id	sid	tid	token	lemma	pos
2v23vt17r9s	1	1	the	the	DET
2v23vt17r9s	1	2	internet	internet	NOUN
2v23vt17r9s	1	3	is	be	AUX
2v23vt17r9s	1	4	a	a	DET
2v23vt17r9s	1	5	vast	vast	ADJ
2v23vt17r9s	1	6	and	and	CCONJ
2v23vt17r9s	1	7	rapidly	rapidly	ADV
2v23vt17r9s	1	8	changing	change	VERB
2v23vt17r9s	1	9	landscape	landscape	NOUN
2v23vt17r9s	1	10	.	.	PUNCT
2v23vt17r9s	2	1	for	for	SCONJ
2v23vt17r9s	2	2	every	every	DET
2v23vt17r9s	2	3	new	new	ADJ
2v23vt17r9s	2	4	algorithm	algorithm	PROPN
2v23vt17r9s	2	5	invented	invent	VERB
2v23vt17r9s	2	6	to	to	PART
2v23vt17r9s	2	7	increase	increase	VERB
2v23vt17r9s	2	8	the	the	DET
2v23vt17r9s	2	9	security	security	NOUN
2v23vt17r9s	2	10	and	and	CCONJ
2v23vt17r9s	2	11	reliability	reliability	NOUN
2v23vt17r9s	2	12	of	of	ADP
2v23vt17r9s	2	13	our	our	PRON
2v23vt17r9s	2	14	society	society	NOUN
2v23vt17r9s	2	15	,	,	PUNCT
2v23vt17r9s	2	16	an	an	DET
2v23vt17r9s	2	17	adversarial	adversarial	ADJ
2v23vt17r9s	2	18	group	group	NOUN
2v23vt17r9s	2	19	coalesces	coalesce	VERB
2v23vt17r9s	2	20	in	in	ADP
2v23vt17r9s	2	21	reaction	reaction	NOUN
2v23vt17r9s	2	22	.	.	PUNCT
2v23vt17r9s	3	1	some	some	PRON
2v23vt17r9s	3	2	may	may	AUX
2v23vt17r9s	3	3	argue	argue	VERB
2v23vt17r9s	3	4	that	that	SCONJ
2v23vt17r9s	3	5	this	this	DET
2v23vt17r9s	3	6	constant	constant	ADJ
2v23vt17r9s	3	7	arms	arm	NOUN
2v23vt17r9s	3	8	race	race	NOUN
2v23vt17r9s	3	9	between	between	ADP
2v23vt17r9s	3	10	defenders	defender	NOUN
2v23vt17r9s	3	11	and	and	CCONJ
2v23vt17r9s	3	12	attackers	attacker	NOUN
2v23vt17r9s	3	13	is	be	AUX
2v23vt17r9s	3	14	currently	currently	ADV
2v23vt17r9s	3	15	being	be	AUX
2v23vt17r9s	3	16	won	win	VERB
2v23vt17r9s	3	17	by	by	ADP
2v23vt17r9s	3	18	the	the	DET
2v23vt17r9s	3	19	latter	latter	ADJ
2v23vt17r9s	3	20	.	.	PUNCT
2v23vt17r9s	4	1	to	to	PART
2v23vt17r9s	4	2	shift	shift	VERB
2v23vt17r9s	4	3	the	the	DET
2v23vt17r9s	4	4	tides	tide	NOUN
2v23vt17r9s	4	5	,	,	PUNCT
2v23vt17r9s	4	6	we	we	PRON
2v23vt17r9s	4	7	need	need	VERB
2v23vt17r9s	4	8	scalable	scalable	ADJ
2v23vt17r9s	4	9	approaches	approach	NOUN
2v23vt17r9s	4	10	to	to	PART
2v23vt17r9s	4	11	counter	counter	VERB
2v23vt17r9s	4	12	their	their	PRON
2v23vt17r9s	4	13	attacks	attack	NOUN
2v23vt17r9s	4	14	,	,	PUNCT
2v23vt17r9s	4	15	whatever	whatever	PRON
2v23vt17r9s	4	16	they	they	PRON
2v23vt17r9s	4	17	may	may	AUX
2v23vt17r9s	4	18	be	be	AUX
2v23vt17r9s	4	19	,	,	PUNCT
2v23vt17r9s	4	20	at	at	ADP
2v23vt17r9s	4	21	the	the	DET
2v23vt17r9s	4	22	scale	scale	NOUN
2v23vt17r9s	4	23	and	and	CCONJ
2v23vt17r9s	4	24	speed	speed	NOUN
2v23vt17r9s	4	25	with	with	ADP
2v23vt17r9s	4	26	which	which	PRON
2v23vt17r9s	4	27	the	the	DET
2v23vt17r9s	4	28	internet	internet	NOUN
2v23vt17r9s	4	29	moves	move	NOUN
2v23vt17r9s	4	30	.	.	PUNCT
2v23vt17r9s	5	1	this	this	DET
2v23vt17r9s	5	2	dissertation	dissertation	NOUN
2v23vt17r9s	5	3	focuses	focus	VERB
2v23vt17r9s	5	4	on	on	ADP
2v23vt17r9s	5	5	developing	develop	VERB
2v23vt17r9s	5	6	computer	computer	NOUN
2v23vt17r9s	5	7	vision	vision	NOUN
2v23vt17r9s	5	8	algorithms	algorithm	NOUN
2v23vt17r9s	5	9	to	to	PART
2v23vt17r9s	5	10	help	help	VERB
2v23vt17r9s	5	11	ensure	ensure	VERB
2v23vt17r9s	5	12	security	security	NOUN
2v23vt17r9s	5	13	within	within	ADP
2v23vt17r9s	5	14	three	three	NUM
2v23vt17r9s	5	15	main	main	ADJ
2v23vt17r9s	5	16	areas	area	NOUN
2v23vt17r9s	5	17	:	:	PUNCT
2v23vt17r9s	5	18	1	1	X
2v23vt17r9s	5	19	)	)	PUNCT
2v23vt17r9s	5	20	in	in	ADP
2v23vt17r9s	5	21	the	the	DET
2v23vt17r9s	5	22	video	video	NOUN
2v23vt17r9s	5	23	surveillance	surveillance	NOUN
2v23vt17r9s	5	24	chapter	chapter	NOUN
2v23vt17r9s	5	25	(	(	PUNCT
2v23vt17r9s	5	26	ch	ch	PROPN
2v23vt17r9s	5	27	.	.	PROPN
2v23vt17r9s	5	28	2	2	NUM
2v23vt17r9s	5	29	)	)	PUNCT
2v23vt17r9s	5	30	,	,	PUNCT
2v23vt17r9s	5	31	we	we	PRON
2v23vt17r9s	5	32	focus	focus	VERB
2v23vt17r9s	5	33	on	on	ADP
2v23vt17r9s	5	34	the	the	DET
2v23vt17r9s	5	35	problem	problem	NOUN
2v23vt17r9s	5	36	of	of	ADP
2v23vt17r9s	5	37	re	re	NOUN
2v23vt17r9s	5	38	-	-	NOUN
2v23vt17r9s	5	39	identification	identification	NOUN
2v23vt17r9s	5	40	,	,	PUNCT
2v23vt17r9s	5	41	or	or	CCONJ
2v23vt17r9s	5	42	finding	find	VERB
2v23vt17r9s	5	43	all	all	DET
2v23vt17r9s	5	44	appearances	appearance	NOUN
2v23vt17r9s	5	45	of	of	ADP
2v23vt17r9s	5	46	a	a	DET
2v23vt17r9s	5	47	given	give	VERB
2v23vt17r9s	5	48	subject	subject	NOUN
2v23vt17r9s	5	49	across	across	ADP
2v23vt17r9s	5	50	a	a	DET
2v23vt17r9s	5	51	given	give	VERB
2v23vt17r9s	5	52	camera	camera	NOUN
2v23vt17r9s	5	53	system	system	NOUN
2v23vt17r9s	5	54	.	.	PUNCT
2v23vt17r9s	6	1	most	most	ADJ
2v23vt17r9s	6	2	all	all	DET
2v23vt17r9s	6	3	existing	exist	VERB
2v23vt17r9s	6	4	reid	reid	PROPN
2v23vt17r9s	6	5	approaches	approaches	AUX
2v23vt17r9s	6	6	employ	employ	VERB
2v23vt17r9s	6	7	local	local	ADJ
2v23vt17r9s	6	8	and	and	CCONJ
2v23vt17r9s	6	9	global	global	ADJ
2v23vt17r9s	6	10	body	body	NOUN
2v23vt17r9s	6	11	features	feature	VERB
2v23vt17r9s	6	12	(	(	PUNCT
2v23vt17r9s	6	13	e.g.	e.g.	ADJ
2v23vt17r9s	6	14	,	,	PUNCT
2v23vt17r9s	6	15	clothing	clothing	NOUN
2v23vt17r9s	6	16	color	color	NOUN
2v23vt17r9s	6	17	and	and	CCONJ
2v23vt17r9s	6	18	pattern	pattern	NOUN
2v23vt17r9s	6	19	,	,	PUNCT
2v23vt17r9s	6	20	body	body	NOUN
2v23vt17r9s	6	21	symmetry	symmetry	NOUN
2v23vt17r9s	6	22	,	,	PUNCT
2v23vt17r9s	6	23	etc	etc	X
2v23vt17r9s	6	24	.	.	X
2v23vt17r9s	6	25	)	)	PUNCT
2v23vt17r9s	6	26	.	.	PUNCT
2v23vt17r9s	7	1	these	these	DET
2v23vt17r9s	7	2	`	`	PUNCT
2v23vt17r9s	7	3	body	body	NOUN
2v23vt17r9s	7	4	reid	reid	NOUN
2v23vt17r9s	7	5	'	'	PART
2v23vt17r9s	7	6	methods	method	NOUN
2v23vt17r9s	7	7	implicitly	implicitly	ADV
2v23vt17r9s	7	8	assume	assume	VERB
2v23vt17r9s	7	9	that	that	SCONJ
2v23vt17r9s	7	10	facial	facial	ADJ
2v23vt17r9s	7	11	resolution	resolution	NOUN
2v23vt17r9s	7	12	is	be	AUX
2v23vt17r9s	7	13	too	too	ADV
2v23vt17r9s	7	14	low	low	ADJ
2v23vt17r9s	7	15	to	to	PART
2v23vt17r9s	7	16	aid	aid	VERB
2v23vt17r9s	7	17	in	in	ADP
2v23vt17r9s	7	18	the	the	DET
2v23vt17r9s	7	19	reid	reid	PROPN
2v23vt17r9s	7	20	process	process	NOUN
2v23vt17r9s	7	21	.	.	PUNCT
2v23vt17r9s	8	1	we	we	PRON
2v23vt17r9s	8	2	assert	assert	VERB
2v23vt17r9s	8	3	that	that	SCONJ
2v23vt17r9s	8	4	faces	face	VERB
2v23vt17r9s	8	5	,	,	PUNCT
2v23vt17r9s	8	6	even	even	ADV
2v23vt17r9s	8	7	when	when	SCONJ
2v23vt17r9s	8	8	captured	capture	VERB
2v23vt17r9s	8	9	in	in	ADP
2v23vt17r9s	8	10	low	low	ADJ
2v23vt17r9s	8	11	resolution	resolution	NOUN
2v23vt17r9s	8	12	environments	environment	NOUN
2v23vt17r9s	8	13	,	,	PUNCT
2v23vt17r9s	8	14	may	may	AUX
2v23vt17r9s	8	15	contain	contain	VERB
2v23vt17r9s	8	16	unique	unique	ADJ
2v23vt17r9s	8	17	and	and	CCONJ
2v23vt17r9s	8	18	stable	stable	ADJ
2v23vt17r9s	8	19	features	feature	NOUN
2v23vt17r9s	8	20	for	for	ADP
2v23vt17r9s	8	21	reid	reid	PROPN
2v23vt17r9s	8	22	.	.	PUNCT
2v23vt17r9s	9	1	such	such	ADJ
2v23vt17r9s	9	2	`	`	PUNCT
2v23vt17r9s	9	3	facial	facial	ADJ
2v23vt17r9s	9	4	reid	reid	NOUN
2v23vt17r9s	9	5	'	'	PART
2v23vt17r9s	9	6	approaches	approach	NOUN
2v23vt17r9s	9	7	are	be	AUX
2v23vt17r9s	9	8	relatively	relatively	ADV
2v23vt17r9s	9	9	unexplored	unexplored	ADJ
2v23vt17r9s	9	10	in	in	ADP
2v23vt17r9s	9	11	the	the	DET
2v23vt17r9s	9	12	literature	literature	NOUN
2v23vt17r9s	9	13	.	.	PUNCT
2v23vt17r9s	10	1	we	we	PRON
2v23vt17r9s	10	2	explore	explore	VERB
2v23vt17r9s	10	3	facial	facial	ADJ
2v23vt17r9s	10	4	reid	reid	NOUN
2v23vt17r9s	10	5	using	use	VERB
2v23vt17r9s	10	6	a	a	DET
2v23vt17r9s	10	7	new	new	ADJ
2v23vt17r9s	10	8	dataset	dataset	NOUN
2v23vt17r9s	10	9	that	that	PRON
2v23vt17r9s	10	10	was	be	AUX
2v23vt17r9s	10	11	collected	collect	VERB
2v23vt17r9s	10	12	from	from	ADP
2v23vt17r9s	10	13	a	a	DET
2v23vt17r9s	10	14	real	real	ADJ
2v23vt17r9s	10	15	surveillance	surveillance	NOUN
2v23vt17r9s	10	16	network	network	NOUN
2v23vt17r9s	10	17	in	in	ADP
2v23vt17r9s	10	18	a	a	DET
2v23vt17r9s	10	19	municipal	municipal	ADJ
2v23vt17r9s	10	20	rapid	rapid	ADJ
2v23vt17r9s	10	21	transit	transit	NOUN
2v23vt17r9s	10	22	system	system	NOUN
2v23vt17r9s	10	23	2	2	NUM
2v23vt17r9s	10	24	)	)	PUNCT
2v23vt17r9s	10	25	in	in	ADP
2v23vt17r9s	10	26	the	the	DET
2v23vt17r9s	10	27	facial	facial	ADJ
2v23vt17r9s	10	28	recognition	recognition	NOUN
2v23vt17r9s	10	29	chapter	chapter	NOUN
2v23vt17r9s	10	30	(	(	PUNCT
2v23vt17r9s	10	31	ch	ch	PROPN
2v23vt17r9s	10	32	.	.	PROPN
2v23vt17r9s	10	33	3	3	NUM
2v23vt17r9s	10	34	)	)	PUNCT
2v23vt17r9s	10	35	,	,	PUNCT
2v23vt17r9s	10	36	we	we	PRON
2v23vt17r9s	10	37	focus	focus	VERB
2v23vt17r9s	10	38	on	on	ADP
2v23vt17r9s	10	39	the	the	DET
2v23vt17r9s	10	40	problem	problem	NOUN
2v23vt17r9s	10	41	of	of	ADP
2v23vt17r9s	10	42	off	off	ADP
2v23vt17r9s	10	43	-	-	PUNCT
2v23vt17r9s	10	44	pose	pose	NOUN
2v23vt17r9s	10	45	faces	face	NOUN
2v23vt17r9s	10	46	.	.	PUNCT
2v23vt17r9s	11	1	if	if	SCONJ
2v23vt17r9s	11	2	a	a	DET
2v23vt17r9s	11	3	convolutional	convolutional	ADJ
2v23vt17r9s	11	4	neural	neural	ADJ
2v23vt17r9s	11	5	network	network	NOUN
2v23vt17r9s	11	6	(	(	PUNCT
2v23vt17r9s	11	7	cnn	cnn	PROPN
2v23vt17r9s	11	8	)	)	PUNCT
2v23vt17r9s	11	9	is	be	AUX
2v23vt17r9s	11	10	intended	intend	VERB
2v23vt17r9s	11	11	to	to	PART
2v23vt17r9s	11	12	tolerate	tolerate	VERB
2v23vt17r9s	11	13	facial	facial	ADJ
2v23vt17r9s	11	14	pose	pose	NOUN
2v23vt17r9s	11	15	,	,	PUNCT
2v23vt17r9s	11	16	then	then	ADV
2v23vt17r9s	11	17	we	we	PRON
2v23vt17r9s	11	18	face	face	VERB
2v23vt17r9s	11	19	an	an	DET
2v23vt17r9s	11	20	important	important	ADJ
2v23vt17r9s	11	21	question	question	NOUN
2v23vt17r9s	11	22	:	:	PUNCT
2v23vt17r9s	11	23	should	should	AUX
2v23vt17r9s	11	24	this	this	DET
2v23vt17r9s	11	25	training	training	NOUN
2v23vt17r9s	11	26	data	datum	NOUN
2v23vt17r9s	11	27	be	be	AUX
2v23vt17r9s	11	28	diverse	diverse	ADJ
2v23vt17r9s	11	29	in	in	ADP
2v23vt17r9s	11	30	its	its	PRON
2v23vt17r9s	11	31	pose	pose	NOUN
2v23vt17r9s	11	32	distribution	distribution	NOUN
2v23vt17r9s	11	33	,	,	PUNCT
2v23vt17r9s	11	34	or	or	CCONJ
2v23vt17r9s	11	35	should	should	AUX
2v23vt17r9s	11	36	face	face	VERB
2v23vt17r9s	11	37	images	image	NOUN
2v23vt17r9s	11	38	be	be	AUX
2v23vt17r9s	11	39	normalized	normalize	VERB
2v23vt17r9s	11	40	to	to	ADP
2v23vt17r9s	11	41	a	a	DET
2v23vt17r9s	11	42	single	single	ADJ
2v23vt17r9s	11	43	pose	pose	NOUN
2v23vt17r9s	11	44	in	in	ADP
2v23vt17r9s	11	45	a	a	DET
2v23vt17r9s	11	46	pre	pre	ADJ
2v23vt17r9s	11	47	-	-	ADJ
2v23vt17r9s	11	48	processing	processing	ADJ
2v23vt17r9s	11	49	step	step	NOUN
2v23vt17r9s	11	50	?	?	PUNCT
2v23vt17r9s	12	1	to	to	PART
2v23vt17r9s	12	2	address	address	VERB
2v23vt17r9s	12	3	this	this	DET
2v23vt17r9s	12	4	question	question	NOUN
2v23vt17r9s	12	5	,	,	PUNCT
2v23vt17r9s	12	6	we	we	PRON
2v23vt17r9s	12	7	evaluate	evaluate	VERB
2v23vt17r9s	12	8	a	a	DET
2v23vt17r9s	12	9	number	number	NOUN
2v23vt17r9s	12	10	of	of	ADP
2v23vt17r9s	12	11	facial	facial	ADJ
2v23vt17r9s	12	12	landmarking	landmarking	NOUN
2v23vt17r9s	12	13	algorithms	algorithm	NOUN
2v23vt17r9s	12	14	and	and	CCONJ
2v23vt17r9s	12	15	a	a	DET
2v23vt17r9s	12	16	popular	popular	ADJ
2v23vt17r9s	12	17	frontalization	frontalization	NOUN
2v23vt17r9s	12	18	method	method	NOUN
2v23vt17r9s	12	19	to	to	PART
2v23vt17r9s	12	20	understand	understand	VERB
2v23vt17r9s	12	21	their	their	PRON
2v23vt17r9s	12	22	effect	effect	NOUN
2v23vt17r9s	12	23	on	on	ADP
2v23vt17r9s	12	24	facial	facial	ADJ
2v23vt17r9s	12	25	recognition	recognition	NOUN
2v23vt17r9s	12	26	performance	performance	NOUN
2v23vt17r9s	12	27	.	.	PUNCT
2v23vt17r9s	13	1	3	3	X
2v23vt17r9s	13	2	)	)	PUNCT
2v23vt17r9s	13	3	in	in	ADP
2v23vt17r9s	13	4	the	the	DET
2v23vt17r9s	13	5	image	image	NOUN
2v23vt17r9s	13	6	forensics	forensic	NOUN
2v23vt17r9s	13	7	and	and	CCONJ
2v23vt17r9s	13	8	retrieval	retrieval	NOUN
2v23vt17r9s	13	9	chapters	chapter	NOUN
2v23vt17r9s	13	10	(	(	PUNCT
2v23vt17r9s	13	11	ch	ch	PROPN
2v23vt17r9s	13	12	.	.	PROPN
2v23vt17r9s	13	13	4	4	NUM
2v23vt17r9s	13	14	,	,	PUNCT
2v23vt17r9s	13	15	5	5	NUM
2v23vt17r9s	13	16	,	,	PUNCT
2v23vt17r9s	13	17	and	and	CCONJ
2v23vt17r9s	13	18	6	6	X
2v23vt17r9s	13	19	)	)	PUNCT
2v23vt17r9s	13	20	,	,	PUNCT
2v23vt17r9s	13	21	we	we	PRON
2v23vt17r9s	13	22	focus	focus	VERB
2v23vt17r9s	13	23	on	on	ADP
2v23vt17r9s	13	24	building	build	VERB
2v23vt17r9s	13	25	novel	novel	ADJ
2v23vt17r9s	13	26	image	image	NOUN
2v23vt17r9s	13	27	retrieval	retrieval	NOUN
2v23vt17r9s	13	28	algorithms	algorithm	NOUN
2v23vt17r9s	13	29	that	that	PRON
2v23vt17r9s	13	30	can	can	AUX
2v23vt17r9s	13	31	retrieve	retrieve	VERB
2v23vt17r9s	13	32	and	and	CCONJ
2v23vt17r9s	13	33	trace	trace	VERB
2v23vt17r9s	13	34	modified	modify	VERB
2v23vt17r9s	13	35	images	image	NOUN
2v23vt17r9s	13	36	back	back	ADV
2v23vt17r9s	13	37	to	to	ADP
2v23vt17r9s	13	38	their	their	PRON
2v23vt17r9s	13	39	origins	origin	NOUN
2v23vt17r9s	13	40	.	.	PUNCT
2v23vt17r9s	14	1	images	image	NOUN
2v23vt17r9s	14	2	from	from	ADP
2v23vt17r9s	14	3	social	social	ADJ
2v23vt17r9s	14	4	media	medium	NOUN
2v23vt17r9s	14	5	can	can	AUX
2v23vt17r9s	14	6	reflect	reflect	VERB
2v23vt17r9s	14	7	diverse	diverse	ADJ
2v23vt17r9s	14	8	viewpoints	viewpoint	NOUN
2v23vt17r9s	14	9	,	,	PUNCT
2v23vt17r9s	14	10	heated	heated	ADJ
2v23vt17r9s	14	11	arguments	argument	NOUN
2v23vt17r9s	14	12	,	,	PUNCT
2v23vt17r9s	14	13	and	and	CCONJ
2v23vt17r9s	14	14	expressions	expression	NOUN
2v23vt17r9s	14	15	of	of	ADP
2v23vt17r9s	14	16	creativity	creativity	NOUN
2v23vt17r9s	14	17	.	.	PUNCT
2v23vt17r9s	15	1	they	they	PRON
2v23vt17r9s	15	2	are	be	AUX
2v23vt17r9s	15	3	often	often	ADV
2v23vt17r9s	15	4	composites	composite	NOUN
2v23vt17r9s	15	5	,	,	PUNCT
2v23vt17r9s	15	6	borrowing	borrow	VERB
2v23vt17r9s	15	7	content	content	NOUN
2v23vt17r9s	15	8	from	from	ADP
2v23vt17r9s	15	9	many	many	ADJ
2v23vt17r9s	15	10	different	different	ADJ
2v23vt17r9s	15	11	sources	source	NOUN
2v23vt17r9s	15	12	to	to	PART
2v23vt17r9s	15	13	create	create	VERB
2v23vt17r9s	15	14	a	a	DET
2v23vt17r9s	15	15	single	single	ADJ
2v23vt17r9s	15	16	image	image	NOUN
2v23vt17r9s	15	17	.	.	PUNCT
2v23vt17r9s	16	1	taking	take	VERB
2v23vt17r9s	16	2	into	into	ADP
2v23vt17r9s	16	3	account	account	NOUN
2v23vt17r9s	16	4	these	these	DET
2v23vt17r9s	16	5	attributes	attribute	NOUN
2v23vt17r9s	16	6	introduces	introduce	VERB
2v23vt17r9s	16	7	new	new	ADJ
2v23vt17r9s	16	8	complexity	complexity	NOUN
2v23vt17r9s	16	9	to	to	ADP
2v23vt17r9s	16	10	image	image	NOUN
2v23vt17r9s	16	11	search	search	NOUN
2v23vt17r9s	16	12	tasks	task	NOUN
2v23vt17r9s	16	13	.	.	PUNCT
2v23vt17r9s	17	1	we	we	PRON
2v23vt17r9s	17	2	propose	propose	VERB
2v23vt17r9s	17	3	multiple	multiple	ADJ
2v23vt17r9s	17	4	new	new	ADJ
2v23vt17r9s	17	5	frameworks	framework	NOUN
2v23vt17r9s	17	6	for	for	ADP
2v23vt17r9s	17	7	image	image	NOUN
2v23vt17r9s	17	8	retrieval	retrieval	NOUN
2v23vt17r9s	17	9	that	that	PRON
2v23vt17r9s	17	10	model	model	NOUN
2v23vt17r9s	17	11	object	object	NOUN
2v23vt17r9s	17	12	-	-	PUNCT
2v23vt17r9s	17	13	level	level	NOUN
2v23vt17r9s	17	14	regions	region	NOUN
2v23vt17r9s	17	15	,	,	PUNCT
2v23vt17r9s	17	16	allowing	allow	VERB
2v23vt17r9s	17	17	for	for	ADP
2v23vt17r9s	17	18	fine	fine	ADV
2v23vt17r9s	17	19	-	-	PUNCT
2v23vt17r9s	17	20	grained	grain	VERB
2v23vt17r9s	17	21	object	object	NOUN
2v23vt17r9s	17	22	level	level	NOUN
2v23vt17r9s	17	23	retrieval	retrieval	NOUN
2v23vt17r9s	17	24	results	result	NOUN
2v23vt17r9s	17	25	.	.	PUNCT
2v23vt17r9s	18	1	these	these	DET
2v23vt17r9s	18	2	retrieval	retrieval	NOUN
2v23vt17r9s	18	3	results	result	NOUN
2v23vt17r9s	18	4	can	can	AUX
2v23vt17r9s	18	5	help	help	VERB
2v23vt17r9s	18	6	aid	aid	NOUN
2v23vt17r9s	18	7	in	in	ADP
2v23vt17r9s	18	8	the	the	DET
2v23vt17r9s	18	9	task	task	NOUN
2v23vt17r9s	18	10	of	of	ADP
2v23vt17r9s	18	11	verifying	verify	VERB
2v23vt17r9s	18	12	the	the	DET
2v23vt17r9s	18	13	forensic	forensic	ADJ
2v23vt17r9s	18	14	integrity	integrity	NOUN
2v23vt17r9s	18	15	of	of	ADP
2v23vt17r9s	18	16	an	an	DET
2v23vt17r9s	18	17	image	image	NOUN
2v23vt17r9s	18	18	.	.	PUNCT
2v23vt17r9s	19	1	we	we	PRON
2v23vt17r9s	19	2	then	then	ADV
2v23vt17r9s	19	3	utilize	utilize	VERB
2v23vt17r9s	19	4	these	these	DET
2v23vt17r9s	19	5	algorithms	algorithm	NOUN
2v23vt17r9s	19	6	to	to	PART
2v23vt17r9s	19	7	perform	perform	VERB
2v23vt17r9s	19	8	real	real	ADJ
2v23vt17r9s	19	9	-	-	PUNCT
2v23vt17r9s	19	10	world	world	NOUN
2v23vt17r9s	19	11	analysis	analysis	NOUN
2v23vt17r9s	19	12	on	on	ADP
2v23vt17r9s	19	13	data	datum	NOUN
2v23vt17r9s	19	14	scraped	scrape	VERB
2v23vt17r9s	19	15	from	from	ADP
2v23vt17r9s	19	16	twitter	twitter	NOUN
2v23vt17r9s	19	17	and	and	CCONJ
2v23vt17r9s	19	18	instagram	instagram	NOUN
2v23vt17r9s	19	19	from	from	ADP
2v23vt17r9s	19	20	the	the	DET
2v23vt17r9s	19	21	2019	2019	NUM
2v23vt17r9s	19	22	indonesian	indonesian	ADJ
2v23vt17r9s	19	23	election	election	NOUN
2v23vt17r9s	19	24	.	.	PUNCT