prompts	prompt	NOUN
easat-5713	233	5	to	to	PART
easat-5713	233	6	capture	capture	VERB
easat-5713	233	7	diverse	diverse	ADJ
easat-5713	233	8	facial	facial	ADJ
easat-5713	233	9	1868	1868	NUM
easat-5713	233	10	edelweiss	edelweiss	PROPN
easat-5713	233	11	applied	apply	VERB
easat-5713	233	12	science	science	NOUN
easat-5713	233	13	and	and	CCONJ
easat-5713	233	14	technology	technology	NOUN
easat-5713	233	15	issn	issn	PROPN
easat-5713	233	16	:	:	PUNCT
easat-5713	233	17	2576	2576	NUM
easat-5713	233	18	-	-	SYM
easat-5713	233	19	8484	8484	NUM
easat-5713	233	20	vol	vol	NOUN
easat-5713	233	21	.	.	PROPN
easat-5713	233	22	9	9	NUM
easat-5713	233	23	,	,	PUNCT
easat-5713	233	24	no	no	INTJ
easat-5713	233	25	.	.	NOUN
easat-5713	233	26	3	3	NUM
easat-5713	233	27	:	:	PUNCT
easat-5713	233	28	1855	1855	NUM
easat-5713	233	29	-	-	SYM
easat-5713	233	30	1869	1869	NUM
easat-5713	233	31	,	,	PUNCT
easat-5713	233	32	2025	2025	NUM
easat-5713	233	33	doi	doi	NOUN
easat-5713	233	34	:	:	PUNCT
easat-5713	233	35	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	233	36	©	©	ADP
easat-5713	233	37	2025	2025	NUM
easat-5713	233	38	by	by	ADP
easat-5713	233	39	the	the	DET
easat-5713	233	40	authors	author	NOUN
easat-5713	233	41	;	;	PUNCT
easat-5713	233	42	licensee	licensee	PROPN
easat-5713	233	43	learning	learn	VERB
easat-5713	233	44	gate	gate	NOUN
easat-5713	233	45	expressions	expression	NOUN
easat-5713	233	46	and	and	CCONJ
easat-5713	233	47	demographic	demographic	ADJ
easat-5713	233	48	variations	variation	NOUN
easat-5713	233	49	,	,	PUNCT
easat-5713	233	50	the	the	DET
easat-5713	233	51	synthetic	synthetic	ADJ
easat-5713	233	52	images	image	NOUN
easat-5713	233	53	may	may	AUX
easat-5713	233	54	not	not	PART
easat-5713	233	55	have	have	AUX
easat-5713	233	56	sufficiently	sufficiently	ADV
easat-5713	233	57	replicated	replicate	VERB
easat-5713	233	58	the	the	DET
easat-5713	233	59	subtle	subtle	ADJ
easat-5713	233	60	complexities	complexity	NOUN
easat-5713	233	61	of	of	ADP
easat-5713	233	62	real	real	ADJ
easat-5713	233	63	-	-	PUNCT
easat-5713	233	64	world	world	NOUN
easat-5713	233	65	facial	facial	ADJ
easat-5713	233	66	expressions	expression	NOUN
easat-5713	233	67	.	.	PUNCT
easat-5713	234	1	this	this	PRON
easat-5713	234	2	suggests	suggest	VERB
easat-5713	234	3	that	that	SCONJ
easat-5713	234	4	,	,	PUNCT
easat-5713	234	5	while	while	SCONJ
easat-5713	234	6	synthetic	synthetic	ADJ
easat-5713	234	7	data	datum	NOUN
easat-5713	234	8	can	can	AUX
easat-5713	234	9	increase	increase	VERB
easat-5713	234	10	dataset	dataset	ADJ
easat-5713	234	11	diversity	diversity	NOUN
easat-5713	234	12	,	,	PUNCT
easat-5713	234	13	it	it	PRON
easat-5713	234	14	may	may	AUX
easat-5713	234	15	also	also	ADV
easat-5713	234	16	introduce	introduce	VERB
easat-5713	234	17	noise	noise	NOUN
easat-5713	234	18	if	if	SCONJ
easat-5713	234	19	the	the	DET
easat-5713	234	20	images	image	NOUN
easat-5713	234	21	lack	lack	VERB
easat-5713	234	22	fidelity	fidelity	NOUN
easat-5713	234	23	to	to	ADP
easat-5713	234	24	natural	natural	ADJ
easat-5713	234	25	human	human	ADJ
easat-5713	234	26	expressions	expression	NOUN
easat-5713	234	27	.	.	PUNCT
easat-5713	235	1	the	the	DET
easat-5713	235	2	training	training	NOUN
easat-5713	235	3	duration	duration	NOUN
easat-5713	235	4	was	be	AUX
easat-5713	235	5	limited	limit	VERB
easat-5713	235	6	to	to	ADP
easat-5713	235	7	25	25	NUM
easat-5713	235	8	epochs	epoch	NOUN
easat-5713	235	9	due	due	ADJ
easat-5713	235	10	to	to	ADP
easat-5713	235	11	hardware	hardware	NOUN
easat-5713	235	12	constraints	constraint	NOUN
easat-5713	235	13	,	,	PUNCT
easat-5713	235	14	which	which	PRON
easat-5713	235	15	may	may	AUX
easat-5713	235	16	have	have	AUX
easat-5713	235	17	affected	affect	VERB
easat-5713	235	18	the	the	DET
easat-5713	235	19	model	model	NOUN
easat-5713	235	20	’s	’s	PART
easat-5713	235	21	ability	ability	NOUN
easat-5713	235	22	to	to	PART
easat-5713	235	23	fully	fully	ADV
easat-5713	235	24	leverage	leverage	VERB
easat-5713	235	25	the	the	DET
easat-5713	235	26	synthetic	synthetic	ADJ
easat-5713	235	27	data	datum	NOUN
easat-5713	235	28	.	.	PUNCT
easat-5713	236	1	the	the	DET
easat-5713	236	2	number	number	NOUN
easat-5713	236	3	of	of	ADP
easat-5713	236	4	synthetic	synthetic	ADJ
easat-5713	236	5	images	image	NOUN
easat-5713	236	6	generated	generate	VERB
easat-5713	236	7	per	per	ADP
easat-5713	236	8	emotion	emotion	NOUN
easat-5713	236	9	category	category	NOUN
easat-5713	236	10	,	,	PUNCT
easat-5713	236	11	while	while	SCONJ
easat-5713	236	12	substantial	substantial	ADJ
easat-5713	236	13	,	,	PUNCT
easat-5713	236	14	may	may	AUX
easat-5713	236	15	also	also	ADV
easat-5713	236	16	not	not	PART
easat-5713	236	17	have	have	AUX
easat-5713	236	18	been	be	AUX
easat-5713	236	19	sufficient	sufficient	ADJ
easat-5713	236	20	to	to	PART
easat-5713	236	21	provide	provide	VERB
easat-5713	236	22	meaningful	meaningful	ADJ
easat-5713	236	23	improvements	improvement	NOUN
easat-5713	236	24	.	.	PUNCT
easat-5713	237	1	future	future	ADJ
easat-5713	237	2	research	research	NOUN
easat-5713	237	3	should	should	AUX
easat-5713	237	4	explore	explore	VERB
easat-5713	237	5	alternative	alternative	ADJ
easat-5713	237	6	generative	generative	ADJ
easat-5713	237	7	models	model	NOUN
easat-5713	237	8	,	,	PUNCT
easat-5713	237	9	such	such	ADJ
easat-5713	237	10	as	as	ADP
easat-5713	237	11	diffusion	diffusion	NOUN
easat-5713	237	12	models	model	NOUN
easat-5713	237	13	or	or	CCONJ
easat-5713	237	14	stylegan	stylegan	NOUN
easat-5713	237	15	,	,	PUNCT
easat-5713	237	16	which	which	PRON
easat-5713	237	17	have	have	AUX
easat-5713	237	18	demonstrated	demonstrate	VERB
easat-5713	237	19	the	the	DET
easat-5713	237	20	ability	ability	NOUN
easat-5713	237	21	to	to	PART
easat-5713	237	22	produce	produce	VERB
easat-5713	237	23	higher	high	ADJ
easat-5713	237	24	-	-	PUNCT
easat-5713	237	25	quality	quality	NOUN
easat-5713	237	26	,	,	PUNCT
easat-5713	237	27	more	more	ADV
easat-5713	237	28	realistic	realistic	ADJ
easat-5713	237	29	images	image	NOUN
easat-5713	237	30	.	.	PUNCT
easat-5713	238	1	optimizing	optimize	VERB
easat-5713	238	2	training	training	NOUN
easat-5713	238	3	strategies	strategy	NOUN
easat-5713	238	4	,	,	PUNCT
easat-5713	238	5	including	include	VERB
easat-5713	238	6	longer	long	ADJ
easat-5713	238	7	training	training	NOUN
easat-5713	238	8	periods	period	NOUN
easat-5713	238	9	,	,	PUNCT
easat-5713	238	10	fine	fine	ADV
easat-5713	238	11	-	-	PUNCT
easat-5713	238	12	tuning	tune	VERB
easat-5713	238	13	approaches	approach	NOUN
easat-5713	238	14	,	,	PUNCT
easat-5713	238	15	and	and	CCONJ
easat-5713	238	16	selective	selective	ADJ
easat-5713	238	17	filtering	filtering	NOUN
easat-5713	238	18	of	of	ADP
easat-5713	238	19	synthetic	synthetic	ADJ
easat-5713	238	20	images	image	NOUN
easat-5713	238	21	based	base	VERB
easat-5713	238	22	on	on	ADP
easat-5713	238	23	quality	quality	NOUN
easat-5713	238	24	assessment	assessment	NOUN
easat-5713	238	25	,	,	PUNCT
easat-5713	238	26	could	could	AUX
easat-5713	238	27	enhance	enhance	VERB
easat-5713	238	28	the	the	DET
easat-5713	238	29	effectiveness	effectiveness	NOUN
easat-5713	238	30	of	of	ADP
easat-5713	238	31	synthetic	synthetic	ADJ
easat-5713	238	32	data	datum	NOUN
easat-5713	238	33	augmentation	augmentation	NOUN
easat-5713	238	34	in	in	ADP
easat-5713	238	35	fer	fer	PROPN
easat-5713	238	36	.	.	PUNCT
easat-5713	239	1	another	another	DET
easat-5713	239	2	promising	promising	ADJ
easat-5713	239	3	avenue	avenue	NOUN
easat-5713	239	4	is	be	AUX
easat-5713	239	5	the	the	DET
easat-5713	239	6	development	development	NOUN
easat-5713	239	7	of	of	ADP
easat-5713	239	8	advanced	advanced	ADJ
easat-5713	239	9	techniques	technique	NOUN
easat-5713	239	10	such	such	ADJ
easat-5713	239	11	as	as	ADP
easat-5713	239	12	dreambooth	dreambooth	NOUN
easat-5713	239	13	to	to	ADP
easat-5713	239	14	fine	fine	ADJ
easat-5713	239	15	-	-	PUNCT
easat-5713	239	16	tune	tune	NOUN
easat-5713	239	17	diffusion	diffusion	NOUN
easat-5713	239	18	models	model	NOUN
easat-5713	239	19	for	for	ADP
easat-5713	239	20	datasets	dataset	NOUN
easat-5713	239	21	.	.	PUNCT
easat-5713	240	1	such	such	ADJ
easat-5713	240	2	methods	method	NOUN
easat-5713	240	3	could	could	AUX
easat-5713	240	4	allow	allow	VERB
easat-5713	240	5	models	model	NOUN
easat-5713	240	6	to	to	PART
easat-5713	240	7	learn	learn	VERB
easat-5713	240	8	from	from	ADP
easat-5713	240	9	a	a	DET
easat-5713	240	10	small	small	ADJ
easat-5713	240	11	number	number	NOUN
easat-5713	240	12	of	of	ADP
easat-5713	240	13	real	real	ADJ
easat-5713	240	14	images	image	NOUN
easat-5713	240	15	while	while	SCONJ
easat-5713	240	16	generating	generate	VERB
easat-5713	240	17	high	high	ADJ
easat-5713	240	18	-	-	PUNCT
easat-5713	240	19	fidelity	fidelity	NOUN
easat-5713	240	20	synthetic	synthetic	ADJ
easat-5713	240	21	data	datum	NOUN
easat-5713	240	22	that	that	PRON
easat-5713	240	23	more	more	ADV
easat-5713	240	24	accurately	accurately	ADV
easat-5713	240	25	mimics	mimic	VERB
easat-5713	240	26	real	real	ADJ
easat-5713	240	27	-	-	PUNCT
easat-5713	240	28	world	world	NOUN
easat-5713	240	29	variations	variation	NOUN
easat-5713	240	30	.	.	PUNCT
easat-5713	241	1	in	in	ADP
easat-5713	241	2	conclusion	conclusion	NOUN
easat-5713	241	3	,	,	PUNCT
easat-5713	241	4	the	the	DET
easat-5713	241	5	findings	finding	NOUN
easat-5713	241	6	from	from	ADP
easat-5713	241	7	this	this	DET
easat-5713	241	8	study	study	NOUN
easat-5713	241	9	suggest	suggest	VERB
easat-5713	241	10	that	that	SCONJ
easat-5713	241	11	further	further	ADJ
easat-5713	241	12	improvements	improvement	NOUN
easat-5713	241	13	in	in	ADP
easat-5713	241	14	synthetic	synthetic	ADJ
easat-5713	241	15	image	image	NOUN
easat-5713	241	16	generation	generation	NOUN
easat-5713	241	17	and	and	CCONJ
easat-5713	241	18	optimization	optimization	NOUN
easat-5713	241	19	are	be	AUX
easat-5713	241	20	necessary	necessary	ADJ
easat-5713	241	21	before	before	SCONJ
easat-5713	241	22	ai	ai	ADJ
easat-5713	241	23	-	-	PUNCT
easat-5713	241	24	generated	generate	VERB
easat-5713	241	25	data	datum	NOUN
easat-5713	241	26	can	can	AUX
easat-5713	241	27	fully	fully	ADV
easat-5713	241	28	replace	replace	VERB
easat-5713	241	29	or	or	CCONJ
easat-5713	241	30	significantly	significantly	ADV
easat-5713	241	31	enhance	enhance	VERB
easat-5713	241	32	traditional	traditional	ADJ
easat-5713	241	33	datasets	dataset	NOUN
easat-5713	241	34	.	.	PUNCT
easat-5713	242	1	transparency	transparency	NOUN
easat-5713	242	2	:	:	PUNCT
easat-5713	242	3	the	the	DET
easat-5713	242	4	authors	author	NOUN
easat-5713	242	5	confirm	confirm	VERB
easat-5713	242	6	that	that	SCONJ
easat-5713	242	7	the	the	DET
easat-5713	242	8	manuscript	manuscript	NOUN
easat-5713	242	9	is	be	AUX
easat-5713	242	10	an	an	DET
easat-5713	242	11	honest	honest	ADJ
easat-5713	242	12	,	,	PUNCT
easat-5713	242	13	accurate	accurate	ADJ
easat-5713	242	14	,	,	PUNCT
easat-5713	242	15	and	and	CCONJ
easat-5713	242	16	transparent	transparent	ADJ
easat-5713	242	17	account	account	NOUN
easat-5713	242	18	of	of	ADP
easat-5713	242	19	the	the	DET
easat-5713	242	20	study	study	NOUN
easat-5713	242	21	;	;	PUNCT
easat-5713	242	22	that	that	SCONJ
easat-5713	242	23	no	no	DET
easat-5713	242	24	vital	vital	ADJ
easat-5713	242	25	features	feature	NOUN
easat-5713	242	26	of	of	ADP
easat-5713	242	27	the	the	DET
easat-5713	242	28	study	study	NOUN
easat-5713	242	29	have	have	AUX
easat-5713	242	30	been	be	AUX
easat-5713	242	31	omitted	omit	VERB
easat-5713	242	32	;	;	PUNCT
easat-5713	242	33	and	and	CCONJ
easat-5713	242	34	that	that	SCONJ
easat-5713	242	35	any	any	DET
easat-5713	242	36	discrepancies	discrepancy	NOUN
easat-5713	242	37	from	from	ADP
easat-5713	242	38	the	the	DET
easat-5713	242	39	study	study	NOUN
easat-5713	242	40	as	as	SCONJ
easat-5713	242	41	planned	plan	VERB
easat-5713	242	42	have	have	AUX
easat-5713	242	43	been	be	AUX
easat-5713	242	44	explained	explain	VERB
easat-5713	242	45	.	.	PUNCT
easat-5713	243	1	this	this	DET
easat-5713	243	2	study	study	NOUN
easat-5713	243	3	followed	follow	VERB
easat-5713	243	4	all	all	DET
easat-5713	243	5	ethical	ethical	ADJ
easat-5713	243	6	practices	practice	NOUN
easat-5713	243	7	during	during	ADP
easat-5713	243	8	writing	writing	NOUN
easat-5713	243	9	.	.	PUNCT
easat-5713	244	1	copyright	copyright	NOUN
easat-5713	244	2	:	:	PUNCT
easat-5713	244	3	©	©	PROPN
easat-5713	244	4	2025	2025	NUM
easat-5713	244	5	by	by	ADP
easat-5713	244	6	the	the	DET
easat-5713	244	7	authors	author	NOUN
easat-5713	244	8	.	.	PUNCT
easat-5713	245	1	this	this	DET
easat-5713	245	2	open	open	ADJ
easat-5713	245	3	-	-	PUNCT
easat-5713	245	4	access	access	NOUN
easat-5713	245	5	article	article	NOUN
easat-5713	245	6	is	be	AUX
easat-5713	245	7	distributed	distribute	VERB
easat-5713	245	8	under	under	ADP
easat-5713	245	9	the	the	DET
easat-5713	245	10	terms	term	NOUN
easat-5713	245	11	and	and	CCONJ
easat-5713	245	12	conditions	condition	NOUN
easat-5713	245	13	of	of	ADP
easat-5713	245	14	the	the	DET
easat-5713	245	15	creative	creative	ADJ
easat-5713	245	16	commons	common	NOUN
easat-5713	245	17	attribution	attribution	NOUN
easat-5713	245	18	(	(	PUNCT
easat-5713	245	19	cc	cc	NOUN
easat-5713	245	20	by	by	ADP
easat-5713	245	21	)	)	PUNCT
easat-5713	245	22	license	license	NOUN
easat-5713	245	23	(	(	PUNCT
easat-5713	245	24	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-5713	245	25	)	)	PUNCT
easat-5713	245	26	.	.	PUNCT
easat-5713	246	1	references	reference	NOUN
easat-5713	246	2	[	[	X
easat-5713	246	3	1	1	NUM
easat-5713	246	4	]	]	PUNCT
easat-5713	246	5	m.	m.	NOUN
easat-5713	246	6	mattioli	mattioli	PROPN
easat-5713	246	7	and	and	CCONJ
easat-5713	246	8	f.	f.	PROPN
easat-5713	246	9	cabitza	cabitza	PROPN
easat-5713	246	10	,	,	PUNCT
easat-5713	246	11	"	"	PUNCT
easat-5713	246	12	not	not	PART
easat-5713	246	13	in	in	ADP
easat-5713	246	14	my	my	PRON
easat-5713	246	15	face	face	NOUN
easat-5713	246	16	:	:	PUNCT
easat-5713	246	17	challenges	challenge	NOUN
easat-5713	246	18	and	and	CCONJ
easat-5713	246	19	ethical	ethical	ADJ
easat-5713	246	20	considerations	consideration	NOUN
easat-5713	246	21	in	in	ADP
easat-5713	246	22	automatic	automatic	ADJ
easat-5713	246	23	face	face	NOUN
easat-5713	246	24	emotion	emotion	NOUN
easat-5713	246	25	recognition	recognition	NOUN
easat-5713	246	26	technology	technology	NOUN
easat-5713	246	27	,	,	PUNCT
easat-5713	246	28	"	"	PUNCT
easat-5713	246	29	machine	machine	NOUN
easat-5713	246	30	learning	learning	NOUN
easat-5713	246	31	and	and	CCONJ
easat-5713	246	32	knowledge	knowledge	NOUN
easat-5713	246	33	extraction	extraction	NOUN
easat-5713	246	34	,	,	PUNCT
easat-5713	246	35	vol	vol	NOUN
easat-5713	246	36	.	.	PROPN
easat-5713	246	37	6	6	NUM
easat-5713	246	38	,	,	PUNCT
easat-5713	246	39	no	no	INTJ
easat-5713	246	40	.	.	NOUN
easat-5713	246	41	4	4	NUM
easat-5713	246	42	,	,	PUNCT
easat-5713	246	43	pp	pp	ADJ
easat-5713	246	44	.	.	PUNCT
easat-5713	247	1	2201	2201	NUM
easat-5713	247	2	-	-	SYM
easat-5713	247	3	2231	2231	NUM
easat-5713	247	4	,	,	PUNCT
easat-5713	247	5	2024	2024	NUM
easat-5713	247	6	.	.	PUNCT
easat-5713	248	1	https://doi.org/10.3390/make6040109	https://doi.org/10.3390/make6040109	PROPN
easat-5713	249	1	[	[	X
easat-5713	249	2	2	2	NUM
easat-5713	249	3	]	]	PUNCT
easat-5713	249	4	j.	j.	PROPN
easat-5713	249	5	le	le	PROPN
easat-5713	249	6	ngwe	ngwe	PROPN
easat-5713	249	7	,	,	PUNCT
easat-5713	249	8	k.	k.	PROPN
easat-5713	249	9	m.	m.	PROPN
easat-5713	249	10	lim	lim	PROPN
easat-5713	249	11	,	,	PUNCT
easat-5713	249	12	c.	c.	PROPN
easat-5713	249	13	p.	p.	PROPN
easat-5713	249	14	lee	lee	PROPN
easat-5713	249	15	,	,	PUNCT
easat-5713	249	16	t.	t.	PROPN
easat-5713	249	17	s.	s.	PROPN
easat-5713	249	18	ong	ong	PROPN
easat-5713	249	19	,	,	PUNCT
easat-5713	249	20	and	and	CCONJ
easat-5713	249	21	a.	a.	PROPN
easat-5713	249	22	alqahtani	alqahtani	PROPN
easat-5713	249	23	,	,	PUNCT
easat-5713	249	24	"	"	PUNCT
easat-5713	249	25	patt	patt	NOUN
easat-5713	249	26	-	-	PUNCT
easat-5713	249	27	lite	lite	NOUN
easat-5713	249	28	:	:	PUNCT
easat-5713	249	29	lightweight	lightweight	ADJ
easat-5713	249	30	patch	patch	NOUN
easat-5713	249	31	and	and	CCONJ
easat-5713	249	32	attention	attention	NOUN
easat-5713	249	33	mobilenet	mobilenet	NOUN
easat-5713	249	34	for	for	ADP
easat-5713	249	35	challenging	challenge	VERB
easat-5713	249	36	facial	facial	ADJ
easat-5713	249	37	expression	expression	NOUN
easat-5713	249	38	recognition	recognition	NOUN
easat-5713	249	39	,	,	PUNCT
easat-5713	249	40	"	"	PUNCT
easat-5713	249	41	ieee	ieee	NOUN
easat-5713	249	42	access	access	NOUN
easat-5713	249	43	,	,	PUNCT
easat-5713	249	44	vol	vol	NOUN
easat-5713	249	45	.	.	PROPN
easat-5713	249	46	12	12	NUM
easat-5713	249	47	,	,	PUNCT
easat-5713	249	48	pp	pp	ADJ
easat-5713	249	49	.	.	PUNCT
easat-5713	249	50	79327–79341	79327–79341	NUM
easat-5713	249	51	,	,	PUNCT
easat-5713	249	52	2024	2024	NUM
easat-5713	249	53	.	.	PUNCT
easat-5713	250	1	https://doi.org/10.1109/access.2024.3407108	https://doi.org/10.1109/access.2024.3407108	PROPN
easat-5713	251	1	[	[	X
easat-5713	251	2	3	3	X
easat-5713	251	3	]	]	PUNCT
easat-5713	251	4	s.	s.	PROPN
easat-5713	251	5	li	li	PROPN
easat-5713	251	6	and	and	CCONJ
easat-5713	251	7	w.	w.	PROPN
easat-5713	251	8	deng	deng	PROPN
easat-5713	251	9	,	,	PUNCT
easat-5713	251	10	"	"	PUNCT
easat-5713	251	11	deep	deep	ADJ
easat-5713	251	12	facial	facial	ADJ
easat-5713	251	13	expression	expression	NOUN
easat-5713	251	14	recognition	recognition	NOUN
easat-5713	251	15	:	:	PUNCT
easat-5713	251	16	a	a	DET
easat-5713	251	17	survey	survey	NOUN
easat-5713	251	18	,	,	PUNCT
easat-5713	251	19	"	"	PUNCT
easat-5713	251	20	ieee	ieee	NOUN
easat-5713	251	21	transactions	transaction	NOUN
easat-5713	251	22	on	on	ADP
easat-5713	251	23	affective	affective	ADJ
easat-5713	251	24	computing	computing	NOUN
easat-5713	251	25	,	,	PUNCT
easat-5713	251	26	vol	vol	NOUN
easat-5713	251	27	.	.	PROPN
easat-5713	251	28	13	13	NUM
easat-5713	251	29	,	,	PUNCT
easat-5713	251	30	no	no	INTJ
easat-5713	251	31	.	.	NOUN
easat-5713	251	32	3	3	NUM
easat-5713	251	33	,	,	PUNCT
easat-5713	251	34	pp	pp	ADJ
easat-5713	251	35	.	.	PUNCT
easat-5713	252	1	1195	1195	NUM
easat-5713	252	2	-	-	SYM
easat-5713	252	3	1215	1215	NUM
easat-5713	252	4	,	,	PUNCT
easat-5713	252	5	2020	2020	NUM
easat-5713	252	6	.	.	PUNCT
easat-5713	253	1	https://doi.org/10.1109/taffc.2020.2981446	https://doi.org/10.1109/taffc.2020.2981446	PROPN
easat-5713	254	1	[	[	X
easat-5713	254	2	4	4	NUM
easat-5713	254	3	]	]	PUNCT
easat-5713	254	4	r.	r.	NOUN
easat-5713	254	5	he	he	PRON
easat-5713	254	6	et	et	PROPN
easat-5713	254	7	al	al	PROPN
easat-5713	254	8	.	.	PROPN
easat-5713	254	9	,	,	PUNCT
easat-5713	254	10	"	"	PUNCT
easat-5713	254	11	is	be	AUX
easat-5713	254	12	synthetic	synthetic	ADJ
easat-5713	254	13	data	datum	NOUN
easat-5713	254	14	from	from	ADP
easat-5713	254	15	generative	generative	ADJ
easat-5713	254	16	models	model	NOUN
easat-5713	254	17	ready	ready	ADJ
easat-5713	254	18	for	for	ADP
easat-5713	254	19	image	image	NOUN
easat-5713	254	20	recognition	recognition	NOUN
easat-5713	254	21	?	?	PUNCT
easat-5713	254	22	,	,	PUNCT
easat-5713	254	23	"	"	PUNCT
easat-5713	254	24	arxiv	arxiv	PROPN
easat-5713	254	25	preprint	preprint	VERB
easat-5713	254	26	arxiv:2210.07574	arxiv:2210.07574	NOUN
easat-5713	254	27	,	,	PUNCT
easat-5713	254	28	2022	2022	NUM
easat-5713	254	29	.	.	PUNCT
easat-5713	255	1	https://doi.org/10.48550/arxiv.2210.07574	https://doi.org/10.48550/arxiv.2210.07574	PROPN
easat-5713	255	2	[	[	X
easat-5713	255	3	5	5	NUM
easat-5713	255	4	]	]	PUNCT
easat-5713	255	5	g.	g.	PROPN
easat-5713	255	6	bae	bae	PROPN
easat-5713	255	7	et	et	PROPN
easat-5713	255	8	al	al	PROPN
easat-5713	255	9	.	.	PROPN
easat-5713	255	10	,	,	PUNCT
easat-5713	255	11	"	"	PUNCT
easat-5713	255	12	digiface-1	digiface-1	PROPN
easat-5713	255	13	m	m	NOUN
easat-5713	255	14	:	:	PUNCT
easat-5713	255	15	1	1	NUM
easat-5713	255	16	million	million	NUM
easat-5713	255	17	digital	digital	ADJ
easat-5713	255	18	face	face	NOUN
easat-5713	255	19	images	image	NOUN
easat-5713	255	20	for	for	ADP
easat-5713	255	21	face	face	NOUN
easat-5713	255	22	recognition	recognition	NOUN
easat-5713	255	23	,	,	PUNCT
easat-5713	255	24	"	"	PUNCT
easat-5713	255	25	in	in	ADP
easat-5713	255	26	proceedings	proceeding	NOUN
easat-5713	255	27	of	of	ADP
easat-5713	255	28	the	the	DET
easat-5713	255	29	ieee	ieee	NOUN
easat-5713	255	30	/	/	SYM
easat-5713	255	31	cvf	cvf	NOUN
easat-5713	255	32	winter	winter	NOUN
easat-5713	255	33	conference	conference	NOUN
easat-5713	255	34	on	on	ADP
easat-5713	255	35	applications	application	NOUN
easat-5713	255	36	of	of	ADP
easat-5713	255	37	computer	computer	NOUN
easat-5713	255	38	vision	vision	NOUN
easat-5713	255	39	,	,	PUNCT
easat-5713	255	40	2023	2023	NUM
easat-5713	255	41	,	,	PUNCT
easat-5713	255	42	pp	pp	ADJ
easat-5713	255	43	.	.	PUNCT
easat-5713	256	1	3526	3526	NUM
easat-5713	256	2	-	-	SYM
easat-5713	256	3	3535	3535	NUM
easat-5713	256	4	.	.	PUNCT
easat-5713	257	1	[	[	X
easat-5713	257	2	6	6	NUM
easat-5713	257	3	]	]	PUNCT
easat-5713	257	4	g.	g.	PROPN
easat-5713	257	5	raut	raut	PROPN
easat-5713	257	6	and	and	CCONJ
easat-5713	257	7	a.	a.	PROPN
easat-5713	257	8	singh	singh	PROPN
easat-5713	257	9	,	,	PUNCT
easat-5713	257	10	"	"	PUNCT
easat-5713	257	11	generative	generative	ADJ
easat-5713	257	12	ai	ai	VERB
easat-5713	257	13	in	in	ADP
easat-5713	257	14	vision	vision	NOUN
easat-5713	257	15	:	:	PUNCT
easat-5713	257	16	a	a	DET
easat-5713	257	17	survey	survey	NOUN
easat-5713	257	18	on	on	ADP
easat-5713	257	19	models	model	NOUN
easat-5713	257	20	,	,	PUNCT
easat-5713	257	21	metrics	metric	NOUN
easat-5713	257	22	and	and	CCONJ
easat-5713	257	23	applications	application	NOUN
easat-5713	257	24	,	,	PUNCT
easat-5713	257	25	"	"	PUNCT
easat-5713	257	26	arxiv	arxiv	PROPN
easat-5713	257	27	preprint	preprint	NOUN
easat-5713	257	28	arxiv:2402.16369	arxiv:2402.16369	PROPN
easat-5713	257	29	,	,	PUNCT
easat-5713	257	30	2024	2024	NUM
easat-5713	257	31	.	.	PUNCT
easat-5713	258	1	https://doi.org/10.48550/arxiv.2402.16369	https://doi.org/10.48550/arxiv.2402.16369	PROPN
easat-5713	258	2	[	[	X
easat-5713	258	3	7	7	NUM
easat-5713	258	4	]	]	X
easat-5713	258	5	i.	i.	PROPN
easat-5713	258	6	goodfellow	goodfellow	PROPN
easat-5713	258	7	et	et	PROPN
easat-5713	258	8	al	al	PROPN
easat-5713	258	9	.	.	PROPN
easat-5713	258	10	,	,	PUNCT
easat-5713	258	11	"	"	PUNCT
easat-5713	258	12	generative	generative	ADJ
easat-5713	258	13	adversarial	adversarial	ADJ
easat-5713	258	14	networks	network	NOUN
easat-5713	258	15	,	,	PUNCT
easat-5713	258	16	"	"	PUNCT
easat-5713	258	17	communications	communication	NOUN
easat-5713	258	18	of	of	ADP
easat-5713	258	19	the	the	DET
easat-5713	258	20	acm	acm	NOUN
easat-5713	258	21	,	,	PUNCT
easat-5713	258	22	vol	vol	NOUN
easat-5713	258	23	.	.	PROPN
easat-5713	258	24	63	63	NUM
easat-5713	258	25	,	,	PUNCT
easat-5713	258	26	no	no	INTJ
easat-5713	258	27	.	.	NOUN
easat-5713	258	28	11	11	NUM
easat-5713	258	29	,	,	PUNCT
easat-5713	258	30	pp	pp	ADJ
easat-5713	258	31	.	.	PUNCT
easat-5713	259	1	139	139	NUM
easat-5713	259	2	-	-	SYM
easat-5713	259	3	144	144	NUM
easat-5713	259	4	,	,	PUNCT
easat-5713	259	5	2020	2020	NUM
easat-5713	259	6	.	.	PUNCT
easat-5713	260	1	https://doi.org/10.1145/3422622	https://doi.org/10.1145/3422622	ADJ
easat-5713	260	2	[	[	X
easat-5713	260	3	8	8	X
easat-5713	260	4	]	]	PUNCT
easat-5713	260	5	t.	t.	PROPN
easat-5713	260	6	karras	karras	PROPN
easat-5713	260	7	,	,	PUNCT
easat-5713	260	8	s.	s.	PROPN
easat-5713	260	9	laine	laine	PROPN
easat-5713	260	10	,	,	PUNCT
easat-5713	260	11	and	and	CCONJ
easat-5713	260	12	t.	t.	PROPN
easat-5713	260	13	aila	aila	PROPN
easat-5713	260	14	,	,	PUNCT
easat-5713	260	15	"	"	PUNCT
easat-5713	260	16	a	a	DET
easat-5713	260	17	style	style	NOUN
easat-5713	260	18	-	-	PUNCT
easat-5713	260	19	based	base	VERB
easat-5713	260	20	generator	generator	NOUN
easat-5713	260	21	architecture	architecture	NOUN
easat-5713	260	22	for	for	ADP
easat-5713	260	23	generative	generative	ADJ
easat-5713	260	24	adversarial	adversarial	ADJ
easat-5713	260	25	networks	network	NOUN
easat-5713	260	26	,	,	PUNCT
easat-5713	260	27	"	"	PUNCT
easat-5713	260	28	in	in	ADP
easat-5713	260	29	proceedings	proceeding	NOUN
easat-5713	260	30	of	of	ADP
easat-5713	260	31	the	the	DET
easat-5713	260	32	ieee	ieee	NOUN
easat-5713	260	33	/	/	SYM
easat-5713	260	34	cvf	cvf	NOUN
easat-5713	260	35	conference	conference	NOUN
easat-5713	260	36	on	on	ADP
easat-5713	260	37	computer	computer	NOUN
easat-5713	260	38	vision	vision	NOUN
easat-5713	260	39	and	and	CCONJ
easat-5713	260	40	pattern	pattern	NOUN
easat-5713	260	41	recognition	recognition	NOUN
easat-5713	260	42	,	,	PUNCT
easat-5713	260	43	2019	2019	NUM
easat-5713	260	44	,	,	PUNCT
easat-5713	260	45	pp	pp	ADJ
easat-5713	260	46	.	.	PUNCT
easat-5713	260	47	4401	4401	NUM
easat-5713	260	48	-	-	SYM
easat-5713	260	49	4410	4410	NUM
easat-5713	260	50	.	.	PUNCT
easat-5713	261	1	[	[	X
easat-5713	261	2	9	9	NUM
easat-5713	261	3	]	]	PUNCT
easat-5713	261	4	t.	t.	PROPN
easat-5713	261	5	karras	karras	PROPN
easat-5713	261	6	,	,	PUNCT
easat-5713	261	7	s.	s.	PROPN
easat-5713	261	8	laine	laine	PROPN
easat-5713	261	9	,	,	PUNCT
easat-5713	261	10	m.	m.	NOUN
easat-5713	261	11	aittala	aittala	PROPN
easat-5713	261	12	,	,	PUNCT
easat-5713	261	13	j.	j.	PROPN
easat-5713	261	14	hellsten	hellsten	PROPN
easat-5713	261	15	,	,	PUNCT
easat-5713	261	16	j.	j.	PROPN
easat-5713	261	17	lehtinen	lehtinen	PROPN
easat-5713	261	18	,	,	PUNCT
easat-5713	261	19	and	and	CCONJ
easat-5713	261	20	t.	t.	PROPN
easat-5713	261	21	aila	aila	PROPN
easat-5713	261	22	,	,	PUNCT
easat-5713	261	23	"	"	PUNCT
easat-5713	261	24	analyzing	analyze	VERB
easat-5713	261	25	and	and	CCONJ
easat-5713	261	26	improving	improve	VERB
easat-5713	261	27	the	the	DET
easat-5713	261	28	image	image	NOUN
easat-5713	261	29	quality	quality	NOUN
easat-5713	261	30	of	of	ADP
easat-5713	261	31	stylegan	stylegan	NOUN
easat-5713	261	32	,	,	PUNCT
easat-5713	261	33	"	"	PUNCT
easat-5713	261	34	in	in	ADP
easat-5713	261	35	proceedings	proceeding	NOUN
easat-5713	261	36	of	of	ADP
easat-5713	261	37	the	the	DET
easat-5713	261	38	ieee	ieee	NOUN
easat-5713	261	39	/	/	SYM
easat-5713	261	40	cvf	cvf	NOUN
easat-5713	261	41	conference	conference	NOUN
easat-5713	261	42	on	on	ADP
easat-5713	261	43	computer	computer	NOUN
easat-5713	261	44	vision	vision	NOUN
easat-5713	261	45	and	and	CCONJ
easat-5713	261	46	pattern	pattern	NOUN
easat-5713	261	47	recognition	recognition	NOUN
easat-5713	261	48	,	,	PUNCT
easat-5713	261	49	2020	2020	NUM
easat-5713	261	50	,	,	PUNCT
easat-5713	261	51	pp	pp	ADP
easat-5713	261	52	.	.	PUNCT
easat-5713	261	53	8110	8110	NUM
easat-5713	261	54	-	-	SYM
easat-5713	261	55	8119	8119	NUM
easat-5713	261	56	.	.	PUNCT
easat-5713	262	1	[	[	X
easat-5713	262	2	10	10	NUM
easat-5713	262	3	]	]	PUNCT
easat-5713	262	4	t.	t.	PROPN
easat-5713	262	5	c.	c.	PROPN
easat-5713	262	6	ma	ma	PROPN
easat-5713	262	7	and	and	CCONJ
easat-5713	262	8	d.	d.	PROPN
easat-5713	262	9	h.	h.	PROPN
easat-5713	262	10	nguyen	nguyen	PROPN
easat-5713	262	11	,	,	PUNCT
easat-5713	262	12	"	"	PUNCT
easat-5713	262	13	vn3dface	vn3dface	ADP
easat-5713	262	14	:	:	PUNCT
easat-5713	262	15	3d	3d	NUM
easat-5713	262	16	facial	facial	ADJ
easat-5713	262	17	expression	expression	NOUN
easat-5713	262	18	database	database	NOUN
easat-5713	262	19	of	of	ADP
easat-5713	262	20	vietnamese	vietnamese	PROPN
easat-5713	262	21	,	,	PUNCT
easat-5713	262	22	"	"	PUNCT
easat-5713	262	23	edelweiss	edelweiss	PROPN
easat-5713	262	24	applied	apply	VERB
easat-5713	262	25	science	science	NOUN
easat-5713	262	26	and	and	CCONJ
easat-5713	262	27	technology	technology	NOUN
easat-5713	262	28	,	,	PUNCT
easat-5713	262	29	vol	vol	NOUN
easat-5713	262	30	.	.	PROPN
easat-5713	262	31	9	9	NUM
easat-5713	262	32	,	,	PUNCT
easat-5713	262	33	no	no	INTJ
easat-5713	262	34	.	.	NOUN
easat-5713	262	35	1	1	NUM
easat-5713	262	36	,	,	PUNCT
easat-5713	262	37	pp	pp	ADJ
easat-5713	262	38	.	.	PUNCT
easat-5713	263	1	161	161	NUM
easat-5713	263	2	-	-	SYM
easat-5713	263	3	175	175	NUM
easat-5713	263	4	,	,	PUNCT
easat-5713	263	5	2025	2025	NUM
easat-5713	263	6	.	.	PUNCT
easat-5713	264	1	https://doi.org/10.55214/25768484.v9i1.3883	https://doi.org/10.55214/25768484.v9i1.3883	NOUN
easat-5713	265	1	[	[	X
easat-5713	265	2	11	11	NUM
easat-5713	265	3	]	]	PUNCT
easat-5713	265	4	x.	x.	NOUN
easat-5713	266	1	he	he	PRON
easat-5713	266	2	et	et	PROPN
easat-5713	266	3	al	al	PROPN
easat-5713	266	4	.	.	PROPN
easat-5713	266	5	,	,	PUNCT
easat-5713	266	6	"	"	PUNCT
easat-5713	266	7	synfer	synfer	NOUN
easat-5713	266	8	:	:	PUNCT
easat-5713	266	9	towards	towards	ADP
easat-5713	266	10	boosting	boost	VERB
easat-5713	266	11	facial	facial	ADJ
easat-5713	266	12	expression	expression	NOUN
easat-5713	266	13	recognition	recognition	NOUN
easat-5713	266	14	with	with	ADP
easat-5713	266	15	synthetic	synthetic	ADJ
easat-5713	266	16	data	datum	NOUN
easat-5713	266	17	,	,	PUNCT
easat-5713	266	18	"	"	PUNCT
easat-5713	266	19	arxiv	arxiv	PROPN
easat-5713	266	20	preprint	preprint	NOUN
easat-5713	266	21	arxiv:2410.09865	arxiv:2410.09865	ADV
easat-5713	266	22	,	,	PUNCT
easat-5713	266	23	2024	2024	NUM
easat-5713	266	24	.	.	PUNCT
easat-5713	267	1	https://doi.org/10.48550/arxiv.2410.09865	https://doi.org/10.48550/arxiv.2410.09865	PROPN
easat-5713	268	1	[	[	X
easat-5713	268	2	12	12	NUM
easat-5713	268	3	]	]	X
easat-5713	268	4	d.	d.	PROPN
easat-5713	268	5	meng	meng	PROPN
easat-5713	268	6	,	,	PUNCT
easat-5713	268	7	x.	x.	PROPN
easat-5713	268	8	peng	peng	PROPN
easat-5713	268	9	,	,	PUNCT
easat-5713	268	10	k.	k.	PROPN
easat-5713	268	11	wang	wang	PROPN
easat-5713	268	12	,	,	PUNCT
easat-5713	268	13	and	and	CCONJ
easat-5713	268	14	y.	y.	PROPN
easat-5713	268	15	qiao	qiao	PROPN
easat-5713	268	16	,	,	PUNCT
easat-5713	268	17	"	"	PUNCT
easat-5713	268	18	frame	frame	NOUN
easat-5713	268	19	attention	attention	NOUN
easat-5713	268	20	networks	network	NOUN
easat-5713	268	21	for	for	ADP
easat-5713	268	22	facial	facial	ADJ
easat-5713	268	23	expression	expression	NOUN
easat-5713	268	24	recognition	recognition	NOUN
easat-5713	268	25	in	in	ADP
easat-5713	268	26	videos	video	NOUN
easat-5713	268	27	,	,	PUNCT
easat-5713	268	28	"	"	PUNCT
easat-5713	268	29	in	in	ADP
easat-5713	268	30	2019	2019	NUM
easat-5713	268	31	ieee	ieee	NOUN
easat-5713	268	32	international	international	ADJ
easat-5713	268	33	conference	conference	NOUN
easat-5713	268	34	on	on	ADP
easat-5713	268	35	image	image	NOUN
easat-5713	268	36	processing	processing	NOUN
easat-5713	268	37	(	(	PUNCT
easat-5713	268	38	icip	icip	PROPN
easat-5713	268	39	)	)	PUNCT
easat-5713	268	40	,	,	PUNCT
easat-5713	268	41	2019	2019	NUM
easat-5713	268	42	:	:	PUNCT
easat-5713	268	43	ieee	ieee	NOUN
easat-5713	268	44	,	,	PUNCT
easat-5713	268	45	pp	pp	ADJ
easat-5713	268	46	.	.	PUNCT
easat-5713	268	47	3866	3866	NUM
easat-5713	268	48	-	-	SYM
easat-5713	268	49	3870	3870	NUM
easat-5713	268	50	.	.	PUNCT
easat-5713	269	1	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-5713	269	2	https://doi.org/10.3390/make6040109	https://doi.org/10.3390/make6040109	PROPN
easat-5713	269	3	https://doi.org/10.1109/access.2024.3407108	https://doi.org/10.1109/access.2024.3407108	PROPN
easat-5713	269	4	https://doi.org/10.1109/taffc.2020.2981446	https://doi.org/10.1109/taffc.2020.2981446	PROPN
easat-5713	269	5	https://doi.org/10.48550/arxiv.2210.07574	https://doi.org/10.48550/arxiv.2210.07574	PROPN
easat-5713	269	6	https://doi.org/10.48550/arxiv.2402.16369	https://doi.org/10.48550/arxiv.2402.16369	PROPN
easat-5713	269	7	https://doi.org/10.1145/3422622	https://doi.org/10.1145/3422622	ADJ
easat-5713	269	8	https://doi.org/10.55214/25768484.v9i1.3883	https://doi.org/10.55214/25768484.v9i1.3883	NOUN
easat-5713	269	9	https://doi.org/10.48550/arxiv.2410.09865	https://doi.org/10.48550/arxiv.2410.09865	PROPN
easat-5713	269	10	1869	1869	NUM
easat-5713	269	11	edelweiss	edelweiss	PROPN
easat-5713	269	12	applied	apply	VERB
easat-5713	269	13	science	science	NOUN
easat-5713	269	14	and	and	CCONJ
easat-5713	269	15	technology	technology	NOUN
easat-5713	269	16	issn	issn	PROPN
easat-5713	269	17	:	:	PUNCT
easat-5713	269	18	2576	2576	NUM
easat-5713	269	19	-	-	SYM
easat-5713	269	20	8484	8484	NUM
easat-5713	269	21	vol	vol	NOUN
easat-5713	269	22	.	.	PROPN
easat-5713	270	1	9	9	NUM
easat-5713	270	2	,	,	PUNCT
easat-5713	270	3	no	no	INTJ
easat-5713	270	4	.	.	NOUN
easat-5713	271	1	3	3	NUM
easat-5713	271	2	:	:	PUNCT
easat-5713	271	3	1855	1855	NUM
easat-5713	271	4	-	-	SYM
easat-5713	271	5	1869	1869	NUM
easat-5713	271	6	,	,	PUNCT
easat-5713	271	7	2025	2025	NUM
easat-5713	271	8	doi	doi	NOUN
easat-5713	271	9	:	:	PUNCT
easat-5713	271	10	10.55214/25768484.v9i3.5713	10.55214/25768484.v9i3.5713	NUM
easat-5713	271	11	©	©	ADP
easat-5713	271	12	2025	2025	NUM
easat-5713	271	13	by	by	ADP
easat-5713	271	14	the	the	DET
easat-5713	271	15	authors	author	NOUN
easat-5713	271	16	;	;	PUNCT
easat-5713	271	17	licensee	licensee	PROPN
easat-5713	271	18	learning	learning	NOUN
easat-5713	271	19	gate	gate	NOUN
easat-5713	272	1	[	[	X
easat-5713	272	2	13	13	NUM
easat-5713	272	3	]	]	PUNCT
easat-5713	272	4	a.	a.	NOUN
easat-5713	272	5	kartbayev	kartbayev	PROPN
easat-5713	272	6	,	,	PUNCT
easat-5713	272	7	"	"	PUNCT
easat-5713	272	8	learning	learn	VERB
easat-5713	272	9	word	word	NOUN
easat-5713	272	10	alignment	alignment	NOUN
easat-5713	272	11	models	model	NOUN
easat-5713	272	12	for	for	ADP
easat-5713	272	13	kazakh	kazakh	ADJ
easat-5713	272	14	-	-	PUNCT
easat-5713	272	15	english	english	ADJ
easat-5713	272	16	machine	machine	NOUN
easat-5713	272	17	translation	translation	NOUN
easat-5713	272	18	,	,	PUNCT
easat-5713	272	19	"	"	PUNCT
easat-5713	272	20	in	in	ADP
easat-5713	272	21	integrated	integrated	ADJ
easat-5713	272	22	uncertainty	uncertainty	NOUN
easat-5713	272	23	in	in	ADP
easat-5713	272	24	knowledge	knowledge	NOUN
easat-5713	272	25	modelling	modelling	NOUN
easat-5713	272	26	and	and	CCONJ
easat-5713	272	27	decision	decision	NOUN
easat-5713	272	28	making	making	NOUN
easat-5713	272	29	:	:	PUNCT
easat-5713	272	30	4th	4th	ADJ
easat-5713	272	31	international	international	ADJ
easat-5713	272	32	symposium	symposium	NOUN
easat-5713	272	33	,	,	PUNCT
easat-5713	272	34	iukm	iukm	NOUN
easat-5713	272	35	2015	2015	NUM
easat-5713	272	36	,	,	PUNCT
easat-5713	272	37	nha	nha	X
easat-5713	272	38	trang	trang	PROPN
easat-5713	272	39	,	,	PUNCT
easat-5713	272	40	vietnam	vietnam	PROPN
easat-5713	272	41	,	,	PUNCT
easat-5713	272	42	october	october	PROPN
easat-5713	272	43	15	15	NUM
easat-5713	272	44	-	-	SYM
easat-5713	272	45	17	17	NUM
easat-5713	272	46	,	,	PUNCT
easat-5713	272	47	2015	2015	NUM
easat-5713	272	48	,	,	PUNCT
easat-5713	272	49	proceedings	proceeding	NOUN
easat-5713	272	50	4	4	NUM
easat-5713	272	51	,	,	PUNCT
easat-5713	272	52	2015	2015	NUM
easat-5713	272	53	:	:	PUNCT
easat-5713	272	54	springer	springer	NOUN
easat-5713	272	55	,	,	PUNCT
easat-5713	272	56	pp	pp	ADJ
easat-5713	272	57	.	.	PUNCT
easat-5713	273	1	326	326	NUM
easat-5713	273	2	-	-	SYM
easat-5713	273	3	335	335	NUM
easat-5713	273	4	.	.	PUNCT
easat-5713	274	1	[	[	X
easat-5713	274	2	14	14	NUM
easat-5713	274	3	]	]	PUNCT
easat-5713	274	4	s.	s.	PROPN
easat-5713	274	5	ammar	ammar	PROPN
easat-5713	274	6	,	,	PUNCT
easat-5713	274	7	t.	t.	PROPN
easat-5713	274	8	bouwmans	bouwman	NOUN
easat-5713	274	9	,	,	PUNCT
easat-5713	274	10	n.	n.	PROPN
easat-5713	274	11	zaghden	zaghden	PROPN
easat-5713	274	12	,	,	PUNCT
easat-5713	274	13	and	and	CCONJ
easat-5713	274	14	m.	m.	NOUN
easat-5713	274	15	neji	neji	PROPN
easat-5713	274	16	,	,	PUNCT
easat-5713	274	17	"	"	PUNCT
easat-5713	274	18	towards	towards	ADP
easat-5713	274	19	an	an	DET
easat-5713	274	20	effective	effective	ADJ
easat-5713	274	21	approach	approach	NOUN
easat-5713	274	22	for	for	ADP
easat-5713	274	23	face	face	NOUN
easat-5713	274	24	recognition	recognition	NOUN
easat-5713	274	25	with	with	ADP
easat-5713	274	26	dcgans	dcgan	NOUN
easat-5713	274	27	data	datum	NOUN
easat-5713	274	28	augmentation	augmentation	NOUN
easat-5713	274	29	,	,	PUNCT
easat-5713	274	30	"	"	PUNCT
easat-5713	274	31	in	in	ADP
easat-5713	274	32	advances	advance	NOUN
easat-5713	274	33	in	in	ADP
easat-5713	274	34	visual	visual	ADJ
easat-5713	274	35	computing	computing	NOUN
easat-5713	274	36	:	:	PUNCT
easat-5713	274	37	15th	15th	ADJ
easat-5713	274	38	international	international	ADJ
easat-5713	274	39	symposium	symposium	NOUN
easat-5713	274	40	,	,	PUNCT
easat-5713	274	41	isvc	isvc	NOUN
easat-5713	274	42	2020	2020	NUM
easat-5713	274	43	,	,	PUNCT
easat-5713	274	44	san	san	PROPN
easat-5713	274	45	diego	diego	PROPN
easat-5713	274	46	,	,	PUNCT
easat-5713	274	47	ca	ca	PROPN
easat-5713	274	48	,	,	PUNCT
easat-5713	274	49	usa	usa	PROPN
easat-5713	274	50	,	,	PUNCT
easat-5713	274	51	october	october	PROPN
easat-5713	274	52	5–7	5–7	NUM
easat-5713	274	53	,	,	PUNCT
easat-5713	274	54	2020	2020	NUM
easat-5713	274	55	,	,	PUNCT
easat-5713	274	56	proceedings	proceeding	NOUN
easat-5713	274	57	,	,	PUNCT
easat-5713	274	58	part	part	NOUN
easat-5713	274	59	i	i	PRON
easat-5713	274	60	15	15	NUM
easat-5713	274	61	,	,	PUNCT
easat-5713	274	62	2020	2020	NUM
easat-5713	274	63	:	:	PUNCT
easat-5713	274	64	springer	springer	NOUN
easat-5713	274	65	,	,	PUNCT
easat-5713	274	66	pp	pp	ADJ
easat-5713	274	67	.	.	PUNCT
easat-5713	275	1	463	463	NUM
easat-5713	275	2	-	-	SYM
easat-5713	275	3	475	475	NUM
easat-5713	275	4	.	.	PUNCT
easat-5713	276	1	[	[	X
easat-5713	276	2	15	15	NUM
easat-5713	276	3	]	]	PUNCT
easat-5713	276	4	k.	k.	PROPN
easat-5713	276	5	kim	kim	PROPN
easat-5713	276	6	and	and	CCONJ
easat-5713	276	7	h.	h.	PROPN
easat-5713	276	8	park	park	PROPN
easat-5713	276	9	,	,	PUNCT
easat-5713	276	10	"	"	PUNCT
easat-5713	276	11	neural	neural	ADJ
easat-5713	276	12	network	network	NOUN
easat-5713	276	13	approach	approach	NOUN
easat-5713	276	14	to	to	PART
easat-5713	276	15	predict	predict	VERB
easat-5713	276	16	the	the	DET
easat-5713	276	17	association	association	NOUN
easat-5713	276	18	between	between	ADP
easat-5713	276	19	blood	blood	NOUN
easat-5713	276	20	cadmium	cadmium	NOUN
easat-5713	276	21	levels	level	NOUN
easat-5713	276	22	and	and	CCONJ
easat-5713	276	23	hypertension	hypertension	NOUN
easat-5713	276	24	,	,	PUNCT
easat-5713	276	25	"	"	PUNCT
easat-5713	276	26	edelweiss	edelweiss	PROPN
easat-5713	276	27	applied	apply	VERB
easat-5713	276	28	science	science	NOUN
easat-5713	276	29	and	and	CCONJ
easat-5713	276	30	technology	technology	NOUN
easat-5713	276	31	,	,	PUNCT
easat-5713	276	32	vol	vol	NOUN
easat-5713	276	33	.	.	PROPN
easat-5713	276	34	8	8	NUM
easat-5713	276	35	,	,	PUNCT
easat-5713	276	36	no	no	INTJ
easat-5713	276	37	.	.	NOUN
easat-5713	276	38	4	4	NUM
easat-5713	276	39	,	,	PUNCT
easat-5713	276	40	pp	pp	ADJ
easat-5713	276	41	.	.	PUNCT
easat-5713	276	42	336	336	NUM
easat-5713	276	43	-	-	SYM
easat-5713	276	44	344	344	NUM
easat-5713	276	45	,	,	PUNCT
easat-5713	276	46	2024	2024	NUM
easat-5713	276	47	.	.	PUNCT
easat-5713	276	48	https://doi.org/10.55214/25768484.v8i4.978	https://doi.org/10.55214/25768484.v8i4.978	X
easat-5713	277	1	[	[	X
easat-5713	277	2	16	16	NUM
easat-5713	277	3	]	]	X
easat-5713	277	4	n.	n.	PROPN
easat-5713	277	5	yalçin	yalçin	PROPN
easat-5713	277	6	and	and	CCONJ
easat-5713	277	7	m.	m.	NOUN
easat-5713	277	8	alisawi	alisawi	PROPN
easat-5713	277	9	,	,	PUNCT
easat-5713	277	10	"	"	PUNCT
easat-5713	277	11	introducing	introduce	VERB
easat-5713	277	12	a	a	DET
easat-5713	277	13	novel	novel	NOUN
easat-5713	277	14	dataset	dataset	NOUN
easat-5713	277	15	for	for	ADP
easat-5713	277	16	facial	facial	ADJ
easat-5713	277	17	emotion	emotion	NOUN
easat-5713	277	18	recognition	recognition	NOUN
easat-5713	277	19	and	and	CCONJ
easat-5713	277	20	demonstrating	demonstrate	VERB
easat-5713	277	21	significant	significant	ADJ
easat-5713	277	22	enhancements	enhancement	NOUN
easat-5713	277	23	in	in	ADP
easat-5713	277	24	deep	deep	ADJ
easat-5713	277	25	learning	learning	NOUN
easat-5713	277	26	performance	performance	NOUN
easat-5713	277	27	through	through	ADP
easat-5713	277	28	pre	pre	ADJ
easat-5713	277	29	-	-	ADJ
easat-5713	277	30	processing	processing	ADJ
easat-5713	277	31	techniques	technique	NOUN
easat-5713	277	32	,	,	PUNCT
easat-5713	277	33	"	"	PUNCT
easat-5713	277	34	heliyon	heliyon	NOUN
easat-5713	277	35	,	,	PUNCT
easat-5713	277	36	vol	vol	NOUN
easat-5713	277	37	.	.	PROPN
easat-5713	277	38	10	10	NUM
easat-5713	277	39	,	,	PUNCT
easat-5713	277	40	no	no	INTJ
easat-5713	277	41	.	.	NOUN
easat-5713	277	42	20	20	NUM
easat-5713	277	43	,	,	PUNCT
easat-5713	277	44	p.	p.	NOUN
easat-5713	277	45	e38913	e38913	PROPN
easat-5713	277	46	,	,	PUNCT
easat-5713	277	47	2024	2024	NUM
easat-5713	277	48	.	.	PUNCT
easat-5713	278	1	https://doi.org/10.1016/j.heliyon.2024.e38913	https://doi.org/10.1016/j.heliyon.2024.e38913	PROPN
easat-5713	279	1	[	[	X
easat-5713	279	2	17	17	NUM
easat-5713	279	3	]	]	PUNCT
easat-5713	279	4	m.	m.	NOUN
easat-5713	279	5	sajjad	sajjad	PROPN
easat-5713	279	6	et	et	PROPN
easat-5713	279	7	al	al	PROPN
easat-5713	279	8	.	.	PROPN
easat-5713	279	9	,	,	PUNCT
easat-5713	279	10	"	"	PUNCT
easat-5713	279	11	a	a	DET
easat-5713	279	12	comprehensive	comprehensive	ADJ
easat-5713	279	13	survey	survey	NOUN
easat-5713	279	14	on	on	ADP
easat-5713	279	15	deep	deep	ADJ
easat-5713	279	16	facial	facial	ADJ
easat-5713	279	17	expression	expression	NOUN
easat-5713	279	18	recognition	recognition	NOUN
easat-5713	279	19	:	:	PUNCT
easat-5713	279	20	challenges	challenge	NOUN
easat-5713	279	21	,	,	PUNCT
easat-5713	279	22	applications	application	NOUN
easat-5713	279	23	,	,	PUNCT
easat-5713	279	24	and	and	CCONJ
easat-5713	279	25	future	future	ADJ
easat-5713	279	26	guidelines	guideline	NOUN
easat-5713	279	27	,	,	PUNCT
easat-5713	279	28	"	"	PUNCT
easat-5713	279	29	alexandria	alexandria	PROPN
easat-5713	279	30	engineering	engineering	PROPN
easat-5713	279	31	journal	journal	PROPN
easat-5713	279	32	,	,	PUNCT
easat-5713	279	33	vol	vol	NOUN
easat-5713	279	34	.	.	PROPN
easat-5713	279	35	68	68	NUM
easat-5713	279	36	,	,	PUNCT
easat-5713	279	37	pp	pp	ADJ
easat-5713	279	38	.	.	PUNCT
easat-5713	279	39	817	817	NUM
easat-5713	279	40	-	-	SYM
easat-5713	279	41	840	840	NUM
easat-5713	279	42	,	,	PUNCT
easat-5713	279	43	2023	2023	NUM
easat-5713	279	44	.	.	PUNCT
easat-5713	280	1	https://doi.org/10.1016/j.aej.2023.01.017	https://doi.org/10.1016/j.aej.2023.01.017	X
easat-5713	280	2	[	[	X
easat-5713	280	3	18	18	NUM
easat-5713	280	4	]	]	PUNCT
easat-5713	280	5	p.	p.	NOUN
easat-5713	280	6	esser	esser	PROPN
easat-5713	280	7	et	et	PROPN
easat-5713	280	8	al	al	PROPN
easat-5713	280	9	.	.	PROPN
easat-5713	280	10	,	,	PUNCT
easat-5713	280	11	"	"	PUNCT
easat-5713	280	12	scaling	scale	VERB
easat-5713	280	13	rectified	rectified	ADJ
easat-5713	280	14	flow	flow	NOUN
easat-5713	280	15	transformers	transformer	NOUN
easat-5713	280	16	for	for	ADP
easat-5713	280	17	high	high	ADJ
easat-5713	280	18	-	-	PUNCT
easat-5713	280	19	resolution	resolution	NOUN
easat-5713	280	20	image	image	NOUN
easat-5713	280	21	synthesis	synthesis	NOUN
easat-5713	280	22	,	,	PUNCT
easat-5713	280	23	"	"	PUNCT
easat-5713	280	24	in	in	ADP
easat-5713	280	25	proceedings	proceeding	NOUN
easat-5713	280	26	of	of	ADP
easat-5713	280	27	the	the	DET
easat-5713	280	28	41st	41st	ADJ
easat-5713	280	29	international	international	ADJ
easat-5713	280	30	conference	conference	NOUN
easat-5713	280	31	on	on	ADP
easat-5713	280	32	machine	machine	NOUN
easat-5713	280	33	learning	learning	NOUN
easat-5713	280	34	(	(	PUNCT
easat-5713	280	35	icml'24	icml'24	PROPN
easat-5713	280	36	)	)	PUNCT
easat-5713	280	37	,	,	PUNCT
easat-5713	280	38	2024	2024	NUM
easat-5713	280	39	.	.	PUNCT
easat-5713	281	1	[	[	X
easat-5713	281	2	19	19	NUM
easat-5713	281	3	]	]	X
easat-5713	281	4	s.	s.	PROPN
easat-5713	281	5	vignesh	vignesh	PROPN
easat-5713	281	6	,	,	PUNCT
easat-5713	281	7	m.	m.	NOUN
easat-5713	281	8	savithadevi	savithadevi	NOUN
easat-5713	281	9	,	,	PUNCT
easat-5713	281	10	m.	m.	NOUN
easat-5713	281	11	sridevi	sridevi	NOUN
easat-5713	281	12	,	,	PUNCT
easat-5713	281	13	and	and	CCONJ
easat-5713	281	14	r.	r.	PROPN
easat-5713	281	15	sridhar	sridhar	PROPN
easat-5713	281	16	,	,	PUNCT
easat-5713	281	17	"	"	PUNCT
easat-5713	281	18	a	a	DET
easat-5713	281	19	novel	novel	ADJ
easat-5713	281	20	facial	facial	ADJ
easat-5713	281	21	emotion	emotion	NOUN
easat-5713	281	22	recognition	recognition	NOUN
easat-5713	281	23	model	model	NOUN
easat-5713	281	24	using	use	VERB
easat-5713	281	25	segmentation	segmentation	NOUN
easat-5713	281	26	vgg-19	vgg-19	NOUN
easat-5713	281	27	architecture	architecture	NOUN
easat-5713	281	28	,	,	PUNCT
easat-5713	281	29	"	"	PUNCT
easat-5713	281	30	international	international	ADJ
easat-5713	281	31	journal	journal	NOUN
easat-5713	281	32	of	of	ADP
easat-5713	281	33	information	information	NOUN
easat-5713	281	34	technology	technology	NOUN
easat-5713	281	35	,	,	PUNCT
easat-5713	281	36	vol	vol	NOUN
easat-5713	281	37	.	.	PROPN
easat-5713	281	38	15	15	NUM
easat-5713	281	39	,	,	PUNCT
easat-5713	281	40	no	no	INTJ
easat-5713	281	41	.	.	NOUN
easat-5713	281	42	4	4	NUM
easat-5713	281	43	,	,	PUNCT
easat-5713	281	44	pp	pp	ADJ
easat-5713	281	45	.	.	PUNCT
easat-5713	282	1	1777	1777	NUM
easat-5713	282	2	-	-	SYM
easat-5713	282	3	1787	1787	NUM
easat-5713	282	4	,	,	PUNCT
easat-5713	282	5	2023	2023	NUM
easat-5713	282	6	.	.	PUNCT
easat-5713	283	1	https://doi.org/10.1007/s41870-023-01184-z	https://doi.org/10.1007/s41870-023-01184-z	PROPN
easat-5713	284	1	[	[	X
easat-5713	284	2	20	20	NUM
easat-5713	284	3	]	]	PUNCT
easat-5713	285	1	p.	p.	NOUN
easat-5713	285	2	k.	k.	PROPN
easat-5713	286	1	sahoo	sahoo	PROPN
easat-5713	286	2	et	et	PROPN
easat-5713	286	3	al	al	PROPN
easat-5713	286	4	.	.	PROPN
easat-5713	286	5	,	,	PUNCT
easat-5713	286	6	"	"	PUNCT
easat-5713	286	7	an	an	DET
easat-5713	286	8	improved	improved	ADJ
easat-5713	286	9	vgg-19	vgg-19	NOUN
easat-5713	286	10	network	network	NOUN
easat-5713	286	11	induced	induce	VERB
easat-5713	286	12	enhanced	enhanced	ADJ
easat-5713	286	13	feature	feature	NOUN
easat-5713	286	14	pooling	pool	VERB
easat-5713	286	15	for	for	ADP
easat-5713	286	16	precise	precise	ADJ
easat-5713	286	17	moving	moving	NOUN
easat-5713	286	18	object	object	NOUN
easat-5713	286	19	detection	detection	NOUN
easat-5713	286	20	in	in	ADP
easat-5713	286	21	complex	complex	ADJ
easat-5713	286	22	video	video	NOUN
easat-5713	286	23	scenes	scene	NOUN
easat-5713	286	24	,	,	PUNCT
easat-5713	286	25	"	"	PUNCT
easat-5713	286	26	ieee	ieee	NOUN
easat-5713	286	27	access	access	NOUN
easat-5713	286	28	,	,	PUNCT
easat-5713	286	29	vol	vol	NOUN
easat-5713	286	30	.	.	PROPN
easat-5713	286	31	12	12	NUM
easat-5713	286	32	,	,	PUNCT
easat-5713	286	33	pp	pp	ADJ
easat-5713	286	34	.	.	PUNCT
easat-5713	287	1	45847–45864	45847–45864	NUM
easat-5713	287	2	,	,	PUNCT
easat-5713	287	3	2024	2024	NUM
easat-5713	287	4	.	.	PUNCT
easat-5713	288	1	https://doi.org/10.1109/access.2024.3381612	https://doi.org/10.1109/access.2024.3381612	PROPN
easat-5713	288	2	[	[	X
easat-5713	288	3	21	21	NUM
easat-5713	288	4	]	]	X
easat-5713	288	5	m.	m.	PROPN
easat-5713	288	6	e.	e.	PROPN
easat-5713	288	7	paoletti	paoletti	PROPN
easat-5713	288	8	,	,	PUNCT
easat-5713	288	9	j.	j.	PROPN
easat-5713	288	10	m.	m.	PROPN
easat-5713	288	11	haut	haut	PROPN
easat-5713	288	12	,	,	PUNCT
easat-5713	288	13	j.	j.	PROPN
easat-5713	288	14	plaza	plaza	PROPN
easat-5713	288	15	,	,	PUNCT
easat-5713	288	16	and	and	CCONJ
easat-5713	288	17	a.	a.	NOUN
easat-5713	288	18	plaza	plaza	PROPN
easat-5713	288	19	,	,	PUNCT
easat-5713	288	20	"	"	PUNCT
easat-5713	288	21	deep	deep	ADJ
easat-5713	288	22	learning	learning	NOUN
easat-5713	288	23	classifiers	classifier	NOUN
easat-5713	288	24	for	for	ADP
easat-5713	288	25	hyperspectral	hyperspectral	ADJ
easat-5713	288	26	imaging	imaging	NOUN
easat-5713	288	27	:	:	PUNCT
easat-5713	288	28	a	a	DET
easat-5713	288	29	review	review	NOUN
easat-5713	288	30	,	,	PUNCT
easat-5713	288	31	"	"	PUNCT
easat-5713	288	32	isprs	isprs	NOUN
easat-5713	288	33	journal	journal	NOUN
easat-5713	288	34	of	of	ADP
easat-5713	288	35	photogrammetry	photogrammetry	NOUN
easat-5713	288	36	and	and	CCONJ
easat-5713	288	37	remote	remote	ADJ
easat-5713	288	38	sensing	sensing	NOUN
easat-5713	288	39	,	,	PUNCT
easat-5713	288	40	vol	vol	NOUN
easat-5713	288	41	.	.	PROPN
easat-5713	288	42	158	158	NUM
easat-5713	288	43	,	,	PUNCT
easat-5713	288	44	pp	pp	ADJ
easat-5713	288	45	.	.	PUNCT
easat-5713	289	1	279	279	NUM
easat-5713	289	2	-	-	SYM
easat-5713	289	3	317	317	NUM
easat-5713	289	4	,	,	PUNCT
easat-5713	289	5	2019	2019	NUM
easat-5713	289	6	.	.	PUNCT
easat-5713	290	1	https://doi.org/10.1016/j.isprsjprs.2019.09.006	https://doi.org/10.1016/j.isprsjprs.2019.09.006	NOUN
easat-5713	290	2	[	[	X
easat-5713	290	3	22	22	NUM
easat-5713	290	4	]	]	PUNCT
easat-5713	290	5	h.	h.	PROPN
easat-5713	290	6	schieber	schieber	PROPN
easat-5713	290	7	,	,	PUNCT
easat-5713	290	8	k.	k.	PROPN
easat-5713	290	9	c.	c.	PROPN
easat-5713	290	10	demir	demir	PROPN
easat-5713	290	11	,	,	PUNCT
easat-5713	290	12	c.	c.	PROPN
easat-5713	290	13	kleinbeck	kleinbeck	PROPN
easat-5713	290	14	,	,	PUNCT
easat-5713	290	15	s.	s.	PROPN
easat-5713	290	16	h.	h.	PROPN
easat-5713	290	17	yang	yang	PROPN
easat-5713	290	18	,	,	PUNCT
easat-5713	290	19	and	and	CCONJ
easat-5713	290	20	d.	d.	PROPN
easat-5713	290	21	roth	roth	PROPN
easat-5713	290	22	,	,	PUNCT
easat-5713	290	23	"	"	PUNCT
easat-5713	290	24	indoor	indoor	ADJ
easat-5713	290	25	synthetic	synthetic	ADJ
easat-5713	290	26	data	datum	NOUN
easat-5713	290	27	generation	generation	NOUN
easat-5713	290	28	:	:	PUNCT
easat-5713	290	29	a	a	DET
easat-5713	290	30	systematic	systematic	ADJ
easat-5713	290	31	review	review	NOUN
easat-5713	290	32	,	,	PUNCT
easat-5713	290	33	"	"	PUNCT
easat-5713	290	34	computer	computer	NOUN
easat-5713	290	35	vision	vision	NOUN
easat-5713	290	36	and	and	CCONJ
easat-5713	290	37	image	image	NOUN
easat-5713	290	38	understanding	understanding	NOUN
easat-5713	290	39	,	,	PUNCT
easat-5713	290	40	vol	vol	NOUN
easat-5713	290	41	.	.	PROPN
easat-5713	290	42	240	240	NUM
easat-5713	290	43	,	,	PUNCT
easat-5713	290	44	p.	p.	NOUN
easat-5713	290	45	103907	103907	NUM
easat-5713	290	46	,	,	PUNCT
easat-5713	290	47	2024	2024	NUM
easat-5713	290	48	.	.	PUNCT
easat-5713	291	1	https://doi.org/10.1016/j.cviu.2023.103907	https://doi.org/10.1016/j.cviu.2023.103907	PROPN
easat-5713	291	2	[	[	X
easat-5713	291	3	23	23	NUM
easat-5713	291	4	]	]	PUNCT
easat-5713	291	5	a.	a.	NOUN
easat-5713	291	6	kartbayev	kartbayev	PROPN
easat-5713	291	7	,	,	PUNCT
easat-5713	291	8	"	"	PUNCT
easat-5713	291	9	smt	smt	PROPN
easat-5713	291	10	:	:	PUNCT
easat-5713	291	11	a	a	DET
easat-5713	291	12	case	case	NOUN
easat-5713	291	13	study	study	NOUN
easat-5713	291	14	of	of	ADP
easat-5713	291	15	kazakh	kazakh	PROPN
easat-5713	291	16	-	-	PUNCT
easat-5713	291	17	english	english	ADJ
easat-5713	291	18	word	word	NOUN
easat-5713	291	19	alignment	alignment	NOUN
easat-5713	291	20	,	,	PUNCT
easat-5713	291	21	"	"	PUNCT
easat-5713	291	22	in	in	ADP
easat-5713	291	23	international	international	ADJ
easat-5713	291	24	conference	conference	NOUN
easat-5713	291	25	on	on	ADP
easat-5713	291	26	web	web	NOUN
easat-5713	291	27	engineering	engineering	NOUN
easat-5713	291	28	,	,	PUNCT
easat-5713	291	29	2015	2015	NUM
easat-5713	291	30	:	:	PUNCT
easat-5713	291	31	springer	springer	NOUN
easat-5713	291	32	,	,	PUNCT
easat-5713	291	33	pp	pp	PROPN
easat-5713	291	34	.	.	PUNCT
easat-5713	292	1	40	40	NUM
easat-5713	292	2	-	-	SYM
easat-5713	292	3	49	49	NUM
easat-5713	292	4	.	.	PUNCT
easat-5713	293	1	[	[	X
easat-5713	293	2	24	24	NUM
easat-5713	293	3	]	]	X
easat-5713	293	4	p.	p.	NOUN
easat-5713	293	5	a.	a.	NOUN
easat-5713	293	6	apellániz	apellániz	PROPN
easat-5713	293	7	,	,	PUNCT
easat-5713	293	8	j.	j.	PROPN
easat-5713	293	9	parras	parras	PROPN
easat-5713	293	10	,	,	PUNCT
easat-5713	293	11	and	and	CCONJ
easat-5713	293	12	s.	s.	PROPN
easat-5713	293	13	zazo	zazo	PROPN
easat-5713	293	14	,	,	PUNCT
easat-5713	293	15	"	"	PUNCT
easat-5713	293	16	improving	improve	VERB
easat-5713	293	17	synthetic	synthetic	ADJ
easat-5713	293	18	data	datum	NOUN
easat-5713	293	19	generation	generation	NOUN
easat-5713	293	20	through	through	ADP
easat-5713	293	21	federated	federated	ADJ
easat-5713	293	22	learning	learning	NOUN
easat-5713	293	23	in	in	ADP
easat-5713	293	24	scarce	scarce	ADJ
easat-5713	293	25	and	and	CCONJ
easat-5713	293	26	heterogeneous	heterogeneous	ADJ
easat-5713	293	27	data	datum	NOUN
easat-5713	293	28	scenarios	scenario	NOUN
easat-5713	293	29	,	,	PUNCT
easat-5713	293	30	"	"	PUNCT
easat-5713	293	31	big	big	ADJ
easat-5713	293	32	data	datum	NOUN
easat-5713	293	33	and	and	CCONJ
easat-5713	293	34	cognitive	cognitive	ADJ
easat-5713	293	35	computing	computing	NOUN
easat-5713	293	36	,	,	PUNCT
easat-5713	293	37	vol	vol	NOUN
easat-5713	293	38	.	.	PROPN
easat-5713	293	39	9	9	NUM
easat-5713	293	40	,	,	PUNCT
easat-5713	293	41	no	no	INTJ
easat-5713	293	42	.	.	NOUN
easat-5713	293	43	2	2	NUM
easat-5713	293	44	,	,	PUNCT
easat-5713	293	45	p.	p.	NOUN
easat-5713	293	46	18	18	NUM
easat-5713	293	47	,	,	PUNCT
easat-5713	293	48	2025	2025	NUM
easat-5713	293	49	.	.	PUNCT
easat-5713	294	1	https://doi.org/10.3390/bdcc9020018	https://doi.org/10.3390/bdcc9020018	NOUN
easat-5713	294	2	[	[	X
easat-5713	294	3	25	25	NUM
easat-5713	294	4	]	]	PUNCT
easat-5713	294	5	n.	n.	NOUN
easat-5713	294	6	smatov	smatov	NOUN
easat-5713	294	7	,	,	PUNCT
easat-5713	294	8	r.	r.	PROPN
easat-5713	294	9	kalashnikov	kalashnikov	PROPN
easat-5713	294	10	,	,	PUNCT
easat-5713	294	11	and	and	CCONJ
easat-5713	294	12	a.	a.	PROPN
easat-5713	294	13	kartbayev	kartbayev	PROPN
easat-5713	294	14	,	,	PUNCT
easat-5713	294	15	"	"	PUNCT
easat-5713	294	16	development	development	NOUN
easat-5713	294	17	of	of	ADP
easat-5713	294	18	context	context	NOUN
easat-5713	294	19	-	-	PUNCT
easat-5713	294	20	based	base	VERB
easat-5713	294	21	sentiment	sentiment	NOUN
easat-5713	294	22	classification	classification	NOUN
easat-5713	294	23	for	for	ADP
easat-5713	294	24	intelligent	intelligent	ADJ
easat-5713	294	25	stock	stock	NOUN
easat-5713	294	26	market	market	NOUN
easat-5713	294	27	prediction	prediction	NOUN
easat-5713	294	28	,	,	PUNCT
easat-5713	294	29	"	"	PUNCT
easat-5713	294	30	big	big	ADJ
easat-5713	294	31	data	datum	NOUN
easat-5713	294	32	and	and	CCONJ
easat-5713	294	33	cognitive	cognitive	ADJ
easat-5713	294	34	computing	computing	NOUN
easat-5713	294	35	,	,	PUNCT
easat-5713	294	36	vol	vol	NOUN
easat-5713	294	37	.	.	PROPN
easat-5713	294	38	8	8	NUM
easat-5713	294	39	,	,	PUNCT
easat-5713	294	40	no	no	INTJ
easat-5713	294	41	.	.	NOUN
easat-5713	294	42	6	6	NUM
easat-5713	294	43	,	,	PUNCT
easat-5713	294	44	p.	p.	NOUN
easat-5713	294	45	51	51	NUM
easat-5713	294	46	,	,	PUNCT
easat-5713	294	47	2024	2024	NUM
easat-5713	294	48	.	.	PUNCT
easat-5713	295	1	https://doi.org/10.3390/bdcc8060051	https://doi.org/10.3390/bdcc8060051	PROPN
easat-5713	295	2	[	[	X
easat-5713	295	3	26	26	NUM
easat-5713	295	4	]	]	PUNCT
easat-5713	295	5	l.	l.	PROPN
easat-5713	295	6	laurier	laurier	PROPN
easat-5713	295	7	,	,	PUNCT
easat-5713	295	8	a.	a.	PROPN
easat-5713	295	9	giulietta	giulietta	PROPN
easat-5713	295	10	,	,	PUNCT
easat-5713	295	11	a.	a.	NOUN
easat-5713	295	12	octavia	octavia	PROPN
easat-5713	295	13	,	,	PUNCT
easat-5713	295	14	and	and	CCONJ
easat-5713	295	15	m.	m.	PROPN
easat-5713	295	16	cleti	cleti	PROPN
easat-5713	295	17	,	,	PUNCT
easat-5713	295	18	"	"	PUNCT
easat-5713	295	19	the	the	DET
easat-5713	295	20	cat	cat	NOUN
easat-5713	295	21	and	and	CCONJ
easat-5713	295	22	mouse	mouse	NOUN
easat-5713	295	23	game	game	NOUN
easat-5713	295	24	:	:	PUNCT
easat-5713	295	25	the	the	DET
easat-5713	295	26	ongoing	ongoing	ADJ
easat-5713	295	27	arms	arm	NOUN
easat-5713	295	28	race	race	NOUN
easat-5713	295	29	between	between	ADP
easat-5713	295	30	diffusion	diffusion	NOUN
easat-5713	295	31	models	model	NOUN
easat-5713	295	32	and	and	CCONJ
easat-5713	295	33	detection	detection	NOUN
easat-5713	295	34	methods	method	NOUN
easat-5713	295	35	,	,	PUNCT
easat-5713	295	36	"	"	PUNCT
easat-5713	295	37	arxiv	arxiv	PROPN
easat-5713	295	38	preprint	preprint	VERB
easat-5713	295	39	arxiv:2410.18866	arxiv:2410.18866	PROPN
easat-5713	295	40	,	,	PUNCT
easat-5713	295	41	2024	2024	NUM
easat-5713	295	42	.	.	PUNCT
easat-5713	296	1	https://doi.org/10.48550/arxiv.2410.18866	https://doi.org/10.48550/arxiv.2410.18866	PROPN
easat-5713	297	1	[	[	X
easat-5713	297	2	27	27	NUM
easat-5713	297	3	]	]	X
easat-5713	297	4	v.	v.	PROPN
easat-5713	297	5	c.	c.	PROPN
easat-5713	297	6	n.	n.	PROPN
easat-5713	297	7	maturana	maturana	PROPN
easat-5713	297	8	,	,	PUNCT
easat-5713	297	9	a.	a.	PROPN
easat-5713	297	10	l.	l.	PROPN
easat-5713	297	11	sandoval	sandoval	PROPN
easat-5713	297	12	orozco	orozco	PROPN
easat-5713	297	13	,	,	PUNCT
easat-5713	297	14	and	and	CCONJ
easat-5713	297	15	l.	l.	PROPN
easat-5713	297	16	j.	j.	PROPN
easat-5713	297	17	garcía	garcía	PROPN
easat-5713	297	18	villalba	villalba	PROPN
easat-5713	297	19	,	,	PUNCT
easat-5713	297	20	"	"	PUNCT
easat-5713	297	21	exploration	exploration	NOUN
easat-5713	297	22	of	of	ADP
easat-5713	297	23	metrics	metric	NOUN
easat-5713	297	24	and	and	CCONJ
easat-5713	297	25	datasets	dataset	NOUN
easat-5713	297	26	to	to	PART
easat-5713	297	27	assess	assess	VERB
easat-5713	297	28	the	the	DET
easat-5713	297	29	fidelity	fidelity	NOUN
easat-5713	297	30	of	of	ADP
easat-5713	297	31	images	image	NOUN
easat-5713	297	32	generated	generate	VERB
easat-5713	297	33	by	by	ADP
easat-5713	297	34	generative	generative	ADJ
easat-5713	297	35	adversarial	adversarial	ADJ
easat-5713	297	36	networks	network	NOUN
easat-5713	297	37	,	,	PUNCT
easat-5713	297	38	"	"	PUNCT
easat-5713	297	39	applied	apply	VERB
easat-5713	297	40	sciences	science	NOUN
easat-5713	297	41	,	,	PUNCT
easat-5713	297	42	vol	vol	NOUN
easat-5713	297	43	.	.	PROPN
easat-5713	297	44	13	13	NUM
easat-5713	297	45	,	,	PUNCT
easat-5713	297	46	no	no	INTJ
easat-5713	297	47	.	.	NOUN
easat-5713	297	48	19	19	NUM
easat-5713	297	49	,	,	PUNCT
easat-5713	297	50	p.	p.	NOUN
easat-5713	297	51	10637	10637	NUM
easat-5713	297	52	,	,	PUNCT
easat-5713	297	53	2023	2023	NUM
easat-5713	297	54	.	.	PUNCT
easat-5713	298	1	https://doi.org/10.3390/app131910637	https://doi.org/10.3390/app131910637	NOUN
easat-5713	299	1	[	[	X
easat-5713	299	2	28	28	NUM
easat-5713	299	3	]	]	X
easat-5713	299	4	n.	n.	PROPN
easat-5713	299	5	ruiz	ruiz	PROPN
easat-5713	299	6	,	,	PUNCT
easat-5713	299	7	y.	y.	PROPN
easat-5713	299	8	li	li	PROPN
easat-5713	299	9	,	,	PUNCT
easat-5713	299	10	v.	v.	PROPN
easat-5713	299	11	jampani	jampani	PROPN
easat-5713	299	12	,	,	PUNCT
easat-5713	299	13	y.	y.	PROPN
easat-5713	299	14	pritch	pritch	PROPN
easat-5713	299	15	,	,	PUNCT
easat-5713	299	16	m.	m.	NOUN
easat-5713	299	17	rubinstein	rubinstein	PROPN
easat-5713	299	18	,	,	PUNCT
easat-5713	299	19	and	and	CCONJ
easat-5713	299	20	k.	k.	PROPN
easat-5713	299	21	aberman	aberman	PROPN
easat-5713	299	22	,	,	PUNCT
easat-5713	299	23	"	"	PUNCT
easat-5713	299	24	dreambooth	dreambooth	NOUN
easat-5713	299	25	:	:	PUNCT
easat-5713	299	26	fine	fine	ADJ
easat-5713	299	27	tuning	tune	VERB
easat-5713	299	28	text	text	NOUN
easat-5713	299	29	-	-	PUNCT
easat-5713	299	30	to	to	ADP
easat-5713	299	31	-	-	PUNCT
easat-5713	299	32	image	image	NOUN
easat-5713	299	33	diffusion	diffusion	NOUN
easat-5713	299	34	models	model	NOUN
easat-5713	299	35	for	for	ADP
easat-5713	299	36	subject	subject	NOUN
easat-5713	299	37	-	-	PUNCT
easat-5713	299	38	driven	drive	VERB
easat-5713	299	39	generation	generation	NOUN
easat-5713	299	40	,	,	PUNCT
easat-5713	299	41	"	"	PUNCT
easat-5713	299	42	in	in	ADP
easat-5713	299	43	2023	2023	NUM
easat-5713	299	44	ieee	ieee	NOUN
easat-5713	299	45	/	/	SYM
easat-5713	299	46	cvf	cvf	NOUN
easat-5713	299	47	conference	conference	NOUN
easat-5713	299	48	on	on	ADP
easat-5713	299	49	computer	computer	NOUN
easat-5713	299	50	vision	vision	NOUN
easat-5713	299	51	and	and	CCONJ
easat-5713	299	52	pattern	pattern	NOUN
easat-5713	299	53	recognition	recognition	NOUN
easat-5713	299	54	(	(	PUNCT
easat-5713	299	55	cvpr	cvpr	NOUN
easat-5713	299	56	)	)	PUNCT
easat-5713	299	57	ieee	ieee	NOUN
easat-5713	299	58	,	,	PUNCT
easat-5713	299	59	2023	2023	NUM
easat-5713	299	60	,	,	PUNCT
easat-5713	299	61	pp	pp	ADJ
easat-5713	299	62	.	.	PUNCT
easat-5713	299	63	22500	22500	NUM
easat-5713	299	64	-	-	SYM
easat-5713	299	65	22510	22510	NUM
easat-5713	299	66	.	.	PUNCT
easat-5713	300	1	[	[	X
easat-5713	300	2	29	29	NUM
easat-5713	300	3	]	]	PUNCT
easat-5713	300	4	s.	s.	PROPN
easat-5713	300	5	f.	f.	PROPN
easat-5713	300	6	ahmed	ahmed	PROPN
easat-5713	300	7	et	et	PROPN
easat-5713	300	8	al	al	PROPN
easat-5713	300	9	.	.	PROPN
easat-5713	300	10	,	,	PUNCT
easat-5713	300	11	"	"	PUNCT
easat-5713	300	12	deep	deep	ADJ
easat-5713	300	13	learning	learning	NOUN
easat-5713	300	14	modelling	model	VERB
easat-5713	300	15	techniques	technique	NOUN
easat-5713	300	16	:	:	PUNCT
easat-5713	300	17	current	current	ADJ
easat-5713	300	18	progress	progress	NOUN
easat-5713	300	19	,	,	PUNCT
easat-5713	300	20	applications	application	NOUN
easat-5713	300	21	,	,	PUNCT
easat-5713	300	22	advantages	advantage	NOUN
easat-5713	300	23	,	,	PUNCT
easat-5713	300	24	and	and	CCONJ
easat-5713	300	25	challenges	challenge	NOUN
easat-5713	300	26	,	,	PUNCT
easat-5713	300	27	"	"	PUNCT
easat-5713	300	28	artificial	artificial	ADJ
easat-5713	300	29	intelligence	intelligence	NOUN
easat-5713	300	30	review	review	NOUN
easat-5713	300	31	,	,	PUNCT
easat-5713	300	32	vol	vol	NOUN
easat-5713	300	33	.	.	PROPN
easat-5713	301	1	56	56	NUM
easat-5713	301	2	,	,	PUNCT
easat-5713	301	3	no	no	INTJ
easat-5713	301	4	.	.	NOUN
easat-5713	301	5	11	11	NUM
easat-5713	301	6	,	,	PUNCT
easat-5713	301	7	pp	pp	ADJ
easat-5713	301	8	.	.	PUNCT
easat-5713	301	9	13521	13521	NUM
easat-5713	301	10	-	-	SYM
easat-5713	301	11	13617	13617	NUM
easat-5713	301	12	,	,	PUNCT
easat-5713	301	13	2023	2023	NUM
easat-5713	301	14	.	.	PUNCT
easat-5713	302	1	https://doi.org/10.1007/s10462-023-10466-8	https://doi.org/10.1007/s10462-023-10466-8	NUM
easat-5713	302	2	https://doi.org/10.55214/25768484.v8i4.978	https://doi.org/10.55214/25768484.v8i4.978	DET
easat-5713	302	3	https://doi.org/10.1016/j.heliyon.2024.e38913	https://doi.org/10.1016/j.heliyon.2024.e38913	PROPN
easat-5713	302	4	https://doi.org/10.1016/j.aej.2023.01.017	https://doi.org/10.1016/j.aej.2023.01.017	NOUN
easat-5713	302	5	https://doi.org/10.1007/s41870-023-01184-z	https://doi.org/10.1007/s41870-023-01184-z	PROPN
easat-5713	302	6	https://doi.org/10.1109/access.2024.3381612	https://doi.org/10.1109/access.2024.3381612	DET
easat-5713	302	7	https://doi.org/10.1016/j.isprsjprs.2019.09.006	https://doi.org/10.1016/j.isprsjprs.2019.09.006	NOUN
easat-5713	302	8	https://doi.org/10.1016/j.cviu.2023.103907	https://doi.org/10.1016/j.cviu.2023.103907	PROPN
easat-5713	302	9	https://doi.org/10.3390/bdcc9020018	https://doi.org/10.3390/bdcc9020018	NOUN
easat-5713	302	10	https://doi.org/10.3390/bdcc8060051	https://doi.org/10.3390/bdcc8060051	VERB
easat-5713	302	11	https://doi.org/10.48550/arxiv.2410.18866	https://doi.org/10.48550/arxiv.2410.18866	PROPN
easat-5713	302	12	https://doi.org/10.3390/app131910637	https://doi.org/10.3390/app131910637	PROPN
easat-5713	302	13	https://doi.org/10.1007/s10462-023-10466-8	https://doi.org/10.1007/s10462-023-10466-8	NUM