id	sid	tid	token	lemma	pos
bracis-28383	1	1	dog	dog	NOUN
bracis-28383	1	2	face	face	NOUN
bracis-28383	1	3	recognition	recognition	NOUN
bracis-28383	1	4	using	use	VERB
bracis-28383	1	5	deep	deep	ADJ
bracis-28383	1	6	features	feature	NOUN
bracis-28383	1	7	embeddings	embedding	NOUN
bracis-28383	1	8	|	|	ADV
bracis-28383	1	9	springer	springer	NOUN
bracis-28383	1	10	nature	nature	NOUN
bracis-28383	1	11	link	link	PROPN
bracis-28383	1	12	(	(	PUNCT
bracis-28383	1	13	formerly	formerly	ADV
bracis-28383	1	14	springerlink	springerlink	NOUN
bracis-28383	1	15	)	)	PUNCT
bracis-28383	1	16	skip	skip	VERB
bracis-28383	1	17	to	to	ADP
bracis-28383	1	18	main	main	ADJ
bracis-28383	1	19	content	content	NOUN
bracis-28383	1	20	advertisement	advertisement	NOUN
bracis-28383	1	21	log	log	NOUN
bracis-28383	1	22	in	in	ADP
bracis-28383	1	23	menu	menu	NOUN
bracis-28383	1	24	find	find	VERB
bracis-28383	1	25	a	a	DET
bracis-28383	1	26	journal	journal	NOUN
bracis-28383	1	27	publish	publish	VERB
bracis-28383	1	28	with	with	ADP
bracis-28383	1	29	us	we	PRON
bracis-28383	1	30	track	track	VERB
bracis-28383	1	31	your	your	PRON
bracis-28383	1	32	research	research	NOUN
bracis-28383	1	33	search	search	NOUN
bracis-28383	1	34	cart	cart	NOUN
bracis-28383	1	35	home	home	NOUN
bracis-28383	1	36	intelligent	intelligent	ADJ
bracis-28383	1	37	systems	system	NOUN
bracis-28383	1	38	conference	conference	NOUN
bracis-28383	1	39	paper	paper	NOUN
bracis-28383	1	40	dog	dog	NOUN
bracis-28383	1	41	face	face	NOUN
bracis-28383	1	42	recognition	recognition	NOUN
bracis-28383	1	43	using	use	VERB
bracis-28383	1	44	deep	deep	ADJ
bracis-28383	1	45	features	feature	NOUN
bracis-28383	1	46	embeddings	embedding	NOUN
bracis-28383	1	47	conference	conference	NOUN
bracis-28383	1	48	paper	paper	NOUN
bracis-28383	1	49	first	first	ADV
bracis-28383	1	50	online	online	ADV
bracis-28383	1	51	:	:	PUNCT
bracis-28383	1	52	12	12	NUM
bracis-28383	1	53	october	october	NOUN
bracis-28383	1	54	2023	2023	NUM
bracis-28383	1	55	pp	pp	PRON
bracis-28383	1	56	125–139	125–139	NUM
bracis-28383	1	57	cite	cite	VERB
bracis-28383	1	58	this	this	DET
bracis-28383	1	59	conference	conference	NOUN
bracis-28383	1	60	paper	paper	NOUN
bracis-28383	1	61	access	access	NOUN
bracis-28383	1	62	provided	provide	VERB
bracis-28383	1	63	by	by	ADP
bracis-28383	1	64	university	university	PROPN
bracis-28383	1	65	of	of	ADP
bracis-28383	1	66	notre	notre	PROPN
bracis-28383	1	67	dame	dame	PROPN
bracis-28383	1	68	hesburgh	hesburgh	PROPN
bracis-28383	1	69	library	library	PROPN
bracis-28383	1	70	download	download	PROPN
bracis-28383	1	71	book	book	NOUN
bracis-28383	1	72	pdf	pdf	PROPN
bracis-28383	1	73	download	download	NOUN
bracis-28383	1	74	book	book	NOUN
bracis-28383	1	75	epub	epub	PROPN
bracis-28383	1	76	intelligent	intelligent	ADJ
bracis-28383	1	77	systems	system	NOUN
bracis-28383	1	78	(	(	PUNCT
bracis-28383	1	79	bracis	bracis	NOUN
bracis-28383	1	80	2023	2023	NUM
bracis-28383	1	81	)	)	PUNCT
bracis-28383	1	82	dog	dog	NOUN
bracis-28383	1	83	face	face	NOUN
bracis-28383	1	84	recognition	recognition	NOUN
bracis-28383	1	85	using	use	VERB
bracis-28383	1	86	deep	deep	ADJ
bracis-28383	1	87	features	feature	NOUN
bracis-28383	1	88	embeddings	embedding	NOUN
bracis-28383	1	89	download	download	NOUN
bracis-28383	1	90	book	book	NOUN
bracis-28383	1	91	pdf	pdf	PROPN
bracis-28383	1	92	download	download	NOUN
bracis-28383	1	93	book	book	PROPN
bracis-28383	1	94	epub	epub	PROPN
bracis-28383	1	95	joão	joão	PROPN
bracis-28383	1	96	p.	p.	PROPN
bracis-28383	1	97	b.	b.	PROPN
bracis-28383	1	98	andrade9	andrade9	PROPN
bracis-28383	1	99	,	,	PUNCT
bracis-28383	1	100	leonardo	leonardo	PROPN
bracis-28383	1	101	f.	f.	PROPN
bracis-28383	1	102	costa9	costa9	PROPN
bracis-28383	1	103	,	,	PUNCT
bracis-28383	1	104	lucas	lucas	PROPN
bracis-28383	1	105	s.	s.	PROPN
bracis-28383	1	106	fernandes9	fernandes9	PROPN
bracis-28383	1	107	,	,	PUNCT
bracis-28383	1	108	paulo	paulo	PROPN
bracis-28383	1	109	a.	a.	PROPN
bracis-28383	1	110	l.	l.	PROPN
bracis-28383	1	111	rego9	rego9	PROPN
bracis-28383	1	112	&	&	CCONJ
bracis-28383	1	113	…	…	PUNCT
bracis-28383	1	114	josé	josé	PROPN
bracis-28383	1	115	g.	g.	PROPN
bracis-28383	1	116	r.	r.	PROPN
bracis-28383	1	117	maia10	maia10	PROPN
bracis-28383	1	118	  	  	SPACE
bracis-28383	1	119	show	show	VERB
bracis-28383	1	120	authors	author	NOUN
bracis-28383	1	121	part	part	NOUN
bracis-28383	1	122	of	of	ADP
bracis-28383	1	123	the	the	DET
bracis-28383	1	124	book	book	NOUN
bracis-28383	1	125	series	series	NOUN
bracis-28383	1	126	:	:	PUNCT
bracis-28383	1	127	lecture	lecture	NOUN
bracis-28383	1	128	notes	note	NOUN
bracis-28383	1	129	in	in	ADP
bracis-28383	1	130	computer	computer	NOUN
bracis-28383	1	131	science	science	NOUN
bracis-28383	1	132	(	(	PUNCT
bracis-28383	1	133	(	(	PUNCT
bracis-28383	1	134	lnai	lnai	ADJ
bracis-28383	1	135	,	,	PUNCT
bracis-28383	1	136	volume	volume	NOUN
bracis-28383	1	137	14196	14196	NUM
bracis-28383	1	138	)	)	PUNCT
bracis-28383	1	139	)	)	PUNCT
bracis-28383	1	140	included	include	VERB
bracis-28383	1	141	in	in	ADP
bracis-28383	1	142	the	the	DET
bracis-28383	1	143	following	follow	VERB
bracis-28383	1	144	conference	conference	NOUN
bracis-28383	1	145	series	series	NOUN
bracis-28383	1	146	:	:	PUNCT
bracis-28383	1	147	brazilian	brazilian	ADJ
bracis-28383	1	148	conference	conference	NOUN
bracis-28383	1	149	on	on	ADP
bracis-28383	1	150	intelligent	intelligent	ADJ
bracis-28383	1	151	systems	system	NOUN
bracis-28383	1	152	589	589	NUM
bracis-28383	1	153	accesses	access	NOUN
bracis-28383	1	154	abstract	abstract	ADJ
bracis-28383	1	155	over	over	ADP
bracis-28383	1	156	470	470	NUM
bracis-28383	1	157	million	million	NUM
bracis-28383	1	158	dogs	dog	NOUN
bracis-28383	1	159	are	be	AUX
bracis-28383	1	160	kept	keep	VERB
bracis-28383	1	161	as	as	ADP
bracis-28383	1	162	pets	pet	NOUN
bracis-28383	1	163	around	around	ADP
bracis-28383	1	164	the	the	DET
bracis-28383	1	165	world	world	NOUN
bracis-28383	1	166	.	.	PUNCT
bracis-28383	2	1	dogs	dog	NOUN
bracis-28383	2	2	are	be	AUX
bracis-28383	2	3	owned	own	VERB
bracis-28383	2	4	at	at	ADP
bracis-28383	2	5	an	an	DET
bracis-28383	2	6	average	average	ADJ
bracis-28383	2	7	number	number	NOUN
bracis-28383	2	8	of	of	ADP
bracis-28383	2	9	\(1.6\%\	\(1.6\%\	NOUN
bracis-28383	2	10	)	)	PUNCT
bracis-28383	2	11	per	per	ADP
bracis-28383	2	12	household	household	NOUN
bracis-28383	2	13	.	.	PUNCT
bracis-28383	3	1	the	the	DET
bracis-28383	3	2	us	us	PROPN
bracis-28383	3	3	has	have	VERB
bracis-28383	3	4	the	the	DET
bracis-28383	3	5	most	most	ADJ
bracis-28383	3	6	dog	dog	NOUN
bracis-28383	3	7	pets	pet	NOUN
bracis-28383	3	8	,	,	PUNCT
bracis-28383	3	9	where	where	SCONJ
bracis-28383	3	10	about	about	ADP
bracis-28383	3	11	\(68\%\	\(68\%\	NOUN
bracis-28383	3	12	)	)	PUNCT
bracis-28383	3	13	of	of	ADP
bracis-28383	3	14	households	household	NOUN
bracis-28383	3	15	own	own	VERB
bracis-28383	3	16	at	at	ADV
bracis-28383	3	17	least	least	ADV
bracis-28383	3	18	one	one	NUM
bracis-28383	3	19	pet	pet	NOUN
bracis-28383	3	20	.	.	PUNCT
bracis-28383	4	1	lost	lose	VERB
bracis-28383	4	2	and	and	CCONJ
bracis-28383	4	3	missing	missing	ADJ
bracis-28383	4	4	dogs	dog	NOUN
bracis-28383	4	5	are	be	AUX
bracis-28383	4	6	a	a	DET
bracis-28383	4	7	severe	severe	ADJ
bracis-28383	4	8	source	source	NOUN
bracis-28383	4	9	of	of	ADP
bracis-28383	4	10	suffering	suffering	NOUN
bracis-28383	4	11	and	and	CCONJ
bracis-28383	4	12	problems	problem	NOUN
bracis-28383	4	13	for	for	ADP
bracis-28383	4	14	their	their	PRON
bracis-28383	4	15	families	family	NOUN
bracis-28383	4	16	.	.	PUNCT
bracis-28383	5	1	so	so	ADV
bracis-28383	5	2	,	,	PUNCT
bracis-28383	5	3	this	this	DET
bracis-28383	5	4	paper	paper	NOUN
bracis-28383	5	5	addresses	address	VERB
bracis-28383	5	6	the	the	DET
bracis-28383	5	7	problem	problem	NOUN
bracis-28383	5	8	of	of	ADP
bracis-28383	5	9	facial	facial	ADJ
bracis-28383	5	10	dog	dog	NOUN
bracis-28383	5	11	identification	identification	NOUN
bracis-28383	5	12	.	.	PUNCT
bracis-28383	6	1	this	this	DET
bracis-28383	6	2	technology	technology	NOUN
bracis-28383	6	3	can	can	AUX
bracis-28383	6	4	benefit	benefit	VERB
bracis-28383	6	5	many	many	ADJ
bracis-28383	6	6	applications	application	NOUN
bracis-28383	6	7	,	,	PUNCT
bracis-28383	6	8	such	such	ADJ
bracis-28383	6	9	as	as	ADP
bracis-28383	6	10	handling	handle	VERB
bracis-28383	6	11	the	the	DET
bracis-28383	6	12	missing	missing	ADJ
bracis-28383	6	13	pet	pet	ADJ
bracis-28383	6	14	problem	problem	NOUN
bracis-28383	6	15	,	,	PUNCT
bracis-28383	6	16	granting	grant	VERB
bracis-28383	6	17	pets	pet	NOUN
bracis-28383	6	18	access	access	NOUN
bracis-28383	6	19	to	to	ADP
bracis-28383	6	20	their	their	PRON
bracis-28383	6	21	houses	house	NOUN
bracis-28383	6	22	,	,	PUNCT
bracis-28383	6	23	more	more	ADV
bracis-28383	6	24	intelligent	intelligent	ADJ
bracis-28383	6	25	zoonosis	zoonosis	NOUN
bracis-28383	6	26	control	control	NOUN
bracis-28383	6	27	,	,	PUNCT
bracis-28383	6	28	pet	pet	ADJ
bracis-28383	6	29	health	health	NOUN
bracis-28383	6	30	care	care	NOUN
bracis-28383	6	31	,	,	PUNCT
bracis-28383	6	32	and	and	CCONJ
bracis-28383	6	33	tracking	track	VERB
bracis-28383	6	34	stray	stray	ADJ
bracis-28383	6	35	pets	pet	NOUN
bracis-28383	6	36	.	.	PUNCT
bracis-28383	7	1	we	we	PRON
bracis-28383	7	2	evaluate	evaluate	VERB
bracis-28383	7	3	a	a	DET
bracis-28383	7	4	residual	residual	ADJ
bracis-28383	7	5	convolutional	convolutional	ADJ
bracis-28383	7	6	neural	neural	ADJ
bracis-28383	7	7	network	network	NOUN
bracis-28383	7	8	,	,	PUNCT
bracis-28383	7	9	specifically	specifically	ADV
bracis-28383	7	10	resnet-34	resnet-34	PROPN
bracis-28383	7	11	,	,	PUNCT
bracis-28383	7	12	for	for	ADP
bracis-28383	7	13	facial	facial	ADJ
bracis-28383	7	14	identification	identification	NOUN
bracis-28383	7	15	in	in	ADP
bracis-28383	7	16	dogs	dog	NOUN
bracis-28383	7	17	.	.	PUNCT
bracis-28383	8	1	we	we	PRON
bracis-28383	8	2	tested	test	VERB
bracis-28383	8	3	in	in	ADP
bracis-28383	8	4	dogfacenet	dogfacenet	NOUN
bracis-28383	8	5	and	and	CCONJ
bracis-28383	8	6	flickr	flickr	ADJ
bracis-28383	8	7	-	-	PUNCT
bracis-28383	8	8	dog	dog	NOUN
bracis-28383	8	9	datasets	dataset	NOUN
bracis-28383	8	10	with	with	ADP
bracis-28383	8	11	and	and	CCONJ
bracis-28383	8	12	without	without	ADP
bracis-28383	8	13	two	two	NUM
bracis-28383	8	14	face	face	NOUN
bracis-28383	8	15	preprocessing	preprocesse	VERB
bracis-28383	8	16	techniques	technique	NOUN
bracis-28383	8	17	:	:	PUNCT
bracis-28383	8	18	a	a	DET
bracis-28383	8	19	central	central	ADJ
bracis-28383	8	20	crop	crop	NOUN
bracis-28383	8	21	and	and	CCONJ
bracis-28383	8	22	an	an	DET
bracis-28383	8	23	aligned	aligned	ADJ
bracis-28383	8	24	facial	facial	ADJ
bracis-28383	8	25	extraction	extraction	NOUN
bracis-28383	8	26	.	.	PUNCT
bracis-28383	9	1	experimental	experimental	ADJ
bracis-28383	9	2	results	result	NOUN
bracis-28383	9	3	show	show	VERB
bracis-28383	9	4	promising	promising	ADJ
bracis-28383	9	5	results	result	NOUN
bracis-28383	9	6	surpassing	surpass	VERB
bracis-28383	9	7	the	the	DET
bracis-28383	9	8	state	state	NOUN
bracis-28383	9	9	-	-	PUNCT
bracis-28383	9	10	of	of	ADP
bracis-28383	9	11	-	-	PUNCT
bracis-28383	9	12	the	the	DET
bracis-28383	9	13	-	-	PUNCT
bracis-28383	9	14	art	art	NOUN
bracis-28383	9	15	:	:	PUNCT
bracis-28383	9	16	\(97.6\%\	\(97.6\%\	X
bracis-28383	9	17	)	)	PUNCT
bracis-28383	9	18	and	and	CCONJ
bracis-28383	9	19	\(82.8\%\	\(82.8\%\	ADJ
bracis-28383	9	20	)	)	PUNCT
bracis-28383	9	21	accuracies	accuracy	NOUN
bracis-28383	9	22	for	for	ADP
bracis-28383	9	23	dogfacenet	dogfacenet	NOUN
bracis-28383	9	24	and	and	CCONJ
bracis-28383	9	25	flickr	flickr	NOUN
bracis-28383	9	26	-	-	PUNCT
bracis-28383	9	27	dog	dog	NOUN
bracis-28383	9	28	,	,	PUNCT
bracis-28383	9	29	respectively	respectively	ADV
bracis-28383	9	30	.	.	PUNCT
bracis-28383	10	1	moreover	moreover	ADV
bracis-28383	10	2	,	,	PUNCT
bracis-28383	10	3	we	we	PRON
bracis-28383	10	4	also	also	ADV
bracis-28383	10	5	provide	provide	VERB
bracis-28383	10	6	recall	recall	NOUN
bracis-28383	10	7	metrics	metric	NOUN
bracis-28383	10	8	for	for	ADP
bracis-28383	10	9	the	the	DET
bracis-28383	10	10	best	good	ADJ
bracis-28383	10	11	models	model	NOUN
bracis-28383	10	12	.	.	PUNCT
bracis-28383	11	1	supported	support	VERB
bracis-28383	11	2	by	by	ADP
bracis-28383	11	3	organization	organization	NOUN
bracis-28383	11	4	funcap	funcap	NOUN
bracis-28383	11	5	.	.	PUNCT
bracis-28383	12	1	access	access	NOUN
bracis-28383	12	2	provided	provide	VERB
bracis-28383	12	3	by	by	ADP
bracis-28383	12	4	university	university	PROPN
bracis-28383	12	5	of	of	ADP
bracis-28383	12	6	notre	notre	PROPN
bracis-28383	12	7	dame	dame	PROPN
bracis-28383	12	8	hesburgh	hesburgh	PROPN
bracis-28383	12	9	library	library	PROPN
bracis-28383	12	10	.	.	PUNCT
bracis-28383	13	1	download	download	PROPN
bracis-28383	13	2	conference	conference	NOUN
bracis-28383	13	3	paper	paper	NOUN
bracis-28383	13	4	pdf	pdf	NOUN
bracis-28383	13	5	similar	similar	ADJ
bracis-28383	13	6	content	content	NOUN
bracis-28383	13	7	being	be	AUX
bracis-28383	13	8	viewed	view	VERB
bracis-28383	13	9	by	by	ADP
bracis-28383	13	10	others	other	NOUN
bracis-28383	13	11	a	a	DET
bracis-28383	13	12	deep	deep	ADJ
bracis-28383	13	13	learning	learning	NOUN
bracis-28383	13	14	approach	approach	NOUN
bracis-28383	13	15	for	for	ADP
bracis-28383	13	16	dog	dog	NOUN
bracis-28383	13	17	face	face	NOUN
bracis-28383	13	18	verification	verification	NOUN
bracis-28383	13	19	and	and	CCONJ
bracis-28383	13	20	recognition	recognition	NOUN
bracis-28383	13	21	chapter	chapter	NOUN
bracis-28383	13	22	©	©	PROPN
bracis-28383	13	23	2019	2019	NUM
bracis-28383	13	24	dogface	dogface	NOUN
bracis-28383	13	25	detection	detection	NOUN
bracis-28383	13	26	and	and	CCONJ
bracis-28383	13	27	localization	localization	NOUN
bracis-28383	13	28	of	of	ADP
bracis-28383	13	29	dogface	dogface	NOUN
bracis-28383	13	30	’s	’s	PART
bracis-28383	13	31	landmarks	landmark	NOUN
bracis-28383	13	32	chapter	chapter	NOUN
bracis-28383	13	33	©	©	PROPN
bracis-28383	13	34	2019	2019	NUM
bracis-28383	13	35	where	where	SCONJ
bracis-28383	13	36	is	be	AUX
bracis-28383	13	37	my	my	PRON
bracis-28383	13	38	puppy	puppy	NOUN
bracis-28383	13	39	?	?	PUNCT
bracis-28383	14	1	retrieving	retrieve	VERB
bracis-28383	14	2	lost	lose	VERB
bracis-28383	14	3	dogs	dog	NOUN
bracis-28383	14	4	by	by	ADP
bracis-28383	14	5	facial	facial	ADJ
bracis-28383	14	6	features	feature	NOUN
bracis-28383	14	7	article	article	NOUN
bracis-28383	14	8	03	03	NUM
bracis-28383	14	9	september	september	PROPN
bracis-28383	14	10	2016	2016	NUM
bracis-28383	14	11	explore	explore	VERB
bracis-28383	14	12	related	relate	VERB
bracis-28383	14	13	subjects	subject	NOUN
bracis-28383	14	14	discover	discover	VERB
bracis-28383	14	15	the	the	DET
bracis-28383	14	16	latest	late	ADJ
bracis-28383	14	17	articles	article	NOUN
bracis-28383	14	18	,	,	PUNCT
bracis-28383	14	19	books	book	NOUN
bracis-28383	14	20	and	and	CCONJ
bracis-28383	14	21	news	news	NOUN
bracis-28383	14	22	in	in	ADP
bracis-28383	14	23	related	related	ADJ
bracis-28383	14	24	subjects	subject	NOUN
bracis-28383	14	25	,	,	PUNCT
bracis-28383	14	26	suggested	suggest	VERB
bracis-28383	14	27	using	use	VERB
bracis-28383	14	28	machine	machine	NOUN
bracis-28383	14	29	learning	learning	NOUN
bracis-28383	14	30	.	.	PUNCT
bracis-28383	15	1	biometrics	biometrics	PROPN
bracis-28383	15	2	computer	computer	NOUN
bracis-28383	15	3	vision	vision	PROPN
bracis-28383	15	4	machine	machine	NOUN
bracis-28383	15	5	learning	learn	VERB
bracis-28383	15	6	neural	neural	ADJ
bracis-28383	15	7	decoding	decode	VERB
bracis-28383	15	8	object	object	NOUN
bracis-28383	15	9	recognition	recognition	NOUN
bracis-28383	15	10	artificial	artificial	ADJ
bracis-28383	15	11	intelligence	intelligence	NOUN
bracis-28383	15	12	1	1	NUM
bracis-28383	15	13	introduction	introduction	NOUN
bracis-28383	15	14	effective	effective	ADJ
bracis-28383	15	15	identification	identification	NOUN
bracis-28383	15	16	and	and	CCONJ
bracis-28383	15	17	tracking	tracking	NOUN
bracis-28383	15	18	of	of	ADP
bracis-28383	15	19	pets	pet	NOUN
bracis-28383	15	20	is	be	AUX
bracis-28383	15	21	a	a	DET
bracis-28383	15	22	valuable	valuable	ADJ
bracis-28383	15	23	technology	technology	NOUN
bracis-28383	15	24	for	for	ADP
bracis-28383	15	25	modern	modern	ADJ
bracis-28383	15	26	society	society	NOUN
bracis-28383	15	27	.	.	PUNCT
bracis-28383	16	1	for	for	ADP
bracis-28383	16	2	example	example	NOUN
bracis-28383	16	3	,	,	PUNCT
bracis-28383	16	4	many	many	ADJ
bracis-28383	16	5	countries	country	NOUN
bracis-28383	16	6	have	have	VERB
bracis-28383	16	7	more	more	ADJ
bracis-28383	16	8	dogs	dog	NOUN
bracis-28383	16	9	than	than	ADP
bracis-28383	16	10	children	child	NOUN
bracis-28383	16	11	,	,	PUNCT
bracis-28383	16	12	according	accord	VERB
bracis-28383	16	13	to	to	ADP
bracis-28383	16	14	recent	recent	ADJ
bracis-28383	16	15	official	official	ADJ
bracis-28383	16	16	data	datum	NOUN
bracis-28383	16	17	releases	release	NOUN
bracis-28383	16	18	by	by	ADP
bracis-28383	16	19	[	[	X
bracis-28383	16	20	15	15	NUM
bracis-28383	16	21	]	]	PUNCT
bracis-28383	16	22	.	.	PUNCT
bracis-28383	17	1	stray	stray	ADJ
bracis-28383	17	2	animals	animal	NOUN
bracis-28383	17	3	,	,	PUNCT
bracis-28383	17	4	in	in	ADP
bracis-28383	17	5	general	general	ADJ
bracis-28383	17	6	,	,	PUNCT
bracis-28383	17	7	could	could	AUX
bracis-28383	17	8	pose	pose	VERB
bracis-28383	17	9	a	a	DET
bracis-28383	17	10	serious	serious	ADJ
bracis-28383	17	11	hazard	hazard	NOUN
bracis-28383	17	12	to	to	ADP
bracis-28383	17	13	other	other	ADJ
bracis-28383	17	14	species	specie	NOUN
bracis-28383	17	15	and	and	CCONJ
bracis-28383	17	16	human	human	ADJ
bracis-28383	17	17	health	health	NOUN
bracis-28383	17	18	in	in	ADP
bracis-28383	17	19	urban	urban	ADJ
bracis-28383	17	20	areas	area	NOUN
bracis-28383	17	21	,	,	PUNCT
bracis-28383	17	22	especially	especially	ADV
bracis-28383	17	23	to	to	ADP
bracis-28383	17	24	children	child	NOUN
bracis-28383	17	25	.	.	PUNCT
bracis-28383	18	1	on	on	ADP
bracis-28383	18	2	the	the	DET
bracis-28383	18	3	other	other	ADJ
bracis-28383	18	4	hand	hand	NOUN
bracis-28383	18	5	,	,	PUNCT
bracis-28383	18	6	missing	miss	VERB
bracis-28383	18	7	pets	pet	NOUN
bracis-28383	18	8	is	be	AUX
bracis-28383	18	9	a	a	DET
bracis-28383	18	10	frequent	frequent	ADJ
bracis-28383	18	11	problem	problem	NOUN
bracis-28383	18	12	whose	whose	DET
bracis-28383	18	13	conditions	condition	NOUN
bracis-28383	18	14	of	of	ADP
bracis-28383	18	15	occurrence	occurrence	NOUN
bracis-28383	18	16	are	be	AUX
bracis-28383	18	17	minimal	minimal	ADJ
bracis-28383	18	18	.	.	PUNCT
bracis-28383	19	1	since	since	SCONJ
bracis-28383	19	2	dogs	dog	NOUN
bracis-28383	19	3	are	be	AUX
bracis-28383	19	4	autonomous	autonomous	ADJ
bracis-28383	19	5	animals	animal	NOUN
bracis-28383	19	6	,	,	PUNCT
bracis-28383	19	7	it	it	PRON
bracis-28383	19	8	only	only	ADV
bracis-28383	19	9	takes	take	VERB
bracis-28383	19	10	a	a	DET
bracis-28383	19	11	little	little	ADJ
bracis-28383	19	12	carelessness	carelessness	NOUN
bracis-28383	19	13	to	to	PART
bracis-28383	19	14	make	make	VERB
bracis-28383	19	15	them	they	PRON
bracis-28383	19	16	run	run	VERB
bracis-28383	19	17	away	away	ADV
bracis-28383	19	18	from	from	ADP
bracis-28383	19	19	home	home	NOUN
bracis-28383	19	20	,	,	PUNCT
bracis-28383	19	21	especially	especially	ADV
bracis-28383	19	22	when	when	SCONJ
bracis-28383	19	23	there	there	PRON
bracis-28383	19	24	are	be	VERB
bracis-28383	19	25	festive	festive	ADJ
bracis-28383	19	26	events	event	NOUN
bracis-28383	19	27	with	with	ADP
bracis-28383	19	28	fireworks	firework	NOUN
bracis-28383	19	29	.	.	PUNCT
bracis-28383	20	1	on	on	ADP
bracis-28383	20	2	top	top	NOUN
bracis-28383	20	3	of	of	ADP
bracis-28383	20	4	that	that	PRON
bracis-28383	20	5	,	,	PUNCT
bracis-28383	20	6	the	the	DET
bracis-28383	20	7	human	human	ADJ
bracis-28383	20	8	ability	ability	NOUN
bracis-28383	20	9	to	to	PART
bracis-28383	20	10	discern	discern	VERB
bracis-28383	20	11	animals	animal	NOUN
bracis-28383	20	12	by	by	ADP
bracis-28383	20	13	their	their	PRON
bracis-28383	20	14	traits	trait	NOUN
bracis-28383	20	15	is	be	AUX
bracis-28383	20	16	reasonably	reasonably	ADV
bracis-28383	20	17	limited	limited	ADJ
bracis-28383	20	18	;	;	PUNCT
bracis-28383	20	19	consequently	consequently	ADV
bracis-28383	20	20	,	,	PUNCT
bracis-28383	20	21	it	it	PRON
bracis-28383	20	22	is	be	AUX
bracis-28383	20	23	difficult	difficult	ADJ
bracis-28383	20	24	to	to	PART
bracis-28383	20	25	identify	identify	VERB
bracis-28383	20	26	and	and	CCONJ
bracis-28383	20	27	track	track	VERB
bracis-28383	20	28	animals	animal	NOUN
bracis-28383	20	29	effectively	effectively	ADV
bracis-28383	20	30	with	with	ADP
bracis-28383	20	31	the	the	DET
bracis-28383	20	32	naked	naked	ADJ
bracis-28383	20	33	eye	eye	NOUN
bracis-28383	20	34	[	[	X
bracis-28383	20	35	4	4	NUM
bracis-28383	20	36	]	]	PUNCT
bracis-28383	20	37	.	.	PUNCT
bracis-28383	21	1	in	in	ADP
bracis-28383	21	2	[	[	X
bracis-28383	21	3	7	7	NUM
bracis-28383	21	4	]	]	PUNCT
bracis-28383	21	5	,	,	PUNCT
bracis-28383	21	6	the	the	DET
bracis-28383	21	7	authors	author	NOUN
bracis-28383	21	8	sought	seek	VERB
bracis-28383	21	9	to	to	PART
bracis-28383	21	10	understand	understand	VERB
bracis-28383	21	11	the	the	DET
bracis-28383	21	12	process	process	NOUN
bracis-28383	21	13	of	of	ADP
bracis-28383	21	14	coping	cope	VERB
bracis-28383	21	15	with	with	ADP
bracis-28383	21	16	grief	grief	NOUN
bracis-28383	21	17	experienced	experience	VERB
bracis-28383	21	18	by	by	ADP
bracis-28383	21	19	people	people	NOUN
bracis-28383	21	20	who	who	PRON
bracis-28383	21	21	experienced	experience	VERB
bracis-28383	21	22	the	the	DET
bracis-28383	21	23	loss	loss	NOUN
bracis-28383	21	24	of	of	ADP
bracis-28383	21	25	their	their	PRON
bracis-28383	21	26	pets	pet	NOUN
bracis-28383	21	27	.	.	PUNCT
bracis-28383	22	1	they	they	PRON
bracis-28383	22	2	concluded	conclude	VERB
bracis-28383	22	3	that	that	SCONJ
bracis-28383	22	4	these	these	DET
bracis-28383	22	5	processes	process	NOUN
bracis-28383	22	6	have	have	VERB
bracis-28383	22	7	characteristics	characteristic	NOUN
bracis-28383	22	8	similar	similar	ADJ
bracis-28383	22	9	to	to	ADP
bracis-28383	22	10	those	those	PRON
bracis-28383	22	11	present	present	ADJ
bracis-28383	22	12	in	in	ADP
bracis-28383	22	13	mourning	mourning	NOUN
bracis-28383	22	14	processes	process	NOUN
bracis-28383	22	15	arising	arise	VERB
bracis-28383	22	16	from	from	ADP
bracis-28383	22	17	the	the	DET
bracis-28383	22	18	loss	loss	NOUN
bracis-28383	22	19	of	of	ADP
bracis-28383	22	20	significant	significant	ADJ
bracis-28383	22	21	people	people	NOUN
bracis-28383	22	22	.	.	PUNCT
bracis-28383	23	1	that	that	PRON
bracis-28383	23	2	said	say	VERB
bracis-28383	23	3	,	,	PUNCT
bracis-28383	23	4	canine	canine	NOUN
bracis-28383	23	5	recognition	recognition	NOUN
bracis-28383	23	6	can	can	AUX
bracis-28383	23	7	be	be	AUX
bracis-28383	23	8	an	an	DET
bracis-28383	23	9	attractive	attractive	ADJ
bracis-28383	23	10	alternative	alternative	NOUN
bracis-28383	23	11	for	for	ADP
bracis-28383	23	12	smart	smart	ADJ
bracis-28383	23	13	cities	city	NOUN
bracis-28383	23	14	to	to	PART
bracis-28383	23	15	deal	deal	VERB
bracis-28383	23	16	with	with	ADP
bracis-28383	23	17	complex	complex	ADJ
bracis-28383	23	18	problems	problem	NOUN
bracis-28383	23	19	related	relate	VERB
bracis-28383	23	20	to	to	PART
bracis-28383	23	21	stray	stray	VERB
bracis-28383	23	22	or	or	CCONJ
bracis-28383	23	23	lost	lose	VERB
bracis-28383	23	24	animals	animal	NOUN
bracis-28383	23	25	,	,	PUNCT
bracis-28383	23	26	access	access	NOUN
bracis-28383	23	27	control	control	NOUN
bracis-28383	23	28	for	for	ADP
bracis-28383	23	29	pets	pet	NOUN
bracis-28383	23	30	,	,	PUNCT
bracis-28383	23	31	dog	dog	NOUN
bracis-28383	23	32	catching	catching	NOUN
bracis-28383	23	33	,	,	PUNCT
bracis-28383	23	34	zoonosis	zoonosis	NOUN
bracis-28383	23	35	control	control	NOUN
bracis-28383	23	36	,	,	PUNCT
bracis-28383	23	37	and	and	CCONJ
bracis-28383	23	38	others	other	NOUN
bracis-28383	23	39	.	.	PUNCT
bracis-28383	24	1	as	as	ADP
bracis-28383	24	2	evidence	evidence	NOUN
bracis-28383	24	3	of	of	ADP
bracis-28383	24	4	the	the	DET
bracis-28383	24	5	interest	interest	NOUN
bracis-28383	24	6	in	in	ADP
bracis-28383	24	7	identifying	identify	VERB
bracis-28383	24	8	dogs	dog	NOUN
bracis-28383	24	9	,	,	PUNCT
bracis-28383	24	10	some	some	DET
bracis-28383	24	11	work	work	NOUN
bracis-28383	24	12	seeks	seek	VERB
bracis-28383	24	13	to	to	PART
bracis-28383	24	14	identify	identify	VERB
bracis-28383	24	15	dogs	dog	NOUN
bracis-28383	24	16	using	use	VERB
bracis-28383	24	17	the	the	DET
bracis-28383	24	18	dog	dog	NOUN
bracis-28383	24	19	’s	’s	PART
bracis-28383	24	20	nose	nose	NOUN
bracis-28383	24	21	print	print	NOUN
bracis-28383	24	22	,	,	PUNCT
bracis-28383	24	23	an	an	DET
bracis-28383	24	24	even	even	ADV
bracis-28383	24	25	more	more	ADV
bracis-28383	24	26	challenging	challenging	ADJ
bracis-28383	24	27	problem	problem	NOUN
bracis-28383	24	28	with	with	ADP
bracis-28383	24	29	the	the	DET
bracis-28383	24	30	images	image	NOUN
bracis-28383	24	31	further	far	ADV
bracis-28383	24	32	reduced	reduce	VERB
bracis-28383	24	33	to	to	ADP
bracis-28383	24	34	just	just	ADV
bracis-28383	24	35	the	the	DET
bracis-28383	24	36	dog	dog	NOUN
bracis-28383	24	37	’s	’s	PART
bracis-28383	24	38	nose	nose	NOUN
bracis-28383	24	39	.	.	PUNCT
bracis-28383	25	1	[	[	X
bracis-28383	25	2	13	13	NUM
bracis-28383	25	3	]	]	PUNCT
bracis-28383	25	4	reaches	reach	VERB
bracis-28383	25	5	auc	auc	NOUN
bracis-28383	25	6	of	of	ADP
bracis-28383	25	7	\(88.81\%\	\(88.81\%\	NOUN
bracis-28383	25	8	)	)	PUNCT
bracis-28383	25	9	with	with	ADP
bracis-28383	25	10	a	a	DET
bracis-28383	25	11	model	model	NOUN
bracis-28383	25	12	with	with	ADP
bracis-28383	25	13	the	the	DET
bracis-28383	25	14	dual	dual	ADJ
bracis-28383	25	15	global	global	ADJ
bracis-28383	25	16	descriptor	descriptor	NOUN
bracis-28383	25	17	,	,	PUNCT
bracis-28383	25	18	which	which	PRON
bracis-28383	25	19	can	can	AUX
bracis-28383	25	20	exploit	exploit	VERB
bracis-28383	25	21	the	the	DET
bracis-28383	25	22	multi	multi	ADJ
bracis-28383	25	23	-	-	ADJ
bracis-28383	25	24	scaled	scaled	ADJ
bracis-28383	25	25	features	feature	NOUN
bracis-28383	25	26	of	of	ADP
bracis-28383	25	27	image	image	NOUN
bracis-28383	25	28	working	work	VERB
bracis-28383	25	29	in	in	ADP
bracis-28383	25	30	a	a	DET
bracis-28383	25	31	dataset	dataset	NOUN
bracis-28383	25	32	with	with	ADP
bracis-28383	25	33	6000	6000	NUM
bracis-28383	25	34	dogs	dog	NOUN
bracis-28383	25	35	with	with	ADP
bracis-28383	25	36	20	20	NUM
bracis-28383	25	37	,	,	PUNCT
bracis-28383	25	38	 	 	SPACE
bracis-28383	25	39	000	000	NUM
bracis-28383	25	40	photographs	photograph	NOUN
bracis-28383	25	41	of	of	ADP
bracis-28383	25	42	their	their	PRON
bracis-28383	25	43	nose	nose	NOUN
bracis-28383	25	44	prints	print	NOUN
bracis-28383	25	45	.	.	PUNCT
bracis-28383	26	1	a	a	DET
bracis-28383	26	2	recent	recent	ADJ
bracis-28383	26	3	study	study	NOUN
bracis-28383	26	4	,	,	PUNCT
bracis-28383	26	5	[	[	X
bracis-28383	26	6	23	23	NUM
bracis-28383	26	7	]	]	PUNCT
bracis-28383	26	8	,	,	PUNCT
bracis-28383	26	9	moved	move	VERB
bracis-28383	26	10	by	by	ADP
bracis-28383	26	11	the	the	DET
bracis-28383	26	12	problem	problem	NOUN
bracis-28383	26	13	that	that	SCONJ
bracis-28383	26	14	in	in	ADP
bracis-28383	26	15	sri	sri	PROPN
bracis-28383	26	16	lanka	lanka	PROPN
bracis-28383	26	17	,	,	PUNCT
bracis-28383	26	18	the	the	DET
bracis-28383	26	19	authors	author	NOUN
bracis-28383	26	20	’	’	PART
bracis-28383	26	21	country	country	NOUN
bracis-28383	26	22	,	,	PUNCT
bracis-28383	26	23	a	a	DET
bracis-28383	26	24	lost	lost	ADJ
bracis-28383	26	25	dog	dog	NOUN
bracis-28383	26	26	being	be	AUX
bracis-28383	26	27	found	find	VERB
bracis-28383	26	28	is	be	AUX
bracis-28383	26	29	an	an	DET
bracis-28383	26	30	infrequent	infrequent	ADJ
bracis-28383	26	31	occurrence	occurrence	NOUN
bracis-28383	26	32	,	,	PUNCT
bracis-28383	26	33	presents	present	VERB
bracis-28383	26	34	a	a	DET
bracis-28383	26	35	model	model	NOUN
bracis-28383	26	36	of	of	ADP
bracis-28383	26	37	convolutional	convolutional	ADJ
bracis-28383	26	38	neural	neural	ADJ
bracis-28383	26	39	network	network	NOUN
bracis-28383	26	40	(	(	PUNCT
bracis-28383	26	41	cnn	cnn	PROPN
bracis-28383	26	42	)	)	PUNCT
bracis-28383	26	43	for	for	ADP
bracis-28383	26	44	classification	classification	NOUN
bracis-28383	26	45	and	and	CCONJ
bracis-28383	26	46	recognition	recognition	NOUN
bracis-28383	26	47	of	of	ADP
bracis-28383	26	48	images	image	NOUN
bracis-28383	26	49	for	for	ADP
bracis-28383	26	50	facial	facial	ADJ
bracis-28383	26	51	recognition	recognition	NOUN
bracis-28383	26	52	of	of	ADP
bracis-28383	26	53	dogs	dog	NOUN
bracis-28383	26	54	.	.	PUNCT
bracis-28383	27	1	the	the	DET
bracis-28383	27	2	cnn	cnn	PROPN
bracis-28383	27	3	used	use	VERB
bracis-28383	27	4	consists	consist	NOUN
bracis-28383	27	5	of	of	ADP
bracis-28383	27	6	a	a	DET
bracis-28383	27	7	customized	customized	ADJ
bracis-28383	27	8	model	model	NOUN
bracis-28383	27	9	vgg-16	vgg-16	NOUN
bracis-28383	27	10	,	,	PUNCT
bracis-28383	27	11	obtaining	obtain	VERB
bracis-28383	27	12	about	about	ADP
bracis-28383	27	13	\(94\%\	\(94\%\	NOUN
bracis-28383	27	14	)	)	PUNCT
bracis-28383	27	15	accuracy	accuracy	NOUN
bracis-28383	27	16	in	in	ADP
bracis-28383	27	17	the	the	DET
bracis-28383	27	18	dog	dog	NOUN
bracis-28383	27	19	recognition	recognition	NOUN
bracis-28383	27	20	task	task	NOUN
bracis-28383	27	21	.	.	PUNCT
bracis-28383	28	1	it	it	PRON
bracis-28383	28	2	is	be	AUX
bracis-28383	28	3	noteworthy	noteworthy	ADJ
bracis-28383	28	4	that	that	SCONJ
bracis-28383	28	5	facial	facial	ADJ
bracis-28383	28	6	recognition	recognition	NOUN
bracis-28383	28	7	is	be	AUX
bracis-28383	28	8	one	one	NUM
bracis-28383	28	9	of	of	ADP
bracis-28383	28	10	the	the	DET
bracis-28383	28	11	least	least	ADJ
bracis-28383	28	12	invasive	invasive	ADJ
bracis-28383	28	13	options	option	NOUN
bracis-28383	28	14	and	and	CCONJ
bracis-28383	28	15	is	be	AUX
bracis-28383	28	16	less	less	ADV
bracis-28383	28	17	subject	subject	ADJ
bracis-28383	28	18	to	to	ADP
bracis-28383	28	19	operational	operational	ADJ
bracis-28383	28	20	problems	problem	NOUN
bracis-28383	28	21	when	when	SCONJ
bracis-28383	28	22	the	the	DET
bracis-28383	28	23	animal	animal	NOUN
bracis-28383	28	24	does	do	AUX
bracis-28383	28	25	not	not	PART
bracis-28383	28	26	cooperate	cooperate	VERB
bracis-28383	28	27	with	with	ADP
bracis-28383	28	28	the	the	DET
bracis-28383	28	29	use	use	NOUN
bracis-28383	28	30	of	of	ADP
bracis-28383	28	31	devices	device	NOUN
bracis-28383	28	32	or	or	CCONJ
bracis-28383	28	33	collars	collar	NOUN
bracis-28383	28	34	[	[	X
bracis-28383	28	35	15	15	NUM
bracis-28383	28	36	]	]	PUNCT
bracis-28383	28	37	.	.	PUNCT
bracis-28383	29	1	of	of	ADP
bracis-28383	29	2	course	course	NOUN
bracis-28383	29	3	,	,	PUNCT
bracis-28383	29	4	in	in	ADP
bracis-28383	29	5	the	the	DET
bracis-28383	29	6	case	case	NOUN
bracis-28383	29	7	of	of	ADP
bracis-28383	29	8	dogs	dog	NOUN
bracis-28383	29	9	,	,	PUNCT
bracis-28383	29	10	the	the	DET
bracis-28383	29	11	effectiveness	effectiveness	NOUN
bracis-28383	29	12	of	of	ADP
bracis-28383	29	13	this	this	DET
bracis-28383	29	14	approach	approach	NOUN
bracis-28383	29	15	may	may	AUX
bracis-28383	29	16	vary	vary	VERB
bracis-28383	29	17	if	if	SCONJ
bracis-28383	29	18	the	the	DET
bracis-28383	29	19	animal	animal	NOUN
bracis-28383	29	20	is	be	AUX
bracis-28383	29	21	not	not	PART
bracis-28383	29	22	cared	care	VERB
bracis-28383	29	23	for	for	ADP
bracis-28383	29	24	similarly	similarly	ADV
bracis-28383	29	25	after	after	SCONJ
bracis-28383	29	26	it	it	PRON
bracis-28383	29	27	is	be	AUX
bracis-28383	29	28	lost	lose	VERB
bracis-28383	29	29	(	(	PUNCT
bracis-28383	29	30	hair	hair	NOUN
bracis-28383	29	31	and	and	CCONJ
bracis-28383	29	32	other	other	ADJ
bracis-28383	29	33	characteristics	characteristic	NOUN
bracis-28383	29	34	may	may	AUX
bracis-28383	29	35	change	change	VERB
bracis-28383	29	36	,	,	PUNCT
bracis-28383	29	37	for	for	ADP
bracis-28383	29	38	example	example	NOUN
bracis-28383	29	39	)	)	PUNCT
bracis-28383	29	40	.	.	PUNCT
bracis-28383	30	1	there	there	PRON
bracis-28383	30	2	are	be	VERB
bracis-28383	30	3	algorithms	algorithm	NOUN
bracis-28383	30	4	in	in	ADP
bracis-28383	30	5	the	the	DET
bracis-28383	30	6	literature	literature	NOUN
bracis-28383	30	7	that	that	PRON
bracis-28383	30	8	deal	deal	VERB
bracis-28383	30	9	with	with	ADP
bracis-28383	30	10	this	this	DET
bracis-28383	30	11	problem	problem	NOUN
bracis-28383	30	12	,	,	PUNCT
bracis-28383	30	13	but	but	CCONJ
bracis-28383	30	14	several	several	ADJ
bracis-28383	30	15	of	of	ADP
bracis-28383	30	16	them	they	PRON
bracis-28383	30	17	are	be	AUX
bracis-28383	30	18	arguably	arguably	ADV
bracis-28383	30	19	overly	overly	ADV
bracis-28383	30	20	complex	complex	ADJ
bracis-28383	30	21	or	or	CCONJ
bracis-28383	30	22	may	may	AUX
bracis-28383	30	23	present	present	VERB
bracis-28383	30	24	a	a	DET
bracis-28383	30	25	still	still	ADV
bracis-28383	30	26	limited	limit	VERB
bracis-28383	30	27	performance	performance	NOUN
bracis-28383	30	28	for	for	ADP
bracis-28383	30	29	such	such	DET
bracis-28383	30	30	a	a	DET
bracis-28383	30	31	challenging	challenging	ADJ
bracis-28383	30	32	but	but	CCONJ
bracis-28383	30	33	significant	significant	ADJ
bracis-28383	30	34	problem	problem	NOUN
bracis-28383	30	35	.	.	PUNCT
bracis-28383	31	1	thus	thus	ADV
bracis-28383	31	2	,	,	PUNCT
bracis-28383	31	3	the	the	DET
bracis-28383	31	4	present	present	ADJ
bracis-28383	31	5	work	work	NOUN
bracis-28383	31	6	seeks	seek	VERB
bracis-28383	31	7	to	to	PART
bracis-28383	31	8	consolidate	consolidate	VERB
bracis-28383	31	9	results	result	NOUN
bracis-28383	31	10	in	in	ADP
bracis-28383	31	11	this	this	DET
bracis-28383	31	12	area	area	NOUN
bracis-28383	31	13	by	by	ADP
bracis-28383	31	14	investigating	investigate	VERB
bracis-28383	31	15	the	the	DET
bracis-28383	31	16	use	use	NOUN
bracis-28383	31	17	of	of	ADP
bracis-28383	31	18	deep	deep	ADJ
bracis-28383	31	19	feature	feature	NOUN
bracis-28383	31	20	embedding	embed	VERB
bracis-28383	31	21	vectors	vector	NOUN
bracis-28383	31	22	for	for	ADP
bracis-28383	31	23	dog	dog	NOUN
bracis-28383	31	24	face	face	NOUN
bracis-28383	31	25	recognition	recognition	NOUN
bracis-28383	31	26	,	,	PUNCT
bracis-28383	31	27	focusing	focus	VERB
bracis-28383	31	28	on	on	ADP
bracis-28383	31	29	the	the	DET
bracis-28383	31	30	facial	facial	ADJ
bracis-28383	31	31	identification	identification	NOUN
bracis-28383	31	32	task	task	NOUN
bracis-28383	31	33	.	.	PUNCT
bracis-28383	32	1	these	these	PRON
bracis-28383	32	2	are	be	AUX
bracis-28383	32	3	the	the	DET
bracis-28383	32	4	main	main	ADJ
bracis-28383	32	5	contributions	contribution	NOUN
bracis-28383	32	6	of	of	ADP
bracis-28383	32	7	this	this	DET
bracis-28383	32	8	paper	paper	NOUN
bracis-28383	32	9	:	:	PUNCT
bracis-28383	32	10	we	we	PRON
bracis-28383	32	11	used	use	VERB
bracis-28383	32	12	resnet-34	resnet-34	PROPN
bracis-28383	32	13	,	,	PUNCT
bracis-28383	32	14	a	a	DET
bracis-28383	32	15	residual	residual	ADJ
bracis-28383	32	16	cnn	cnn	NOUN
bracis-28383	32	17	for	for	ADP
bracis-28383	32	18	the	the	DET
bracis-28383	32	19	face	face	NOUN
bracis-28383	32	20	identification	identification	NOUN
bracis-28383	32	21	task	task	NOUN
bracis-28383	32	22	in	in	ADP
bracis-28383	32	23	dogs	dog	NOUN
bracis-28383	32	24	,	,	PUNCT
bracis-28383	32	25	on	on	ADP
bracis-28383	32	26	two	two	NUM
bracis-28383	32	27	datasets	dataset	NOUN
bracis-28383	32	28	in	in	ADP
bracis-28383	32	29	which	which	PRON
bracis-28383	32	30	each	each	DET
bracis-28383	32	31	dataset	dataset	NOUN
bracis-28383	32	32	was	be	AUX
bracis-28383	32	33	evaluated	evaluate	VERB
bracis-28383	32	34	in	in	ADP
bracis-28383	32	35	its	its	PRON
bracis-28383	32	36	standard	standard	ADJ
bracis-28383	32	37	form	form	NOUN
bracis-28383	32	38	and	and	CCONJ
bracis-28383	32	39	applied	apply	VERB
bracis-28383	32	40	two	two	NUM
bracis-28383	32	41	preprocessing	preprocessing	NOUN
bracis-28383	32	42	techniques	technique	NOUN
bracis-28383	32	43	for	for	ADP
bracis-28383	32	44	each	each	DET
bracis-28383	32	45	dataset	dataset	NOUN
bracis-28383	32	46	.	.	PUNCT
bracis-28383	33	1	we	we	PRON
bracis-28383	33	2	evaluate	evaluate	VERB
bracis-28383	33	3	the	the	DET
bracis-28383	33	4	use	use	NOUN
bracis-28383	33	5	of	of	ADP
bracis-28383	33	6	this	this	DET
bracis-28383	33	7	computation	computation	NOUN
bracis-28383	33	8	-	-	PUNCT
bracis-28383	33	9	intensive	intensive	ADJ
bracis-28383	33	10	method	method	NOUN
bracis-28383	33	11	,	,	PUNCT
bracis-28383	33	12	its	its	PRON
bracis-28383	33	13	benefits	benefit	NOUN
bracis-28383	33	14	,	,	PUNCT
bracis-28383	33	15	and	and	CCONJ
bracis-28383	33	16	whether	whether	SCONJ
bracis-28383	33	17	they	they	PRON
bracis-28383	33	18	are	be	AUX
bracis-28383	33	19	justifiable	justifiable	ADJ
bracis-28383	33	20	,	,	PUNCT
bracis-28383	33	21	taking	take	VERB
bracis-28383	33	22	its	its	PRON
bracis-28383	33	23	accuracy	accuracy	NOUN
bracis-28383	33	24	into	into	ADP
bracis-28383	33	25	account	account	NOUN
bracis-28383	33	26	.	.	PUNCT
bracis-28383	34	1	we	we	PRON
bracis-28383	34	2	compare	compare	VERB
bracis-28383	34	3	three	three	NUM
bracis-28383	34	4	dog	dog	NOUN
bracis-28383	34	5	face	face	NOUN
bracis-28383	34	6	preprocessing	preprocesse	VERB
bracis-28383	34	7	methods	method	NOUN
bracis-28383	34	8	using	use	VERB
bracis-28383	34	9	two	two	NUM
bracis-28383	34	10	public	public	ADJ
bracis-28383	34	11	datasets	dataset	NOUN
bracis-28383	34	12	[	[	X
bracis-28383	34	13	15	15	NUM
bracis-28383	34	14	,	,	PUNCT
bracis-28383	34	15	16	16	NUM
bracis-28383	34	16	]	]	PUNCT
bracis-28383	34	17	found	find	VERB
bracis-28383	34	18	in	in	ADP
bracis-28383	34	19	similar	similar	ADJ
bracis-28383	34	20	works	work	NOUN
bracis-28383	34	21	.	.	PUNCT
bracis-28383	35	1	while	while	SCONJ
bracis-28383	35	2	most	most	ADJ
bracis-28383	35	3	papers	paper	NOUN
bracis-28383	35	4	only	only	ADV
bracis-28383	35	5	provide	provide	VERB
bracis-28383	35	6	performance	performance	NOUN
bracis-28383	35	7	analysis	analysis	NOUN
bracis-28383	35	8	considering	consider	VERB
bracis-28383	35	9	accuracy	accuracy	NOUN
bracis-28383	35	10	,	,	PUNCT
bracis-28383	35	11	this	this	DET
bracis-28383	35	12	paper	paper	NOUN
bracis-28383	35	13	also	also	ADV
bracis-28383	35	14	includes	include	VERB
bracis-28383	35	15	recall	recall	NOUN
bracis-28383	35	16	.	.	PUNCT
bracis-28383	36	1	we	we	PRON
bracis-28383	36	2	obtained	obtain	VERB
bracis-28383	36	3	state	state	NOUN
bracis-28383	36	4	-	-	PUNCT
bracis-28383	36	5	of	of	ADP
bracis-28383	36	6	-	-	PUNCT
bracis-28383	36	7	the	the	DET
bracis-28383	36	8	-	-	PUNCT
bracis-28383	36	9	art	art	NOUN
bracis-28383	36	10	performance	performance	NOUN
bracis-28383	36	11	using	use	VERB
bracis-28383	36	12	a	a	DET
bracis-28383	36	13	fast	fast	ADJ
bracis-28383	36	14	approach	approach	NOUN
bracis-28383	36	15	that	that	PRON
bracis-28383	36	16	outperforms	outperform	VERB
bracis-28383	36	17	previous	previous	ADJ
bracis-28383	36	18	methods	method	NOUN
bracis-28383	36	19	,	,	PUNCT
bracis-28383	36	20	which	which	PRON
bracis-28383	36	21	usually	usually	ADV
bracis-28383	36	22	resort	resort	VERB
bracis-28383	36	23	to	to	ADP
bracis-28383	36	24	more	more	ADV
bracis-28383	36	25	complex	complex	ADJ
bracis-28383	36	26	,	,	PUNCT
bracis-28383	36	27	computationally	computationally	ADV
bracis-28383	36	28	expensive	expensive	ADJ
bracis-28383	36	29	models	model	NOUN
bracis-28383	36	30	.	.	PUNCT
bracis-28383	37	1	the	the	DET
bracis-28383	37	2	remainder	remainder	NOUN
bracis-28383	37	3	of	of	ADP
bracis-28383	37	4	this	this	DET
bracis-28383	37	5	paper	paper	NOUN
bracis-28383	37	6	is	be	AUX
bracis-28383	37	7	organized	organize	VERB
bracis-28383	37	8	as	as	SCONJ
bracis-28383	37	9	follows	follow	VERB
bracis-28383	37	10	.	.	PUNCT
bracis-28383	38	1	in	in	ADP
bracis-28383	38	2	sect	sect	NOUN
bracis-28383	38	3	.	.	PUNCT
bracis-28383	38	4	 	 	SPACE
bracis-28383	38	5	2	2	NUM
bracis-28383	38	6	,	,	PUNCT
bracis-28383	38	7	we	we	PRON
bracis-28383	38	8	presented	present	VERB
bracis-28383	38	9	the	the	DET
bracis-28383	38	10	related	relate	VERB
bracis-28383	38	11	works	work	NOUN
bracis-28383	38	12	.	.	PUNCT
bracis-28383	39	1	the	the	DET
bracis-28383	39	2	datasets	dataset	NOUN
bracis-28383	39	3	that	that	PRON
bracis-28383	39	4	address	address	VERB
bracis-28383	39	5	the	the	DET
bracis-28383	39	6	problem	problem	NOUN
bracis-28383	39	7	of	of	ADP
bracis-28383	39	8	facial	facial	ADJ
bracis-28383	39	9	identification	identification	NOUN
bracis-28383	39	10	in	in	ADP
bracis-28383	39	11	dogs	dog	NOUN
bracis-28383	39	12	are	be	AUX
bracis-28383	39	13	described	describe	VERB
bracis-28383	39	14	in	in	ADP
bracis-28383	39	15	sect	sect	NOUN
bracis-28383	39	16	.	.	PUNCT
bracis-28383	39	17	 	 	SPACE
bracis-28383	40	1	3	3	X
bracis-28383	40	2	.	.	PUNCT
bracis-28383	40	3	in	in	ADP
bracis-28383	40	4	sect	sect	NOUN
bracis-28383	40	5	.	.	PUNCT
bracis-28383	40	6	 	 	SPACE
bracis-28383	41	1	4	4	NUM
bracis-28383	41	2	,	,	PUNCT
bracis-28383	41	3	we	we	PRON
bracis-28383	41	4	described	describe	VERB
bracis-28383	41	5	the	the	DET
bracis-28383	41	6	approach	approach	NOUN
bracis-28383	41	7	presented	present	VERB
bracis-28383	41	8	in	in	ADP
bracis-28383	41	9	this	this	DET
bracis-28383	41	10	work	work	NOUN
bracis-28383	41	11	.	.	PUNCT
bracis-28383	42	1	in	in	ADP
bracis-28383	42	2	sect	sect	NOUN
bracis-28383	42	3	.	.	PUNCT
bracis-28383	42	4	 	 	SPACE
bracis-28383	42	5	5	5	NUM
bracis-28383	42	6	,	,	PUNCT
bracis-28383	42	7	we	we	PRON
bracis-28383	42	8	present	present	VERB
bracis-28383	42	9	the	the	DET
bracis-28383	42	10	methodology	methodology	NOUN
bracis-28383	42	11	for	for	ADP
bracis-28383	42	12	the	the	DET
bracis-28383	42	13	experimental	experimental	ADJ
bracis-28383	42	14	evaluation	evaluation	NOUN
bracis-28383	42	15	of	of	ADP
bracis-28383	42	16	our	our	PRON
bracis-28383	42	17	approach	approach	NOUN
bracis-28383	42	18	.	.	PUNCT
bracis-28383	43	1	in	in	ADP
bracis-28383	43	2	sect	sect	NOUN
bracis-28383	43	3	.	.	PUNCT
bracis-28383	43	4	 	 	SPACE
bracis-28383	43	5	6	6	NUM
bracis-28383	43	6	,	,	PUNCT
bracis-28383	43	7	we	we	PRON
bracis-28383	43	8	presented	present	VERB
bracis-28383	43	9	the	the	DET
bracis-28383	43	10	results	result	NOUN
bracis-28383	43	11	.	.	PUNCT
bracis-28383	44	1	moreover	moreover	ADV
bracis-28383	44	2	,	,	PUNCT
bracis-28383	44	3	finally	finally	ADV
bracis-28383	44	4	,	,	PUNCT
bracis-28383	44	5	in	in	ADP
bracis-28383	44	6	sect	sect	NOUN
bracis-28383	44	7	.	.	PUNCT
bracis-28383	44	8	 	 	SPACE
bracis-28383	44	9	7	7	NUM
bracis-28383	44	10	,	,	PUNCT
bracis-28383	44	11	we	we	PRON
bracis-28383	44	12	presented	present	VERB
bracis-28383	44	13	final	final	ADJ
bracis-28383	44	14	remarks	remark	NOUN
bracis-28383	44	15	and	and	CCONJ
bracis-28383	44	16	future	future	ADJ
bracis-28383	44	17	works	work	NOUN
bracis-28383	44	18	.	.	PUNCT
bracis-28383	45	1	2	2	NUM
bracis-28383	45	2	related	relate	VERB
bracis-28383	45	3	work	work	NOUN
bracis-28383	45	4	it	it	PRON
bracis-28383	45	5	is	be	AUX
bracis-28383	45	6	worth	worth	ADJ
bracis-28383	45	7	mentioning	mention	VERB
bracis-28383	45	8	that	that	SCONJ
bracis-28383	45	9	most	most	ADJ
bracis-28383	45	10	of	of	ADP
bracis-28383	45	11	the	the	DET
bracis-28383	45	12	works	work	NOUN
bracis-28383	45	13	found	find	VERB
bracis-28383	45	14	in	in	ADP
bracis-28383	45	15	the	the	DET
bracis-28383	45	16	literature	literature	NOUN
bracis-28383	45	17	are	be	AUX
bracis-28383	45	18	related	relate	VERB
bracis-28383	45	19	to	to	ADP
bracis-28383	45	20	the	the	DET
bracis-28383	45	21	task	task	NOUN
bracis-28383	45	22	of	of	ADP
bracis-28383	45	23	dog	dog	NOUN
bracis-28383	45	24	breed	breed	NOUN
bracis-28383	45	25	classification	classification	NOUN
bracis-28383	45	26	.	.	PUNCT
bracis-28383	46	1	this	this	PRON
bracis-28383	46	2	is	be	AUX
bracis-28383	46	3	a	a	DET
bracis-28383	46	4	challenging	challenging	ADJ
bracis-28383	46	5	problem	problem	NOUN
bracis-28383	46	6	due	due	ADP
bracis-28383	46	7	to	to	ADP
bracis-28383	46	8	the	the	DET
bracis-28383	46	9	diversity	diversity	NOUN
bracis-28383	46	10	of	of	ADP
bracis-28383	46	11	the	the	DET
bracis-28383	46	12	existing	exist	VERB
bracis-28383	46	13	native	native	ADJ
bracis-28383	46	14	breeds	breed	NOUN
bracis-28383	46	15	and	and	CCONJ
bracis-28383	46	16	the	the	DET
bracis-28383	46	17	breeds	breed	NOUN
bracis-28383	46	18	resulting	result	VERB
bracis-28383	46	19	from	from	ADP
bracis-28383	46	20	crossbreeding	crossbreed	VERB
bracis-28383	46	21	and	and	CCONJ
bracis-28383	46	22	genetic	genetic	ADJ
bracis-28383	46	23	modifications	modification	NOUN
bracis-28383	46	24	,	,	PUNCT
bracis-28383	46	25	where	where	SCONJ
bracis-28383	46	26	many	many	ADJ
bracis-28383	46	27	of	of	ADP
bracis-28383	46	28	these	these	PRON
bracis-28383	46	29	are	be	AUX
bracis-28383	46	30	visually	visually	ADV
bracis-28383	46	31	similar	similar	ADJ
bracis-28383	46	32	to	to	ADP
bracis-28383	46	33	each	each	DET
bracis-28383	46	34	other	other	ADJ
bracis-28383	46	35	.	.	PUNCT
bracis-28383	47	1	[	[	X
bracis-28383	47	2	14	14	NUM
bracis-28383	47	3	]	]	PUNCT
bracis-28383	47	4	presented	present	VERB
bracis-28383	47	5	a	a	DET
bracis-28383	47	6	method	method	NOUN
bracis-28383	47	7	for	for	ADP
bracis-28383	47	8	fine	fine	ADV
bracis-28383	47	9	-	-	PUNCT
bracis-28383	47	10	grained	grain	VERB
bracis-28383	47	11	image	image	NOUN
bracis-28383	47	12	classification	classification	NOUN
bracis-28383	47	13	based	base	VERB
bracis-28383	47	14	on	on	ADP
bracis-28383	47	15	using	use	VERB
bracis-28383	47	16	key	key	ADJ
bracis-28383	47	17	regions	region	NOUN
bracis-28383	47	18	during	during	ADP
bracis-28383	47	19	the	the	DET
bracis-28383	47	20	feature	feature	NOUN
bracis-28383	47	21	extraction	extraction	NOUN
bracis-28383	47	22	process	process	NOUN
bracis-28383	47	23	to	to	PART
bracis-28383	47	24	improve	improve	VERB
bracis-28383	47	25	the	the	DET
bracis-28383	47	26	performance	performance	NOUN
bracis-28383	47	27	of	of	ADP
bracis-28383	47	28	the	the	DET
bracis-28383	47	29	classification	classification	NOUN
bracis-28383	47	30	.	.	PUNCT
bracis-28383	48	1	this	this	DET
bracis-28383	48	2	method	method	NOUN
bracis-28383	48	3	was	be	AUX
bracis-28383	48	4	evaluated	evaluate	VERB
bracis-28383	48	5	on	on	ADP
bracis-28383	48	6	the	the	DET
bracis-28383	48	7	dog	dog	NOUN
bracis-28383	48	8	breed	breed	NOUN
bracis-28383	48	9	classification	classification	NOUN
bracis-28383	48	10	problem	problem	NOUN
bracis-28383	48	11	using	use	VERB
bracis-28383	48	12	a	a	DET
bracis-28383	48	13	dataset	dataset	NOUN
bracis-28383	48	14	of	of	ADP
bracis-28383	48	15	images	image	NOUN
bracis-28383	48	16	downloaded	download	VERB
bracis-28383	48	17	from	from	ADP
bracis-28383	48	18	google	google	NOUN
bracis-28383	48	19	,	,	PUNCT
bracis-28383	48	20	image	image	NOUN
bracis-28383	48	21	-	-	PUNCT
bracis-28383	48	22	net	net	NOUN
bracis-28383	48	23	,	,	PUNCT
bracis-28383	48	24	and	and	CCONJ
bracis-28383	48	25	flickr	flickr	PROPN
bracis-28383	48	26	,	,	PUNCT
bracis-28383	48	27	which	which	PRON
bracis-28383	48	28	comprises	comprise	VERB
bracis-28383	48	29	133	133	NUM
bracis-28383	48	30	dog	dog	NOUN
bracis-28383	48	31	breeds	breed	NOUN
bracis-28383	48	32	and	and	CCONJ
bracis-28383	48	33	8	8	NUM
bracis-28383	48	34	,	,	PUNCT
bracis-28383	48	35	 	 	SPACE
bracis-28383	48	36	351	351	NUM
bracis-28383	48	37	images	image	NOUN
bracis-28383	48	38	.	.	PUNCT
bracis-28383	49	1	the	the	DET
bracis-28383	49	2	proposed	propose	VERB
bracis-28383	49	3	method	method	NOUN
bracis-28383	49	4	achieved	achieve	VERB
bracis-28383	49	5	a	a	DET
bracis-28383	49	6	recognition	recognition	NOUN
bracis-28383	49	7	rate	rate	NOUN
bracis-28383	49	8	of	of	ADP
bracis-28383	49	9	\(67\%\	\(67\%\	PROPN
bracis-28383	49	10	)	)	PUNCT
bracis-28383	49	11	and	and	CCONJ
bracis-28383	49	12	used	use	VERB
bracis-28383	49	13	conventional	conventional	ADJ
bracis-28383	49	14	tools	tool	NOUN
bracis-28383	49	15	such	such	ADJ
bracis-28383	49	16	as	as	ADP
bracis-28383	49	17	sift	sift	ADJ
bracis-28383	49	18	and	and	CCONJ
bracis-28383	49	19	color	color	NOUN
bracis-28383	49	20	histograms	histogram	NOUN
bracis-28383	49	21	of	of	ADP
bracis-28383	49	22	the	the	DET
bracis-28383	49	23	animal	animal	NOUN
bracis-28383	49	24	’s	’s	PART
bracis-28383	49	25	face	face	NOUN
bracis-28383	49	26	region	region	NOUN
bracis-28383	49	27	,	,	PUNCT
bracis-28383	49	28	excluding	exclude	VERB
bracis-28383	49	29	everything	everything	PRON
bracis-28383	49	30	else	else	ADV
bracis-28383	49	31	in	in	ADP
bracis-28383	49	32	the	the	DET
bracis-28383	49	33	image	image	NOUN
bracis-28383	49	34	.	.	PUNCT
bracis-28383	50	1	[	[	X
bracis-28383	50	2	17	17	NUM
bracis-28383	50	3	]	]	PUNCT
bracis-28383	50	4	used	use	VERB
bracis-28383	50	5	dense	dense	ADJ
bracis-28383	50	6	-	-	PUNCT
bracis-28383	50	7	sift	sift	NOUN
bracis-28383	50	8	to	to	PART
bracis-28383	50	9	extract	extract	VERB
bracis-28383	50	10	features	feature	NOUN
bracis-28383	50	11	over	over	ADP
bracis-28383	50	12	the	the	DET
bracis-28383	50	13	input	input	NOUN
bracis-28383	50	14	image	image	NOUN
bracis-28383	50	15	.	.	PUNCT
bracis-28383	51	1	these	these	DET
bracis-28383	51	2	features	feature	NOUN
bracis-28383	51	3	are	be	AUX
bracis-28383	51	4	split	split	VERB
bracis-28383	51	5	and	and	CCONJ
bracis-28383	51	6	combined	combine	VERB
bracis-28383	51	7	before	before	ADP
bracis-28383	51	8	feeding	feed	VERB
bracis-28383	51	9	a	a	DET
bracis-28383	51	10	cnn	cnn	NOUN
bracis-28383	51	11	classifier	classifier	NOUN
bracis-28383	51	12	.	.	PUNCT
bracis-28383	52	1	this	this	DET
bracis-28383	52	2	method	method	NOUN
bracis-28383	52	3	resulted	result	VERB
bracis-28383	52	4	in	in	ADP
bracis-28383	52	5	\(94.2\%\	\(94.2\%\	ADJ
bracis-28383	52	6	)	)	PUNCT
bracis-28383	52	7	accuracy	accuracy	NOUN
bracis-28383	52	8	for	for	ADP
bracis-28383	52	9	121	121	NUM
bracis-28383	52	10	dog	dog	NOUN
bracis-28383	52	11	breeds	breed	NOUN
bracis-28383	52	12	,	,	PUNCT
bracis-28383	52	13	using	use	VERB
bracis-28383	52	14	a	a	DET
bracis-28383	52	15	database	database	NOUN
bracis-28383	52	16	obtained	obtain	VERB
bracis-28383	52	17	by	by	ADP
bracis-28383	52	18	crawl	crawl	NOUN
bracis-28383	52	19	image	image	NOUN
bracis-28383	52	20	data	datum	NOUN
bracis-28383	52	21	methods	method	NOUN
bracis-28383	52	22	on	on	ADP
bracis-28383	52	23	the	the	DET
bracis-28383	52	24	web	web	NOUN
bracis-28383	52	25	.	.	PUNCT
bracis-28383	53	1	similarly	similarly	ADV
bracis-28383	53	2	,	,	PUNCT
bracis-28383	53	3	[	[	X
bracis-28383	53	4	4	4	X
bracis-28383	53	5	]	]	PUNCT
bracis-28383	53	6	created	create	VERB
bracis-28383	53	7	a	a	DET
bracis-28383	53	8	(	(	PUNCT
bracis-28383	53	9	closed	closed	ADJ
bracis-28383	53	10	)	)	PUNCT
bracis-28383	53	11	dataset	dataset	NOUN
bracis-28383	53	12	for	for	ADP
bracis-28383	53	13	dog	dog	NOUN
bracis-28383	53	14	breed	breed	NOUN
bracis-28383	53	15	and	and	CCONJ
bracis-28383	53	16	wild	wild	ADJ
bracis-28383	53	17	wolf	wolf	PROPN
bracis-28383	53	18	classification	classification	NOUN
bracis-28383	53	19	.	.	PUNCT
bracis-28383	54	1	these	these	DET
bracis-28383	54	2	authors	author	NOUN
bracis-28383	54	3	compared	compare	VERB
bracis-28383	54	4	different	different	ADJ
bracis-28383	54	5	cnn	cnn	PROPN
bracis-28383	54	6	architectures	architecture	NOUN
bracis-28383	54	7	using	use	VERB
bracis-28383	54	8	transfer	transfer	NOUN
bracis-28383	54	9	learning	learning	NOUN
bracis-28383	54	10	and	and	CCONJ
bracis-28383	54	11	obtained	obtain	VERB
bracis-28383	54	12	\(92.48\%\	\(92.48\%\	NOUN
bracis-28383	54	13	)	)	PUNCT
bracis-28383	54	14	accuracy	accuracy	NOUN
bracis-28383	54	15	.	.	PUNCT
bracis-28383	55	1	[	[	X
bracis-28383	55	2	19	19	NUM
bracis-28383	55	3	]	]	PUNCT
bracis-28383	55	4	proposed	propose	VERB
bracis-28383	55	5	another	another	DET
bracis-28383	55	6	method	method	NOUN
bracis-28383	55	7	based	base	VERB
bracis-28383	55	8	on	on	ADP
bracis-28383	55	9	deep	deep	ADJ
bracis-28383	55	10	learning	learning	NOUN
bracis-28383	55	11	,	,	PUNCT
bracis-28383	55	12	using	use	VERB
bracis-28383	55	13	supervised	supervised	ADJ
bracis-28383	55	14	clustering	clustering	NOUN
bracis-28383	55	15	based	base	VERB
bracis-28383	55	16	on	on	ADP
bracis-28383	55	17	a	a	DET
bracis-28383	55	18	multi	multi	ADJ
bracis-28383	55	19	-	-	ADJ
bracis-28383	55	20	part	part	ADJ
bracis-28383	55	21	-	-	PUNCT
bracis-28383	55	22	convolutional	convolutional	ADJ
bracis-28383	55	23	neural	neural	ADJ
bracis-28383	55	24	network	network	NOUN
bracis-28383	55	25	and	and	CCONJ
bracis-28383	55	26	expectation	expectation	NOUN
bracis-28383	55	27	-	-	PUNCT
bracis-28383	55	28	maximization	maximization	NOUN
bracis-28383	55	29	.	.	PUNCT
bracis-28383	56	1	the	the	DET
bracis-28383	56	2	method	method	NOUN
bracis-28383	56	3	was	be	AUX
bracis-28383	56	4	tested	test	VERB
bracis-28383	56	5	on	on	ADP
bracis-28383	56	6	many	many	ADJ
bracis-28383	56	7	datasets	dataset	NOUN
bracis-28383	56	8	,	,	PUNCT
bracis-28383	56	9	including	include	VERB
bracis-28383	56	10	the	the	DET
bracis-28383	56	11	snapshot	snapshot	ADJ
bracis-28383	56	12	serengeti	serengeti	PROPN
bracis-28383	56	13	dataset	dataset	VERB
bracis-28383	57	1	[	[	PUNCT
bracis-28383	57	2	20	20	NUM
bracis-28383	57	3	]	]	PUNCT
bracis-28383	57	4	,	,	PUNCT
bracis-28383	57	5	which	which	PRON
bracis-28383	57	6	contains	contain	VERB
bracis-28383	57	7	approximately	approximately	ADV
bracis-28383	57	8	78	78	NUM
bracis-28383	58	1	,	,	PUNCT
bracis-28383	58	2	 	 	SPACE
bracis-28383	58	3	000	000	NUM
bracis-28383	58	4	labeled	label	VERB
bracis-28383	58	5	images	image	NOUN
bracis-28383	58	6	of	of	ADP
bracis-28383	58	7	animals	animal	NOUN
bracis-28383	58	8	in	in	ADP
bracis-28383	58	9	africa	africa	PROPN
bracis-28383	58	10	,	,	PUNCT
bracis-28383	58	11	obtaining	obtain	VERB
bracis-28383	58	12	an	an	DET
bracis-28383	58	13	accuracy	accuracy	NOUN
bracis-28383	58	14	of	of	ADP
bracis-28383	58	15	\(98.4\%\	\(98.4\%\	NOUN
bracis-28383	58	16	)	)	PUNCT
bracis-28383	58	17	,	,	PUNCT
bracis-28383	58	18	and	and	CCONJ
bracis-28383	58	19	the	the	DET
bracis-28383	58	20	authors	author	NOUN
bracis-28383	58	21	reinforce	reinforce	VERB
bracis-28383	58	22	the	the	DET
bracis-28383	58	23	idea	idea	NOUN
bracis-28383	58	24	that	that	SCONJ
bracis-28383	58	25	using	use	VERB
bracis-28383	58	26	cnns	cnn	NOUN
bracis-28383	58	27	with	with	ADP
bracis-28383	58	28	supervised	supervised	ADJ
bracis-28383	58	29	training	training	NOUN
bracis-28383	58	30	works	work	NOUN
bracis-28383	58	31	very	very	ADV
bracis-28383	58	32	well	well	ADV
bracis-28383	58	33	in	in	ADP
bracis-28383	58	34	extracting	extract	VERB
bracis-28383	58	35	features	feature	NOUN
bracis-28383	58	36	from	from	ADP
bracis-28383	58	37	images	image	NOUN
bracis-28383	58	38	in	in	ADP
bracis-28383	58	39	such	such	DET
bracis-28383	58	40	a	a	DET
bracis-28383	58	41	type	type	NOUN
bracis-28383	58	42	of	of	ADP
bracis-28383	58	43	problem	problem	NOUN
bracis-28383	58	44	.	.	PUNCT
bracis-28383	59	1	other	other	ADJ
bracis-28383	59	2	works	work	NOUN
bracis-28383	59	3	relate	relate	VERB
bracis-28383	59	4	to	to	ADP
bracis-28383	59	5	using	use	VERB
bracis-28383	59	6	different	different	ADJ
bracis-28383	59	7	techniques	technique	NOUN
bracis-28383	59	8	to	to	PART
bracis-28383	59	9	track	track	VERB
bracis-28383	59	10	and	and	CCONJ
bracis-28383	59	11	find	find	VERB
bracis-28383	59	12	lost	lose	VERB
bracis-28383	59	13	pets	pet	NOUN
bracis-28383	59	14	,	,	PUNCT
bracis-28383	59	15	whereas	whereas	SCONJ
bracis-28383	59	16	many	many	ADJ
bracis-28383	59	17	approaches	approach	NOUN
bracis-28383	59	18	resort	resort	VERB
bracis-28383	59	19	to	to	ADP
bracis-28383	59	20	installing	instal	VERB
bracis-28383	59	21	gps	gps	PROPN
bracis-28383	59	22	and	and	CCONJ
bracis-28383	59	23	other	other	ADJ
bracis-28383	59	24	kinds	kind	NOUN
bracis-28383	59	25	of	of	ADP
bracis-28383	59	26	sensors	sensor	NOUN
bracis-28383	59	27	.	.	PUNCT
bracis-28383	60	1	[	[	X
bracis-28383	60	2	21	21	NUM
bracis-28383	60	3	]	]	PUNCT
bracis-28383	60	4	proposed	propose	VERB
bracis-28383	60	5	bokk	bokk	PROPN
bracis-28383	60	6	meow	meow	PROPN
bracis-28383	60	7	,	,	PUNCT
bracis-28383	60	8	a	a	DET
bracis-28383	60	9	mobile	mobile	ADJ
bracis-28383	60	10	application	application	NOUN
bracis-28383	60	11	whose	whose	DET
bracis-28383	60	12	goal	goal	NOUN
bracis-28383	60	13	is	be	AUX
bracis-28383	60	14	to	to	PART
bracis-28383	60	15	help	help	VERB
bracis-28383	60	16	users	user	NOUN
bracis-28383	60	17	locate	locate	VERB
bracis-28383	60	18	their	their	PRON
bracis-28383	60	19	animals	animal	NOUN
bracis-28383	60	20	using	use	VERB
bracis-28383	60	21	gps	gps	PROPN
bracis-28383	60	22	once	once	SCONJ
bracis-28383	60	23	the	the	DET
bracis-28383	60	24	animals	animal	NOUN
bracis-28383	60	25	might	might	AUX
bracis-28383	60	26	have	have	VERB
bracis-28383	60	27	this	this	DET
bracis-28383	60	28	type	type	NOUN
bracis-28383	60	29	of	of	ADP
bracis-28383	60	30	sensor	sensor	NOUN
bracis-28383	60	31	installed	instal	VERB
bracis-28383	60	32	on	on	ADP
bracis-28383	60	33	their	their	PRON
bracis-28383	60	34	collars	collar	NOUN
bracis-28383	60	35	.	.	PUNCT
bracis-28383	61	1	the	the	DET
bracis-28383	61	2	authors	author	NOUN
bracis-28383	61	3	argue	argue	VERB
bracis-28383	61	4	that	that	SCONJ
bracis-28383	61	5	their	their	PRON
bracis-28383	61	6	application	application	NOUN
bracis-28383	61	7	contributes	contribute	VERB
bracis-28383	61	8	to	to	ADP
bracis-28383	61	9	lesser	less	ADJ
bracis-28383	61	10	stray	stray	VERB
bracis-28383	61	11	/	/	SYM
bracis-28383	61	12	lost	lose	VERB
bracis-28383	61	13	pets	pet	NOUN
bracis-28383	61	14	.	.	PUNCT
bracis-28383	62	1	however	however	ADV
bracis-28383	62	2	,	,	PUNCT
bracis-28383	62	3	bokk	bokk	PROPN
bracis-28383	62	4	meow	meow	VERB
bracis-28383	62	5	demands	demand	VERB
bracis-28383	62	6	an	an	DET
bracis-28383	62	7	active	active	ADJ
bracis-28383	62	8	wi	wi	PROPN
bracis-28383	62	9	-	-	PUNCT
bracis-28383	62	10	fi	fi	NOUN
bracis-28383	62	11	connection	connection	NOUN
bracis-28383	62	12	,	,	PUNCT
bracis-28383	62	13	and	and	CCONJ
bracis-28383	62	14	its	its	PRON
bracis-28383	62	15	tracking	tracking	NOUN
bracis-28383	62	16	features	feature	NOUN
bracis-28383	62	17	do	do	AUX
bracis-28383	62	18	not	not	PART
bracis-28383	62	19	work	work	VERB
bracis-28383	62	20	appropriately	appropriately	ADV
bracis-28383	62	21	in	in	ADP
bracis-28383	62	22	dense	dense	ADJ
bracis-28383	62	23	urban	urban	ADJ
bracis-28383	62	24	building	building	NOUN
bracis-28383	62	25	areas	area	NOUN
bracis-28383	62	26	.	.	PUNCT
bracis-28383	63	1	[	[	X
bracis-28383	63	2	1	1	X
bracis-28383	63	3	]	]	PUNCT
bracis-28383	63	4	proposed	propose	VERB
bracis-28383	63	5	a	a	DET
bracis-28383	63	6	methodology	methodology	NOUN
bracis-28383	63	7	for	for	ADP
bracis-28383	63	8	noseprint	noseprint	NOUN
bracis-28383	63	9	-	-	PUNCT
bracis-28383	63	10	based	base	VERB
bracis-28383	63	11	dog	dog	NOUN
bracis-28383	63	12	recognition	recognition	NOUN
bracis-28383	63	13	using	use	VERB
bracis-28383	63	14	a	a	DET
bracis-28383	63	15	deep	deep	ADJ
bracis-28383	63	16	,	,	PUNCT
bracis-28383	63	17	siamese	siamese	ADJ
bracis-28383	63	18	convolutional	convolutional	ADJ
bracis-28383	63	19	network	network	NOUN
bracis-28383	63	20	called	call	VERB
bracis-28383	63	21	dnnet	dnnet	PROPN
bracis-28383	63	22	.	.	PUNCT
bracis-28383	64	1	this	this	DET
bracis-28383	64	2	cnn	cnn	PROPN
bracis-28383	64	3	is	be	AUX
bracis-28383	64	4	composed	compose	VERB
bracis-28383	64	5	of	of	ADP
bracis-28383	64	6	feature	feature	NOUN
bracis-28383	64	7	extraction	extraction	NOUN
bracis-28383	64	8	and	and	CCONJ
bracis-28383	64	9	(	(	PUNCT
bracis-28383	64	10	expensive	expensive	ADJ
bracis-28383	64	11	)	)	PUNCT
bracis-28383	64	12	self	self	NOUN
bracis-28383	64	13	-	-	PUNCT
bracis-28383	64	14	attention	attention	NOUN
bracis-28383	64	15	modules	module	NOUN
bracis-28383	64	16	.	.	PUNCT
bracis-28383	65	1	their	their	PRON
bracis-28383	65	2	method	method	NOUN
bracis-28383	65	3	seems	seem	VERB
bracis-28383	65	4	promising	promise	VERB
bracis-28383	65	5	,	,	PUNCT
bracis-28383	65	6	presenting	present	VERB
bracis-28383	65	7	an	an	DET
bracis-28383	65	8	accuracy	accuracy	NOUN
bracis-28383	65	9	above	above	ADP
bracis-28383	65	10	\(98.97\%\	\(98.97\%\	NOUN
bracis-28383	65	11	)	)	PUNCT
bracis-28383	65	12	and	and	CCONJ
bracis-28383	65	13	resorts	resort	NOUN
bracis-28383	65	14	to	to	ADP
bracis-28383	65	15	affordable	affordable	ADJ
bracis-28383	65	16	equipment	equipment	NOUN
bracis-28383	65	17	for	for	ADP
bracis-28383	65	18	image	image	NOUN
bracis-28383	65	19	acquisition	acquisition	NOUN
bracis-28383	65	20	.	.	PUNCT
bracis-28383	66	1	however	however	ADV
bracis-28383	66	2	,	,	PUNCT
bracis-28383	66	3	such	such	DET
bracis-28383	66	4	an	an	DET
bracis-28383	66	5	approach	approach	NOUN
bracis-28383	66	6	sounds	sound	VERB
bracis-28383	66	7	hard	hard	ADJ
bracis-28383	66	8	to	to	PART
bracis-28383	66	9	automate	automate	VERB
bracis-28383	66	10	.	.	PUNCT
bracis-28383	67	1	moreover	moreover	ADV
bracis-28383	67	2	,	,	PUNCT
bracis-28383	67	3	human	human	ADJ
bracis-28383	67	4	operators	operator	NOUN
bracis-28383	67	5	should	should	AUX
bracis-28383	67	6	handle	handle	VERB
bracis-28383	67	7	dogs	dog	NOUN
bracis-28383	67	8	physically	physically	ADV
bracis-28383	67	9	,	,	PUNCT
bracis-28383	67	10	which	which	PRON
bracis-28383	67	11	may	may	AUX
bracis-28383	67	12	cause	cause	VERB
bracis-28383	67	13	risks	risk	NOUN
bracis-28383	67	14	or	or	CCONJ
bracis-28383	67	15	discomfort	discomfort	VERB
bracis-28383	67	16	to	to	ADP
bracis-28383	67	17	the	the	DET
bracis-28383	67	18	operator	operator	NOUN
bracis-28383	67	19	and	and	CCONJ
bracis-28383	67	20	the	the	DET
bracis-28383	67	21	dog	dog	NOUN
bracis-28383	67	22	when	when	SCONJ
bracis-28383	67	23	there	there	PRON
bracis-28383	67	24	is	be	VERB
bracis-28383	67	25	no	no	DET
bracis-28383	67	26	collaboration	collaboration	NOUN
bracis-28383	67	27	from	from	ADP
bracis-28383	67	28	the	the	DET
bracis-28383	67	29	animal	animal	NOUN
bracis-28383	67	30	.	.	PUNCT
bracis-28383	68	1	the	the	DET
bracis-28383	68	2	work	work	NOUN
bracis-28383	68	3	of	of	ADP
bracis-28383	68	4	[	[	X
bracis-28383	68	5	22	22	NUM
bracis-28383	68	6	]	]	PUNCT
bracis-28383	68	7	presents	present	VERB
bracis-28383	68	8	a	a	DET
bracis-28383	68	9	smartphone	smartphone	NOUN
bracis-28383	68	10	application	application	NOUN
bracis-28383	68	11	for	for	ADP
bracis-28383	68	12	dog	dog	NOUN
bracis-28383	68	13	face	face	NOUN
bracis-28383	68	14	detection	detection	NOUN
bracis-28383	68	15	.	.	PUNCT
bracis-28383	69	1	the	the	DET
bracis-28383	69	2	system	system	NOUN
bracis-28383	69	3	uses	use	VERB
bracis-28383	69	4	the	the	DET
bracis-28383	69	5	yolo	yolo	ADJ
bracis-28383	69	6	deep	deep	ADJ
bracis-28383	69	7	convolutional	convolutional	ADJ
bracis-28383	69	8	neural	neural	ADJ
bracis-28383	69	9	network	network	NOUN
bracis-28383	69	10	to	to	PART
bracis-28383	69	11	simultaneously	simultaneously	ADV
bracis-28383	69	12	predict	predict	VERB
bracis-28383	69	13	face	face	NOUN
bracis-28383	69	14	bounding	bound	VERB
bracis-28383	69	15	boxes	box	NOUN
bracis-28383	69	16	and	and	CCONJ
bracis-28383	69	17	class	class	NOUN
bracis-28383	69	18	confidences	confidence	NOUN
bracis-28383	69	19	.	.	PUNCT
bracis-28383	70	1	the	the	DET
bracis-28383	70	2	doggie	doggie	NOUN
bracis-28383	70	3	smile	smile	VERB
bracis-28383	70	4	app	app	PROPN
bracis-28383	70	5	helps	help	VERB
bracis-28383	70	6	to	to	PART
bracis-28383	70	7	take	take	VERB
bracis-28383	70	8	pictures	picture	NOUN
bracis-28383	70	9	of	of	ADP
bracis-28383	70	10	the	the	DET
bracis-28383	70	11	dogs	dog	NOUN
bracis-28383	70	12	as	as	SCONJ
bracis-28383	70	13	they	they	PRON
bracis-28383	70	14	face	face	VERB
bracis-28383	70	15	the	the	DET
bracis-28383	70	16	camera	camera	NOUN
bracis-28383	70	17	.	.	PUNCT
bracis-28383	71	1	the	the	DET
bracis-28383	71	2	proposed	propose	VERB
bracis-28383	71	3	mobile	mobile	ADJ
bracis-28383	71	4	app	app	NOUN
bracis-28383	71	5	can	can	AUX
bracis-28383	71	6	simultaneously	simultaneously	ADV
bracis-28383	71	7	assess	assess	VERB
bracis-28383	71	8	the	the	DET
bracis-28383	71	9	gaze	gaze	ADJ
bracis-28383	71	10	directions	direction	NOUN
bracis-28383	71	11	of	of	ADP
bracis-28383	71	12	three	three	NUM
bracis-28383	71	13	dogs	dog	NOUN
bracis-28383	71	14	in	in	ADP
bracis-28383	71	15	a	a	DET
bracis-28383	71	16	scene	scene	NOUN
bracis-28383	71	17	more	more	ADJ
bracis-28383	71	18	than	than	ADP
bracis-28383	71	19	13	13	NUM
bracis-28383	71	20	times	time	NOUN
bracis-28383	71	21	per	per	ADP
bracis-28383	71	22	second	second	NOUN
bracis-28383	71	23	,	,	PUNCT
bracis-28383	71	24	measured	measure	VERB
bracis-28383	71	25	on	on	ADP
bracis-28383	71	26	the	the	DET
bracis-28383	71	27	iphone	iphone	NOUN
bracis-28383	71	28	xr	xr	PROPN
bracis-28383	71	29	.	.	PUNCT
bracis-28383	72	1	the	the	DET
bracis-28383	72	2	reported	report	VERB
bracis-28383	72	3	average	average	ADJ
bracis-28383	72	4	accuracy	accuracy	NOUN
bracis-28383	72	5	of	of	ADP
bracis-28383	72	6	the	the	DET
bracis-28383	72	7	dog	dog	NOUN
bracis-28383	72	8	face	face	NOUN
bracis-28383	72	9	detection	detection	NOUN
bracis-28383	72	10	system	system	NOUN
bracis-28383	72	11	is	be	AUX
bracis-28383	72	12	\(92\%\	\(92\%\	NOUN
bracis-28383	72	13	)	)	PUNCT
bracis-28383	72	14	.	.	PUNCT
bracis-28383	73	1	3	3	NUM
bracis-28383	73	2	dog	dog	NOUN
bracis-28383	73	3	face	face	NOUN
bracis-28383	73	4	recognition	recognition	NOUN
bracis-28383	73	5	datasets	dataset	NOUN
bracis-28383	73	6	and	and	CCONJ
bracis-28383	73	7	 	 	SPACE
bracis-28383	73	8	methods	method	NOUN
bracis-28383	73	9	there	there	PRON
bracis-28383	73	10	are	be	VERB
bracis-28383	73	11	datasets	dataset	NOUN
bracis-28383	73	12	for	for	ADP
bracis-28383	73	13	many	many	ADJ
bracis-28383	73	14	tasks	task	NOUN
bracis-28383	73	15	related	relate	VERB
bracis-28383	73	16	to	to	ADP
bracis-28383	73	17	fine	fine	ADV
bracis-28383	73	18	-	-	PUNCT
bracis-28383	73	19	grained	grain	VERB
bracis-28383	73	20	dog	dog	NOUN
bracis-28383	73	21	image	image	NOUN
bracis-28383	73	22	classification	classification	NOUN
bracis-28383	73	23	,	,	PUNCT
bracis-28383	73	24	such	such	ADJ
bracis-28383	73	25	as	as	ADP
bracis-28383	73	26	columbia	columbia	PROPN
bracis-28383	73	27	dogs	dog	NOUN
bracis-28383	73	28	and	and	CCONJ
bracis-28383	73	29	stanford	stanford	PROPN
bracis-28383	73	30	dogs	dog	NOUN
bracis-28383	73	31	,	,	PUNCT
bracis-28383	73	32	that	that	PRON
bracis-28383	73	33	are	be	AUX
bracis-28383	73	34	used	use	VERB
bracis-28383	73	35	in	in	ADP
bracis-28383	73	36	[	[	X
bracis-28383	73	37	9	9	NUM
bracis-28383	73	38	,	,	PUNCT
bracis-28383	73	39	14	14	NUM
bracis-28383	73	40	]	]	PUNCT
bracis-28383	73	41	,	,	PUNCT
bracis-28383	73	42	respectively	respectively	ADV
bracis-28383	73	43	.	.	PUNCT
bracis-28383	74	1	those	those	DET
bracis-28383	74	2	datasets	dataset	NOUN
bracis-28383	74	3	were	be	AUX
bracis-28383	74	4	used	use	VERB
bracis-28383	74	5	in	in	ADP
bracis-28383	74	6	previous	previous	ADJ
bracis-28383	74	7	works	work	NOUN
bracis-28383	74	8	for	for	ADP
bracis-28383	74	9	augmenting	augment	VERB
bracis-28383	74	10	data	datum	NOUN
bracis-28383	74	11	and	and	CCONJ
bracis-28383	74	12	for	for	ADP
bracis-28383	74	13	fine	fine	ADV
bracis-28383	74	14	-	-	PUNCT
bracis-28383	74	15	tuning	tune	VERB
bracis-28383	74	16	pre	pre	ADJ
bracis-28383	74	17	-	-	ADJ
bracis-28383	74	18	trained	train	VERB
bracis-28383	74	19	models	model	NOUN
bracis-28383	74	20	before	before	SCONJ
bracis-28383	74	21	the	the	DET
bracis-28383	74	22	dog	dog	NOUN
bracis-28383	74	23	recognition	recognition	NOUN
bracis-28383	74	24	task	task	NOUN
bracis-28383	74	25	was	be	AUX
bracis-28383	74	26	carried	carry	VERB
bracis-28383	74	27	out	out	ADP
bracis-28383	74	28	.	.	PUNCT
bracis-28383	75	1	for	for	ADP
bracis-28383	75	2	example	example	NOUN
bracis-28383	75	3	,	,	PUNCT
bracis-28383	75	4	[	[	X
bracis-28383	75	5	16	16	NUM
bracis-28383	75	6	]	]	PUNCT
bracis-28383	75	7	also	also	ADV
bracis-28383	75	8	resorted	resort	VERB
bracis-28383	75	9	to	to	ADP
bracis-28383	75	10	these	these	DET
bracis-28383	75	11	datasets	dataset	NOUN
bracis-28383	75	12	to	to	PART
bracis-28383	75	13	train	train	VERB
bracis-28383	75	14	and	and	CCONJ
bracis-28383	75	15	evaluate	evaluate	VERB
bracis-28383	75	16	fog	fog	NOUN
bracis-28383	75	17	face	face	NOUN
bracis-28383	75	18	detectors	detector	NOUN
bracis-28383	75	19	and	and	CCONJ
bracis-28383	75	20	aligners	aligner	NOUN
bracis-28383	75	21	.	.	PUNCT
bracis-28383	76	1	3.1	3.1	NUM
bracis-28383	76	2	dog	dog	NOUN
bracis-28383	76	3	face	face	NOUN
bracis-28383	76	4	recognition	recognition	NOUN
bracis-28383	76	5	datasets	dataset	NOUN
bracis-28383	76	6	table	table	NOUN
bracis-28383	76	7	 	 	SPACE
bracis-28383	76	8	1	1	NUM
bracis-28383	76	9	summarizes	summarize	NOUN
bracis-28383	76	10	the	the	DET
bracis-28383	76	11	datasets	dataset	NOUN
bracis-28383	76	12	known	know	VERB
bracis-28383	76	13	in	in	ADP
bracis-28383	76	14	the	the	DET
bracis-28383	76	15	literature	literature	NOUN
bracis-28383	76	16	for	for	ADP
bracis-28383	76	17	the	the	DET
bracis-28383	76	18	face	face	NOUN
bracis-28383	76	19	identification	identification	NOUN
bracis-28383	76	20	task	task	NOUN
bracis-28383	76	21	in	in	ADP
bracis-28383	76	22	dogs	dog	NOUN
bracis-28383	76	23	,	,	PUNCT
bracis-28383	76	24	their	their	PRON
bracis-28383	76	25	main	main	ADJ
bracis-28383	76	26	characteristics	characteristic	NOUN
bracis-28383	76	27	,	,	PUNCT
bracis-28383	76	28	and	and	CCONJ
bracis-28383	76	29	which	which	DET
bracis-28383	76	30	ones	one	NOUN
bracis-28383	76	31	we	we	PRON
bracis-28383	76	32	used	use	VERB
bracis-28383	76	33	in	in	ADP
bracis-28383	76	34	our	our	PRON
bracis-28383	76	35	experiments	experiment	NOUN
bracis-28383	76	36	.	.	PUNCT
bracis-28383	77	1	our	our	PRON
bracis-28383	77	2	only	only	ADJ
bracis-28383	77	3	motivation	motivation	NOUN
bracis-28383	77	4	for	for	ADP
bracis-28383	77	5	choosing	choose	VERB
bracis-28383	77	6	these	these	DET
bracis-28383	77	7	datasets	dataset	NOUN
bracis-28383	77	8	was	be	AUX
bracis-28383	77	9	that	that	SCONJ
bracis-28383	77	10	they	they	PRON
bracis-28383	77	11	were	be	AUX
bracis-28383	77	12	public	public	ADJ
bracis-28383	77	13	and	and	CCONJ
bracis-28383	77	14	freely	freely	ADV
bracis-28383	77	15	available	available	ADJ
bracis-28383	77	16	for	for	ADP
bracis-28383	77	17	download	download	NOUN
bracis-28383	77	18	.	.	PUNCT
bracis-28383	78	1	table	table	NOUN
bracis-28383	78	2	1	1	NUM
bracis-28383	78	3	.	.	PUNCT
bracis-28383	79	1	datasets	dataset	NOUN
bracis-28383	79	2	found	find	VERB
bracis-28383	79	3	in	in	ADP
bracis-28383	79	4	the	the	DET
bracis-28383	79	5	literature	literature	NOUN
bracis-28383	79	6	for	for	ADP
bracis-28383	79	7	the	the	DET
bracis-28383	79	8	task	task	NOUN
bracis-28383	79	9	of	of	ADP
bracis-28383	79	10	facial	facial	ADJ
bracis-28383	79	11	identification	identification	NOUN
bracis-28383	79	12	in	in	ADP
bracis-28383	79	13	dogs.full	dogs.full	DET
bracis-28383	79	14	size	size	NOUN
bracis-28383	79	15	table	table	NOUN
bracis-28383	79	16	the	the	DET
bracis-28383	79	17	dogfacenet	dogfacenet	NOUN
bracis-28383	79	18	dataset	dataset	NOUN
bracis-28383	79	19	[	[	X
bracis-28383	79	20	16	16	NUM
bracis-28383	79	21	]	]	PUNCT
bracis-28383	79	22	,	,	PUNCT
bracis-28383	79	23	contains	contain	VERB
bracis-28383	79	24	8	8	NUM
bracis-28383	79	25	,	,	PUNCT
bracis-28383	79	26	 	 	SPACE
bracis-28383	79	27	363	363	NUM
bracis-28383	79	28	images	image	NOUN
bracis-28383	79	29	from	from	ADP
bracis-28383	79	30	1	1	NUM
bracis-28383	79	31	,	,	PUNCT
bracis-28383	79	32	 	 	SPACE
bracis-28383	79	33	393	393	NUM
bracis-28383	79	34	dog	dog	NOUN
bracis-28383	79	35	identities	identity	NOUN
bracis-28383	79	36	.	.	PUNCT
bracis-28383	80	1	each	each	DET
bracis-28383	80	2	dog	dog	NOUN
bracis-28383	80	3	has	have	VERB
bracis-28383	80	4	at	at	ADV
bracis-28383	80	5	least	least	ADV
bracis-28383	80	6	two	two	NUM
bracis-28383	80	7	jpeg	jpeg	NOUN
bracis-28383	80	8	images	image	NOUN
bracis-28383	80	9	with	with	ADP
bracis-28383	80	10	\(224\times	\(224\time	NOUN
bracis-28383	80	11	224\	224\	NOUN
bracis-28383	80	12	)	)	PUNCT
bracis-28383	80	13	rgb	rgb	PROPN
bracis-28383	80	14	pixels	pixel	NOUN
bracis-28383	80	15	.	.	PUNCT
bracis-28383	81	1	despite	despite	SCONJ
bracis-28383	81	2	the	the	DET
bracis-28383	81	3	previous	previous	ADJ
bracis-28383	81	4	works	work	NOUN
bracis-28383	81	5	of	of	ADP
bracis-28383	81	6	[	[	X
bracis-28383	81	7	15	15	NUM
bracis-28383	81	8	,	,	PUNCT
bracis-28383	81	9	16	16	NUM
bracis-28383	81	10	]	]	PUNCT
bracis-28383	81	11	claim	claim	NOUN
bracis-28383	81	12	that	that	SCONJ
bracis-28383	81	13	dogfacenet	dogfacenet	NOUN
bracis-28383	81	14	is	be	AUX
bracis-28383	81	15	the	the	DET
bracis-28383	81	16	first	first	ADJ
bracis-28383	81	17	public	public	ADJ
bracis-28383	81	18	dataset	dataset	NOUN
bracis-28383	81	19	for	for	ADP
bracis-28383	81	20	dog	dog	NOUN
bracis-28383	81	21	face	face	NOUN
bracis-28383	81	22	recognition	recognition	NOUN
bracis-28383	81	23	and	and	CCONJ
bracis-28383	81	24	verification	verification	NOUN
bracis-28383	81	25	.	.	PUNCT
bracis-28383	82	1	examples	example	NOUN
bracis-28383	82	2	of	of	ADP
bracis-28383	82	3	the	the	DET
bracis-28383	82	4	dogfacenetfootnote	dogfacenetfootnote	NOUN
bracis-28383	82	5	1	1	NUM
bracis-28383	82	6	are	be	AUX
bracis-28383	82	7	shown	show	VERB
bracis-28383	82	8	at	at	ADP
bracis-28383	82	9	the	the	DET
bracis-28383	82	10	top	top	NOUN
bracis-28383	82	11	of	of	ADP
bracis-28383	82	12	fig	fig	NOUN
bracis-28383	82	13	.	.	PUNCT
bracis-28383	82	14	 	 	SPACE
bracis-28383	83	1	1	1	NUM
bracis-28383	83	2	,	,	PUNCT
bracis-28383	83	3	with	with	ADP
bracis-28383	83	4	3	3	NUM
bracis-28383	83	5	identities	identity	NOUN
bracis-28383	83	6	with	with	ADP
bracis-28383	83	7	5	5	NUM
bracis-28383	83	8	images	image	NOUN
bracis-28383	83	9	each	each	PRON
bracis-28383	83	10	.	.	PUNCT
bracis-28383	84	1	it	it	PRON
bracis-28383	84	2	is	be	AUX
bracis-28383	84	3	essential	essential	ADJ
bracis-28383	84	4	to	to	PART
bracis-28383	84	5	highlight	highlight	VERB
bracis-28383	84	6	that	that	SCONJ
bracis-28383	84	7	the	the	DET
bracis-28383	84	8	images	image	NOUN
bracis-28383	84	9	were	be	AUX
bracis-28383	84	10	cropped	crop	VERB
bracis-28383	84	11	and	and	CCONJ
bracis-28383	84	12	aligned	align	VERB
bracis-28383	84	13	beforehand	beforehand	ADV
bracis-28383	84	14	.	.	PUNCT
bracis-28383	85	1	however	however	ADV
bracis-28383	85	2	,	,	PUNCT
bracis-28383	85	3	this	this	DET
bracis-28383	85	4	process	process	NOUN
bracis-28383	85	5	introduces	introduce	VERB
bracis-28383	85	6	black	black	ADJ
bracis-28383	85	7	borders	border	NOUN
bracis-28383	85	8	in	in	ADP
bracis-28383	85	9	the	the	DET
bracis-28383	85	10	images	image	NOUN
bracis-28383	85	11	.	.	PUNCT
bracis-28383	86	1	moreover	moreover	ADV
bracis-28383	86	2	,	,	PUNCT
bracis-28383	86	3	there	there	PRON
bracis-28383	86	4	are	be	VERB
bracis-28383	86	5	many	many	ADJ
bracis-28383	86	6	samples	sample	NOUN
bracis-28383	86	7	where	where	SCONJ
bracis-28383	86	8	the	the	DET
bracis-28383	86	9	dog	dog	NOUN
bracis-28383	86	10	appears	appear	VERB
bracis-28383	86	11	in	in	ADP
bracis-28383	86	12	profile	profile	NOUN
bracis-28383	86	13	or	or	CCONJ
bracis-28383	86	14	their	their	PRON
bracis-28383	86	15	face	face	NOUN
bracis-28383	86	16	is	be	AUX
bracis-28383	86	17	partially	partially	ADV
bracis-28383	86	18	occluded	occlude	VERB
bracis-28383	86	19	.	.	PUNCT
bracis-28383	87	1	the	the	DET
bracis-28383	87	2	flickr	flickr	PROPN
bracis-28383	87	3	-	-	PUNCT
bracis-28383	87	4	dog	dog	NOUN
bracis-28383	87	5	dataset	dataset	NOUN
bracis-28383	87	6	[	[	X
bracis-28383	87	7	15	15	NUM
bracis-28383	87	8	]	]	PUNCT
bracis-28383	87	9	is	be	AUX
bracis-28383	87	10	22.3	22.3	NUM
bracis-28383	87	11	times	time	NOUN
bracis-28383	87	12	smaller	small	ADJ
bracis-28383	87	13	than	than	ADP
bracis-28383	87	14	dogfacenet	dogfacenet	NOUN
bracis-28383	87	15	,	,	PUNCT
bracis-28383	87	16	containing	contain	VERB
bracis-28383	87	17	only	only	ADV
bracis-28383	87	18	374	374	NUM
bracis-28383	87	19	images	image	NOUN
bracis-28383	87	20	of	of	ADP
bracis-28383	87	21	dogs	dog	NOUN
bracis-28383	87	22	from	from	ADP
bracis-28383	87	23	two	two	NUM
bracis-28383	87	24	breeds	breed	NOUN
bracis-28383	87	25	:	:	PUNCT
bracis-28383	87	26	pugs	pug	NOUN
bracis-28383	87	27	and	and	CCONJ
bracis-28383	87	28	huskies	husky	NOUN
bracis-28383	87	29	.	.	PUNCT
bracis-28383	88	1	the	the	DET
bracis-28383	88	2	dataset	dataset	NOUN
bracis-28383	88	3	includes	include	VERB
bracis-28383	88	4	21	21	NUM
bracis-28383	88	5	identities	identity	NOUN
bracis-28383	88	6	for	for	ADP
bracis-28383	88	7	each	each	DET
bracis-28383	88	8	dog	dog	NOUN
bracis-28383	88	9	breed	breed	NOUN
bracis-28383	88	10	,	,	PUNCT
bracis-28383	88	11	totaling	total	VERB
bracis-28383	88	12	42	42	NUM
bracis-28383	88	13	for	for	ADP
bracis-28383	88	14	individual	individual	ADJ
bracis-28383	88	15	dogs	dog	NOUN
bracis-28383	88	16	and	and	CCONJ
bracis-28383	88	17	thus	thus	ADV
bracis-28383	88	18	5	5	NUM
bracis-28383	88	19	images	image	NOUN
bracis-28383	88	20	per	per	ADP
bracis-28383	88	21	identity	identity	NOUN
bracis-28383	88	22	.	.	PUNCT
bracis-28383	89	1	this	this	PRON
bracis-28383	89	2	means	mean	VERB
bracis-28383	89	3	that	that	SCONJ
bracis-28383	89	4	flickr	flickr	NOUN
bracis-28383	89	5	-	-	PUNCT
bracis-28383	89	6	dog	dog	NOUN
bracis-28383	89	7	provides	provide	VERB
bracis-28383	89	8	33.1	33.1	NUM
bracis-28383	89	9	times	time	NOUN
bracis-28383	89	10	fewer	few	ADJ
bracis-28383	89	11	identities	identity	NOUN
bracis-28383	89	12	than	than	ADP
bracis-28383	89	13	dogfacenet	dogfacenet	NOUN
bracis-28383	89	14	,	,	PUNCT
bracis-28383	89	15	so	so	SCONJ
bracis-28383	89	16	the	the	DET
bracis-28383	89	17	task	task	NOUN
bracis-28383	89	18	of	of	ADP
bracis-28383	89	19	face	face	NOUN
bracis-28383	89	20	recognition	recognition	NOUN
bracis-28383	89	21	is	be	AUX
bracis-28383	89	22	also	also	ADV
bracis-28383	89	23	related	relate	VERB
bracis-28383	89	24	to	to	ADP
bracis-28383	89	25	the	the	DET
bracis-28383	89	26	challenge	challenge	NOUN
bracis-28383	89	27	of	of	ADP
bracis-28383	89	28	discerning	discern	VERB
bracis-28383	89	29	from	from	ADP
bracis-28383	89	30	multiple	multiple	ADJ
bracis-28383	89	31	dogs	dog	NOUN
bracis-28383	89	32	of	of	ADP
bracis-28383	89	33	the	the	DET
bracis-28383	89	34	same	same	ADJ
bracis-28383	89	35	breed	breed	NOUN
bracis-28383	89	36	.	.	PUNCT
bracis-28383	90	1	each	each	DET
bracis-28383	90	2	image	image	NOUN
bracis-28383	90	3	in	in	ADP
bracis-28383	90	4	the	the	DET
bracis-28383	90	5	dataset	dataset	NOUN
bracis-28383	90	6	has	have	VERB
bracis-28383	90	7	\(250\times	\(250\time	NOUN
bracis-28383	90	8	250\	250\	NUM
bracis-28383	90	9	)	)	PUNCT
bracis-28383	90	10	rgb	rgb	PROPN
bracis-28383	90	11	pixels	pixel	NOUN
bracis-28383	90	12	containing	contain	VERB
bracis-28383	90	13	a	a	DET
bracis-28383	90	14	horizontally	horizontally	ADV
bracis-28383	90	15	aligned	align	VERB
bracis-28383	90	16	face	face	NOUN
bracis-28383	90	17	of	of	ADP
bracis-28383	90	18	a	a	DET
bracis-28383	90	19	dog	dog	NOUN
bracis-28383	90	20	.	.	PUNCT
bracis-28383	91	1	also	also	ADV
bracis-28383	91	2	,	,	PUNCT
bracis-28383	91	3	the	the	DET
bracis-28383	91	4	images	image	NOUN
bracis-28383	91	5	are	be	AUX
bracis-28383	91	6	normalized	normalize	VERB
bracis-28383	91	7	such	such	ADJ
bracis-28383	91	8	that	that	SCONJ
bracis-28383	91	9	all	all	DET
bracis-28383	91	10	images	image	NOUN
bracis-28383	91	11	are	be	AUX
bracis-28383	91	12	cropped	crop	VERB
bracis-28383	91	13	,	,	PUNCT
bracis-28383	91	14	(	(	PUNCT
bracis-28383	91	15	loosely	loosely	ADV
bracis-28383	91	16	)	)	PUNCT
bracis-28383	91	17	aligned	align	VERB
bracis-28383	91	18	,	,	PUNCT
bracis-28383	91	19	and	and	CCONJ
bracis-28383	91	20	resized	resize	VERB
bracis-28383	91	21	.	.	PUNCT
bracis-28383	92	1	five	five	NUM
bracis-28383	92	2	sample	sample	NOUN
bracis-28383	92	3	images	image	NOUN
bracis-28383	92	4	are	be	AUX
bracis-28383	92	5	shown	show	VERB
bracis-28383	92	6	at	at	ADP
bracis-28383	92	7	the	the	DET
bracis-28383	92	8	bottom	bottom	NOUN
bracis-28383	92	9	of	of	ADP
bracis-28383	92	10	fig	fig	NOUN
bracis-28383	92	11	.	.	PUNCT
bracis-28383	92	12	 	 	SPACE
bracis-28383	93	1	1	1	NUM
bracis-28383	93	2	for	for	ADP
bracis-28383	93	3	three	three	NUM
bracis-28383	93	4	individuals	individual	NOUN
bracis-28383	93	5	:	:	PUNCT
bracis-28383	93	6	lighting	lighting	NOUN
bracis-28383	93	7	conditions	condition	NOUN
bracis-28383	93	8	vary	vary	VERB
bracis-28383	93	9	significantly	significantly	ADV
bracis-28383	93	10	,	,	PUNCT
bracis-28383	93	11	there	there	PRON
bracis-28383	93	12	are	be	VERB
bracis-28383	93	13	artifacts	artifact	NOUN
bracis-28383	93	14	,	,	PUNCT
bracis-28383	93	15	and	and	CCONJ
bracis-28383	93	16	also	also	ADV
bracis-28383	93	17	black	black	ADJ
bracis-28383	93	18	borders	border	NOUN
bracis-28383	93	19	are	be	AUX
bracis-28383	93	20	evident	evident	ADJ
bracis-28383	93	21	.	.	PUNCT
bracis-28383	94	1	fig	fig	NOUN
bracis-28383	94	2	.	.	PUNCT
bracis-28383	95	1	1	1	X
bracis-28383	95	2	.	.	X
bracis-28383	95	3	public	public	ADJ
bracis-28383	95	4	dog	dog	NOUN
bracis-28383	95	5	face	face	NOUN
bracis-28383	95	6	recognition	recognition	NOUN
bracis-28383	95	7	datasets	dataset	NOUN
bracis-28383	95	8	:	:	PUNCT
bracis-28383	95	9	dogfacenet	dogfacenet	NOUN
bracis-28383	95	10	(	(	PUNCT
bracis-28383	95	11	top	top	ADJ
bracis-28383	95	12	)	)	PUNCT
bracis-28383	95	13	and	and	CCONJ
bracis-28383	95	14	flickr	flickr	NOUN
bracis-28383	95	15	-	-	PUNCT
bracis-28383	95	16	dog	dog	NOUN
bracis-28383	95	17	(	(	PUNCT
bracis-28383	95	18	bottom	bottom	NOUN
bracis-28383	95	19	)	)	PUNCT
bracis-28383	95	20	.	.	PUNCT
bracis-28383	96	1	many	many	ADJ
bracis-28383	96	2	faces	face	NOUN
bracis-28383	96	3	are	be	AUX
bracis-28383	96	4	misaligned	misalign	VERB
bracis-28383	96	5	and/or	and/or	CCONJ
bracis-28383	96	6	appear	appear	VERB
bracis-28383	96	7	at	at	ADP
bracis-28383	96	8	challenging	challenge	VERB
bracis-28383	96	9	poses	pose	NOUN
bracis-28383	96	10	in	in	ADP
bracis-28383	96	11	both	both	DET
bracis-28383	96	12	datasets	dataset	NOUN
bracis-28383	96	13	.	.	PUNCT
bracis-28383	97	1	full	full	ADJ
bracis-28383	97	2	size	size	NOUN
bracis-28383	97	3	image	image	NOUN
bracis-28383	97	4	3.2	3.2	NUM
bracis-28383	97	5	dog	dog	NOUN
bracis-28383	97	6	face	face	NOUN
bracis-28383	97	7	recognition	recognition	NOUN
bracis-28383	97	8	methods	method	NOUN
bracis-28383	97	9	in	in	ADP
bracis-28383	97	10	the	the	DET
bracis-28383	97	11	work	work	NOUN
bracis-28383	97	12	of	of	ADP
bracis-28383	97	13	[	[	X
bracis-28383	97	14	23	23	NUM
bracis-28383	97	15	]	]	PUNCT
bracis-28383	97	16	,	,	PUNCT
bracis-28383	97	17	a	a	DET
bracis-28383	97	18	vgg16	vgg16	NOUN
bracis-28383	97	19	-	-	PUNCT
bracis-28383	97	20	based	base	VERB
bracis-28383	97	21	[	[	X
bracis-28383	97	22	18	18	NUM
bracis-28383	97	23	]	]	X
bracis-28383	97	24	cnn	cnn	PROPN
bracis-28383	97	25	is	be	AUX
bracis-28383	97	26	trained	train	VERB
bracis-28383	97	27	to	to	PART
bracis-28383	97	28	classify	classify	VERB
bracis-28383	97	29	dogs	dog	NOUN
bracis-28383	97	30	in	in	ADP
bracis-28383	97	31	5	5	NUM
bracis-28383	97	32	categories	category	NOUN
bracis-28383	97	33	using	use	VERB
bracis-28383	97	34	transfer	transfer	NOUN
bracis-28383	97	35	learning	learning	NOUN
bracis-28383	97	36	.	.	PUNCT
bracis-28383	98	1	then	then	ADV
bracis-28383	98	2	,	,	PUNCT
bracis-28383	98	3	128	128	NUM
bracis-28383	98	4	-	-	PUNCT
bracis-28383	98	5	dimensional	dimensional	ADJ
bracis-28383	98	6	embeddings	embedding	NOUN
bracis-28383	98	7	are	be	AUX
bracis-28383	98	8	extracted	extract	VERB
bracis-28383	98	9	from	from	ADP
bracis-28383	98	10	this	this	DET
bracis-28383	98	11	cnn	cnn	NOUN
bracis-28383	98	12	and	and	CCONJ
bracis-28383	98	13	used	use	VERB
bracis-28383	98	14	for	for	ADP
bracis-28383	98	15	identification	identification	NOUN
bracis-28383	98	16	by	by	ADP
bracis-28383	98	17	training	train	VERB
bracis-28383	98	18	an	an	DET
bracis-28383	98	19	unsupervised	unsupervised	ADJ
bracis-28383	98	20	model	model	NOUN
bracis-28383	98	21	for	for	ADP
bracis-28383	98	22	distance	distance	NOUN
bracis-28383	98	23	calculation	calculation	NOUN
bracis-28383	98	24	.	.	PUNCT
bracis-28383	99	1	according	accord	VERB
bracis-28383	99	2	to	to	ADP
bracis-28383	99	3	the	the	DET
bracis-28383	99	4	authors	author	NOUN
bracis-28383	99	5	,	,	PUNCT
bracis-28383	99	6	experimental	experimental	ADJ
bracis-28383	99	7	evaluation	evaluation	NOUN
bracis-28383	99	8	resulted	result	VERB
bracis-28383	99	9	in	in	ADP
bracis-28383	99	10	more	more	ADJ
bracis-28383	99	11	than	than	ADP
bracis-28383	99	12	\(90\%\	\(90\%\	NOUN
bracis-28383	99	13	)	)	PUNCT
bracis-28383	99	14	accuracy	accuracy	NOUN
bracis-28383	99	15	using	use	VERB
bracis-28383	99	16	their	their	PRON
bracis-28383	99	17	dataset	dataset	NOUN
bracis-28383	99	18	where	where	SCONJ
bracis-28383	99	19	dogs	dog	NOUN
bracis-28383	99	20	appear	appear	VERB
bracis-28383	99	21	in	in	ADP
bracis-28383	99	22	various	various	ADJ
bracis-28383	99	23	poses	pose	NOUN
bracis-28383	99	24	and	and	CCONJ
bracis-28383	99	25	sizes	size	NOUN
bracis-28383	99	26	,	,	PUNCT
bracis-28383	99	27	but	but	CCONJ
bracis-28383	99	28	the	the	DET
bracis-28383	99	29	work	work	NOUN
bracis-28383	99	30	is	be	AUX
bracis-28383	99	31	missing	miss	VERB
bracis-28383	99	32	evaluation	evaluation	NOUN
bracis-28383	99	33	details	detail	NOUN
bracis-28383	99	34	of	of	ADP
bracis-28383	99	35	how	how	SCONJ
bracis-28383	99	36	dog	dog	NOUN
bracis-28383	99	37	faces	face	NOUN
bracis-28383	99	38	are	be	AUX
bracis-28383	99	39	processed	process	VERB
bracis-28383	99	40	.	.	PUNCT
bracis-28383	100	1	moreover	moreover	ADV
bracis-28383	100	2	,	,	PUNCT
bracis-28383	100	3	the	the	DET
bracis-28383	100	4	dataset	dataset	NOUN
bracis-28383	100	5	is	be	AUX
bracis-28383	100	6	private	private	ADJ
bracis-28383	100	7	.	.	PUNCT
bracis-28383	101	1	[	[	X
bracis-28383	101	2	15	15	NUM
bracis-28383	101	3	]	]	PUNCT
bracis-28383	101	4	proposed	propose	VERB
bracis-28383	101	5	a	a	DET
bracis-28383	101	6	pioneer	pioneer	NOUN
bracis-28383	101	7	work	work	NOUN
bracis-28383	101	8	in	in	ADP
bracis-28383	101	9	dog	dog	NOUN
bracis-28383	101	10	face	face	NOUN
bracis-28383	101	11	recognition	recognition	NOUN
bracis-28383	101	12	.	.	PUNCT
bracis-28383	102	1	the	the	DET
bracis-28383	102	2	authors	author	NOUN
bracis-28383	102	3	evaluated	evaluate	VERB
bracis-28383	102	4	four	four	NUM
bracis-28383	102	5	classic	classic	ADJ
bracis-28383	102	6	approaches	approach	NOUN
bracis-28383	102	7	(	(	PUNCT
bracis-28383	102	8	eigenfaces	eigenface	NOUN
bracis-28383	102	9	,	,	PUNCT
bracis-28383	102	10	fisherfaces	fisherface	NOUN
bracis-28383	102	11	,	,	PUNCT
bracis-28383	102	12	lbph	lbph	NOUN
bracis-28383	102	13	,	,	PUNCT
bracis-28383	102	14	and	and	CCONJ
bracis-28383	102	15	sparse	sparse	ADJ
bracis-28383	102	16	)	)	PUNCT
bracis-28383	102	17	against	against	ADP
bracis-28383	102	18	two	two	NUM
bracis-28383	102	19	cnns	cnn	NOUN
bracis-28383	102	20	,	,	PUNCT
bracis-28383	102	21	bark	bark	NOUN
bracis-28383	102	22	and	and	CCONJ
bracis-28383	102	23	woof	woof	VERB
bracis-28383	102	24	,	,	PUNCT
bracis-28383	102	25	for	for	ADP
bracis-28383	102	26	retrieving	retrieve	VERB
bracis-28383	102	27	lost	lose	VERB
bracis-28383	102	28	dogs	dog	NOUN
bracis-28383	102	29	based	base	VERB
bracis-28383	102	30	on	on	ADP
bracis-28383	102	31	facial	facial	ADJ
bracis-28383	102	32	features	feature	NOUN
bracis-28383	102	33	.	.	PUNCT
bracis-28383	103	1	these	these	DET
bracis-28383	103	2	cnns	cnn	NOUN
bracis-28383	103	3	significantly	significantly	ADV
bracis-28383	103	4	outperformed	outperform	VERB
bracis-28383	103	5	the	the	DET
bracis-28383	103	6	other	other	ADJ
bracis-28383	103	7	methods	method	NOUN
bracis-28383	103	8	,	,	PUNCT
bracis-28383	103	9	obtaining	obtain	VERB
bracis-28383	103	10	an	an	DET
bracis-28383	103	11	accuracy	accuracy	NOUN
bracis-28383	103	12	of	of	ADP
bracis-28383	103	13	\(89.4\%\	\(89.4\%\	PROPN
bracis-28383	103	14	)	)	PUNCT
bracis-28383	103	15	and	and	CCONJ
bracis-28383	103	16	\(81.1\%\	\(81.1\%\	PROPN
bracis-28383	103	17	)	)	PUNCT
bracis-28383	103	18	for	for	ADP
bracis-28383	103	19	woof	woof	NOUN
bracis-28383	103	20	and	and	CCONJ
bracis-28383	103	21	bark	bark	NOUN
bracis-28383	103	22	,	,	PUNCT
bracis-28383	103	23	respectively	respectively	ADV
bracis-28383	103	24	,	,	PUNCT
bracis-28383	103	25	for	for	ADP
bracis-28383	103	26	the	the	DET
bracis-28383	103	27	single	single	ADJ
bracis-28383	103	28	-	-	PUNCT
bracis-28383	103	29	breed	breed	NOUN
bracis-28383	103	30	snoopybook	snoopybook	NOUN
bracis-28383	103	31	dataset	dataset	NOUN
bracis-28383	103	32	,	,	PUNCT
bracis-28383	103	33	which	which	PRON
bracis-28383	103	34	is	be	AUX
bracis-28383	103	35	not	not	PART
bracis-28383	103	36	publicly	publicly	ADV
bracis-28383	103	37	available	available	ADJ
bracis-28383	103	38	.	.	PUNCT
bracis-28383	104	1	moreover	moreover	ADV
bracis-28383	104	2	,	,	PUNCT
bracis-28383	104	3	the	the	DET
bracis-28383	104	4	accuracy	accuracy	NOUN
bracis-28383	104	5	is	be	AUX
bracis-28383	104	6	considerably	considerably	ADV
bracis-28383	104	7	lower	low	ADJ
bracis-28383	104	8	considering	consider	VERB
bracis-28383	104	9	the	the	DET
bracis-28383	104	10	evaluation	evaluation	NOUN
bracis-28383	104	11	on	on	ADP
bracis-28383	104	12	the	the	DET
bracis-28383	104	13	flickr	flickr	ADJ
bracis-28383	104	14	-	-	PUNCT
bracis-28383	104	15	dog	dog	NOUN
bracis-28383	104	16	dataset	dataset	NOUN
bracis-28383	104	17	:	:	PUNCT
bracis-28383	104	18	\(66.9\%\	\(66.9\%\	X
bracis-28383	104	19	)	)	PUNCT
bracis-28383	104	20	and	and	CCONJ
bracis-28383	104	21	\(67.6\%\	\(67.6\%\	PROPN
bracis-28383	104	22	)	)	PUNCT
bracis-28383	104	23	for	for	ADP
bracis-28383	104	24	woof	woof	NOUN
bracis-28383	104	25	and	and	CCONJ
bracis-28383	104	26	bark	bark	NOUN
bracis-28383	104	27	,	,	PUNCT
bracis-28383	104	28	respectively	respectively	ADV
bracis-28383	104	29	.	.	PUNCT
bracis-28383	105	1	in	in	ADP
bracis-28383	105	2	[	[	X
bracis-28383	105	3	3	3	NUM
bracis-28383	105	4	]	]	PUNCT
bracis-28383	105	5	,	,	PUNCT
bracis-28383	105	6	in	in	ADP
bracis-28383	105	7	order	order	NOUN
bracis-28383	105	8	to	to	PART
bracis-28383	105	9	identify	identify	VERB
bracis-28383	105	10	animals	animal	NOUN
bracis-28383	105	11	,	,	PUNCT
bracis-28383	105	12	the	the	DET
bracis-28383	105	13	authors	author	NOUN
bracis-28383	105	14	compared	compare	VERB
bracis-28383	105	15	two	two	NUM
bracis-28383	105	16	different	different	ADJ
bracis-28383	105	17	cnn	cnn	PROPN
bracis-28383	105	18	architectures	architecture	NOUN
bracis-28383	105	19	(	(	PUNCT
bracis-28383	105	20	vgg-16	vgg-16	NOUN
bracis-28383	105	21	and	and	CCONJ
bracis-28383	105	22	inception	inception	PROPN
bracis-28383	105	23	-	-	PUNCT
bracis-28383	105	24	v3	v3	NOUN
bracis-28383	105	25	)	)	PUNCT
bracis-28383	105	26	originally	originally	ADV
bracis-28383	105	27	applied	apply	VERB
bracis-28383	105	28	in	in	ADP
bracis-28383	105	29	object	object	NOUN
bracis-28383	105	30	detection	detection	NOUN
bracis-28383	105	31	.	.	PUNCT
bracis-28383	106	1	these	these	DET
bracis-28383	106	2	two	two	NUM
bracis-28383	106	3	architectures	architecture	NOUN
bracis-28383	106	4	are	be	AUX
bracis-28383	106	5	applied	apply	VERB
bracis-28383	106	6	to	to	ADP
bracis-28383	106	7	the	the	DET
bracis-28383	106	8	flickr	flickr	NOUN
bracis-28383	106	9	-	-	PUNCT
bracis-28383	106	10	dog	dog	NOUN
bracis-28383	106	11	and	and	CCONJ
bracis-28383	106	12	dogsvscats	dogsvscat	NOUN
bracis-28383	106	13	datasets	dataset	NOUN
bracis-28383	106	14	.	.	PUNCT
bracis-28383	107	1	the	the	DET
bracis-28383	107	2	accuracy	accuracy	NOUN
bracis-28383	107	3	achieved	achieve	VERB
bracis-28383	107	4	by	by	ADP
bracis-28383	107	5	vgg-16	vgg-16	NOUN
bracis-28383	107	6	on	on	ADP
bracis-28383	107	7	flickr	flickr	NOUN
bracis-28383	107	8	-	-	PUNCT
bracis-28383	107	9	dog	dog	NOUN
bracis-28383	107	10	was	be	AUX
bracis-28383	107	11	\(95\%\	\(95\%\	PRON
bracis-28383	107	12	)	)	PUNCT
bracis-28383	107	13	and	and	CCONJ
bracis-28383	107	14	\(98\%\	\(98\%\	NOUN
bracis-28383	107	15	)	)	PUNCT
bracis-28383	107	16	on	on	ADP
bracis-28383	107	17	dogsvscats	dogsvscat	NOUN
bracis-28383	107	18	,	,	PUNCT
bracis-28383	107	19	while	while	SCONJ
bracis-28383	107	20	inception	inception	NOUN
bracis-28383	107	21	-	-	PUNCT
bracis-28383	107	22	v3	v3	PROPN
bracis-28383	107	23	reached	reach	VERB
bracis-28383	107	24	\(87\%\	\(87\%\	PRON
bracis-28383	107	25	)	)	PUNCT
bracis-28383	107	26	on	on	ADP
bracis-28383	107	27	flickr	flickr	NOUN
bracis-28383	107	28	-	-	PUNCT
bracis-28383	107	29	dog	dog	NOUN
bracis-28383	107	30	and	and	CCONJ
bracis-28383	107	31	\(85\%\	\(85\%\	NOUN
bracis-28383	107	32	)	)	PUNCT
bracis-28383	107	33	on	on	ADP
bracis-28383	107	34	the	the	DET
bracis-28383	107	35	dogsvscats	dogsvscat	NOUN
bracis-28383	107	36	dataset	dataset	VERB
bracis-28383	107	37	.	.	PUNCT
bracis-28383	108	1	dogsvscats	dogsvscat	NOUN
bracis-28383	108	2	dataset	dataset	VERB
bracis-28383	108	3	is	be	AUX
bracis-28383	108	4	a	a	DET
bracis-28383	108	5	dataset	dataset	NOUN
bracis-28383	108	6	with	with	ADP
bracis-28383	108	7	a	a	DET
bracis-28383	108	8	different	different	ADJ
bracis-28383	108	9	purpose	purpose	NOUN
bracis-28383	108	10	,	,	PUNCT
bracis-28383	108	11	which	which	PRON
bracis-28383	108	12	includes	include	VERB
bracis-28383	108	13	images	image	NOUN
bracis-28383	108	14	of	of	ADP
bracis-28383	108	15	cats	cat	NOUN
bracis-28383	108	16	,	,	PUNCT
bracis-28383	108	17	so	so	SCONJ
bracis-28383	108	18	it	it	PRON
bracis-28383	108	19	was	be	AUX
bracis-28383	108	20	not	not	PART
bracis-28383	108	21	considered	consider	VERB
bracis-28383	108	22	among	among	ADP
bracis-28383	108	23	the	the	DET
bracis-28383	108	24	datasets	dataset	NOUN
bracis-28383	108	25	in	in	ADP
bracis-28383	108	26	this	this	DET
bracis-28383	108	27	work	work	NOUN
bracis-28383	108	28	.	.	PUNCT
bracis-28383	109	1	[	[	X
bracis-28383	109	2	16	16	NUM
bracis-28383	109	3	]	]	PUNCT
bracis-28383	109	4	proposed	propose	VERB
bracis-28383	109	5	a	a	DET
bracis-28383	109	6	deep	deep	ADJ
bracis-28383	109	7	learning	learning	NOUN
bracis-28383	109	8	approach	approach	NOUN
bracis-28383	109	9	for	for	ADP
bracis-28383	109	10	dog	dog	NOUN
bracis-28383	109	11	face	face	NOUN
bracis-28383	109	12	verification	verification	NOUN
bracis-28383	109	13	and	and	CCONJ
bracis-28383	109	14	introduced	introduce	VERB
bracis-28383	109	15	the	the	DET
bracis-28383	109	16	dogfacenet	dogfacenet	NOUN
bracis-28383	109	17	dataset	dataset	NOUN
bracis-28383	109	18	.	.	PUNCT
bracis-28383	110	1	they	they	PRON
bracis-28383	110	2	evaluated	evaluate	VERB
bracis-28383	110	3	two	two	NUM
bracis-28383	110	4	models	model	NOUN
bracis-28383	110	5	based	base	VERB
bracis-28383	110	6	on	on	ADP
bracis-28383	110	7	vgg	vgg	NOUN
bracis-28383	110	8	and	and	CCONJ
bracis-28383	110	9	resnet	resnet	VERB
bracis-28383	110	10	using	use	VERB
bracis-28383	110	11	\(224\times	\(224\time	NOUN
bracis-28383	110	12	224\	224\	NOUN
bracis-28383	110	13	)	)	PUNCT
bracis-28383	110	14	rgb	rgb	PROPN
bracis-28383	110	15	images	image	NOUN
bracis-28383	110	16	as	as	ADP
bracis-28383	110	17	input	input	NOUN
bracis-28383	110	18	.	.	PUNCT
bracis-28383	111	1	deep	deep	ADJ
bracis-28383	111	2	feature	feature	NOUN
bracis-28383	111	3	embedding	embed	VERB
bracis-28383	111	4	vectors	vector	NOUN
bracis-28383	111	5	were	be	AUX
bracis-28383	111	6	generated	generate	VERB
bracis-28383	111	7	using	use	VERB
bracis-28383	111	8	triplet	triplet	NOUN
bracis-28383	111	9	loss	loss	NOUN
bracis-28383	111	10	with	with	ADP
bracis-28383	111	11	soft	soft	ADJ
bracis-28383	111	12	and	and	CCONJ
bracis-28383	111	13	hard	hard	ADJ
bracis-28383	111	14	triplet	triplet	NOUN
bracis-28383	111	15	mining	mining	NOUN
bracis-28383	111	16	,	,	PUNCT
bracis-28383	111	17	generated	generate	VERB
bracis-28383	111	18	both	both	CCONJ
bracis-28383	111	19	online	online	ADJ
bracis-28383	111	20	and	and	CCONJ
bracis-28383	111	21	offline	offline	ADJ
bracis-28383	111	22	.	.	PUNCT
bracis-28383	112	1	these	these	DET
bracis-28383	112	2	authors	author	NOUN
bracis-28383	112	3	obtained	obtain	VERB
bracis-28383	112	4	a	a	DET
bracis-28383	112	5	verification	verification	NOUN
bracis-28383	112	6	accuracy	accuracy	NOUN
bracis-28383	112	7	of	of	ADP
bracis-28383	112	8	\(86\%\	\(86\%\	NOUN
bracis-28383	112	9	)	)	PUNCT
bracis-28383	112	10	,	,	PUNCT
bracis-28383	112	11	considered	consider	VERB
bracis-28383	112	12	poor	poor	ADJ
bracis-28383	112	13	compared	compare	VERB
bracis-28383	112	14	to	to	ADP
bracis-28383	112	15	the	the	DET
bracis-28383	112	16	typical	typical	ADJ
bracis-28383	112	17	performance	performance	NOUN
bracis-28383	112	18	of	of	ADP
bracis-28383	112	19	similar	similar	ADJ
bracis-28383	112	20	approaches	approach	NOUN
bracis-28383	112	21	for	for	ADP
bracis-28383	112	22	human	human	ADJ
bracis-28383	112	23	face	face	NOUN
bracis-28383	112	24	recognition	recognition	NOUN
bracis-28383	112	25	.	.	PUNCT
bracis-28383	113	1	they	they	PRON
bracis-28383	113	2	stated	state	VERB
bracis-28383	113	3	that	that	SCONJ
bracis-28383	113	4	errors	error	NOUN
bracis-28383	113	5	arose	arise	VERB
bracis-28383	113	6	from	from	ADP
bracis-28383	113	7	frequent	frequent	ADJ
bracis-28383	113	8	occlusions	occlusion	NOUN
bracis-28383	113	9	in	in	ADP
bracis-28383	113	10	the	the	DET
bracis-28383	113	11	test	test	NOUN
bracis-28383	113	12	dataset	dataset	NOUN
bracis-28383	113	13	,	,	PUNCT
bracis-28383	113	14	dog	dog	NOUN
bracis-28383	113	15	’s	’s	PART
bracis-28383	113	16	pose	pose	NOUN
bracis-28383	113	17	,	,	PUNCT
bracis-28383	113	18	similarity	similarity	NOUN
bracis-28383	113	19	of	of	ADP
bracis-28383	113	20	different	different	ADJ
bracis-28383	113	21	dogs	dog	NOUN
bracis-28383	113	22	of	of	ADP
bracis-28383	113	23	the	the	DET
bracis-28383	113	24	same	same	ADJ
bracis-28383	113	25	breed	breed	NOUN
bracis-28383	113	26	,	,	PUNCT
bracis-28383	113	27	and	and	CCONJ
bracis-28383	113	28	light	light	ADJ
bracis-28383	113	29	exposure	exposure	NOUN
bracis-28383	113	30	.	.	PUNCT
bracis-28383	114	1	[	[	X
bracis-28383	114	2	12	12	NUM
bracis-28383	114	3	]	]	PUNCT
bracis-28383	114	4	proposed	propose	VERB
bracis-28383	114	5	the	the	DET
bracis-28383	114	6	use	use	NOUN
bracis-28383	114	7	of	of	ADP
bracis-28383	114	8	soft	soft	ADJ
bracis-28383	114	9	biometrics	biometric	NOUN
bracis-28383	114	10	for	for	ADP
bracis-28383	114	11	improving	improve	VERB
bracis-28383	114	12	identification	identification	NOUN
bracis-28383	114	13	.	.	PUNCT
bracis-28383	115	1	based	base	VERB
bracis-28383	115	2	on	on	ADP
bracis-28383	115	3	a	a	DET
bracis-28383	115	4	fast	fast	ADJ
bracis-28383	115	5	r	r	NOUN
bracis-28383	115	6	-	-	PUNCT
bracis-28383	115	7	cnn	cnn	PROPN
bracis-28383	115	8	,	,	PUNCT
bracis-28383	115	9	their	their	PRON
bracis-28383	115	10	method	method	NOUN
bracis-28383	115	11	performs	perform	VERB
bracis-28383	115	12	a	a	DET
bracis-28383	115	13	coarse	coarse	NOUN
bracis-28383	115	14	-	-	PUNCT
bracis-28383	115	15	to	to	ADP
bracis-28383	115	16	-	-	PUNCT
bracis-28383	115	17	fine	fine	ADJ
bracis-28383	115	18	identification	identification	NOUN
bracis-28383	115	19	,	,	PUNCT
bracis-28383	115	20	i.e.	i.e.	X
bracis-28383	115	21	,	,	PUNCT
bracis-28383	115	22	starting	start	VERB
bracis-28383	115	23	from	from	ADP
bracis-28383	115	24	the	the	DET
bracis-28383	115	25	breed	breed	NOUN
bracis-28383	115	26	filtering	filtering	NOUN
bracis-28383	115	27	,	,	PUNCT
bracis-28383	115	28	then	then	ADV
bracis-28383	115	29	carrying	carry	VERB
bracis-28383	115	30	out	out	ADP
bracis-28383	115	31	the	the	DET
bracis-28383	115	32	identification	identification	NOUN
bracis-28383	115	33	.	.	PUNCT
bracis-28383	116	1	their	their	PRON
bracis-28383	116	2	results	result	NOUN
bracis-28383	116	3	suggest	suggest	VERB
bracis-28383	116	4	that	that	SCONJ
bracis-28383	116	5	breed	breed	NOUN
bracis-28383	116	6	information	information	NOUN
bracis-28383	116	7	improved	improve	VERB
bracis-28383	116	8	accuracy	accuracy	NOUN
bracis-28383	116	9	for	for	ADP
bracis-28383	116	10	different	different	ADJ
bracis-28383	116	11	dog	dog	NOUN
bracis-28383	116	12	breeds	breed	NOUN
bracis-28383	116	13	,	,	PUNCT
bracis-28383	116	14	and	and	CCONJ
bracis-28383	116	15	overall	overall	ADJ
bracis-28383	116	16	accuracy	accuracy	NOUN
bracis-28383	116	17	is	be	AUX
bracis-28383	116	18	\(76.53\%\	\(76.53\%\	PROPN
bracis-28383	116	19	)	)	PUNCT
bracis-28383	116	20	for	for	SCONJ
bracis-28383	116	21	the	the	DET
bracis-28383	116	22	flickr	flickr	NOUN
bracis-28383	116	23	-	-	PUNCT
bracis-28383	116	24	dog	dog	NOUN
bracis-28383	116	25	using	use	VERB
bracis-28383	116	26	an	an	DET
bracis-28383	116	27	xception	xception	NOUN
bracis-28383	116	28	back	back	ADJ
bracis-28383	116	29	-	-	PUNCT
bracis-28383	116	30	end	end	NOUN
bracis-28383	116	31	[	[	X
bracis-28383	116	32	5	5	NUM
bracis-28383	116	33	]	]	PUNCT
bracis-28383	116	34	.	.	PUNCT
bracis-28383	117	1	the	the	DET
bracis-28383	117	2	authors	author	NOUN
bracis-28383	117	3	improved	improve	VERB
bracis-28383	117	4	this	this	DET
bracis-28383	117	5	result	result	NOUN
bracis-28383	117	6	to	to	ADP
bracis-28383	117	7	\(84.94\%\	\(84.94\%\	PROPN
bracis-28383	117	8	)	)	PUNCT
bracis-28383	117	9	by	by	ADP
bracis-28383	117	10	using	use	VERB
bracis-28383	117	11	a	a	DET
bracis-28383	117	12	likelihood	likelihood	NOUN
bracis-28383	117	13	-	-	PUNCT
bracis-28383	117	14	adjusted	adjust	VERB
bracis-28383	117	15	decision	decision	NOUN
bracis-28383	117	16	network	network	NOUN
bracis-28383	117	17	.	.	PUNCT
bracis-28383	118	1	[	[	X
bracis-28383	118	2	2	2	X
bracis-28383	118	3	]	]	PUNCT
bracis-28383	118	4	proposed	propose	VERB
bracis-28383	118	5	and	and	CCONJ
bracis-28383	118	6	evaluated	evaluate	VERB
bracis-28383	118	7	a	a	DET
bracis-28383	118	8	deep	deep	ADJ
bracis-28383	118	9	learning	learning	NOUN
bracis-28383	118	10	-	-	PUNCT
bracis-28383	118	11	based	base	VERB
bracis-28383	118	12	approach	approach	NOUN
bracis-28383	118	13	to	to	ADP
bracis-28383	118	14	pet	pet	ADJ
bracis-28383	118	15	detection	detection	NOUN
bracis-28383	118	16	and	and	CCONJ
bracis-28383	118	17	recognition	recognition	NOUN
bracis-28383	118	18	,	,	PUNCT
bracis-28383	118	19	named	name	VERB
bracis-28383	118	20	pet2netid	pet2netid	NUM
bracis-28383	118	21	,	,	PUNCT
bracis-28383	118	22	based	base	VERB
bracis-28383	118	23	on	on	ADP
bracis-28383	118	24	a	a	DET
bracis-28383	118	25	dataset	dataset	NOUN
bracis-28383	118	26	provided	provide	VERB
bracis-28383	118	27	by	by	ADP
bracis-28383	118	28	pet2network	pet2network	PROPN
bracis-28383	118	29	,	,	PUNCT
bracis-28383	118	30	a	a	DET
bracis-28383	118	31	social	social	ADJ
bracis-28383	118	32	network	network	NOUN
bracis-28383	118	33	for	for	ADP
bracis-28383	118	34	pets	pet	NOUN
bracis-28383	118	35	and	and	CCONJ
bracis-28383	118	36	their	their	PRON
bracis-28383	118	37	owners	owner	NOUN
bracis-28383	118	38	.	.	PUNCT
bracis-28383	119	1	they	they	PRON
bracis-28383	119	2	achieved	achieve	VERB
bracis-28383	119	3	\(94.59\%\	\(94.59\%\	NOUN
bracis-28383	119	4	)	)	PUNCT
bracis-28383	119	5	accuracy	accuracy	NOUN
bracis-28383	119	6	by	by	ADP
bracis-28383	119	7	combining	combine	VERB
bracis-28383	119	8	transfer	transfer	NOUN
bracis-28383	119	9	learning	learning	NOUN
bracis-28383	119	10	and	and	CCONJ
bracis-28383	119	11	object	object	VERB
bracis-28383	119	12	detection	detection	NOUN
bracis-28383	119	13	approaches	approach	NOUN
bracis-28383	119	14	with	with	ADP
bracis-28383	119	15	inception	inception	PROPN
bracis-28383	119	16	v3	v3	PROPN
bracis-28383	119	17	and	and	CCONJ
bracis-28383	119	18	ssd	ssd	PROPN
bracis-28383	119	19	inception	inception	ADJ
bracis-28383	119	20	v2	v2	PROPN
bracis-28383	119	21	on	on	ADP
bracis-28383	119	22	the	the	DET
bracis-28383	119	23	provided	provide	VERB
bracis-28383	119	24	pet2net	pet2net	NOUN
bracis-28383	119	25	dataset	dataset	NOUN
bracis-28383	119	26	,	,	PUNCT
bracis-28383	119	27	and	and	CCONJ
bracis-28383	119	28	\(77.19\%\	\(77.19\%\	NOUN
bracis-28383	119	29	)	)	PUNCT
bracis-28383	119	30	was	be	AUX
bracis-28383	119	31	achieved	achieve	VERB
bracis-28383	119	32	when	when	SCONJ
bracis-28383	119	33	training	train	VERB
bracis-28383	119	34	with	with	ADP
bracis-28383	119	35	the	the	DET
bracis-28383	119	36	pet2net	pet2net	NOUN
bracis-28383	119	37	dataset	dataset	NOUN
bracis-28383	119	38	on	on	ADP
bracis-28383	119	39	the	the	DET
bracis-28383	119	40	flickr	flickr	ADJ
bracis-28383	119	41	-	-	PUNCT
bracis-28383	119	42	dog	dog	NOUN
bracis-28383	119	43	dataset	dataset	NOUN
bracis-28383	119	44	.	.	PUNCT
bracis-28383	120	1	the	the	DET
bracis-28383	120	2	approach	approach	NOUN
bracis-28383	120	3	proposes	propose	VERB
bracis-28383	120	4	to	to	PART
bracis-28383	120	5	work	work	VERB
bracis-28383	120	6	with	with	ADP
bracis-28383	120	7	wild	wild	ADJ
bracis-28383	120	8	images	image	NOUN
bracis-28383	120	9	without	without	ADP
bracis-28383	120	10	any	any	DET
bracis-28383	120	11	pre	pre	ADJ
bracis-28383	120	12	-	-	NOUN
bracis-28383	120	13	processing	processing	NOUN
bracis-28383	120	14	,	,	PUNCT
bracis-28383	120	15	including	include	VERB
bracis-28383	120	16	other	other	ADJ
bracis-28383	120	17	objects	object	NOUN
bracis-28383	120	18	besides	besides	SCONJ
bracis-28383	120	19	the	the	DET
bracis-28383	120	20	pets	pet	NOUN
bracis-28383	120	21	themselves	themselves	PRON
bracis-28383	120	22	,	,	PUNCT
bracis-28383	120	23	and	and	CCONJ
bracis-28383	120	24	the	the	DET
bracis-28383	120	25	identification	identification	NOUN
bracis-28383	120	26	uses	use	VERB
bracis-28383	120	27	information	information	NOUN
bracis-28383	120	28	from	from	ADP
bracis-28383	120	29	the	the	DET
bracis-28383	120	30	whole	whole	ADJ
bracis-28383	120	31	animal	animal	NOUN
bracis-28383	120	32	,	,	PUNCT
bracis-28383	120	33	not	not	PART
bracis-28383	120	34	only	only	ADV
bracis-28383	120	35	the	the	DET
bracis-28383	120	36	face	face	NOUN
bracis-28383	120	37	.	.	PUNCT
bracis-28383	121	1	in	in	ADP
bracis-28383	121	2	general	general	ADJ
bracis-28383	121	3	,	,	PUNCT
bracis-28383	121	4	as	as	SCONJ
bracis-28383	121	5	seen	see	VERB
bracis-28383	121	6	in	in	ADP
bracis-28383	121	7	table	table	NOUN
bracis-28383	121	8	 	 	SPACE
bracis-28383	121	9	2	2	NUM
bracis-28383	121	10	,	,	PUNCT
bracis-28383	121	11	these	these	DET
bracis-28383	121	12	works	work	NOUN
bracis-28383	121	13	use	use	VERB
bracis-28383	121	14	deep	deep	ADJ
bracis-28383	121	15	learning	learning	NOUN
bracis-28383	121	16	methods	method	NOUN
bracis-28383	121	17	without	without	ADP
bracis-28383	121	18	preprocessing	preprocesse	VERB
bracis-28383	121	19	experimentation	experimentation	NOUN
bracis-28383	121	20	(	(	PUNCT
bracis-28383	121	21	e.g.	e.g.	ADV
bracis-28383	121	22	,	,	PUNCT
bracis-28383	121	23	face	face	NOUN
bracis-28383	121	24	detection	detection	NOUN
bracis-28383	121	25	)	)	PUNCT
bracis-28383	121	26	and	and	CCONJ
bracis-28383	121	27	do	do	AUX
bracis-28383	121	28	not	not	PART
bracis-28383	121	29	provide	provide	VERB
bracis-28383	121	30	recall	recall	NOUN
bracis-28383	121	31	results	result	NOUN
bracis-28383	121	32	,	,	PUNCT
bracis-28383	121	33	only	only	ADV
bracis-28383	121	34	accuracy	accuracy	NOUN
bracis-28383	121	35	.	.	PUNCT
bracis-28383	122	1	we	we	PRON
bracis-28383	122	2	observed	observe	VERB
bracis-28383	122	3	this	this	PRON
bracis-28383	122	4	and	and	CCONJ
bracis-28383	122	5	included	include	VERB
bracis-28383	122	6	it	it	PRON
bracis-28383	122	7	in	in	ADP
bracis-28383	122	8	our	our	PRON
bracis-28383	122	9	experiments	experiment	NOUN
bracis-28383	122	10	,	,	PUNCT
bracis-28383	122	11	described	describe	VERB
bracis-28383	122	12	in	in	ADP
bracis-28383	122	13	the	the	DET
bracis-28383	122	14	following	follow	VERB
bracis-28383	122	15	sections	section	NOUN
bracis-28383	122	16	,	,	PUNCT
bracis-28383	122	17	developing	develop	VERB
bracis-28383	122	18	a	a	DET
bracis-28383	122	19	unique	unique	ADJ
bracis-28383	122	20	approach	approach	NOUN
bracis-28383	122	21	to	to	ADP
bracis-28383	122	22	the	the	DET
bracis-28383	122	23	problem	problem	NOUN
bracis-28383	122	24	with	with	ADP
bracis-28383	122	25	a	a	DET
bracis-28383	122	26	more	more	ADV
bracis-28383	122	27	robust	robust	ADJ
bracis-28383	122	28	evaluation	evaluation	NOUN
bracis-28383	122	29	than	than	ADP
bracis-28383	122	30	the	the	DET
bracis-28383	122	31	related	relate	VERB
bracis-28383	122	32	works	work	NOUN
bracis-28383	122	33	in	in	ADP
bracis-28383	122	34	this	this	DET
bracis-28383	122	35	section	section	NOUN
bracis-28383	122	36	.	.	PUNCT
bracis-28383	123	1	table	table	NOUN
bracis-28383	123	2	2	2	NUM
bracis-28383	123	3	.	.	X
bracis-28383	123	4	comparison	comparison	NOUN
bracis-28383	123	5	between	between	ADP
bracis-28383	123	6	related	related	ADJ
bracis-28383	123	7	works.full	works.full	NOUN
bracis-28383	123	8	size	size	NOUN
bracis-28383	123	9	table	table	NOUN
bracis-28383	123	10	4	4	NUM
bracis-28383	123	11	proposed	propose	VERB
bracis-28383	123	12	dog	dog	NOUN
bracis-28383	123	13	face	face	NOUN
bracis-28383	123	14	recognition	recognition	NOUN
bracis-28383	123	15	method	method	NOUN
bracis-28383	123	16	since	since	SCONJ
bracis-28383	123	17	many	many	ADJ
bracis-28383	123	18	works	work	NOUN
bracis-28383	123	19	in	in	ADP
bracis-28383	123	20	the	the	DET
bracis-28383	123	21	literature	literature	NOUN
bracis-28383	123	22	already	already	ADV
bracis-28383	123	23	resorted	resort	VERB
bracis-28383	123	24	to	to	ADP
bracis-28383	123	25	transferring	transfer	VERB
bracis-28383	123	26	learning	learning	NOUN
bracis-28383	123	27	and	and	CCONJ
bracis-28383	123	28	fine	fine	ADV
bracis-28383	123	29	-	-	PUNCT
bracis-28383	123	30	tuning	tune	VERB
bracis-28383	123	31	existing	exist	VERB
bracis-28383	123	32	models	model	NOUN
bracis-28383	123	33	,	,	PUNCT
bracis-28383	123	34	our	our	PRON
bracis-28383	123	35	investigation	investigation	NOUN
bracis-28383	123	36	prioritized	prioritize	VERB
bracis-28383	123	37	following	follow	VERB
bracis-28383	123	38	a	a	DET
bracis-28383	123	39	different	different	ADJ
bracis-28383	123	40	strategy	strategy	NOUN
bracis-28383	123	41	,	,	PUNCT
bracis-28383	123	42	so	so	ADV
bracis-28383	123	43	we	we	PRON
bracis-28383	123	44	trained	train	VERB
bracis-28383	123	45	models	model	NOUN
bracis-28383	123	46	from	from	ADP
bracis-28383	123	47	scratch	scratch	NOUN
bracis-28383	123	48	.	.	PUNCT
bracis-28383	124	1	as	as	SCONJ
bracis-28383	124	2	stated	state	VERB
bracis-28383	124	3	before	before	ADV
bracis-28383	124	4	,	,	PUNCT
bracis-28383	124	5	the	the	DET
bracis-28383	124	6	incoming	incoming	ADJ
bracis-28383	124	7	images	image	NOUN
bracis-28383	124	8	are	be	AUX
bracis-28383	124	9	subjected	subject	VERB
bracis-28383	124	10	to	to	ADP
bracis-28383	124	11	optional	optional	ADJ
bracis-28383	124	12	preprocessing	preprocessing	NOUN
bracis-28383	124	13	,	,	PUNCT
bracis-28383	124	14	which	which	PRON
bracis-28383	124	15	can	can	AUX
bracis-28383	124	16	occur	occur	VERB
bracis-28383	124	17	in	in	ADP
bracis-28383	124	18	two	two	NUM
bracis-28383	124	19	ways	way	NOUN
bracis-28383	124	20	.	.	PUNCT
bracis-28383	125	1	however	however	ADV
bracis-28383	125	2	,	,	PUNCT
bracis-28383	125	3	only	only	ADV
bracis-28383	125	4	cropping	cropping	NOUN
bracis-28383	125	5	is	be	AUX
bracis-28383	125	6	allowed	allow	VERB
bracis-28383	125	7	in	in	ADP
bracis-28383	125	8	this	this	DET
bracis-28383	125	9	step	step	NOUN
bracis-28383	125	10	,	,	PUNCT
bracis-28383	125	11	so	so	ADV
bracis-28383	125	12	no	no	DET
bracis-28383	125	13	further	far	ADV
bracis-28383	125	14	filtering	filtering	NOUN
bracis-28383	125	15	or	or	CCONJ
bracis-28383	125	16	image	image	NOUN
bracis-28383	125	17	enhancement	enhancement	NOUN
bracis-28383	125	18	is	be	AUX
bracis-28383	125	19	applied	apply	VERB
bracis-28383	125	20	in	in	ADP
bracis-28383	125	21	favor	favor	NOUN
bracis-28383	125	22	of	of	ADP
bracis-28383	125	23	an	an	DET
bracis-28383	125	24	honest	honest	ADJ
bracis-28383	125	25	comparison	comparison	NOUN
bracis-28383	125	26	against	against	ADP
bracis-28383	125	27	other	other	ADJ
bracis-28383	125	28	methods	method	NOUN
bracis-28383	125	29	.	.	PUNCT
bracis-28383	126	1	4.1	4.1	NUM
bracis-28383	126	2	preprocessing	preprocessing	NOUN
bracis-28383	126	3	and	and	CCONJ
bracis-28383	126	4	 	 	SPACE
bracis-28383	126	5	training	training	NOUN
bracis-28383	126	6	pipeline	pipeline	NOUN
bracis-28383	126	7	the	the	DET
bracis-28383	126	8	pipeline	pipeline	NOUN
bracis-28383	126	9	of	of	ADP
bracis-28383	126	10	evaluation	evaluation	NOUN
bracis-28383	126	11	using	use	VERB
bracis-28383	126	12	the	the	DET
bracis-28383	126	13	methods	method	NOUN
bracis-28383	126	14	is	be	AUX
bracis-28383	126	15	illustrated	illustrate	VERB
bracis-28383	126	16	in	in	ADP
bracis-28383	126	17	fig	fig	NOUN
bracis-28383	126	18	.	.	PUNCT
bracis-28383	126	19	 	 	SPACE
bracis-28383	127	1	2	2	X
bracis-28383	127	2	.	.	X
bracis-28383	128	1	we	we	PRON
bracis-28383	128	2	took	take	VERB
bracis-28383	128	3	three	three	NUM
bracis-28383	128	4	different	different	ADJ
bracis-28383	128	5	preprocessing	preprocessing	NOUN
bracis-28383	128	6	paths	path	NOUN
bracis-28383	128	7	and	and	CCONJ
bracis-28383	128	8	evaluated	evaluate	VERB
bracis-28383	128	9	them	they	PRON
bracis-28383	128	10	separately	separately	ADV
bracis-28383	128	11	,	,	PUNCT
bracis-28383	128	12	these	these	PRON
bracis-28383	128	13	being	be	AUX
bracis-28383	128	14	:	:	PUNCT
bracis-28383	128	15	(	(	PUNCT
bracis-28383	128	16	a	a	X
bracis-28383	128	17	)	)	PUNCT
bracis-28383	128	18	the	the	DET
bracis-28383	128	19	dataset	dataset	NOUN
bracis-28383	128	20	in	in	ADP
bracis-28383	128	21	its	its	PRON
bracis-28383	128	22	pure	pure	ADJ
bracis-28383	128	23	form	form	NOUN
bracis-28383	128	24	;	;	PUNCT
bracis-28383	128	25	(	(	PUNCT
bracis-28383	128	26	b	b	X
bracis-28383	128	27	)	)	PUNCT
bracis-28383	128	28	a	a	DET
bracis-28383	128	29	version	version	NOUN
bracis-28383	128	30	where	where	SCONJ
bracis-28383	128	31	each	each	DET
bracis-28383	128	32	image	image	NOUN
bracis-28383	128	33	was	be	AUX
bracis-28383	128	34	cropped	crop	VERB
bracis-28383	128	35	in	in	ADP
bracis-28383	128	36	the	the	DET
bracis-28383	128	37	center	center	NOUN
bracis-28383	128	38	,	,	PUNCT
bracis-28383	128	39	which	which	PRON
bracis-28383	128	40	we	we	PRON
bracis-28383	128	41	call	call	VERB
bracis-28383	128	42	central	central	ADJ
bracis-28383	128	43	crop	crop	NOUN
bracis-28383	128	44	(	(	PUNCT
bracis-28383	128	45	cc	cc	NOUN
bracis-28383	128	46	)	)	PUNCT
bracis-28383	128	47	;	;	PUNCT
bracis-28383	128	48	and	and	CCONJ
bracis-28383	128	49	(	(	PUNCT
bracis-28383	128	50	c	c	X
bracis-28383	128	51	)	)	PUNCT
bracis-28383	128	52	another	another	DET
bracis-28383	128	53	path	path	NOUN
bracis-28383	128	54	using	use	VERB
bracis-28383	128	55	facial	facial	ADJ
bracis-28383	128	56	detection	detection	NOUN
bracis-28383	128	57	(	(	PUNCT
bracis-28383	128	58	fd	fd	PROPN
bracis-28383	128	59	)	)	PUNCT
bracis-28383	128	60	.	.	PUNCT
bracis-28383	129	1	the	the	DET
bracis-28383	129	2	use	use	NOUN
bracis-28383	129	3	of	of	ADP
bracis-28383	129	4	each	each	DET
bracis-28383	129	5	technique	technique	NOUN
bracis-28383	129	6	is	be	AUX
bracis-28383	129	7	detailed	detail	VERB
bracis-28383	129	8	below	below	ADV
bracis-28383	129	9	.	.	PUNCT
bracis-28383	130	1	a	a	DET
bracis-28383	130	2	centralized	centralized	ADJ
bracis-28383	130	3	crop	crop	NOUN
bracis-28383	130	4	is	be	AUX
bracis-28383	130	5	applied	apply	VERB
bracis-28383	130	6	to	to	PART
bracis-28383	130	7	assume	assume	VERB
bracis-28383	130	8	the	the	DET
bracis-28383	130	9	images	image	NOUN
bracis-28383	130	10	are	be	AUX
bracis-28383	130	11	satisfactorily	satisfactorily	ADV
bracis-28383	130	12	aligned	align	VERB
bracis-28383	130	13	,	,	PUNCT
bracis-28383	130	14	i.e.	i.e.	X
bracis-28383	130	15	,	,	PUNCT
bracis-28383	130	16	the	the	DET
bracis-28383	130	17	dog	dog	NOUN
bracis-28383	130	18	’s	’s	PART
bracis-28383	130	19	eyes	eye	NOUN
bracis-28383	130	20	lie	lie	VERB
bracis-28383	130	21	near	near	ADP
bracis-28383	130	22	the	the	DET
bracis-28383	130	23	center	center	NOUN
bracis-28383	130	24	of	of	ADP
bracis-28383	130	25	the	the	DET
bracis-28383	130	26	image	image	NOUN
bracis-28383	130	27	,	,	PUNCT
bracis-28383	130	28	but	but	CCONJ
bracis-28383	130	29	assuming	assume	VERB
bracis-28383	130	30	there	there	PRON
bracis-28383	130	31	is	be	VERB
bracis-28383	130	32	a	a	DET
bracis-28383	130	33	border	border	NOUN
bracis-28383	130	34	whose	whose	DET
bracis-28383	130	35	existence	existence	NOUN
bracis-28383	130	36	is	be	AUX
bracis-28383	130	37	questionable	questionable	ADJ
bracis-28383	130	38	for	for	ADP
bracis-28383	130	39	facial	facial	ADJ
bracis-28383	130	40	recognition	recognition	NOUN
bracis-28383	130	41	purposes	purpose	NOUN
bracis-28383	130	42	.	.	PUNCT
bracis-28383	131	1	the	the	DET
bracis-28383	131	2	cc	cc	ADP
bracis-28383	131	3	preprocessing	preprocesse	VERB
bracis-28383	131	4	algorithm	algorithm	NOUN
bracis-28383	131	5	extracts	extract	NOUN
bracis-28383	131	6	the	the	DET
bracis-28383	131	7	central	central	ADJ
bracis-28383	131	8	pixels	pixel	NOUN
bracis-28383	131	9	from	from	ADP
bracis-28383	131	10	a	a	DET
bracis-28383	131	11	given	give	VERB
bracis-28383	131	12	dimension	dimension	NOUN
bracis-28383	131	13	,	,	PUNCT
bracis-28383	131	14	thus	thus	ADV
bracis-28383	131	15	reducing	reduce	VERB
bracis-28383	131	16	the	the	DET
bracis-28383	131	17	input	input	NOUN
bracis-28383	131	18	image	image	NOUN
bracis-28383	131	19	size	size	NOUN
bracis-28383	131	20	.	.	PUNCT
bracis-28383	132	1	fd	fd	PROPN
bracis-28383	132	2	includes	include	VERB
bracis-28383	132	3	alignment	alignment	NOUN
bracis-28383	132	4	and	and	CCONJ
bracis-28383	132	5	performs	perform	VERB
bracis-28383	132	6	as	as	SCONJ
bracis-28383	132	7	follows	follow	VERB
bracis-28383	132	8	.	.	PUNCT
bracis-28383	133	1	first	first	ADV
bracis-28383	133	2	,	,	PUNCT
bracis-28383	133	3	the	the	DET
bracis-28383	133	4	dog	dog	NOUN
bracis-28383	133	5	face	face	NOUN
bracis-28383	133	6	is	be	AUX
bracis-28383	133	7	detected	detect	VERB
bracis-28383	133	8	using	use	VERB
bracis-28383	133	9	a	a	DET
bracis-28383	133	10	cnn	cnn	PROPN
bracis-28383	133	11	-	-	PUNCT
bracis-28383	133	12	based	base	VERB
bracis-28383	133	13	object	object	NOUN
bracis-28383	133	14	detector	detector	NOUN
bracis-28383	133	15	using	use	VERB
bracis-28383	133	16	dlib	dlib	NOUN
bracis-28383	133	17	’s	’s	PART
bracis-28383	134	1	[	[	X
bracis-28383	134	2	10	10	NUM
bracis-28383	134	3	]	]	X
bracis-28383	134	4	max	max	ADJ
bracis-28383	134	5	-	-	PUNCT
bracis-28383	134	6	margin	margin	NOUN
bracis-28383	134	7	object	object	NOUN
bracis-28383	134	8	detection	detection	NOUN
bracis-28383	134	9	(	(	PUNCT
bracis-28383	134	10	mmod	mmod	NOUN
bracis-28383	134	11	)	)	PUNCT
bracis-28383	134	12	loss	loss	NOUN
bracis-28383	134	13	layer	layer	NOUN
bracis-28383	134	14	that	that	PRON
bracis-28383	134	15	implements	implement	VERB
bracis-28383	134	16	the	the	DET
bracis-28383	134	17	object	object	NOUN
bracis-28383	134	18	detection	detection	NOUN
bracis-28383	134	19	scheme	scheme	NOUN
bracis-28383	134	20	as	as	SCONJ
bracis-28383	134	21	described	describe	VERB
bracis-28383	134	22	in	in	ADP
bracis-28383	134	23	[	[	X
bracis-28383	134	24	11	11	NUM
bracis-28383	134	25	]	]	PUNCT
bracis-28383	134	26	.	.	PUNCT
bracis-28383	135	1	despite	despite	SCONJ
bracis-28383	135	2	using	use	VERB
bracis-28383	135	3	a	a	DET
bracis-28383	135	4	concise	concise	ADJ
bracis-28383	135	5	representation	representation	NOUN
bracis-28383	135	6	,	,	PUNCT
bracis-28383	135	7	i.e.	i.e.	X
bracis-28383	135	8	,	,	PUNCT
bracis-28383	135	9	this	this	DET
bracis-28383	135	10	cnn	cnn	PROPN
bracis-28383	135	11	’s	’s	PART
bracis-28383	135	12	weights	weight	NOUN
bracis-28383	135	13	require	require	VERB
bracis-28383	135	14	only	only	ADV
bracis-28383	135	15	about	about	ADV
bracis-28383	135	16	700	700	NUM
bracis-28383	135	17	 	 	SPACE
bracis-28383	135	18	kb	kb	PROPN
bracis-28383	135	19	.	.	PUNCT
bracis-28383	136	1	however	however	ADV
bracis-28383	136	2	,	,	PUNCT
bracis-28383	136	3	mmod	mmod	PROPN
bracis-28383	136	4	inference	inference	PROPN
bracis-28383	136	5	is	be	AUX
bracis-28383	136	6	the	the	DET
bracis-28383	136	7	most	most	ADJ
bracis-28383	136	8	time	time	NOUN
bracis-28383	136	9	-	-	PUNCT
bracis-28383	136	10	consuming	consume	VERB
bracis-28383	136	11	operation	operation	NOUN
bracis-28383	136	12	in	in	ADP
bracis-28383	136	13	this	this	DET
bracis-28383	136	14	pipeline	pipeline	NOUN
bracis-28383	136	15	.	.	PUNCT
bracis-28383	137	1	second	second	ADJ
bracis-28383	137	2	,	,	PUNCT
bracis-28383	137	3	six	six	NUM
bracis-28383	137	4	facial	facial	ADJ
bracis-28383	137	5	landmarks	landmark	NOUN
bracis-28383	137	6	are	be	AUX
bracis-28383	137	7	then	then	ADV
bracis-28383	137	8	estimated	estimate	VERB
bracis-28383	137	9	using	use	VERB
bracis-28383	137	10	an	an	DET
bracis-28383	137	11	ensemble	ensemble	NOUN
bracis-28383	137	12	of	of	ADP
bracis-28383	137	13	regression	regression	NOUN
bracis-28383	137	14	trees	tree	NOUN
bracis-28383	137	15	[	[	X
bracis-28383	137	16	8	8	NUM
bracis-28383	137	17	]	]	PUNCT
bracis-28383	137	18	with	with	ADP
bracis-28383	137	19	cascade	cascade	NOUN
bracis-28383	137	20	depth	depth	NOUN
bracis-28383	137	21	\(=20\	\(=20\	NOUN
bracis-28383	137	22	)	)	PUNCT
bracis-28383	137	23	and	and	CCONJ
bracis-28383	137	24	tree	tree	NOUN
bracis-28383	137	25	depth	depth	NOUN
bracis-28383	137	26	\(=5\	\(=5\	NOUN
bracis-28383	137	27	)	)	PUNCT
bracis-28383	137	28	.	.	PUNCT
bracis-28383	138	1	finally	finally	ADV
bracis-28383	138	2	,	,	PUNCT
bracis-28383	138	3	these	these	DET
bracis-28383	138	4	points	point	NOUN
bracis-28383	138	5	(	(	PUNCT
bracis-28383	138	6	shown	show	VERB
bracis-28383	138	7	in	in	ADP
bracis-28383	138	8	fig	fig	NOUN
bracis-28383	138	9	.	.	PUNCT
bracis-28383	138	10	 	 	SPACE
bracis-28383	139	1	3	3	X
bracis-28383	139	2	)	)	PUNCT
bracis-28383	139	3	are	be	AUX
bracis-28383	139	4	converted	convert	VERB
bracis-28383	139	5	into	into	ADP
bracis-28383	139	6	another	another	DET
bracis-28383	139	7	layout	layout	NOUN
bracis-28383	139	8	before	before	SCONJ
bracis-28383	139	9	they	they	PRON
bracis-28383	139	10	are	be	AUX
bracis-28383	139	11	used	use	VERB
bracis-28383	139	12	to	to	PART
bracis-28383	139	13	extract	extract	VERB
bracis-28383	139	14	aligned	aligned	ADJ
bracis-28383	139	15	dog	dog	NOUN
bracis-28383	139	16	faces	face	NOUN
bracis-28383	139	17	with	with	ADP
bracis-28383	139	18	padding	padding	NOUN
bracis-28383	139	19	\(=0.2\	\(=0.2\	NOUN
bracis-28383	139	20	)	)	PUNCT
bracis-28383	139	21	since	since	SCONJ
bracis-28383	139	22	mapping	map	VERB
bracis-28383	139	23	this	this	DET
bracis-28383	139	24	5	5	NUM
bracis-28383	139	25	-	-	PUNCT
bracis-28383	139	26	point	point	NOUN
bracis-28383	139	27	into	into	ADP
bracis-28383	139	28	an	an	DET
bracis-28383	139	29	aligned	aligned	ADJ
bracis-28383	139	30	box	box	NOUN
bracis-28383	139	31	is	be	AUX
bracis-28383	139	32	less	less	ADV
bracis-28383	139	33	prone	prone	ADJ
bracis-28383	139	34	to	to	ADP
bracis-28383	139	35	errors	error	NOUN
bracis-28383	139	36	.	.	PUNCT
bracis-28383	140	1	such	such	ADJ
bracis-28383	140	2	conversion	conversion	NOUN
bracis-28383	140	3	is	be	AUX
bracis-28383	140	4	based	base	VERB
bracis-28383	140	5	on	on	ADP
bracis-28383	140	6	the	the	DET
bracis-28383	140	7	proportions	proportion	NOUN
bracis-28383	140	8	of	of	ADP
bracis-28383	140	9	the	the	DET
bracis-28383	140	10	eye	eye	NOUN
bracis-28383	140	11	and	and	CCONJ
bracis-28383	140	12	nose	nose	NOUN
bracis-28383	140	13	landmarks	landmark	NOUN
bracis-28383	140	14	.	.	PUNCT
bracis-28383	141	1	fig	fig	NOUN
bracis-28383	141	2	.	.	PUNCT
bracis-28383	142	1	2	2	X
bracis-28383	142	2	.	.	X
bracis-28383	142	3	in	in	ADP
bracis-28383	142	4	order	order	NOUN
bracis-28383	142	5	to	to	PART
bracis-28383	142	6	study	study	VERB
bracis-28383	142	7	the	the	DET
bracis-28383	142	8	impact	impact	NOUN
bracis-28383	142	9	of	of	ADP
bracis-28383	142	10	using	use	VERB
bracis-28383	142	11	different	different	ADJ
bracis-28383	142	12	preprocessing	preprocessing	NOUN
bracis-28383	142	13	techniques	technique	NOUN
bracis-28383	142	14	,	,	PUNCT
bracis-28383	142	15	we	we	PRON
bracis-28383	142	16	performed	perform	VERB
bracis-28383	142	17	tests	test	NOUN
bracis-28383	142	18	without	without	ADP
bracis-28383	142	19	preprocessing	preprocesse	VERB
bracis-28383	142	20	(	(	PUNCT
bracis-28383	142	21	a	a	NOUN
bracis-28383	142	22	)	)	PUNCT
bracis-28383	142	23	,	,	PUNCT
bracis-28383	142	24	with	with	ADP
bracis-28383	142	25	alignment	alignment	NOUN
bracis-28383	142	26	and	and	CCONJ
bracis-28383	142	27	face	face	VERB
bracis-28383	142	28	detection	detection	NOUN
bracis-28383	142	29	preprocessing	preprocessing	NOUN
bracis-28383	142	30	(	(	PUNCT
bracis-28383	142	31	b	b	NOUN
bracis-28383	142	32	)	)	PUNCT
bracis-28383	142	33	,	,	PUNCT
bracis-28383	142	34	and	and	CCONJ
bracis-28383	142	35	with	with	ADP
bracis-28383	142	36	central	central	ADJ
bracis-28383	142	37	crop	crop	NOUN
bracis-28383	142	38	(	(	PUNCT
bracis-28383	142	39	c	c	NOUN
bracis-28383	142	40	)	)	PUNCT
bracis-28383	142	41	before	before	ADP
bracis-28383	142	42	using	use	VERB
bracis-28383	142	43	residual	residual	ADJ
bracis-28383	142	44	cnn	cnn	NOUN
bracis-28383	142	45	for	for	ADP
bracis-28383	142	46	feature	feature	NOUN
bracis-28383	142	47	generation	generation	NOUN
bracis-28383	142	48	.	.	PUNCT
bracis-28383	143	1	full	full	ADJ
bracis-28383	143	2	size	size	NOUN
bracis-28383	143	3	image	image	NOUN
bracis-28383	143	4	fig	fig	NOUN
bracis-28383	143	5	.	.	PUNCT
bracis-28383	144	1	3	3	X
bracis-28383	144	2	.	.	X
bracis-28383	144	3	dog	dog	NOUN
bracis-28383	144	4	face	face	NOUN
bracis-28383	144	5	landmark	landmark	NOUN
bracis-28383	144	6	detection	detection	NOUN
bracis-28383	144	7	returns	return	VERB
bracis-28383	144	8	6	6	NUM
bracis-28383	144	9	points	point	NOUN
bracis-28383	144	10	(	(	PUNCT
bracis-28383	144	11	left	left	ADV
bracis-28383	144	12	)	)	PUNCT
bracis-28383	144	13	,	,	PUNCT
bracis-28383	144	14	which	which	PRON
bracis-28383	144	15	are	be	AUX
bracis-28383	144	16	converted	convert	VERB
bracis-28383	144	17	to	to	ADP
bracis-28383	144	18	the	the	DET
bracis-28383	144	19	5	5	NUM
bracis-28383	144	20	-	-	PUNCT
bracis-28383	144	21	point	point	NOUN
bracis-28383	144	22	layout	layout	NOUN
bracis-28383	144	23	(	(	PUNCT
bracis-28383	144	24	right	right	ADJ
bracis-28383	144	25	)	)	PUNCT
bracis-28383	144	26	before	before	ADP
bracis-28383	144	27	alignment	alignment	NOUN
bracis-28383	144	28	and	and	CCONJ
bracis-28383	144	29	cropping	cropping	NOUN
bracis-28383	144	30	.	.	PUNCT
bracis-28383	145	1	full	full	ADJ
bracis-28383	145	2	size	size	NOUN
bracis-28383	145	3	image	image	NOUN
bracis-28383	145	4	4.2	4.2	NUM
bracis-28383	145	5	preprocessing	preprocesse	VERB
bracis-28383	145	6	the	the	DET
bracis-28383	145	7	 	 	SPACE
bracis-28383	145	8	datasets	dataset	NOUN
bracis-28383	145	9	once	once	SCONJ
bracis-28383	145	10	we	we	PRON
bracis-28383	145	11	have	have	AUX
bracis-28383	145	12	clarified	clarify	VERB
bracis-28383	145	13	the	the	DET
bracis-28383	145	14	two	two	NUM
bracis-28383	145	15	preprocessing	preprocessing	NOUN
bracis-28383	145	16	techniques	technique	NOUN
bracis-28383	145	17	,	,	PUNCT
bracis-28383	145	18	in	in	ADP
bracis-28383	145	19	this	this	DET
bracis-28383	145	20	subsection	subsection	NOUN
bracis-28383	145	21	,	,	PUNCT
bracis-28383	145	22	we	we	PRON
bracis-28383	145	23	detail	detail	VERB
bracis-28383	145	24	the	the	DET
bracis-28383	145	25	generation	generation	NOUN
bracis-28383	145	26	of	of	ADP
bracis-28383	145	27	datasets	dataset	NOUN
bracis-28383	145	28	’	'	PUNCT
bracis-28383	145	29	versions	version	NOUN
bracis-28383	145	30	with	with	ADP
bracis-28383	145	31	each	each	DET
bracis-28383	145	32	preprocessing	preprocessing	NOUN
bracis-28383	145	33	and	and	CCONJ
bracis-28383	145	34	the	the	DET
bracis-28383	145	35	details	detail	NOUN
bracis-28383	145	36	that	that	PRON
bracis-28383	145	37	occurred	occur	VERB
bracis-28383	145	38	in	in	ADP
bracis-28383	145	39	each	each	DET
bracis-28383	145	40	technique	technique	NOUN
bracis-28383	145	41	and	and	CCONJ
bracis-28383	145	42	dataset	dataset	NOUN
bracis-28383	145	43	.	.	PUNCT
bracis-28383	146	1	dogfacenet	dogfacenet	NOUN
bracis-28383	146	2	cc	cc	NOUN
bracis-28383	146	3	results	result	NOUN
bracis-28383	146	4	from	from	ADP
bracis-28383	146	5	a	a	DET
bracis-28383	146	6	central	central	ADJ
bracis-28383	146	7	crop	crop	NOUN
bracis-28383	146	8	of	of	ADP
bracis-28383	146	9	\(180\times	\(180\time	NOUN
bracis-28383	146	10	180\	180\	NUM
bracis-28383	146	11	)	)	PUNCT
bracis-28383	146	12	pixels	pixel	NOUN
bracis-28383	146	13	of	of	ADP
bracis-28383	146	14	each	each	DET
bracis-28383	146	15	image	image	NOUN
bracis-28383	146	16	in	in	ADP
bracis-28383	146	17	dogfacenet	dogfacenet	NOUN
bracis-28383	146	18	,	,	PUNCT
bracis-28383	146	19	so	so	CCONJ
bracis-28383	146	20	the	the	DET
bracis-28383	146	21	total	total	ADJ
bracis-28383	146	22	individuals	individual	NOUN
bracis-28383	146	23	and	and	CCONJ
bracis-28383	146	24	images	image	NOUN
bracis-28383	146	25	are	be	AUX
bracis-28383	146	26	kept	keep	VERB
bracis-28383	146	27	.	.	PUNCT
bracis-28383	147	1	dogfacenet	dogfacenet	NOUN
bracis-28383	147	2	fd	fd	PROPN
bracis-28383	147	3	corresponds	correspond	VERB
bracis-28383	147	4	to	to	ADP
bracis-28383	147	5	the	the	DET
bracis-28383	147	6	dogfacenet	dogfacenet	NOUN
bracis-28383	147	7	dataset	dataset	NOUN
bracis-28383	147	8	after	after	ADP
bracis-28383	147	9	face	face	NOUN
bracis-28383	147	10	and	and	CCONJ
bracis-28383	147	11	alignment	alignment	NOUN
bracis-28383	147	12	.	.	PUNCT
bracis-28383	148	1	we	we	PRON
bracis-28383	148	2	then	then	ADV
bracis-28383	148	3	used	use	VERB
bracis-28383	148	4	the	the	DET
bracis-28383	148	5	same	same	ADJ
bracis-28383	148	6	\(80	\(80	NOUN
bracis-28383	148	7	-	-	PUNCT
bracis-28383	148	8	20\	20\	NOUN
bracis-28383	148	9	)	)	PUNCT
bracis-28383	148	10	split	split	NOUN
bracis-28383	148	11	used	use	VERB
bracis-28383	148	12	in	in	ADP
bracis-28383	148	13	the	the	DET
bracis-28383	148	14	original	original	ADJ
bracis-28383	148	15	dogfacenet	dogfacenet	NOUN
bracis-28383	148	16	and	and	CCONJ
bracis-28383	148	17	dogfacenet	dogfacenet	NOUN
bracis-28383	148	18	cc	cc	PROPN
bracis-28383	148	19	,	,	PUNCT
bracis-28383	148	20	but	but	CCONJ
bracis-28383	148	21	some	some	DET
bracis-28383	148	22	images	image	NOUN
bracis-28383	148	23	are	be	AUX
bracis-28383	148	24	missing	miss	VERB
bracis-28383	148	25	due	due	ADJ
bracis-28383	148	26	to	to	PART
bracis-28383	148	27	face	face	VERB
bracis-28383	148	28	detection	detection	NOUN
bracis-28383	148	29	errors	error	NOUN
bracis-28383	148	30	.	.	PUNCT
bracis-28383	149	1	this	this	PRON
bracis-28383	149	2	resulted	result	VERB
bracis-28383	149	3	in	in	ADP
bracis-28383	149	4	1	1	NUM
bracis-28383	149	5	,	,	PUNCT
bracis-28383	149	6	 	 	SPACE
bracis-28383	149	7	108	108	NUM
bracis-28383	149	8	identities	identity	NOUN
bracis-28383	149	9	for	for	ADP
bracis-28383	149	10	training	training	NOUN
bracis-28383	149	11	,	,	PUNCT
bracis-28383	149	12	corresponding	correspond	VERB
bracis-28383	149	13	to	to	ADP
bracis-28383	149	14	5	5	NUM
bracis-28383	149	15	,	,	PUNCT
bracis-28383	149	16	 	 	SPACE
bracis-28383	149	17	978	978	NUM
bracis-28383	149	18	images	image	NOUN
bracis-28383	149	19	.	.	PUNCT
bracis-28383	150	1	there	there	PRON
bracis-28383	150	2	are	be	VERB
bracis-28383	150	3	276	276	NUM
bracis-28383	150	4	identities	identity	NOUN
bracis-28383	150	5	for	for	ADP
bracis-28383	150	6	testing	testing	NOUN
bracis-28383	150	7	,	,	PUNCT
bracis-28383	150	8	corresponding	correspond	VERB
bracis-28383	150	9	to	to	ADP
bracis-28383	150	10	1	1	NUM
bracis-28383	150	11	,	,	PUNCT
bracis-28383	150	12	 	 	SPACE
bracis-28383	150	13	694	694	NUM
bracis-28383	150	14	images	image	NOUN
bracis-28383	150	15	.	.	PUNCT
bracis-28383	151	1	this	this	DET
bracis-28383	151	2	process	process	NOUN
bracis-28383	151	3	lost	lose	VERB
bracis-28383	151	4	9	9	NUM
bracis-28383	151	5	identities	identity	NOUN
bracis-28383	151	6	corresponding	correspond	VERB
bracis-28383	151	7	to	to	ADP
bracis-28383	151	8	\(0.65\%\	\(0.65\%\	PROPN
bracis-28383	151	9	)	)	PUNCT
bracis-28383	151	10	of	of	ADP
bracis-28383	151	11	individuals	individual	NOUN
bracis-28383	151	12	(	(	PUNCT
bracis-28383	151	13	ids	id	NOUN
bracis-28383	151	14	44	44	NUM
bracis-28383	151	15	,	,	PUNCT
bracis-28383	151	16	67	67	NUM
bracis-28383	151	17	,	,	PUNCT
bracis-28383	151	18	287	287	NUM
bracis-28383	151	19	,	,	PUNCT
bracis-28383	151	20	449	449	NUM
bracis-28383	151	21	,	,	PUNCT
bracis-28383	151	22	494	494	NUM
bracis-28383	151	23	,	,	PUNCT
bracis-28383	151	24	948	948	NUM
bracis-28383	151	25	,	,	PUNCT
bracis-28383	151	26	1131	1131	NUM
bracis-28383	151	27	,	,	PUNCT
bracis-28383	151	28	1274	1274	NUM
bracis-28383	151	29	,	,	PUNCT
bracis-28383	151	30	and	and	CCONJ
bracis-28383	151	31	1310	1310	NUM
bracis-28383	151	32	)	)	PUNCT
bracis-28383	151	33	.	.	PUNCT
bracis-28383	152	1	on	on	ADP
bracis-28383	152	2	the	the	DET
bracis-28383	152	3	other	other	ADJ
bracis-28383	152	4	hand	hand	NOUN
bracis-28383	152	5	,	,	PUNCT
bracis-28383	152	6	the	the	DET
bracis-28383	152	7	691	691	NUM
bracis-28383	152	8	images	image	NOUN
bracis-28383	152	9	lost	lose	VERB
bracis-28383	152	10	correspond	correspond	ADV
bracis-28383	152	11	to	to	ADP
bracis-28383	152	12	\(8.26\%\	\(8.26\%\	PROPN
bracis-28383	152	13	)	)	PUNCT
bracis-28383	152	14	of	of	ADP
bracis-28383	152	15	the	the	DET
bracis-28383	152	16	images	image	NOUN
bracis-28383	152	17	from	from	ADP
bracis-28383	152	18	the	the	DET
bracis-28383	152	19	original	original	ADJ
bracis-28383	152	20	dataset	dataset	NOUN
bracis-28383	152	21	.	.	PUNCT
bracis-28383	153	1	it	it	PRON
bracis-28383	153	2	is	be	AUX
bracis-28383	153	3	worth	worth	ADJ
bracis-28383	153	4	highlighting	highlight	VERB
bracis-28383	153	5	that	that	SCONJ
bracis-28383	153	6	missing	miss	VERB
bracis-28383	153	7	images	image	NOUN
bracis-28383	153	8	can	can	AUX
bracis-28383	153	9	not	not	PART
bracis-28383	153	10	be	be	AUX
bracis-28383	153	11	included	include	VERB
bracis-28383	153	12	in	in	ADP
bracis-28383	153	13	the	the	DET
bracis-28383	153	14	training	training	NOUN
bracis-28383	153	15	-	-	PUNCT
bracis-28383	153	16	test	test	NOUN
bracis-28383	153	17	split	split	NOUN
bracis-28383	153	18	.	.	PUNCT
bracis-28383	154	1	flickr	flickr	NOUN
bracis-28383	154	2	-	-	PUNCT
bracis-28383	154	3	dog	dog	NOUN
bracis-28383	154	4	cc	cc	NOUN
bracis-28383	154	5	corresponds	correspond	NOUN
bracis-28383	154	6	to	to	ADP
bracis-28383	154	7	the	the	DET
bracis-28383	154	8	flickr	flickr	ADJ
bracis-28383	154	9	-	-	PUNCT
bracis-28383	154	10	dog	dog	NOUN
bracis-28383	154	11	dataset	dataset	NOUN
bracis-28383	154	12	after	after	ADP
bracis-28383	154	13	a	a	DET
bracis-28383	154	14	central	central	ADJ
bracis-28383	154	15	crop	crop	NOUN
bracis-28383	154	16	for	for	ADP
bracis-28383	154	17	the	the	DET
bracis-28383	154	18	\(200\times	\(200\time	NOUN
bracis-28383	154	19	200\	200\	NUM
bracis-28383	154	20	)	)	PUNCT
bracis-28383	154	21	inner	inner	ADJ
bracis-28383	154	22	pixels	pixel	NOUN
bracis-28383	154	23	.	.	PUNCT
bracis-28383	155	1	in	in	ADP
bracis-28383	155	2	turn	turn	NOUN
bracis-28383	155	3	,	,	PUNCT
bracis-28383	155	4	flickr	flickr	NOUN
bracis-28383	155	5	-	-	PUNCT
bracis-28383	155	6	dog	dog	NOUN
bracis-28383	155	7	fd	fd	NOUN
bracis-28383	155	8	corresponds	correspond	NOUN
bracis-28383	155	9	to	to	ADP
bracis-28383	155	10	flickr	flickr	NOUN
bracis-28383	155	11	-	-	PUNCT
bracis-28383	155	12	dog	dog	NOUN
bracis-28383	155	13	after	after	ADP
bracis-28383	155	14	face	face	NOUN
bracis-28383	155	15	detection	detection	NOUN
bracis-28383	155	16	and	and	CCONJ
bracis-28383	155	17	alignment	alignment	NOUN
bracis-28383	155	18	.	.	PUNCT
bracis-28383	156	1	this	this	DET
bracis-28383	156	2	process	process	NOUN
bracis-28383	156	3	resulted	result	VERB
bracis-28383	156	4	in	in	ADP
bracis-28383	156	5	41	41	NUM
bracis-28383	156	6	of	of	ADP
bracis-28383	156	7	the	the	DET
bracis-28383	156	8	42	42	NUM
bracis-28383	156	9	identities	identity	NOUN
bracis-28383	156	10	,	,	PUNCT
bracis-28383	156	11	totaling	total	VERB
bracis-28383	156	12	218	218	NUM
bracis-28383	156	13	images	image	NOUN
bracis-28383	156	14	.	.	PUNCT
bracis-28383	157	1	this	this	DET
bracis-28383	157	2	process	process	NOUN
bracis-28383	157	3	lost	lose	VERB
bracis-28383	157	4	one	one	NUM
bracis-28383	157	5	identity	identity	NOUN
bracis-28383	157	6	corresponding	correspond	VERB
bracis-28383	157	7	to	to	ADP
bracis-28383	157	8	\(2.38\%\	\(2.38\%\	NOUN
bracis-28383	157	9	)	)	PUNCT
bracis-28383	157	10	of	of	ADP
bracis-28383	157	11	individuals	individual	NOUN
bracis-28383	157	12	since	since	SCONJ
bracis-28383	157	13	the	the	DET
bracis-28383	157	14	face	face	NOUN
bracis-28383	157	15	of	of	ADP
bracis-28383	157	16	“	"	PUNCT
bracis-28383	157	17	wilco	wilco	PROPN
bracis-28383	157	18	”	"	PUNCT
bracis-28383	157	19	could	could	AUX
bracis-28383	157	20	not	not	PART
bracis-28383	157	21	be	be	AUX
bracis-28383	157	22	detected	detect	VERB
bracis-28383	157	23	in	in	ADP
bracis-28383	157	24	any	any	PRON
bracis-28383	157	25	of	of	ADP
bracis-28383	157	26	its	its	PRON
bracis-28383	157	27	images	image	NOUN
bracis-28383	157	28	(	(	PUNCT
bracis-28383	157	29	first	first	ADJ
bracis-28383	157	30	row	row	NOUN
bracis-28383	157	31	of	of	ADP
bracis-28383	157	32	fig	fig	NOUN
bracis-28383	157	33	.	.	PUNCT
bracis-28383	157	34	 	 	SPACE
bracis-28383	157	35	1	1	NUM
bracis-28383	157	36	)	)	PUNCT
bracis-28383	157	37	.	.	PUNCT
bracis-28383	158	1	on	on	ADP
bracis-28383	158	2	the	the	DET
bracis-28383	158	3	other	other	ADJ
bracis-28383	158	4	hand	hand	NOUN
bracis-28383	158	5	,	,	PUNCT
bracis-28383	158	6	156	156	NUM
bracis-28383	158	7	images	image	NOUN
bracis-28383	158	8	lost	lose	VERB
bracis-28383	158	9	correspond	correspond	ADV
bracis-28383	158	10	to	to	ADP
bracis-28383	158	11	\(41.71\%\	\(41.71\%\	NOUN
bracis-28383	158	12	)	)	PUNCT
bracis-28383	158	13	of	of	ADP
bracis-28383	158	14	the	the	DET
bracis-28383	158	15	images	image	NOUN
bracis-28383	158	16	from	from	ADP
bracis-28383	158	17	the	the	DET
bracis-28383	158	18	original	original	ADJ
bracis-28383	158	19	dataset	dataset	NOUN
bracis-28383	158	20	.	.	PUNCT
bracis-28383	159	1	such	such	DET
bracis-28383	159	2	a	a	DET
bracis-28383	159	3	fact	fact	NOUN
bracis-28383	159	4	is	be	AUX
bracis-28383	159	5	quite	quite	DET
bracis-28383	159	6	a	a	DET
bracis-28383	159	7	loss	loss	NOUN
bracis-28383	159	8	in	in	ADP
bracis-28383	159	9	terms	term	NOUN
bracis-28383	159	10	of	of	ADP
bracis-28383	159	11	images	image	NOUN
bracis-28383	159	12	.	.	PUNCT
bracis-28383	160	1	however	however	ADV
bracis-28383	160	2	,	,	PUNCT
bracis-28383	160	3	it	it	PRON
bracis-28383	160	4	is	be	AUX
bracis-28383	160	5	essential	essential	ADJ
bracis-28383	160	6	to	to	PART
bracis-28383	160	7	remember	remember	VERB
bracis-28383	160	8	that	that	SCONJ
bracis-28383	160	9	all	all	DET
bracis-28383	160	10	the	the	DET
bracis-28383	160	11	218	218	NUM
bracis-28383	160	12	remaining	remain	VERB
bracis-28383	160	13	images	image	NOUN
bracis-28383	160	14	will	will	AUX
bracis-28383	160	15	be	be	AUX
bracis-28383	160	16	used	use	VERB
bracis-28383	160	17	for	for	ADP
bracis-28383	160	18	model	model	NOUN
bracis-28383	160	19	evaluation	evaluation	NOUN
bracis-28383	160	20	.	.	PUNCT
bracis-28383	161	1	in	in	ADP
bracis-28383	161	2	fig	fig	NOUN
bracis-28383	161	3	.	.	PUNCT
bracis-28383	161	4	 	 	SPACE
bracis-28383	161	5	4	4	NUM
bracis-28383	161	6	,	,	PUNCT
bracis-28383	161	7	there	there	PRON
bracis-28383	161	8	are	be	VERB
bracis-28383	161	9	some	some	DET
bracis-28383	161	10	examples	example	NOUN
bracis-28383	161	11	of	of	ADP
bracis-28383	161	12	removed	remove	VERB
bracis-28383	161	13	images	image	NOUN
bracis-28383	161	14	where	where	SCONJ
bracis-28383	161	15	the	the	DET
bracis-28383	161	16	algorithm	algorithm	NOUN
bracis-28383	161	17	could	could	AUX
bracis-28383	161	18	not	not	PART
bracis-28383	161	19	detect	detect	VERB
bracis-28383	161	20	the	the	DET
bracis-28383	161	21	face	face	NOUN
bracis-28383	161	22	and	and	CCONJ
bracis-28383	161	23	examples	example	NOUN
bracis-28383	161	24	of	of	ADP
bracis-28383	161	25	dogs	dog	NOUN
bracis-28383	161	26	that	that	PRON
bracis-28383	161	27	lost	lose	VERB
bracis-28383	161	28	all	all	DET
bracis-28383	161	29	images	image	NOUN
bracis-28383	161	30	and	and	CCONJ
bracis-28383	161	31	consequently	consequently	ADV
bracis-28383	161	32	lost	lose	VERB
bracis-28383	161	33	their	their	PRON
bracis-28383	161	34	identity	identity	NOUN
bracis-28383	161	35	for	for	ADP
bracis-28383	161	36	both	both	DET
bracis-28383	161	37	datasets	dataset	NOUN
bracis-28383	161	38	.	.	PUNCT
bracis-28383	162	1	the	the	DET
bracis-28383	162	2	final	final	ADJ
bracis-28383	162	3	numbers	number	NOUN
bracis-28383	162	4	for	for	ADP
bracis-28383	162	5	the	the	DET
bracis-28383	162	6	datasets	dataset	NOUN
bracis-28383	162	7	are	be	AUX
bracis-28383	162	8	described	describe	VERB
bracis-28383	162	9	in	in	ADP
bracis-28383	162	10	table	table	NOUN
bracis-28383	162	11	 	 	SPACE
bracis-28383	162	12	3	3	NUM
bracis-28383	162	13	.	.	PUNCT
bracis-28383	162	14	fig	fig	NOUN
bracis-28383	162	15	.	.	PUNCT
bracis-28383	163	1	4	4	X
bracis-28383	163	2	.	.	X
bracis-28383	163	3	examples	example	NOUN
bracis-28383	163	4	of	of	ADP
bracis-28383	163	5	images	image	NOUN
bracis-28383	163	6	removed	remove	VERB
bracis-28383	163	7	from	from	ADP
bracis-28383	163	8	datasets	dataset	NOUN
bracis-28383	163	9	because	because	SCONJ
bracis-28383	163	10	the	the	DET
bracis-28383	163	11	algorithm	algorithm	NOUN
bracis-28383	163	12	could	could	AUX
bracis-28383	163	13	not	not	PART
bracis-28383	163	14	detect	detect	VERB
bracis-28383	163	15	the	the	DET
bracis-28383	163	16	dog	dog	NOUN
bracis-28383	163	17	’s	’s	PART
bracis-28383	163	18	face	face	NOUN
bracis-28383	163	19	full	full	ADJ
bracis-28383	163	20	size	size	NOUN
bracis-28383	163	21	image	image	NOUN
bracis-28383	163	22	table	table	NOUN
bracis-28383	163	23	3	3	NUM
bracis-28383	163	24	.	.	X
bracis-28383	163	25	number	number	NOUN
bracis-28383	163	26	of	of	ADP
bracis-28383	163	27	identities	identity	NOUN
bracis-28383	163	28	and	and	CCONJ
bracis-28383	163	29	images	image	NOUN
bracis-28383	163	30	of	of	ADP
bracis-28383	163	31	each	each	PRON
bracis-28383	163	32	of	of	ADP
bracis-28383	163	33	the	the	DET
bracis-28383	163	34	dog	dog	NOUN
bracis-28383	163	35	face	face	VERB
bracis-28383	163	36	datasets.full	datasets.full	NUM
bracis-28383	163	37	size	size	NOUN
bracis-28383	163	38	table	table	NOUN
bracis-28383	163	39	5	5	NUM
bracis-28383	163	40	experimental	experimental	ADJ
bracis-28383	163	41	evaluation	evaluation	NOUN
bracis-28383	163	42	in	in	ADP
bracis-28383	163	43	order	order	NOUN
bracis-28383	163	44	to	to	PART
bracis-28383	163	45	evaluate	evaluate	VERB
bracis-28383	163	46	the	the	DET
bracis-28383	163	47	use	use	NOUN
bracis-28383	163	48	of	of	ADP
bracis-28383	163	49	the	the	DET
bracis-28383	163	50	residual	residual	ADJ
bracis-28383	163	51	network	network	NOUN
bracis-28383	163	52	in	in	ADP
bracis-28383	163	53	the	the	DET
bracis-28383	163	54	dataset	dataset	NOUN
bracis-28383	163	55	and	and	CCONJ
bracis-28383	163	56	each	each	PRON
bracis-28383	163	57	of	of	ADP
bracis-28383	163	58	the	the	DET
bracis-28383	163	59	generated	generate	VERB
bracis-28383	163	60	versions	version	NOUN
bracis-28383	163	61	,	,	PUNCT
bracis-28383	163	62	we	we	PRON
bracis-28383	163	63	performed	perform	VERB
bracis-28383	163	64	the	the	DET
bracis-28383	163	65	identification	identification	NOUN
bracis-28383	163	66	experiments	experiment	NOUN
bracis-28383	163	67	evaluating	evaluate	VERB
bracis-28383	163	68	different	different	ADJ
bracis-28383	163	69	metrics	metric	NOUN
bracis-28383	163	70	of	of	ADP
bracis-28383	163	71	interest	interest	NOUN
bracis-28383	163	72	in	in	ADP
bracis-28383	163	73	this	this	DET
bracis-28383	163	74	context	context	NOUN
bracis-28383	163	75	.	.	PUNCT
bracis-28383	164	1	we	we	PRON
bracis-28383	164	2	carried	carry	VERB
bracis-28383	164	3	out	out	ADP
bracis-28383	164	4	the	the	DET
bracis-28383	164	5	experiments	experiment	NOUN
bracis-28383	164	6	on	on	ADP
bracis-28383	164	7	a	a	DET
bracis-28383	164	8	computer	computer	NOUN
bracis-28383	164	9	running	run	VERB
bracis-28383	164	10	linux	linux	PROPN
bracis-28383	164	11	operating	operating	NOUN
bracis-28383	164	12	system	system	NOUN
bracis-28383	164	13	(	(	PUNCT
bracis-28383	164	14	ubuntu	ubuntu	ADJ
bracis-28383	164	15	20.04.2	20.04.2	NUM
bracis-28383	164	16	lts	lts	NOUN
bracis-28383	164	17	)	)	PUNCT
bracis-28383	164	18	,	,	PUNCT
bracis-28383	164	19	equipped	equip	VERB
bracis-28383	164	20	with	with	ADP
bracis-28383	164	21	an	an	DET
bracis-28383	164	22	intel(r	intel(r	NOUN
bracis-28383	164	23	)	)	PUNCT
bracis-28383	164	24	core(tm	core(tm	NOUN
bracis-28383	164	25	)	)	PUNCT
bracis-28383	164	26	i7	i7	NOUN
bracis-28383	164	27	-	-	PUNCT
bracis-28383	164	28	9700k	9700k	NUM
bracis-28383	164	29	cpu	cpu	NOUN
bracis-28383	164	30	@	@	ADP
bracis-28383	164	31	3.60	3.60	NUM
bracis-28383	164	32	 	 	SPACE
bracis-28383	164	33	ghz	ghz	NOUN
bracis-28383	164	34	,	,	PUNCT
bracis-28383	164	35	32	32	NUM
bracis-28383	164	36	 	 	SPACE
bracis-28383	164	37	gb	gb	NOUN
bracis-28383	164	38	ram	ram	NOUN
bracis-28383	164	39	,	,	PUNCT
bracis-28383	164	40	a	a	DET
bracis-28383	164	41	1	1	NUM
bracis-28383	164	42	 	 	SPACE
bracis-28383	164	43	tb	tb	NOUN
bracis-28383	164	44	sdd	sdd	NOUN
bracis-28383	164	45	,	,	PUNCT
bracis-28383	164	46	and	and	CCONJ
bracis-28383	164	47	an	an	DET
bracis-28383	164	48	nvidia	nvidia	PROPN
bracis-28383	164	49	geforce	geforce	NOUN
bracis-28383	164	50	rtx	rtx	PROPN
bracis-28383	164	51	2080ti	2080ti	PROPN
bracis-28383	164	52	graphics	graphic	NOUN
bracis-28383	164	53	card	card	NOUN
bracis-28383	164	54	.	.	PUNCT
bracis-28383	165	1	the	the	DET
bracis-28383	165	2	resnet-34	resnet-34	PROPN
bracis-28383	165	3	convolutional	convolutional	ADJ
bracis-28383	165	4	neural	neural	ADJ
bracis-28383	165	5	network	network	NOUN
bracis-28383	165	6	was	be	AUX
bracis-28383	165	7	used	use	VERB
bracis-28383	165	8	for	for	ADP
bracis-28383	165	9	the	the	DET
bracis-28383	165	10	experiments	experiment	NOUN
bracis-28383	165	11	,	,	PUNCT
bracis-28383	165	12	as	as	SCONJ
bracis-28383	165	13	it	it	PRON
bracis-28383	165	14	is	be	AUX
bracis-28383	165	15	a	a	DET
bracis-28383	165	16	good	good	ADJ
bracis-28383	165	17	trade	trade	NOUN
bracis-28383	165	18	-	-	PUNCT
bracis-28383	165	19	off	off	NOUN
bracis-28383	165	20	between	between	ADP
bracis-28383	165	21	performance	performance	NOUN
bracis-28383	165	22	and	and	CCONJ
bracis-28383	165	23	computational	computational	ADJ
bracis-28383	165	24	cost	cost	NOUN
bracis-28383	165	25	.	.	PUNCT
bracis-28383	166	1	it	it	PRON
bracis-28383	166	2	is	be	AUX
bracis-28383	166	3	a	a	DET
bracis-28383	166	4	residual	residual	ADJ
bracis-28383	166	5	neural	neural	ADJ
bracis-28383	166	6	network	network	NOUN
bracis-28383	166	7	with	with	ADP
bracis-28383	166	8	about	about	ADV
bracis-28383	166	9	21	21	NUM
bracis-28383	166	10	,	,	PUNCT
bracis-28383	166	11	 	 	SPACE
bracis-28383	166	12	282	282	NUM
bracis-28383	166	13	,	,	PUNCT
bracis-28383	166	14	 	 	SPACE
bracis-28383	166	15	000	000	NUM
bracis-28383	167	1	parameters	parameter	NOUN
bracis-28383	167	2	,	,	PUNCT
bracis-28383	167	3	requiring	require	VERB
bracis-28383	167	4	about	about	ADP
bracis-28383	167	5	4	4	NUM
bracis-28383	167	6	gflops	gflop	NOUN
bracis-28383	167	7	for	for	ADP
bracis-28383	167	8	activation	activation	NOUN
bracis-28383	167	9	.	.	PUNCT
bracis-28383	168	1	figure	figure	NOUN
bracis-28383	168	2	 	 	SPACE
bracis-28383	168	3	5	5	NUM
bracis-28383	168	4	shows	show	VERB
bracis-28383	168	5	the	the	DET
bracis-28383	168	6	resnet-34	resnet-34	PROPN
bracis-28383	168	7	architecture	architecture	NOUN
bracis-28383	168	8	.	.	PUNCT
bracis-28383	169	1	the	the	DET
bracis-28383	169	2	implementation	implementation	NOUN
bracis-28383	169	3	of	of	ADP
bracis-28383	169	4	resnet-34	resnet-34	PROPN
bracis-28383	169	5	present	present	NOUN
bracis-28383	169	6	in	in	ADP
bracis-28383	169	7	the	the	DET
bracis-28383	169	8	machine	machine	NOUN
bracis-28383	169	9	learning	learn	VERB
bracis-28383	169	10	dlib	dlib	NOUN
bracis-28383	169	11	c++	c++	NOUN
bracis-28383	169	12	library	library	PROPN
bracis-28383	169	13	was	be	AUX
bracis-28383	169	14	used	use	VERB
bracis-28383	169	15	[	[	PUNCT
bracis-28383	169	16	10	10	NUM
bracis-28383	169	17	]	]	PUNCT
bracis-28383	169	18	.	.	PUNCT
bracis-28383	170	1	the	the	DET
bracis-28383	170	2	split	split	ADJ
bracis-28383	170	3	training	training	NOUN
bracis-28383	170	4	and	and	CCONJ
bracis-28383	170	5	testing	testing	NOUN
bracis-28383	170	6	procedures	procedure	NOUN
bracis-28383	170	7	were	be	AUX
bracis-28383	170	8	explained	explain	VERB
bracis-28383	170	9	individually	individually	ADV
bracis-28383	170	10	in	in	ADP
bracis-28383	170	11	specific	specific	ADJ
bracis-28383	170	12	topics	topic	NOUN
bracis-28383	170	13	for	for	ADP
bracis-28383	170	14	each	each	DET
bracis-28383	170	15	dataset	dataset	NOUN
bracis-28383	170	16	.	.	PUNCT
bracis-28383	171	1	we	we	PRON
bracis-28383	171	2	defined	define	VERB
bracis-28383	171	3	that	that	SCONJ
bracis-28383	171	4	all	all	DET
bracis-28383	171	5	training	training	NOUN
bracis-28383	171	6	sessions	session	NOUN
bracis-28383	171	7	would	would	AUX
bracis-28383	171	8	have	have	VERB
bracis-28383	171	9	80	80	NUM
bracis-28383	171	10	,	,	PUNCT
bracis-28383	171	11	 	 	SPACE
bracis-28383	171	12	000	000	NUM
bracis-28383	171	13	steps	step	NOUN
bracis-28383	171	14	,	,	PUNCT
bracis-28383	171	15	to	to	PART
bracis-28383	171	16	standardize	standardize	VERB
bracis-28383	171	17	all	all	DET
bracis-28383	171	18	procedures	procedure	NOUN
bracis-28383	171	19	for	for	ADP
bracis-28383	171	20	each	each	DET
bracis-28383	171	21	dataset	dataset	NOUN
bracis-28383	171	22	.	.	PUNCT
bracis-28383	172	1	we	we	PRON
bracis-28383	172	2	chose	choose	VERB
bracis-28383	172	3	this	this	DET
bracis-28383	172	4	number	number	NOUN
bracis-28383	172	5	because	because	SCONJ
bracis-28383	172	6	it	it	PRON
bracis-28383	172	7	is	be	AUX
bracis-28383	172	8	the	the	DET
bracis-28383	172	9	average	average	ADJ
bracis-28383	172	10	value	value	NOUN
bracis-28383	172	11	where	where	SCONJ
bracis-28383	172	12	the	the	DET
bracis-28383	172	13	loss	loss	NOUN
bracis-28383	172	14	curve	curve	NOUN
bracis-28383	172	15	is	be	AUX
bracis-28383	172	16	already	already	ADV
bracis-28383	172	17	relatively	relatively	ADV
bracis-28383	172	18	stabilized	stabilize	VERB
bracis-28383	172	19	.	.	PUNCT
bracis-28383	173	1	fig	fig	NOUN
bracis-28383	173	2	.	.	PUNCT
bracis-28383	174	1	5.(adapted	5.(adapted	NUM
bracis-28383	174	2	from	from	ADP
bracis-28383	174	3	[	[	X
bracis-28383	174	4	6	6	NUM
bracis-28383	174	5	]	]	NUM
bracis-28383	174	6	)	)	PUNCT
bracis-28383	174	7	.	.	PUNCT
bracis-28383	175	1	resnet34	resnet34	NOUN
bracis-28383	175	2	architecture	architecture	VERB
bracis-28383	175	3	full	full	ADJ
bracis-28383	175	4	size	size	NOUN
bracis-28383	175	5	image	image	NOUN
bracis-28383	175	6	5.1	5.1	NUM
bracis-28383	175	7	performance	performance	NOUN
bracis-28383	175	8	metrics	metric	NOUN
bracis-28383	175	9	the	the	DET
bracis-28383	175	10	metrics	metric	NOUN
bracis-28383	175	11	used	use	VERB
bracis-28383	175	12	in	in	ADP
bracis-28383	175	13	this	this	DET
bracis-28383	175	14	work	work	NOUN
bracis-28383	175	15	are	be	AUX
bracis-28383	175	16	accuracy	accuracy	NOUN
bracis-28383	175	17	and	and	CCONJ
bracis-28383	175	18	recall	recall	NOUN
bracis-28383	175	19	.	.	PUNCT
bracis-28383	176	1	both	both	PRON
bracis-28383	176	2	are	be	AUX
bracis-28383	176	3	defined	define	VERB
bracis-28383	176	4	in	in	ADP
bracis-28383	176	5	terms	term	NOUN
bracis-28383	176	6	of	of	ADP
bracis-28383	176	7	true	true	ADJ
bracis-28383	176	8	positives	positive	NOUN
bracis-28383	176	9	(	(	PUNCT
bracis-28383	176	10	tp	tp	NOUN
bracis-28383	176	11	)	)	PUNCT
bracis-28383	176	12	,	,	PUNCT
bracis-28383	176	13	false	false	ADJ
bracis-28383	176	14	positives	positive	NOUN
bracis-28383	176	15	(	(	PUNCT
bracis-28383	176	16	fp	fp	NOUN
bracis-28383	176	17	)	)	PUNCT
bracis-28383	176	18	,	,	PUNCT
bracis-28383	176	19	true	true	ADJ
bracis-28383	176	20	negative	negative	ADJ
bracis-28383	176	21	(	(	PUNCT
bracis-28383	176	22	tn	tn	NOUN
bracis-28383	176	23	)	)	PUNCT
bracis-28383	176	24	and	and	CCONJ
bracis-28383	176	25	false	false	ADJ
bracis-28383	176	26	negative	negative	ADJ
bracis-28383	176	27	(	(	PUNCT
bracis-28383	176	28	fn	fn	NOUN
bracis-28383	176	29	)	)	PUNCT
bracis-28383	176	30	samples	sample	NOUN
bracis-28383	176	31	.	.	PUNCT
bracis-28383	177	1	accuracy	accuracy	NOUN
bracis-28383	177	2	is	be	AUX
bracis-28383	177	3	defined	define	VERB
bracis-28383	177	4	in	in	ADP
bracis-28383	177	5	eq	eq	PROPN
bracis-28383	177	6	.	.	PUNCT
bracis-28383	177	7	 	 	SPACE
bracis-28383	177	8	1	1	NUM
bracis-28383	177	9	and	and	CCONJ
bracis-28383	177	10	recall	recall	VERB
bracis-28383	177	11	in	in	ADP
bracis-28383	177	12	eq	eq	PROPN
bracis-28383	177	13	.	.	PUNCT
bracis-28383	177	14	 	 	SPACE
bracis-28383	178	1	2	2	NUM
bracis-28383	178	2	.	.	X
bracis-28383	178	3	$	$	SYM
bracis-28383	178	4	$	$	SYM
bracis-28383	178	5	\begin{aligned	\begin{aligne	VERB
bracis-28383	178	6	}	}	PUNCT
bracis-28383	178	7	accuracy	accuracy	NOUN
bracis-28383	178	8	=	=	SYM
bracis-28383	178	9	\frac{tp+tn}{tp+tn+fp+fn	\frac{tp+tn}{tp+tn+fp+fn	NOUN
bracis-28383	178	10	}	}	PUNCT
bracis-28383	178	11	\end{aligned}$$	\end{aligned}$$	X
bracis-28383	178	12	(	(	PUNCT
bracis-28383	178	13	1	1	NUM
bracis-28383	178	14	)	)	PUNCT
bracis-28383	178	15	$	$	SYM
bracis-28383	178	16	$	$	SYM
bracis-28383	178	17	\begin{aligned	\begin{aligne	VERB
bracis-28383	178	18	}	}	PUNCT
bracis-28383	178	19	recall	recall	NOUN
bracis-28383	178	20	=	=	PUNCT
bracis-28383	178	21	\frac{tp}{tp+fn	\frac{tp}{tp+fn	NOUN
bracis-28383	178	22	}	}	PUNCT
bracis-28383	178	23	\end{aligned}$$	\end{aligned}$$	X
bracis-28383	178	24	(	(	PUNCT
bracis-28383	178	25	2	2	NUM
bracis-28383	178	26	)	)	PUNCT
bracis-28383	178	27	precision	precision	NOUN
bracis-28383	178	28	and	and	CCONJ
bracis-28383	178	29	f1	f1	NOUN
bracis-28383	178	30	-	-	PUNCT
bracis-28383	178	31	score	score	NOUN
bracis-28383	178	32	sound	sound	NOUN
bracis-28383	178	33	virtually	virtually	ADV
bracis-28383	178	34	meaningless	meaningless	ADJ
bracis-28383	178	35	for	for	ADP
bracis-28383	178	36	our	our	PRON
bracis-28383	178	37	experiments	experiment	NOUN
bracis-28383	178	38	since	since	SCONJ
bracis-28383	178	39	the	the	DET
bracis-28383	178	40	evaluation	evaluation	NOUN
bracis-28383	178	41	considers	consider	VERB
bracis-28383	178	42	an	an	DET
bracis-28383	178	43	upper	upper	ADJ
bracis-28383	178	44	triangular	triangular	NOUN
bracis-28383	178	45	matrix	matrix	NOUN
bracis-28383	178	46	where	where	SCONJ
bracis-28383	178	47	most	most	ADJ
bracis-28383	178	48	elements	element	NOUN
bracis-28383	178	49	in	in	ADP
bracis-28383	178	50	that	that	DET
bracis-28383	178	51	matrix	matrix	NOUN
bracis-28383	178	52	represent	represent	VERB
bracis-28383	178	53	different	different	ADJ
bracis-28383	178	54	individuals	individual	NOUN
bracis-28383	178	55	.	.	PUNCT
bracis-28383	179	1	in	in	ADP
bracis-28383	179	2	other	other	ADJ
bracis-28383	179	3	words	word	NOUN
bracis-28383	179	4	,	,	PUNCT
bracis-28383	179	5	we	we	PRON
bracis-28383	179	6	would	would	AUX
bracis-28383	179	7	be	be	AUX
bracis-28383	179	8	roughly	roughly	ADV
bracis-28383	179	9	counting	count	VERB
bracis-28383	179	10	the	the	DET
bracis-28383	179	11	main	main	ADJ
bracis-28383	179	12	diagonal	diagonal	NOUN
bracis-28383	179	13	versus	versus	ADP
bracis-28383	179	14	the	the	DET
bracis-28383	179	15	rest	rest	NOUN
bracis-28383	179	16	of	of	ADP
bracis-28383	179	17	the	the	DET
bracis-28383	179	18	matrix	matrix	NOUN
bracis-28383	179	19	in	in	ADP
bracis-28383	179	20	the	the	DET
bracis-28383	179	21	limit	limit	NOUN
bracis-28383	179	22	,	,	PUNCT
bracis-28383	179	23	so	so	SCONJ
bracis-28383	179	24	these	these	DET
bracis-28383	179	25	performance	performance	NOUN
bracis-28383	179	26	metrics	metric	NOUN
bracis-28383	179	27	are	be	AUX
bracis-28383	179	28	influenced	influence	VERB
bracis-28383	179	29	mainly	mainly	ADV
bracis-28383	179	30	by	by	ADP
bracis-28383	179	31	the	the	DET
bracis-28383	179	32	model	model	NOUN
bracis-28383	179	33	’s	’s	PART
bracis-28383	179	34	ability	ability	NOUN
bracis-28383	179	35	to	to	PART
bracis-28383	179	36	discard	discard	VERB
bracis-28383	179	37	dissimilar	dissimilar	ADJ
bracis-28383	179	38	individuals	individual	NOUN
bracis-28383	179	39	.	.	PUNCT
bracis-28383	180	1	5.2	5.2	NUM
bracis-28383	180	2	preparation	preparation	NOUN
bracis-28383	180	3	of	of	ADP
bracis-28383	180	4	 	 	SPACE
bracis-28383	180	5	the	the	DET
bracis-28383	180	6	 	 	SPACE
bracis-28383	180	7	datasets	dataset	NOUN
bracis-28383	180	8	notice	notice	VERB
bracis-28383	180	9	that	that	SCONJ
bracis-28383	180	10	the	the	DET
bracis-28383	180	11	baseline	baseline	NOUN
bracis-28383	180	12	work	work	NOUN
bracis-28383	180	13	[	[	X
bracis-28383	180	14	16	16	NUM
bracis-28383	180	15	]	]	PUNCT
bracis-28383	180	16	adopts	adopt	VERB
bracis-28383	180	17	a	a	DET
bracis-28383	180	18	\(90	\(90	NUM
bracis-28383	180	19	-	-	PUNCT
bracis-28383	180	20	10\	10\	NUM
bracis-28383	180	21	)	)	PUNCT
bracis-28383	180	22	split	split	NOUN
bracis-28383	180	23	,	,	PUNCT
bracis-28383	180	24	i.e.	i.e.	X
bracis-28383	180	25	,	,	PUNCT
bracis-28383	180	26	only	only	ADV
bracis-28383	180	27	697	697	NUM
bracis-28383	180	28	images	image	NOUN
bracis-28383	180	29	(	(	PUNCT
bracis-28383	180	30	8.33	8.33	NUM
bracis-28383	180	31	%	%	NOUN
bracis-28383	180	32	of	of	ADP
bracis-28383	180	33	the	the	DET
bracis-28383	180	34	dataset	dataset	NOUN
bracis-28383	180	35	)	)	PUNCT
bracis-28383	180	36	are	be	AUX
bracis-28383	180	37	used	use	VERB
bracis-28383	180	38	for	for	ADP
bracis-28383	180	39	testing	testing	NOUN
bracis-28383	180	40	.	.	PUNCT
bracis-28383	181	1	to	to	PART
bracis-28383	181	2	better	well	ADV
bracis-28383	181	3	generalize	generalize	VERB
bracis-28383	181	4	the	the	DET
bracis-28383	181	5	model	model	NOUN
bracis-28383	181	6	used	use	VERB
bracis-28383	181	7	here	here	ADV
bracis-28383	181	8	,	,	PUNCT
bracis-28383	181	9	we	we	PRON
bracis-28383	181	10	adopted	adopt	VERB
bracis-28383	181	11	the	the	DET
bracis-28383	181	12	\(80	\(80	NOUN
bracis-28383	181	13	-	-	PUNCT
bracis-28383	181	14	20\	20\	NOUN
bracis-28383	181	15	)	)	PUNCT
bracis-28383	181	16	split	split	NOUN
bracis-28383	181	17	,	,	PUNCT
bracis-28383	181	18	making	make	VERB
bracis-28383	181	19	the	the	DET
bracis-28383	181	20	problem	problem	NOUN
bracis-28383	181	21	even	even	ADV
bracis-28383	181	22	more	more	ADV
bracis-28383	181	23	challenging	challenging	ADJ
bracis-28383	181	24	by	by	ADP
bracis-28383	181	25	having	have	VERB
bracis-28383	181	26	fewer	few	ADJ
bracis-28383	181	27	images	image	NOUN
bracis-28383	181	28	for	for	ADP
bracis-28383	181	29	training	training	NOUN
bracis-28383	181	30	and	and	CCONJ
bracis-28383	181	31	providing	provide	VERB
bracis-28383	181	32	a	a	DET
bracis-28383	181	33	more	more	ADV
bracis-28383	181	34	extensive	extensive	ADJ
bracis-28383	181	35	test	test	NOUN
bracis-28383	181	36	set	set	VERB
bracis-28383	181	37	.	.	PUNCT
bracis-28383	182	1	for	for	ADP
bracis-28383	182	2	the	the	DET
bracis-28383	182	3	dogfacenet	dogfacenet	NOUN
bracis-28383	182	4	case	case	NOUN
bracis-28383	182	5	,	,	PUNCT
bracis-28383	182	6	5	5	NUM
bracis-28383	182	7	,	,	PUNCT
bracis-28383	182	8	 	 	SPACE
bracis-28383	182	9	978	978	NUM
bracis-28383	182	10	images	image	NOUN
bracis-28383	182	11	for	for	ADP
bracis-28383	182	12	training	training	NOUN
bracis-28383	182	13	in	in	ADP
bracis-28383	182	14	each	each	DET
bracis-28383	182	15	case	case	NOUN
bracis-28383	182	16	(	(	PUNCT
bracis-28383	182	17	cc	cc	NOUN
bracis-28383	182	18	and	and	CCONJ
bracis-28383	182	19	fd	fd	PROPN
bracis-28383	182	20	)	)	PUNCT
bracis-28383	182	21	,	,	PUNCT
bracis-28383	182	22	excluding	exclude	VERB
bracis-28383	182	23	all	all	DET
bracis-28383	182	24	the	the	DET
bracis-28383	182	25	fd	fd	PROPN
bracis-28383	182	26	algorithm	algorithm	NOUN
bracis-28383	182	27	images	image	NOUN
bracis-28383	182	28	,	,	PUNCT
bracis-28383	182	29	did	do	AUX
bracis-28383	182	30	not	not	PART
bracis-28383	182	31	detect	detect	VERB
bracis-28383	182	32	a	a	DET
bracis-28383	182	33	face	face	NOUN
bracis-28383	182	34	,	,	PUNCT
bracis-28383	182	35	and	and	CCONJ
bracis-28383	182	36	1	1	NUM
bracis-28383	182	37	,	,	PUNCT
bracis-28383	182	38	 	 	SPACE
bracis-28383	182	39	794	794	NUM
bracis-28383	182	40	test	test	NOUN
bracis-28383	182	41	images	image	NOUN
bracis-28383	182	42	for	for	ADP
bracis-28383	182	43	original	original	ADJ
bracis-28383	182	44	dogfacnenet	dogfacnenet	NOUN
bracis-28383	182	45	,	,	PUNCT
bracis-28383	182	46	1	1	NUM
bracis-28383	182	47	,	,	PUNCT
bracis-28383	182	48	 	 	SPACE
bracis-28383	182	49	704	704	NUM
bracis-28383	182	50	images	image	NOUN
bracis-28383	182	51	in	in	ADP
bracis-28383	182	52	cc	cc	NOUN
bracis-28383	182	53	,	,	PUNCT
bracis-28383	182	54	and	and	CCONJ
bracis-28383	182	55	1	1	NUM
bracis-28383	182	56	,	,	PUNCT
bracis-28383	182	57	 	 	SPACE
bracis-28383	182	58	520	520	NUM
bracis-28383	182	59	in	in	ADP
bracis-28383	182	60	dogfacenet	dogfacenet	NOUN
bracis-28383	182	61	fd	fd	PROPN
bracis-28383	182	62	.	.	PUNCT
bracis-28383	183	1	the	the	DET
bracis-28383	183	2	experiment	experiment	NOUN
bracis-28383	183	3	performed	perform	VERB
bracis-28383	183	4	with	with	ADP
bracis-28383	183	5	flickr	flickr	NOUN
bracis-28383	183	6	-	-	PUNCT
bracis-28383	183	7	dog	dog	NOUN
bracis-28383	183	8	proceeded	proceed	VERB
bracis-28383	183	9	as	as	SCONJ
bracis-28383	183	10	follows	follow	VERB
bracis-28383	183	11	.	.	PUNCT
bracis-28383	184	1	since	since	SCONJ
bracis-28383	184	2	this	this	PRON
bracis-28383	184	3	is	be	AUX
bracis-28383	184	4	a	a	DET
bracis-28383	184	5	small	small	ADJ
bracis-28383	184	6	-	-	PUNCT
bracis-28383	184	7	scale	scale	NOUN
bracis-28383	184	8	dataset	dataset	NOUN
bracis-28383	184	9	,	,	PUNCT
bracis-28383	184	10	the	the	DET
bracis-28383	184	11	method	method	NOUN
bracis-28383	184	12	heavily	heavily	ADV
bracis-28383	184	13	depends	depend	VERB
bracis-28383	184	14	on	on	ADP
bracis-28383	184	15	some	some	DET
bracis-28383	184	16	data	datum	NOUN
bracis-28383	184	17	augmentation	augmentation	NOUN
bracis-28383	184	18	strategy	strategy	NOUN
bracis-28383	184	19	or	or	CCONJ
bracis-28383	184	20	external	external	ADJ
bracis-28383	184	21	data	datum	NOUN
bracis-28383	184	22	.	.	PUNCT
bracis-28383	185	1	the	the	DET
bracis-28383	185	2	results	result	NOUN
bracis-28383	185	3	are	be	AUX
bracis-28383	185	4	inferior	inferior	ADJ
bracis-28383	185	5	without	without	ADP
bracis-28383	185	6	training	train	VERB
bracis-28383	185	7	with	with	ADP
bracis-28383	185	8	external	external	ADJ
bracis-28383	185	9	data	datum	NOUN
bracis-28383	185	10	and	and	CCONJ
bracis-28383	185	11	only	only	ADV
bracis-28383	185	12	with	with	ADP
bracis-28383	185	13	data	datum	NOUN
bracis-28383	185	14	augmentation	augmentation	NOUN
bracis-28383	185	15	:	:	PUNCT
bracis-28383	185	16	an	an	DET
bracis-28383	185	17	accuracy	accuracy	NOUN
bracis-28383	185	18	limited	limit	VERB
bracis-28383	185	19	to	to	ADP
bracis-28383	185	20	76.55	76.55	NUM
bracis-28383	185	21	%	%	NOUN
bracis-28383	185	22	and	and	CCONJ
bracis-28383	185	23	recall	recall	NOUN
bracis-28383	185	24	limited	limit	VERB
bracis-28383	185	25	to	to	ADP
bracis-28383	185	26	65.68	65.68	NUM
bracis-28383	185	27	%	%	NOUN
bracis-28383	185	28	,	,	PUNCT
bracis-28383	185	29	also	also	ADV
bracis-28383	185	30	performed	perform	VERB
bracis-28383	185	31	using	use	VERB
bracis-28383	185	32	the	the	DET
bracis-28383	185	33	\(80	\(80	NOUN
bracis-28383	185	34	-	-	PUNCT
bracis-28383	185	35	20\	20\	NOUN
bracis-28383	185	36	)	)	PUNCT
bracis-28383	185	37	split	split	NOUN
bracis-28383	185	38	.	.	PUNCT
bracis-28383	186	1	we	we	PRON
bracis-28383	186	2	decided	decide	VERB
bracis-28383	186	3	to	to	PART
bracis-28383	186	4	use	use	VERB
bracis-28383	186	5	the	the	DET
bracis-28383	186	6	model	model	NOUN
bracis-28383	186	7	trained	train	VERB
bracis-28383	186	8	with	with	ADP
bracis-28383	186	9	the	the	DET
bracis-28383	186	10	different	different	ADJ
bracis-28383	186	11	versions	version	NOUN
bracis-28383	186	12	of	of	ADP
bracis-28383	186	13	the	the	DET
bracis-28383	186	14	dogfacenet	dogfacenet	NOUN
bracis-28383	186	15	dataset	dataset	NOUN
bracis-28383	186	16	(	(	PUNCT
bracis-28383	186	17	original	original	ADJ
bracis-28383	186	18	,	,	PUNCT
bracis-28383	186	19	cc	cc	NOUN
bracis-28383	186	20	,	,	PUNCT
bracis-28383	186	21	and	and	CCONJ
bracis-28383	186	22	fd	fd	VERB
bracis-28383	186	23	)	)	PUNCT
bracis-28383	186	24	and	and	CCONJ
bracis-28383	186	25	test	test	VERB
bracis-28383	186	26	on	on	ADP
bracis-28383	186	27	a	a	DET
bracis-28383	186	28	range	range	NOUN
bracis-28383	186	29	corresponding	correspond	VERB
bracis-28383	186	30	to	to	ADP
bracis-28383	186	31	20	20	NUM
bracis-28383	186	32	%	%	NOUN
bracis-28383	186	33	of	of	ADP
bracis-28383	186	34	the	the	DET
bracis-28383	186	35	flickr	flickr	ADJ
bracis-28383	186	36	-	-	PUNCT
bracis-28383	186	37	dog	dog	NOUN
bracis-28383	186	38	dataset	dataset	NOUN
bracis-28383	186	39	,	,	PUNCT
bracis-28383	186	40	which	which	PRON
bracis-28383	186	41	was	be	AUX
bracis-28383	186	42	also	also	ADV
bracis-28383	186	43	submitted	submit	VERB
bracis-28383	186	44	and	and	CCONJ
bracis-28383	186	45	generated	generate	VERB
bracis-28383	186	46	versions	version	NOUN
bracis-28383	186	47	with	with	ADP
bracis-28383	186	48	cc	cc	NOUN
bracis-28383	186	49	and	and	CCONJ
bracis-28383	186	50	fd	fd	PROPN
bracis-28383	186	51	preprocessing	preprocessing	NOUN
bracis-28383	186	52	.	.	PUNCT
bracis-28383	187	1	for	for	ADP
bracis-28383	187	2	this	this	DET
bracis-28383	187	3	case	case	NOUN
bracis-28383	187	4	,	,	PUNCT
bracis-28383	187	5	the	the	DET
bracis-28383	187	6	original	original	NOUN
bracis-28383	187	7	with	with	ADP
bracis-28383	187	8	65	65	NUM
bracis-28383	187	9	images	image	NOUN
bracis-28383	187	10	,	,	PUNCT
bracis-28383	187	11	flickr	flickr	NOUN
bracis-28383	187	12	-	-	PUNCT
bracis-28383	187	13	dog	dog	NOUN
bracis-28383	187	14	cc	cc	NOUN
bracis-28383	187	15	with	with	ADP
bracis-28383	187	16	69	69	NUM
bracis-28383	187	17	,	,	PUNCT
bracis-28383	187	18	and	and	CCONJ
bracis-28383	187	19	fd	fd	VERB
bracis-28383	187	20	with	with	ADP
bracis-28383	187	21	45	45	NUM
bracis-28383	187	22	images	image	NOUN
bracis-28383	187	23	.	.	PUNCT
bracis-28383	188	1	6	6	NUM
bracis-28383	188	2	results	result	NOUN
bracis-28383	188	3	preliminarily	preliminarily	ADV
bracis-28383	188	4	,	,	PUNCT
bracis-28383	188	5	we	we	PRON
bracis-28383	188	6	demonstrated	demonstrate	VERB
bracis-28383	188	7	the	the	DET
bracis-28383	188	8	loss	loss	NOUN
bracis-28383	188	9	for	for	ADP
bracis-28383	188	10	each	each	PRON
bracis-28383	188	11	of	of	ADP
bracis-28383	188	12	the	the	DET
bracis-28383	188	13	training	training	NOUN
bracis-28383	188	14	variations	variation	NOUN
bracis-28383	188	15	of	of	ADP
bracis-28383	188	16	the	the	DET
bracis-28383	188	17	dogfacenet	dogfacenet	NOUN
bracis-28383	188	18	dataset	dataset	NOUN
bracis-28383	188	19	in	in	ADP
bracis-28383	188	20	fig	fig	NOUN
bracis-28383	188	21	.	.	PUNCT
bracis-28383	188	22	 	 	SPACE
bracis-28383	188	23	6	6	NUM
bracis-28383	188	24	and	and	CCONJ
bracis-28383	188	25	noticed	notice	VERB
bracis-28383	188	26	that	that	SCONJ
bracis-28383	188	27	the	the	DET
bracis-28383	188	28	curves	curve	NOUN
bracis-28383	188	29	are	be	AUX
bracis-28383	188	30	very	very	ADV
bracis-28383	188	31	similar	similar	ADJ
bracis-28383	188	32	,	,	PUNCT
bracis-28383	188	33	reaching	reach	VERB
bracis-28383	188	34	very	very	ADV
bracis-28383	188	35	close	close	ADJ
bracis-28383	188	36	values	value	NOUN
bracis-28383	188	37	at	at	ADP
bracis-28383	188	38	the	the	DET
bracis-28383	188	39	end	end	NOUN
bracis-28383	188	40	of	of	ADP
bracis-28383	188	41	the	the	DET
bracis-28383	188	42	training	training	NOUN
bracis-28383	188	43	.	.	PUNCT
bracis-28383	189	1	the	the	DET
bracis-28383	189	2	curve	curve	NOUN
bracis-28383	189	3	of	of	ADP
bracis-28383	189	4	the	the	DET
bracis-28383	189	5	fd	fd	PROPN
bracis-28383	189	6	training	training	NOUN
bracis-28383	189	7	stabilized	stabilize	VERB
bracis-28383	189	8	faster	fast	ADV
bracis-28383	189	9	,	,	PUNCT
bracis-28383	189	10	showing	show	VERB
bracis-28383	189	11	that	that	SCONJ
bracis-28383	189	12	it	it	PRON
bracis-28383	189	13	would	would	AUX
bracis-28383	189	14	not	not	PART
bracis-28383	189	15	require	require	VERB
bracis-28383	189	16	pervasive	pervasive	ADJ
bracis-28383	189	17	training	training	NOUN
bracis-28383	189	18	for	for	ADP
bracis-28383	189	19	this	this	DET
bracis-28383	189	20	case	case	NOUN
bracis-28383	189	21	,	,	PUNCT
bracis-28383	189	22	unlike	unlike	ADP
bracis-28383	189	23	the	the	DET
bracis-28383	189	24	other	other	ADJ
bracis-28383	189	25	two	two	NUM
bracis-28383	189	26	pieces	piece	NOUN
bracis-28383	189	27	of	of	ADP
bracis-28383	189	28	training	training	NOUN
bracis-28383	189	29	that	that	PRON
bracis-28383	189	30	stabilized	stabilize	VERB
bracis-28383	189	31	closer	close	ADV
bracis-28383	189	32	to	to	ADP
bracis-28383	189	33	the	the	DET
bracis-28383	189	34	end	end	NOUN
bracis-28383	189	35	of	of	ADP
bracis-28383	189	36	the	the	DET
bracis-28383	189	37	training	training	NOUN
bracis-28383	189	38	.	.	PUNCT
bracis-28383	190	1	still	still	ADV
bracis-28383	190	2	regarding	regard	VERB
bracis-28383	190	3	the	the	DET
bracis-28383	190	4	training	training	NOUN
bracis-28383	190	5	,	,	PUNCT
bracis-28383	190	6	in	in	ADP
bracis-28383	190	7	fig	fig	NOUN
bracis-28383	190	8	.	.	PUNCT
bracis-28383	190	9	 	 	SPACE
bracis-28383	190	10	7	7	NUM
bracis-28383	190	11	,	,	PUNCT
bracis-28383	190	12	we	we	PRON
bracis-28383	190	13	showed	show	VERB
bracis-28383	190	14	through	through	ADP
bracis-28383	190	15	the	the	DET
bracis-28383	190	16	embeddings	embedding	NOUN
bracis-28383	190	17	extracted	extract	VERB
bracis-28383	190	18	from	from	ADP
bracis-28383	190	19	each	each	DET
bracis-28383	190	20	dataset	dataset	NOUN
bracis-28383	190	21	that	that	PRON
bracis-28383	190	22	resnet-34	resnet-34	PROPN
bracis-28383	190	23	produced	produce	VERB
bracis-28383	190	24	an	an	DET
bracis-28383	190	25	effective	effective	ADJ
bracis-28383	190	26	separation	separation	NOUN
bracis-28383	190	27	between	between	ADP
bracis-28383	190	28	classes	class	NOUN
bracis-28383	190	29	,	,	PUNCT
bracis-28383	190	30	where	where	SCONJ
bracis-28383	190	31	classes	class	NOUN
bracis-28383	190	32	are	be	AUX
bracis-28383	190	33	the	the	DET
bracis-28383	190	34	“	"	PUNCT
bracis-28383	190	35	same	same	ADJ
bracis-28383	190	36	dog	dog	NOUN
bracis-28383	190	37	”	"	PUNCT
bracis-28383	190	38	and	and	CCONJ
bracis-28383	190	39	“	"	PUNCT
bracis-28383	190	40	different	different	ADJ
bracis-28383	190	41	dogs	dog	NOUN
bracis-28383	190	42	”	"	PUNCT
bracis-28383	190	43	,	,	PUNCT
bracis-28383	190	44	with	with	ADP
bracis-28383	190	45	euclidean	euclidean	ADJ
bracis-28383	190	46	distance	distance	NOUN
bracis-28383	190	47	and	and	CCONJ
bracis-28383	190	48	relative	relative	ADJ
bracis-28383	190	49	frequency	frequency	NOUN
bracis-28383	190	50	as	as	ADP
bracis-28383	190	51	dimensions	dimension	NOUN
bracis-28383	190	52	.	.	PUNCT
bracis-28383	191	1	it	it	PRON
bracis-28383	191	2	is	be	AUX
bracis-28383	191	3	also	also	ADV
bracis-28383	191	4	possible	possible	ADJ
bracis-28383	191	5	to	to	PART
bracis-28383	191	6	see	see	VERB
bracis-28383	191	7	in	in	ADP
bracis-28383	191	8	these	these	DET
bracis-28383	191	9	distinction	distinction	NOUN
bracis-28383	191	10	curves	curve	NOUN
bracis-28383	191	11	that	that	SCONJ
bracis-28383	191	12	the	the	DET
bracis-28383	191	13	overlap	overlap	NOUN
bracis-28383	191	14	area	area	NOUN
bracis-28383	191	15	of	of	ADP
bracis-28383	191	16	the	the	DET
bracis-28383	191	17	cc	cc	PROPN
bracis-28383	191	18	is	be	AUX
bracis-28383	191	19	visually	visually	ADV
bracis-28383	191	20	smaller	small	ADJ
bracis-28383	191	21	than	than	ADP
bracis-28383	191	22	the	the	DET
bracis-28383	191	23	other	other	ADJ
bracis-28383	191	24	two	two	NUM
bracis-28383	191	25	pieces	piece	NOUN
bracis-28383	191	26	of	of	ADP
bracis-28383	191	27	training	training	NOUN
bracis-28383	191	28	.	.	PUNCT
bracis-28383	192	1	fig	fig	NOUN
bracis-28383	192	2	.	.	PUNCT
bracis-28383	193	1	6	6	NUM
bracis-28383	193	2	.	.	X
bracis-28383	193	3	loss	loss	NOUN
bracis-28383	193	4	curves	curve	NOUN
bracis-28383	193	5	for	for	ADP
bracis-28383	193	6	dogfacenet	dogfacenet	NOUN
bracis-28383	193	7	.	.	PUNCT
bracis-28383	194	1	full	full	ADJ
bracis-28383	194	2	size	size	NOUN
bracis-28383	194	3	image	image	NOUN
bracis-28383	194	4	fig	fig	NOUN
bracis-28383	194	5	.	.	PUNCT
bracis-28383	195	1	7	7	X
bracis-28383	195	2	.	.	NOUN
bracis-28383	195	3	distance	distance	NOUN
bracis-28383	195	4	distribution	distribution	NOUN
bracis-28383	195	5	curves	curve	NOUN
bracis-28383	195	6	for	for	ADP
bracis-28383	195	7	dogfacenet	dogfacenet	NOUN
bracis-28383	195	8	.	.	PUNCT
bracis-28383	196	1	full	full	ADJ
bracis-28383	196	2	size	size	NOUN
bracis-28383	196	3	image	image	NOUN
bracis-28383	196	4	the	the	DET
bracis-28383	196	5	results	result	NOUN
bracis-28383	196	6	are	be	AUX
bracis-28383	196	7	presented	present	VERB
bracis-28383	196	8	in	in	ADP
bracis-28383	196	9	table	table	NOUN
bracis-28383	196	10	 	 	SPACE
bracis-28383	196	11	4	4	NUM
bracis-28383	196	12	,	,	PUNCT
bracis-28383	196	13	containing	contain	VERB
bracis-28383	196	14	each	each	DET
bracis-28383	196	15	dataset	dataset	NOUN
bracis-28383	196	16	variation	variation	NOUN
bracis-28383	196	17	,	,	PUNCT
bracis-28383	196	18	accuracy	accuracy	NOUN
bracis-28383	196	19	,	,	PUNCT
bracis-28383	196	20	and	and	CCONJ
bracis-28383	196	21	recall	recall	NOUN
bracis-28383	196	22	of	of	ADP
bracis-28383	196	23	the	the	DET
bracis-28383	196	24	method	method	NOUN
bracis-28383	196	25	.	.	PUNCT
bracis-28383	197	1	the	the	DET
bracis-28383	197	2	best	good	ADJ
bracis-28383	197	3	result	result	NOUN
bracis-28383	197	4	in	in	ADP
bracis-28383	197	5	the	the	DET
bracis-28383	197	6	dogfacenet	dogfacenet	NOUN
bracis-28383	197	7	dataset	dataset	NOUN
bracis-28383	197	8	is	be	AUX
bracis-28383	197	9	98.43	98.43	NUM
bracis-28383	197	10	%	%	NOUN
bracis-28383	197	11	of	of	ADP
bracis-28383	197	12	accuracy	accuracy	NOUN
bracis-28383	197	13	with	with	ADP
bracis-28383	197	14	dogfacenet	dogfacenet	NOUN
bracis-28383	197	15	with	with	ADP
bracis-28383	197	16	facial	facial	ADJ
bracis-28383	197	17	detection	detection	NOUN
bracis-28383	197	18	for	for	ADP
bracis-28383	197	19	training	training	NOUN
bracis-28383	197	20	and	and	CCONJ
bracis-28383	197	21	testing	testing	NOUN
bracis-28383	197	22	.	.	PUNCT
bracis-28383	198	1	in	in	ADP
bracis-28383	198	2	the	the	DET
bracis-28383	198	3	central	central	ADJ
bracis-28383	198	4	cropped	crop	VERB
bracis-28383	198	5	dataset	dataset	NOUN
bracis-28383	198	6	version	version	NOUN
bracis-28383	198	7	,	,	PUNCT
bracis-28383	198	8	the	the	DET
bracis-28383	198	9	method	method	NOUN
bracis-28383	198	10	achieves	achieve	VERB
bracis-28383	198	11	85.64	85.64	NUM
bracis-28383	198	12	%	%	NOUN
bracis-28383	198	13	of	of	ADP
bracis-28383	198	14	recall	recall	NOUN
bracis-28383	198	15	without	without	ADP
bracis-28383	198	16	preprocessing	preprocesse	VERB
bracis-28383	198	17	in	in	ADP
bracis-28383	198	18	training	training	NOUN
bracis-28383	198	19	.	.	PUNCT
bracis-28383	199	1	furthermore	furthermore	ADV
bracis-28383	199	2	,	,	PUNCT
bracis-28383	199	3	the	the	DET
bracis-28383	199	4	best	good	ADJ
bracis-28383	199	5	result	result	NOUN
bracis-28383	199	6	in	in	ADP
bracis-28383	199	7	the	the	DET
bracis-28383	199	8	flickr	flickr	NOUN
bracis-28383	199	9	-	-	PUNCT
bracis-28383	199	10	dog	dog	NOUN
bracis-28383	199	11	is	be	AUX
bracis-28383	199	12	91.53	91.53	NUM
bracis-28383	199	13	%	%	NOUN
bracis-28383	199	14	of	of	ADP
bracis-28383	199	15	accuracy	accuracy	NOUN
bracis-28383	199	16	with	with	ADP
bracis-28383	199	17	facial	facial	ADJ
bracis-28383	199	18	detection	detection	NOUN
bracis-28383	199	19	in	in	ADP
bracis-28383	199	20	training	training	NOUN
bracis-28383	199	21	and	and	CCONJ
bracis-28383	199	22	testing	testing	NOUN
bracis-28383	199	23	.	.	PUNCT
bracis-28383	200	1	overall	overall	ADV
bracis-28383	200	2	,	,	PUNCT
bracis-28383	200	3	the	the	DET
bracis-28383	200	4	facial	facial	ADJ
bracis-28383	200	5	alignment	alignment	NOUN
bracis-28383	200	6	fd	fd	PROPN
bracis-28383	200	7	allowed	allow	VERB
bracis-28383	200	8	for	for	ADP
bracis-28383	200	9	achieving	achieve	VERB
bracis-28383	200	10	the	the	DET
bracis-28383	200	11	best	good	ADJ
bracis-28383	200	12	accuracy	accuracy	NOUN
bracis-28383	200	13	for	for	ADP
bracis-28383	200	14	both	both	DET
bracis-28383	200	15	datasets	dataset	NOUN
bracis-28383	200	16	used	use	VERB
bracis-28383	200	17	in	in	ADP
bracis-28383	200	18	this	this	DET
bracis-28383	200	19	work	work	NOUN
bracis-28383	200	20	,	,	PUNCT
bracis-28383	200	21	which	which	PRON
bracis-28383	200	22	probably	probably	ADV
bracis-28383	200	23	makes	make	VERB
bracis-28383	200	24	the	the	DET
bracis-28383	200	25	network	network	NOUN
bracis-28383	200	26	more	more	ADV
bracis-28383	200	27	objective	objective	ADJ
bracis-28383	200	28	,	,	PUNCT
bracis-28383	200	29	not	not	PART
bracis-28383	200	30	paying	pay	VERB
bracis-28383	200	31	attention	attention	NOUN
bracis-28383	200	32	to	to	ADP
bracis-28383	200	33	details	detail	NOUN
bracis-28383	200	34	that	that	PRON
bracis-28383	200	35	are	be	AUX
bracis-28383	200	36	not	not	PART
bracis-28383	200	37	part	part	NOUN
bracis-28383	200	38	of	of	ADP
bracis-28383	200	39	the	the	DET
bracis-28383	200	40	dog	dog	NOUN
bracis-28383	200	41	’s	’s	PART
bracis-28383	200	42	face	face	NOUN
bracis-28383	200	43	but	but	CCONJ
bracis-28383	200	44	are	be	AUX
bracis-28383	200	45	in	in	ADP
bracis-28383	200	46	the	the	DET
bracis-28383	200	47	image	image	NOUN
bracis-28383	200	48	.	.	PUNCT
bracis-28383	201	1	table	table	NOUN
bracis-28383	201	2	4	4	NUM
bracis-28383	201	3	.	.	PUNCT
bracis-28383	202	1	results	result	NOUN
bracis-28383	202	2	of	of	ADP
bracis-28383	202	3	accuracy	accuracy	NOUN
bracis-28383	202	4	and	and	CCONJ
bracis-28383	202	5	recall	recall	NOUN
bracis-28383	202	6	using	use	VERB
bracis-28383	202	7	resnet-34	resnet-34	PROPN
bracis-28383	202	8	modelfull	modelfull	ADJ
bracis-28383	202	9	size	size	NOUN
bracis-28383	202	10	table	table	NOUN
bracis-28383	202	11	table	table	NOUN
bracis-28383	202	12	5	5	NUM
bracis-28383	202	13	.	.	PUNCT
bracis-28383	202	14	comparison	comparison	NOUN
bracis-28383	202	15	with	with	ADP
bracis-28383	202	16	other	other	ADJ
bracis-28383	202	17	methodsfull	methodsfull	NOUN
bracis-28383	202	18	size	size	NOUN
bracis-28383	202	19	table	table	NOUN
bracis-28383	202	20	to	to	PART
bracis-28383	202	21	compare	compare	VERB
bracis-28383	202	22	more	more	ADV
bracis-28383	202	23	fairly	fairly	ADV
bracis-28383	202	24	,	,	PUNCT
bracis-28383	202	25	we	we	PRON
bracis-28383	202	26	do	do	AUX
bracis-28383	202	27	not	not	PART
bracis-28383	202	28	use	use	VERB
bracis-28383	202	29	our	our	PRON
bracis-28383	202	30	results	result	NOUN
bracis-28383	202	31	obtained	obtain	VERB
bracis-28383	202	32	in	in	ADP
bracis-28383	202	33	tests	test	NOUN
bracis-28383	202	34	on	on	ADP
bracis-28383	202	35	the	the	DET
bracis-28383	202	36	fd	fd	PROPN
bracis-28383	202	37	version	version	NOUN
bracis-28383	202	38	with	with	ADP
bracis-28383	202	39	the	the	DET
bracis-28383	202	40	other	other	ADJ
bracis-28383	202	41	works	work	NOUN
bracis-28383	202	42	in	in	ADP
bracis-28383	202	43	the	the	DET
bracis-28383	202	44	literature	literature	NOUN
bracis-28383	202	45	.	.	PUNCT
bracis-28383	203	1	we	we	PRON
bracis-28383	203	2	consider	consider	VERB
bracis-28383	203	3	that	that	SCONJ
bracis-28383	203	4	the	the	DET
bracis-28383	203	5	modification	modification	NOUN
bracis-28383	203	6	made	make	VERB
bracis-28383	203	7	and	and	CCONJ
bracis-28383	203	8	the	the	DET
bracis-28383	203	9	loss	loss	NOUN
bracis-28383	203	10	of	of	ADP
bracis-28383	203	11	images	image	NOUN
bracis-28383	203	12	generate	generate	VERB
bracis-28383	203	13	an	an	DET
bracis-28383	203	14	impact	impact	NOUN
bracis-28383	203	15	,	,	PUNCT
bracis-28383	203	16	making	make	VERB
bracis-28383	203	17	the	the	DET
bracis-28383	203	18	data	datum	NOUN
bracis-28383	203	19	set	set	VERB
bracis-28383	203	20	different	different	ADJ
bracis-28383	203	21	.	.	PUNCT
bracis-28383	204	1	using	use	VERB
bracis-28383	204	2	the	the	DET
bracis-28383	204	3	cc	cc	NOUN
bracis-28383	204	4	technique	technique	NOUN
bracis-28383	204	5	with	with	ADP
bracis-28383	204	6	no	no	DET
bracis-28383	204	7	image	image	NOUN
bracis-28383	204	8	loss	loss	NOUN
bracis-28383	204	9	,	,	PUNCT
bracis-28383	204	10	we	we	PRON
bracis-28383	204	11	demonstrated	demonstrate	VERB
bracis-28383	204	12	better	well	ADJ
bracis-28383	204	13	results	result	NOUN
bracis-28383	204	14	than	than	ADP
bracis-28383	204	15	all	all	DET
bracis-28383	204	16	the	the	DET
bracis-28383	204	17	other	other	ADJ
bracis-28383	204	18	works	work	NOUN
bracis-28383	204	19	by	by	ADP
bracis-28383	204	20	using	use	VERB
bracis-28383	204	21	fd	fd	PROPN
bracis-28383	204	22	for	for	ADP
bracis-28383	204	23	training	training	NOUN
bracis-28383	204	24	and	and	CCONJ
bracis-28383	204	25	cc	cc	NOUN
bracis-28383	204	26	for	for	ADP
bracis-28383	204	27	testing	testing	NOUN
bracis-28383	204	28	,	,	PUNCT
bracis-28383	204	29	as	as	SCONJ
bracis-28383	204	30	shown	show	VERB
bracis-28383	204	31	in	in	ADP
bracis-28383	204	32	table	table	NOUN
bracis-28383	204	33	 	 	SPACE
bracis-28383	204	34	5	5	NUM
bracis-28383	204	35	.	.	NOUN
bracis-28383	204	36	7	7	NUM
bracis-28383	204	37	conclusion	conclusion	NOUN
bracis-28383	204	38	this	this	DET
bracis-28383	204	39	work	work	NOUN
bracis-28383	204	40	investigated	investigate	VERB
bracis-28383	204	41	the	the	DET
bracis-28383	204	42	problem	problem	NOUN
bracis-28383	204	43	of	of	ADP
bracis-28383	204	44	dog	dog	NOUN
bracis-28383	204	45	face	face	NOUN
bracis-28383	204	46	recognition	recognition	NOUN
bracis-28383	204	47	.	.	PUNCT
bracis-28383	205	1	we	we	PRON
bracis-28383	205	2	compared	compare	VERB
bracis-28383	205	3	three	three	NUM
bracis-28383	205	4	strategies	strategy	NOUN
bracis-28383	205	5	to	to	PART
bracis-28383	205	6	preprocess	preprocess	VERB
bracis-28383	205	7	the	the	DET
bracis-28383	205	8	input	input	NOUN
bracis-28383	205	9	images	image	NOUN
bracis-28383	205	10	:	:	PUNCT
bracis-28383	205	11	no	no	DET
bracis-28383	205	12	preprocessing	preprocessing	NOUN
bracis-28383	205	13	,	,	PUNCT
bracis-28383	205	14	central	central	ADJ
bracis-28383	205	15	crop	crop	NOUN
bracis-28383	205	16	,	,	PUNCT
bracis-28383	205	17	and	and	CCONJ
bracis-28383	205	18	face	face	VERB
bracis-28383	205	19	detection	detection	NOUN
bracis-28383	205	20	with	with	ADP
bracis-28383	205	21	alignment	alignment	NOUN
bracis-28383	205	22	.	.	PUNCT
bracis-28383	206	1	we	we	PRON
bracis-28383	206	2	explored	explore	VERB
bracis-28383	206	3	the	the	DET
bracis-28383	206	4	performance	performance	NOUN
bracis-28383	206	5	of	of	ADP
bracis-28383	206	6	a	a	DET
bracis-28383	206	7	robust	robust	ADJ
bracis-28383	206	8	baseline	baseline	NOUN
bracis-28383	206	9	model	model	NOUN
bracis-28383	206	10	,	,	PUNCT
bracis-28383	206	11	resnet-34	resnet-34	PROPN
bracis-28383	206	12	,	,	PUNCT
bracis-28383	206	13	showing	show	VERB
bracis-28383	206	14	that	that	SCONJ
bracis-28383	206	15	it	it	PRON
bracis-28383	206	16	can	can	AUX
bracis-28383	206	17	be	be	AUX
bracis-28383	206	18	competitive	competitive	ADJ
bracis-28383	206	19	or	or	CCONJ
bracis-28383	206	20	even	even	ADV
bracis-28383	206	21	outperform	outperform	VERB
bracis-28383	206	22	the	the	DET
bracis-28383	206	23	results	result	NOUN
bracis-28383	206	24	found	find	VERB
bracis-28383	206	25	in	in	ADP
bracis-28383	206	26	the	the	DET
bracis-28383	206	27	literature	literature	NOUN
bracis-28383	206	28	.	.	PUNCT
bracis-28383	207	1	moreover	moreover	ADV
bracis-28383	207	2	,	,	PUNCT
bracis-28383	207	3	this	this	DET
bracis-28383	207	4	paper	paper	NOUN
bracis-28383	207	5	also	also	ADV
bracis-28383	207	6	includes	include	VERB
bracis-28383	207	7	recall	recall	NOUN
bracis-28383	207	8	as	as	ADP
bracis-28383	207	9	an	an	DET
bracis-28383	207	10	essential	essential	ADJ
bracis-28383	207	11	performance	performance	NOUN
bracis-28383	207	12	metric	metric	NOUN
bracis-28383	207	13	regarding	regard	VERB
bracis-28383	207	14	the	the	DET
bracis-28383	207	15	verification	verification	NOUN
bracis-28383	207	16	that	that	PRON
bracis-28383	207	17	is	be	AUX
bracis-28383	207	18	missing	miss	VERB
bracis-28383	207	19	in	in	ADP
bracis-28383	207	20	most	most	ADV
bracis-28383	207	21	related	relate	VERB
bracis-28383	207	22	works	work	NOUN
bracis-28383	207	23	,	,	PUNCT
bracis-28383	207	24	especially	especially	ADV
bracis-28383	207	25	considering	consider	VERB
bracis-28383	207	26	the	the	DET
bracis-28383	207	27	method	method	NOUN
bracis-28383	207	28	’s	’s	PART
bracis-28383	207	29	ability	ability	NOUN
bracis-28383	207	30	not	not	PART
bracis-28383	207	31	to	to	PART
bracis-28383	207	32	repudiate	repudiate	VERB
bracis-28383	207	33	the	the	DET
bracis-28383	207	34	dog	dog	NOUN
bracis-28383	207	35	.	.	PUNCT
bracis-28383	208	1	experimental	experimental	ADJ
bracis-28383	208	2	results	result	NOUN
bracis-28383	208	3	show	show	VERB
bracis-28383	208	4	that	that	SCONJ
bracis-28383	208	5	face	face	NOUN
bracis-28383	208	6	alignment	alignment	NOUN
bracis-28383	208	7	does	do	AUX
bracis-28383	208	8	not	not	PART
bracis-28383	208	9	play	play	VERB
bracis-28383	208	10	an	an	DET
bracis-28383	208	11	important	important	ADJ
bracis-28383	208	12	role	role	NOUN
bracis-28383	208	13	for	for	ADP
bracis-28383	208	14	improving	improve	VERB
bracis-28383	208	15	recall	recall	NOUN
bracis-28383	208	16	in	in	ADP
bracis-28383	208	17	dog	dog	NOUN
bracis-28383	208	18	face	face	NOUN
bracis-28383	208	19	recognition	recognition	NOUN
bracis-28383	208	20	,	,	PUNCT
bracis-28383	208	21	where	where	SCONJ
bracis-28383	208	22	the	the	DET
bracis-28383	208	23	results	result	NOUN
bracis-28383	208	24	on	on	ADP
bracis-28383	208	25	datasets	dataset	NOUN
bracis-28383	208	26	with	with	ADP
bracis-28383	208	27	preprocessing	preprocessing	NOUN
bracis-28383	208	28	were	be	AUX
bracis-28383	208	29	slightly	slightly	ADV
bracis-28383	208	30	better	well	ADJ
bracis-28383	208	31	than	than	ADP
bracis-28383	208	32	the	the	DET
bracis-28383	208	33	others	other	NOUN
bracis-28383	208	34	.	.	PUNCT
bracis-28383	209	1	finally	finally	ADV
bracis-28383	209	2	,	,	PUNCT
bracis-28383	209	3	it	it	PRON
bracis-28383	209	4	is	be	AUX
bracis-28383	209	5	worth	worth	ADJ
bracis-28383	209	6	emphasizing	emphasize	VERB
bracis-28383	209	7	that	that	SCONJ
bracis-28383	209	8	the	the	DET
bracis-28383	209	9	alignment	alignment	NOUN
bracis-28383	209	10	process	process	NOUN
bracis-28383	209	11	pays	pay	VERB
bracis-28383	209	12	off	off	ADP
bracis-28383	209	13	on	on	ADP
bracis-28383	209	14	flickr	flickr	NOUN
bracis-28383	209	15	-	-	PUNCT
bracis-28383	209	16	dog	dog	NOUN
bracis-28383	209	17	since	since	SCONJ
bracis-28383	209	18	(	(	PUNCT
bracis-28383	209	19	1	1	X
bracis-28383	209	20	)	)	PUNCT
bracis-28383	209	21	this	this	DET
bracis-28383	209	22	strategy	strategy	NOUN
bracis-28383	209	23	improved	improve	VERB
bracis-28383	209	24	the	the	DET
bracis-28383	209	25	state	state	NOUN
bracis-28383	209	26	-	-	PUNCT
bracis-28383	209	27	of	of	ADP
bracis-28383	209	28	-	-	PUNCT
bracis-28383	209	29	the	the	DET
bracis-28383	209	30	-	-	PUNCT
bracis-28383	209	31	art	art	NOUN
bracis-28383	209	32	considering	consider	VERB
bracis-28383	209	33	both	both	CCONJ
bracis-28383	209	34	accuracy	accuracy	NOUN
bracis-28383	209	35	and	and	CCONJ
bracis-28383	209	36	recall	recall	NOUN
bracis-28383	209	37	,	,	PUNCT
bracis-28383	209	38	and	and	CCONJ
bracis-28383	209	39	(	(	PUNCT
bracis-28383	209	40	2	2	X
bracis-28383	209	41	)	)	PUNCT
bracis-28383	209	42	our	our	PRON
bracis-28383	209	43	experiments	experiment	NOUN
bracis-28383	209	44	evaluated	evaluate	VERB
bracis-28383	209	45	the	the	DET
bracis-28383	209	46	entire	entire	ADJ
bracis-28383	209	47	flickr	flickr	NOUN
bracis-28383	209	48	-	-	PUNCT
bracis-28383	209	49	dog	dog	NOUN
bracis-28383	209	50	since	since	SCONJ
bracis-28383	209	51	the	the	DET
bracis-28383	209	52	model	model	NOUN
bracis-28383	209	53	was	be	AUX
bracis-28383	209	54	trained	train	VERB
bracis-28383	209	55	with	with	ADP
bracis-28383	209	56	external	external	ADJ
bracis-28383	209	57	data	datum	NOUN
bracis-28383	209	58	.	.	PUNCT
bracis-28383	210	1	the	the	DET
bracis-28383	210	2	latter	latter	ADJ
bracis-28383	210	3	observation	observation	NOUN
bracis-28383	210	4	is	be	AUX
bracis-28383	210	5	consistent	consistent	ADJ
bracis-28383	210	6	with	with	ADP
bracis-28383	210	7	the	the	DET
bracis-28383	210	8	hypothesis	hypothesis	NOUN
bracis-28383	210	9	found	find	VERB
bracis-28383	210	10	in	in	ADP
bracis-28383	210	11	previous	previous	ADJ
bracis-28383	210	12	works	work	NOUN
bracis-28383	210	13	[	[	X
bracis-28383	210	14	23	23	NUM
bracis-28383	210	15	]	]	PUNCT
bracis-28383	210	16	that	that	SCONJ
bracis-28383	210	17	augmentation	augmentation	NOUN
bracis-28383	210	18	of	of	ADP
bracis-28383	210	19	dog	dog	NOUN
bracis-28383	210	20	breeds	breed	NOUN
bracis-28383	210	21	yields	yield	VERB
bracis-28383	210	22	better	well	ADJ
bracis-28383	210	23	accuracy	accuracy	NOUN
bracis-28383	210	24	when	when	SCONJ
bracis-28383	210	25	the	the	DET
bracis-28383	210	26	literature	literature	NOUN
bracis-28383	210	27	focus	focus	VERB
bracis-28383	210	28	on	on	ADP
bracis-28383	210	29	classification	classification	NOUN
bracis-28383	210	30	and	and	CCONJ
bracis-28383	210	31	lacks	lack	VERB
bracis-28383	210	32	a	a	DET
bracis-28383	210	33	proper	proper	ADJ
bracis-28383	210	34	analysis	analysis	NOUN
bracis-28383	210	35	for	for	ADP
bracis-28383	210	36	identification	identification	NOUN
bracis-28383	210	37	.	.	PUNCT
bracis-28383	211	1	future	future	ADJ
bracis-28383	211	2	work	work	NOUN
bracis-28383	211	3	includes	include	VERB
bracis-28383	211	4	extending	extend	VERB
bracis-28383	211	5	the	the	DET
bracis-28383	211	6	existing	exist	VERB
bracis-28383	211	7	dog	dog	NOUN
bracis-28383	211	8	identification	identification	NOUN
bracis-28383	211	9	datasets	dataset	NOUN
bracis-28383	211	10	in	in	ADP
bracis-28383	211	11	terms	term	NOUN
bracis-28383	211	12	of	of	ADP
bracis-28383	211	13	both	both	CCONJ
bracis-28383	211	14	new	new	ADJ
bracis-28383	211	15	identities	identity	NOUN
bracis-28383	211	16	and	and	CCONJ
bracis-28383	211	17	additional	additional	ADJ
bracis-28383	211	18	annotations	annotation	NOUN
bracis-28383	211	19	,	,	PUNCT
bracis-28383	211	20	adding	add	VERB
bracis-28383	211	21	support	support	NOUN
bracis-28383	211	22	for	for	ADP
bracis-28383	211	23	cat	cat	NOUN
bracis-28383	211	24	face	face	NOUN
bracis-28383	211	25	recognition	recognition	NOUN
bracis-28383	211	26	,	,	PUNCT
bracis-28383	211	27	investigating	investigate	VERB
bracis-28383	211	28	effective	effective	ADJ
bracis-28383	211	29	approaches	approach	NOUN
bracis-28383	211	30	for	for	ADP
bracis-28383	211	31	animal	animal	NOUN
bracis-28383	211	32	identification	identification	NOUN
bracis-28383	211	33	and	and	CCONJ
bracis-28383	211	34	re	re	NOUN
bracis-28383	211	35	-	-	NOUN
bracis-28383	211	36	identification	identification	NOUN
bracis-28383	211	37	,	,	PUNCT
bracis-28383	211	38	and	and	CCONJ
bracis-28383	211	39	the	the	DET
bracis-28383	211	40	assessment	assessment	NOUN
bracis-28383	211	41	of	of	ADP
bracis-28383	211	42	other	other	ADJ
bracis-28383	211	43	techniques	technique	NOUN
bracis-28383	211	44	for	for	ADP
bracis-28383	211	45	the	the	DET
bracis-28383	211	46	problem	problem	NOUN
bracis-28383	211	47	of	of	ADP
bracis-28383	211	48	animal	animal	NOUN
bracis-28383	211	49	biometric	biometric	ADJ
bracis-28383	211	50	recognition	recognition	NOUN
bracis-28383	211	51	.	.	PUNCT
bracis-28383	212	1	notes	note	VERB
bracis-28383	212	2	1.https://github.com/guillaumemougeot/dogfacenet	1.https://github.com/guillaumemougeot/dogfacenet	NUM
bracis-28383	212	3	.	.	PUNCT
bracis-28383	213	1	references	reference	NOUN
bracis-28383	213	2	bae	bae	PROPN
bracis-28383	213	3	,	,	PUNCT
bracis-28383	213	4	h.b	h.b	PROPN
bracis-28383	213	5	.	.	PROPN
bracis-28383	213	6	,	,	PUNCT
bracis-28383	213	7	pak	pak	PROPN
bracis-28383	213	8	,	,	PUNCT
bracis-28383	213	9	d.	d.	PROPN
bracis-28383	213	10	,	,	PUNCT
bracis-28383	213	11	lee	lee	PROPN
bracis-28383	213	12	,	,	PUNCT
bracis-28383	213	13	s.	s.	PROPN
bracis-28383	213	14	:	:	PUNCT
bracis-28383	213	15	dog	dog	PROPN
bracis-28383	213	16	nose	nose	NOUN
bracis-28383	213	17	-	-	PUNCT
bracis-28383	213	18	print	print	NOUN
bracis-28383	213	19	identification	identification	NOUN
bracis-28383	213	20	using	use	VERB
bracis-28383	213	21	deep	deep	ADJ
bracis-28383	213	22	neural	neural	ADJ
bracis-28383	213	23	networks	network	NOUN
bracis-28383	213	24	.	.	PUNCT
bracis-28383	214	1	ieee	ieee	NOUN
bracis-28383	214	2	access	access	NOUN
bracis-28383	214	3	9	9	NUM
bracis-28383	214	4	,	,	PUNCT
bracis-28383	214	5	49141–49153	49141–49153	NUM
bracis-28383	214	6	(	(	PUNCT
bracis-28383	214	7	2021	2021	NUM
bracis-28383	214	8	)	)	PUNCT
bracis-28383	214	9	article	article	NOUN
bracis-28383	214	10	  	  	SPACE
bracis-28383	214	11	google	google	PROPN
bracis-28383	214	12	scholar	scholar	NOUN
bracis-28383	214	13	  	  	SPACE
bracis-28383	214	14	batic	batic	PROPN
bracis-28383	214	15	,	,	PUNCT
bracis-28383	214	16	d.	d.	PROPN
bracis-28383	214	17	,	,	PUNCT
bracis-28383	214	18	culibrk	culibrk	PROPN
bracis-28383	214	19	,	,	PUNCT
bracis-28383	214	20	d.	d.	PROPN
bracis-28383	214	21	:	:	PUNCT
bracis-28383	214	22	identifying	identify	VERB
bracis-28383	214	23	individual	individual	ADJ
bracis-28383	214	24	dogs	dog	NOUN
bracis-28383	214	25	in	in	ADP
bracis-28383	214	26	social	social	ADJ
bracis-28383	214	27	media	medium	NOUN
bracis-28383	214	28	images	image	NOUN
bracis-28383	214	29	.	.	PUNCT
bracis-28383	215	1	arxiv	arxiv	PROPN
bracis-28383	215	2	preprint	preprint	VERB
bracis-28383	215	3	arxiv:2003.06705	arxiv:2003.06705	NOUN
bracis-28383	215	4	(	(	PUNCT
bracis-28383	215	5	2020	2020	NUM
bracis-28383	215	6	)	)	PUNCT
bracis-28383	216	1	capone	capone	PROPN
bracis-28383	216	2	,	,	PUNCT
bracis-28383	216	3	v.	v.	PROPN
bracis-28383	216	4	,	,	PUNCT
bracis-28383	216	5	figueiredo	figueiredo	PROPN
bracis-28383	216	6	,	,	PUNCT
bracis-28383	216	7	c.	c.	NOUN
bracis-28383	216	8	,	,	PUNCT
bracis-28383	216	9	valle	valle	PROPN
bracis-28383	216	10	,	,	PUNCT
bracis-28383	216	11	e.	e.	PROPN
bracis-28383	216	12	,	,	PUNCT
bracis-28383	216	13	andaló	andaló	PROPN
bracis-28383	216	14	,	,	PUNCT
bracis-28383	216	15	f.	f.	PROPN
bracis-28383	216	16	:	:	PUNCT
bracis-28383	216	17	crowdpet	crowdpet	NOUN
bracis-28383	216	18	:	:	PUNCT
bracis-28383	216	19	deep	deep	ADJ
bracis-28383	216	20	learning	learning	NOUN
bracis-28383	216	21	applied	apply	VERB
bracis-28383	216	22	to	to	ADP
bracis-28383	216	23	the	the	DET
bracis-28383	216	24	detection	detection	NOUN
bracis-28383	216	25	of	of	ADP
bracis-28383	216	26	dogs	dog	NOUN
bracis-28383	216	27	in	in	ADP
bracis-28383	216	28	the	the	DET
bracis-28383	216	29	wild	wild	NOUN
bracis-28383	216	30	.	.	PUNCT
bracis-28383	217	1	multimedia	multimedia	NOUN
bracis-28383	217	2	tools	tool	NOUN
bracis-28383	217	3	appl	appl	NOUN
bracis-28383	217	4	.	.	PUNCT
bracis-28383	218	1	76(14	76(14	NOUN
bracis-28383	218	2	)	)	PUNCT
bracis-28383	218	3	,	,	PUNCT
bracis-28383	218	4	15325–15340	15325–15340	NUM
bracis-28383	218	5	(	(	PUNCT
bracis-28383	218	6	2017	2017	NUM
bracis-28383	218	7	)	)	PUNCT
bracis-28383	218	8	google	google	PROPN
bracis-28383	218	9	scholar	scholar	NOUN
bracis-28383	218	10	  	  	SPACE
bracis-28383	218	11	chaturvedi	chaturvedi	PROPN
bracis-28383	218	12	,	,	PUNCT
bracis-28383	218	13	k.	k.	PROPN
bracis-28383	218	14	:	:	PUNCT
bracis-28383	218	15	wolf	wolf	PROPN
bracis-28383	218	16	and	and	CCONJ
bracis-28383	218	17	dog	dog	NOUN
bracis-28383	218	18	breed	breed	NOUN
bracis-28383	218	19	image	image	NOUN
bracis-28383	218	20	classification	classification	NOUN
bracis-28383	218	21	using	use	VERB
bracis-28383	218	22	deep	deep	ADJ
bracis-28383	218	23	learning	learning	NOUN
bracis-28383	218	24	techniques	technique	NOUN
bracis-28383	218	25	.	.	PUNCT
bracis-28383	219	1	ph.d	ph.d	PROPN
bracis-28383	219	2	.	.	PUNCT
bracis-28383	220	1	thesis	thesis	PROPN
bracis-28383	220	2	,	,	PUNCT
bracis-28383	220	3	dublin	dublin	PROPN
bracis-28383	220	4	,	,	PUNCT
bracis-28383	220	5	national	national	PROPN
bracis-28383	220	6	college	college	PROPN
bracis-28383	220	7	of	of	ADP
bracis-28383	220	8	ireland	ireland	PROPN
bracis-28383	220	9	(	(	PUNCT
bracis-28383	220	10	2020	2020	NUM
bracis-28383	220	11	)	)	PUNCT
bracis-28383	220	12	google	google	PROPN
bracis-28383	220	13	scholar	scholar	NOUN
bracis-28383	220	14	  	  	SPACE
bracis-28383	220	15	chollet	chollet	NOUN
bracis-28383	220	16	,	,	PUNCT
bracis-28383	220	17	f.	f.	PROPN
bracis-28383	220	18	:	:	PUNCT
bracis-28383	220	19	xception	xception	NOUN
bracis-28383	220	20	:	:	PUNCT
bracis-28383	220	21	deep	deep	ADJ
bracis-28383	220	22	learning	learn	VERB
bracis-28383	220	23	with	with	ADP
bracis-28383	220	24	depthwise	depthwise	NOUN
bracis-28383	220	25	separable	separable	ADJ
bracis-28383	220	26	convolutions	convolution	NOUN
bracis-28383	220	27	.	.	PUNCT
bracis-28383	221	1	in	in	ADP
bracis-28383	221	2	:	:	PUNCT
bracis-28383	221	3	proceedings	proceeding	NOUN
bracis-28383	221	4	of	of	ADP
bracis-28383	221	5	the	the	DET
bracis-28383	221	6	ieee	ieee	NOUN
bracis-28383	221	7	conference	conference	NOUN
bracis-28383	221	8	on	on	ADP
bracis-28383	221	9	computer	computer	NOUN
bracis-28383	221	10	vision	vision	NOUN
bracis-28383	221	11	and	and	CCONJ
bracis-28383	221	12	pattern	pattern	NOUN
bracis-28383	221	13	recognition	recognition	NOUN
bracis-28383	221	14	,	,	PUNCT
bracis-28383	221	15	pp	pp	PROPN
bracis-28383	221	16	.	.	PUNCT
bracis-28383	222	1	1251–1258	1251–1258	NUM
bracis-28383	222	2	(	(	PUNCT
bracis-28383	222	3	2017	2017	NUM
bracis-28383	222	4	)	)	PUNCT
bracis-28383	222	5	google	google	NOUN
bracis-28383	222	6	scholar	scholar	NOUN
bracis-28383	222	7	  	  	SPACE
bracis-28383	222	8	he	he	PRON
bracis-28383	222	9	,	,	PUNCT
bracis-28383	222	10	k.	k.	PROPN
bracis-28383	222	11	,	,	PUNCT
bracis-28383	222	12	zhang	zhang	PROPN
bracis-28383	222	13	,	,	PUNCT
bracis-28383	222	14	x.	x.	PROPN
bracis-28383	222	15	,	,	PUNCT
bracis-28383	222	16	ren	ren	PROPN
bracis-28383	222	17	,	,	PUNCT
bracis-28383	222	18	s.	s.	PROPN
bracis-28383	222	19	,	,	PUNCT
bracis-28383	222	20	sun	sun	PROPN
bracis-28383	222	21	,	,	PUNCT
bracis-28383	222	22	j.	j.	PROPN
bracis-28383	222	23	:	:	PUNCT
bracis-28383	222	24	deep	deep	ADJ
bracis-28383	222	25	residual	residual	ADJ
bracis-28383	222	26	learning	learning	NOUN
bracis-28383	222	27	for	for	ADP
bracis-28383	222	28	image	image	NOUN
bracis-28383	222	29	recognition	recognition	NOUN
bracis-28383	222	30	.	.	PUNCT
bracis-28383	223	1	in	in	ADP
bracis-28383	223	2	:	:	PUNCT
bracis-28383	223	3	proceedings	proceeding	NOUN
bracis-28383	223	4	of	of	ADP
bracis-28383	223	5	the	the	DET
bracis-28383	223	6	ieee	ieee	NOUN
bracis-28383	223	7	conference	conference	NOUN
bracis-28383	223	8	on	on	ADP
bracis-28383	223	9	computer	computer	NOUN
bracis-28383	223	10	vision	vision	NOUN
bracis-28383	223	11	and	and	CCONJ
bracis-28383	223	12	pattern	pattern	NOUN
bracis-28383	223	13	recognition	recognition	NOUN
bracis-28383	223	14	,	,	PUNCT
bracis-28383	223	15	pp	pp	ADP
bracis-28383	223	16	.	.	PUNCT
bracis-28383	224	1	770–778	770–778	NUM
bracis-28383	224	2	(	(	PUNCT
bracis-28383	224	3	2016	2016	NUM
bracis-28383	224	4	)	)	PUNCT
bracis-28383	224	5	google	google	PROPN
bracis-28383	224	6	scholar	scholar	NOUN
bracis-28383	224	7	  	  	SPACE
bracis-28383	224	8	karasu	karasu	PROPN
bracis-28383	224	9	,	,	PUNCT
bracis-28383	224	10	s.	s.	PROPN
bracis-28383	224	11	,	,	PUNCT
bracis-28383	224	12	alkar	alkar	PROPN
bracis-28383	224	13	,	,	PUNCT
bracis-28383	224	14	ö.	ö.	PROPN
bracis-28383	224	15	,	,	PUNCT
bracis-28383	224	16	et	et	PROPN
bracis-28383	224	17	al	al	PROPN
bracis-28383	224	18	.	.	PUNCT
bracis-28383	224	19	:	:	PUNCT
bracis-28383	225	1	a	a	DET
bracis-28383	225	2	qualitative	qualitative	ADJ
bracis-28383	225	3	investigation	investigation	NOUN
bracis-28383	225	4	on	on	ADP
bracis-28383	225	5	the	the	DET
bracis-28383	225	6	mourning	mourning	NOUN
bracis-28383	225	7	period	period	NOUN
bracis-28383	225	8	that	that	PRON
bracis-28383	225	9	occurs	occur	VERB
bracis-28383	225	10	after	after	ADP
bracis-28383	225	11	the	the	DET
bracis-28383	225	12	loss	loss	NOUN
bracis-28383	225	13	of	of	ADP
bracis-28383	225	14	pets	pet	NOUN
bracis-28383	225	15	.	.	PUNCT
bracis-28383	226	1	veteriner	veteriner	PROPN
bracis-28383	226	2	hekimler	hekimler	PROPN
bracis-28383	226	3	derneği	derneği	PROPN
bracis-28383	226	4	dergisi	dergisi	PROPN
bracis-28383	226	5	/	/	SYM
bracis-28383	226	6	j.	j.	PROPN
bracis-28383	226	7	turk	turk	PROPN
bracis-28383	226	8	.	.	PUNCT
bracis-28383	227	1	vet	vet	PROPN
bracis-28383	227	2	.	.	PUNCT
bracis-28383	228	1	med	med	PROPN
bracis-28383	228	2	.	.	PUNCT
bracis-28383	228	3	soc	soc	PROPN
bracis-28383	228	4	.	.	PUNCT
bracis-28383	229	1	91(2	91(2	NUM
bracis-28383	229	2	)	)	PUNCT
bracis-28383	229	3	,	,	PUNCT
bracis-28383	229	4	86–97	86–97	NUM
bracis-28383	229	5	(	(	PUNCT
bracis-28383	229	6	2020	2020	NUM
bracis-28383	229	7	)	)	PUNCT
bracis-28383	229	8	article	article	NOUN
bracis-28383	229	9	  	  	SPACE
bracis-28383	229	10	google	google	PROPN
bracis-28383	229	11	scholar	scholar	NOUN
bracis-28383	229	12	  	  	SPACE
bracis-28383	229	13	kazemi	kazemi	PROPN
bracis-28383	229	14	,	,	PUNCT
bracis-28383	229	15	v.	v.	PROPN
bracis-28383	229	16	,	,	PUNCT
bracis-28383	229	17	sullivan	sullivan	PROPN
bracis-28383	229	18	,	,	PUNCT
bracis-28383	229	19	j.	j.	PROPN
bracis-28383	229	20	:	:	PUNCT
bracis-28383	229	21	one	one	NUM
bracis-28383	229	22	millisecond	millisecond	NOUN
bracis-28383	229	23	face	face	NOUN
bracis-28383	229	24	alignment	alignment	NOUN
bracis-28383	229	25	with	with	ADP
bracis-28383	229	26	an	an	DET
bracis-28383	229	27	ensemble	ensemble	NOUN
bracis-28383	229	28	of	of	ADP
bracis-28383	229	29	regression	regression	NOUN
bracis-28383	229	30	trees	tree	NOUN
bracis-28383	229	31	.	.	PUNCT
bracis-28383	230	1	in	in	ADP
bracis-28383	230	2	:	:	PUNCT
bracis-28383	230	3	proceedings	proceeding	NOUN
bracis-28383	230	4	of	of	ADP
bracis-28383	230	5	the	the	DET
bracis-28383	230	6	ieee	ieee	NOUN
bracis-28383	230	7	conference	conference	NOUN
bracis-28383	230	8	on	on	ADP
bracis-28383	230	9	computer	computer	NOUN
bracis-28383	230	10	vision	vision	NOUN
bracis-28383	230	11	and	and	CCONJ
bracis-28383	230	12	pattern	pattern	NOUN
bracis-28383	230	13	recognition	recognition	NOUN
bracis-28383	230	14	,	,	PUNCT
bracis-28383	230	15	pp	pp	ADV
bracis-28383	230	16	.	.	PUNCT
bracis-28383	231	1	1867–1874	1867–1874	NUM
bracis-28383	231	2	(	(	PUNCT
bracis-28383	231	3	2014	2014	NUM
bracis-28383	231	4	)	)	PUNCT
bracis-28383	231	5	google	google	PROPN
bracis-28383	231	6	scholar	scholar	NOUN
bracis-28383	231	7	  	  	SPACE
bracis-28383	231	8	khosla	khosla	PROPN
bracis-28383	231	9	,	,	PUNCT
bracis-28383	231	10	a.	a.	NOUN
bracis-28383	231	11	,	,	PUNCT
bracis-28383	231	12	jayadevaprakash	jayadevaprakash	ADV
bracis-28383	231	13	,	,	PUNCT
bracis-28383	231	14	n.	n.	NOUN
bracis-28383	231	15	,	,	PUNCT
bracis-28383	231	16	yao	yao	PROPN
bracis-28383	231	17	,	,	PUNCT
bracis-28383	231	18	b.	b.	PROPN
bracis-28383	231	19	,	,	PUNCT
bracis-28383	231	20	li	li	PROPN
bracis-28383	231	21	,	,	PUNCT
bracis-28383	231	22	f.f	f.f	PROPN
bracis-28383	231	23	.	.	PROPN
bracis-28383	231	24	:	:	PUNCT
bracis-28383	231	25	novel	novel	NOUN
bracis-28383	231	26	dataset	dataset	NOUN
bracis-28383	231	27	for	for	ADP
bracis-28383	231	28	fine	fine	ADV
bracis-28383	231	29	-	-	PUNCT
bracis-28383	231	30	grained	grain	VERB
bracis-28383	231	31	image	image	NOUN
bracis-28383	231	32	categorization	categorization	NOUN
bracis-28383	231	33	:	:	PUNCT
bracis-28383	231	34	stanford	stanford	PROPN
bracis-28383	231	35	dogs	dog	NOUN
bracis-28383	231	36	.	.	PUNCT
bracis-28383	232	1	in	in	ADP
bracis-28383	232	2	:	:	PUNCT
bracis-28383	232	3	proceedings	proceeding	NOUN
bracis-28383	232	4	of	of	ADP
bracis-28383	232	5	cvpr	cvpr	NOUN
bracis-28383	232	6	workshop	workshop	NOUN
bracis-28383	232	7	on	on	ADP
bracis-28383	232	8	fine	fine	ADV
bracis-28383	232	9	-	-	PUNCT
bracis-28383	232	10	grained	grain	VERB
bracis-28383	232	11	visual	visual	ADJ
bracis-28383	232	12	categorization	categorization	NOUN
bracis-28383	232	13	(	(	PUNCT
bracis-28383	232	14	fgvc	fgvc	NOUN
bracis-28383	232	15	)	)	PUNCT
bracis-28383	232	16	,	,	PUNCT
bracis-28383	232	17	vol	vol	NOUN
bracis-28383	232	18	.	.	PROPN
bracis-28383	232	19	2	2	NUM
bracis-28383	232	20	.	.	X
bracis-28383	232	21	citeseer	citeseer	NOUN
bracis-28383	232	22	(	(	PUNCT
bracis-28383	232	23	2011	2011	NUM
bracis-28383	232	24	)	)	PUNCT
bracis-28383	232	25	google	google	PROPN
bracis-28383	232	26	scholar	scholar	NOUN
bracis-28383	232	27	  	  	SPACE
bracis-28383	232	28	king	king	NOUN
bracis-28383	232	29	,	,	PUNCT
bracis-28383	232	30	d.e	d.e	PROPN
bracis-28383	232	31	.	.	PROPN
bracis-28383	232	32	:	:	PUNCT
bracis-28383	233	1	dlib	dlib	NOUN
bracis-28383	233	2	-	-	PUNCT
bracis-28383	233	3	ml	ml	NOUN
bracis-28383	233	4	:	:	PUNCT
bracis-28383	233	5	a	a	DET
bracis-28383	233	6	machine	machine	NOUN
bracis-28383	233	7	learning	learn	VERB
bracis-28383	233	8	toolkit	toolkit	NOUN
bracis-28383	233	9	.	.	PUNCT
bracis-28383	234	1	j.	j.	PROPN
bracis-28383	234	2	mach	mach	PROPN
bracis-28383	234	3	.	.	PUNCT
bracis-28383	235	1	learn	learn	VERB
bracis-28383	235	2	.	.	PUNCT
bracis-28383	236	1	res	re	NOUN
bracis-28383	236	2	.	.	PROPN
bracis-28383	237	1	10	10	NUM
bracis-28383	237	2	,	,	PUNCT
bracis-28383	237	3	1755–1758	1755–1758	NUM
bracis-28383	237	4	(	(	PUNCT
bracis-28383	237	5	2009	2009	NUM
bracis-28383	237	6	)	)	PUNCT
bracis-28383	237	7	google	google	PROPN
bracis-28383	237	8	scholar	scholar	NOUN
bracis-28383	237	9	  	  	SPACE
bracis-28383	237	10	king	king	NOUN
bracis-28383	237	11	,	,	PUNCT
bracis-28383	237	12	d.e	d.e	PROPN
bracis-28383	237	13	.	.	PROPN
bracis-28383	237	14	:	:	PUNCT
bracis-28383	238	1	max	max	ADJ
bracis-28383	238	2	-	-	PUNCT
bracis-28383	238	3	margin	margin	NOUN
bracis-28383	238	4	object	object	NOUN
bracis-28383	238	5	detection	detection	NOUN
bracis-28383	238	6	.	.	PUNCT
bracis-28383	239	1	corr	corr	PROPN
bracis-28383	239	2	abs/1502.00046	abs/1502.00046	PROPN
bracis-28383	239	3	(	(	PUNCT
bracis-28383	239	4	2015	2015	NUM
bracis-28383	239	5	)	)	PUNCT
bracis-28383	239	6	.	.	PUNCT
bracis-28383	240	1	http://arxiv.org/abs/1502.00046	http://arxiv.org/abs/1502.00046	PROPN
bracis-28383	240	2	lai	lai	PROPN
bracis-28383	240	3	,	,	PUNCT
bracis-28383	240	4	k.	k.	PROPN
bracis-28383	240	5	,	,	PUNCT
bracis-28383	240	6	tu	tu	PROPN
bracis-28383	240	7	,	,	PUNCT
bracis-28383	240	8	x.	x.	PROPN
bracis-28383	240	9	,	,	PUNCT
bracis-28383	240	10	yanushkevich	yanushkevich	PROPN
bracis-28383	240	11	,	,	PUNCT
bracis-28383	240	12	s.	s.	PROPN
bracis-28383	240	13	:	:	PUNCT
bracis-28383	240	14	dog	dog	NOUN
bracis-28383	240	15	identification	identification	NOUN
bracis-28383	240	16	using	use	VERB
bracis-28383	240	17	soft	soft	ADJ
bracis-28383	240	18	biometrics	biometric	NOUN
bracis-28383	240	19	and	and	CCONJ
bracis-28383	240	20	neural	neural	ADJ
bracis-28383	240	21	networks	network	NOUN
bracis-28383	240	22	.	.	PUNCT
bracis-28383	241	1	in	in	ADP
bracis-28383	241	2	:	:	PUNCT
bracis-28383	241	3	2019	2019	NUM
bracis-28383	241	4	international	international	ADJ
bracis-28383	241	5	joint	joint	ADJ
bracis-28383	241	6	conference	conference	NOUN
bracis-28383	241	7	on	on	ADP
bracis-28383	241	8	neural	neural	ADJ
bracis-28383	241	9	networks	network	NOUN
bracis-28383	241	10	(	(	PUNCT
bracis-28383	241	11	ijcnn	ijcnn	PROPN
bracis-28383	241	12	)	)	PUNCT
bracis-28383	241	13	,	,	PUNCT
bracis-28383	241	14	pp	pp	ADP
bracis-28383	241	15	.	.	PUNCT
bracis-28383	242	1	1–8	1–8	X
bracis-28383	242	2	.	.	PUNCT
bracis-28383	242	3	ieee	ieee	PROPN
bracis-28383	242	4	(	(	PUNCT
bracis-28383	242	5	2019	2019	NUM
bracis-28383	242	6	)	)	PUNCT
bracis-28383	242	7	google	google	PROPN
bracis-28383	242	8	scholar	scholar	NOUN
bracis-28383	242	9	  	  	SPACE
bracis-28383	242	10	li	li	PROPN
bracis-28383	242	11	,	,	PUNCT
bracis-28383	242	12	b.	b.	PROPN
bracis-28383	242	13	,	,	PUNCT
bracis-28383	242	14	wang	wang	PROPN
bracis-28383	242	15	,	,	PUNCT
bracis-28383	242	16	z.	z.	PROPN
bracis-28383	242	17	,	,	PUNCT
bracis-28383	242	18	wu	wu	PROPN
bracis-28383	242	19	,	,	PUNCT
bracis-28383	242	20	n.	n.	PROPN
bracis-28383	242	21	,	,	PUNCT
bracis-28383	242	22	shi	shi	PROPN
bracis-28383	242	23	,	,	PUNCT
bracis-28383	242	24	s.	s.	PROPN
bracis-28383	242	25	,	,	PUNCT
bracis-28383	242	26	ma	ma	PROPN
bracis-28383	242	27	,	,	PUNCT
bracis-28383	242	28	q.	q.	PROPN
bracis-28383	242	29	:	:	PUNCT
bracis-28383	242	30	dog	dog	PROPN
bracis-28383	242	31	nose	nose	NOUN
bracis-28383	242	32	print	print	NOUN
bracis-28383	242	33	matching	match	VERB
bracis-28383	242	34	with	with	ADP
bracis-28383	242	35	dual	dual	ADJ
bracis-28383	242	36	global	global	ADJ
bracis-28383	242	37	descriptor	descriptor	NOUN
bracis-28383	242	38	based	base	VERB
bracis-28383	242	39	on	on	ADP
bracis-28383	242	40	contrastive	contrastive	ADJ
bracis-28383	242	41	learning	learning	NOUN
bracis-28383	242	42	.	.	PUNCT
bracis-28383	243	1	arxiv	arxiv	PROPN
bracis-28383	243	2	preprint	preprint	VERB
bracis-28383	243	3	arxiv:2206.00580	arxiv:2206.00580	PROPN
bracis-28383	243	4	(	(	PUNCT
bracis-28383	243	5	2022	2022	NUM
bracis-28383	243	6	)	)	PUNCT
bracis-28383	244	1	liu	liu	PROPN
bracis-28383	244	2	,	,	PUNCT
bracis-28383	244	3	j.	j.	PROPN
bracis-28383	244	4	,	,	PUNCT
bracis-28383	244	5	kanazawa	kanazawa	PROPN
bracis-28383	244	6	,	,	PUNCT
bracis-28383	244	7	a.	a.	PROPN
bracis-28383	244	8	,	,	PUNCT
bracis-28383	244	9	jacobs	jacobs	PROPN
bracis-28383	244	10	,	,	PUNCT
bracis-28383	244	11	d.	d.	PROPN
bracis-28383	244	12	,	,	PUNCT
bracis-28383	244	13	belhumeur	belhumeur	PROPN
bracis-28383	244	14	,	,	PUNCT
bracis-28383	244	15	p.	p.	NOUN
bracis-28383	244	16	:	:	PUNCT
bracis-28383	244	17	dog	dog	NOUN
bracis-28383	244	18	breed	breed	NOUN
bracis-28383	244	19	classification	classification	NOUN
bracis-28383	244	20	using	use	VERB
bracis-28383	244	21	part	part	NOUN
bracis-28383	244	22	localization	localization	NOUN
bracis-28383	244	23	.	.	PUNCT
bracis-28383	245	1	in	in	ADP
bracis-28383	245	2	:	:	PUNCT
bracis-28383	245	3	fitzgibbon	fitzgibbon	NOUN
bracis-28383	245	4	,	,	PUNCT
bracis-28383	245	5	a.	a.	NOUN
bracis-28383	245	6	,	,	PUNCT
bracis-28383	245	7	lazebnik	lazebnik	PROPN
bracis-28383	245	8	,	,	PUNCT
bracis-28383	245	9	s.	s.	PROPN
bracis-28383	245	10	,	,	PUNCT
bracis-28383	245	11	perona	perona	PROPN
bracis-28383	245	12	,	,	PUNCT
bracis-28383	245	13	p.	p.	NOUN
bracis-28383	245	14	,	,	PUNCT
bracis-28383	245	15	sato	sato	PROPN
bracis-28383	245	16	,	,	PUNCT
bracis-28383	245	17	y.	y.	PROPN
bracis-28383	245	18	,	,	PUNCT
bracis-28383	245	19	schmid	schmid	PROPN
bracis-28383	245	20	,	,	PUNCT
bracis-28383	245	21	c.	c.	PROPN
bracis-28383	245	22	(	(	PUNCT
bracis-28383	245	23	eds	eds	PROPN
bracis-28383	245	24	.	.	PUNCT
bracis-28383	245	25	)	)	PUNCT
bracis-28383	245	26	eccv	eccv	ADV
bracis-28383	245	27	2012	2012	NUM
bracis-28383	245	28	.	.	PUNCT
bracis-28383	246	1	lncs	lncs	PROPN
bracis-28383	246	2	,	,	PUNCT
bracis-28383	246	3	vol	vol	NOUN
bracis-28383	246	4	.	.	PROPN
bracis-28383	246	5	7572	7572	NUM
bracis-28383	246	6	,	,	PUNCT
bracis-28383	246	7	pp	pp	ADJ
bracis-28383	246	8	.	.	PUNCT
bracis-28383	247	1	172–185	172–185	NUM
bracis-28383	247	2	.	.	PUNCT
bracis-28383	247	3	springer	springer	NOUN
bracis-28383	247	4	,	,	PUNCT
bracis-28383	247	5	heidelberg	heidelberg	PROPN
bracis-28383	247	6	(	(	PUNCT
bracis-28383	247	7	2012	2012	NUM
bracis-28383	247	8	)	)	PUNCT
bracis-28383	247	9	.	.	PUNCT
bracis-28383	248	1	https://doi.org/10.1007/978-3-642-33718-5_13	https://doi.org/10.1007/978-3-642-33718-5_13	NOUN
bracis-28383	248	2	chapter	chapter	NOUN
bracis-28383	248	3	  	  	SPACE
bracis-28383	248	4	google	google	PROPN
bracis-28383	248	5	scholar	scholar	NOUN
bracis-28383	248	6	  	  	SPACE
bracis-28383	248	7	moreira	moreira	PROPN
bracis-28383	248	8	,	,	PUNCT
bracis-28383	248	9	t.p	t.p	PROPN
bracis-28383	248	10	.	.	PROPN
bracis-28383	248	11	,	,	PUNCT
bracis-28383	248	12	perez	perez	PROPN
bracis-28383	248	13	,	,	PUNCT
bracis-28383	248	14	m.l	m.l	PROPN
bracis-28383	248	15	.	.	PROPN
bracis-28383	248	16	,	,	PUNCT
bracis-28383	248	17	de	de	PROPN
bracis-28383	248	18	oliveira	oliveira	PROPN
bracis-28383	248	19	werneck	werneck	PROPN
bracis-28383	248	20	,	,	PUNCT
bracis-28383	248	21	r.	r.	PROPN
bracis-28383	248	22	,	,	PUNCT
bracis-28383	248	23	valle	valle	PROPN
bracis-28383	248	24	,	,	PUNCT
bracis-28383	248	25	e.	e.	PROPN
bracis-28383	248	26	:	:	PUNCT
bracis-28383	248	27	where	where	SCONJ
bracis-28383	248	28	is	be	AUX
bracis-28383	248	29	my	my	PRON
bracis-28383	248	30	puppy	puppy	NOUN
bracis-28383	248	31	?	?	PUNCT
bracis-28383	249	1	retrieving	retrieve	VERB
bracis-28383	249	2	lost	lose	VERB
bracis-28383	249	3	dogs	dog	NOUN
bracis-28383	249	4	by	by	ADP
bracis-28383	249	5	facial	facial	ADJ
bracis-28383	249	6	features	feature	NOUN
bracis-28383	249	7	.	.	PUNCT
bracis-28383	250	1	multimedia	multimedia	NOUN
bracis-28383	250	2	tools	tool	NOUN
bracis-28383	250	3	appl	appl	NOUN
bracis-28383	250	4	.	.	PUNCT
bracis-28383	251	1	76(14	76(14	NOUN
bracis-28383	251	2	)	)	PUNCT
bracis-28383	251	3	,	,	PUNCT
bracis-28383	251	4	15325–15340	15325–15340	NUM
bracis-28383	251	5	(	(	PUNCT
bracis-28383	251	6	2017	2017	NUM
bracis-28383	251	7	)	)	PUNCT
bracis-28383	251	8	article	article	NOUN
bracis-28383	251	9	  	  	SPACE
bracis-28383	251	10	google	google	PROPN
bracis-28383	251	11	scholar	scholar	NOUN
bracis-28383	251	12	  	  	SPACE
bracis-28383	251	13	mougeot	mougeot	NOUN
bracis-28383	251	14	,	,	PUNCT
bracis-28383	251	15	g.	g.	PROPN
bracis-28383	251	16	,	,	PUNCT
bracis-28383	251	17	li	li	PROPN
bracis-28383	251	18	,	,	PUNCT
bracis-28383	251	19	d.	d.	PROPN
bracis-28383	251	20	,	,	PUNCT
bracis-28383	251	21	jia	jia	PROPN
bracis-28383	251	22	,	,	PUNCT
bracis-28383	251	23	s.	s.	PROPN
bracis-28383	251	24	:	:	PUNCT
bracis-28383	251	25	a	a	DET
bracis-28383	251	26	deep	deep	ADJ
bracis-28383	251	27	learning	learning	NOUN
bracis-28383	251	28	approach	approach	NOUN
bracis-28383	251	29	for	for	ADP
bracis-28383	251	30	dog	dog	NOUN
bracis-28383	251	31	face	face	NOUN
bracis-28383	251	32	verification	verification	NOUN
bracis-28383	251	33	and	and	CCONJ
bracis-28383	251	34	recognition	recognition	NOUN
bracis-28383	251	35	.	.	PUNCT
bracis-28383	252	1	in	in	ADP
bracis-28383	252	2	:	:	PUNCT
bracis-28383	252	3	nayak	nayak	PROPN
bracis-28383	252	4	,	,	PUNCT
bracis-28383	252	5	a.c	a.c	PROPN
bracis-28383	252	6	.	.	PROPN
bracis-28383	252	7	,	,	PUNCT
bracis-28383	252	8	sharma	sharma	PROPN
bracis-28383	252	9	,	,	PUNCT
bracis-28383	252	10	a.	a.	NOUN
bracis-28383	252	11	(	(	PUNCT
bracis-28383	252	12	eds	eds	PROPN
bracis-28383	252	13	.	.	PUNCT
bracis-28383	252	14	)	)	PUNCT
bracis-28383	252	15	pricai	pricai	PROPN
bracis-28383	252	16	2019	2019	NUM
bracis-28383	252	17	.	.	PUNCT
bracis-28383	253	1	lncs	lncs	PROPN
bracis-28383	253	2	(	(	PUNCT
bracis-28383	253	3	lnai	lnai	ADJ
bracis-28383	253	4	)	)	PUNCT
bracis-28383	253	5	,	,	PUNCT
bracis-28383	253	6	vol	vol	NOUN
bracis-28383	253	7	.	.	PROPN
bracis-28383	253	8	11672	11672	NUM
bracis-28383	253	9	,	,	PUNCT
bracis-28383	253	10	pp	pp	ADV
bracis-28383	253	11	.	.	PUNCT
bracis-28383	254	1	418–430	418–430	NUM
bracis-28383	254	2	.	.	PUNCT
bracis-28383	254	3	springer	springer	NOUN
bracis-28383	254	4	,	,	PUNCT
bracis-28383	254	5	cham	cham	PROPN
bracis-28383	254	6	(	(	PUNCT
bracis-28383	254	7	2019	2019	NUM
bracis-28383	254	8	)	)	PUNCT
bracis-28383	254	9	.	.	PUNCT
bracis-28383	255	1	https://doi.org/10.1007/978-3-030-29894-4_34	https://doi.org/10.1007/978-3-030-29894-4_34	PROPN
bracis-28383	255	2	chapter	chapter	NOUN
bracis-28383	255	3	  	  	SPACE
bracis-28383	255	4	google	google	PROPN
bracis-28383	255	5	scholar	scholar	NOUN
bracis-28383	255	6	  	  	SPACE
bracis-28383	255	7	ouyang	ouyang	PROPN
bracis-28383	255	8	,	,	PUNCT
bracis-28383	255	9	j.	j.	PROPN
bracis-28383	255	10	,	,	PUNCT
bracis-28383	255	11	he	he	PRON
bracis-28383	255	12	,	,	PUNCT
bracis-28383	255	13	h.	h.	PROPN
bracis-28383	255	14	,	,	PUNCT
bracis-28383	255	15	he	he	PRON
bracis-28383	255	16	,	,	PUNCT
bracis-28383	255	17	y.	y.	PROPN
bracis-28383	255	18	,	,	PUNCT
bracis-28383	255	19	tang	tang	PROPN
bracis-28383	255	20	,	,	PUNCT
bracis-28383	255	21	h.	h.	NOUN
bracis-28383	255	22	:	:	PUNCT
bracis-28383	255	23	dog	dog	NOUN
bracis-28383	255	24	recognition	recognition	NOUN
bracis-28383	255	25	in	in	ADP
bracis-28383	255	26	public	public	ADJ
bracis-28383	255	27	places	place	NOUN
bracis-28383	255	28	based	base	VERB
bracis-28383	255	29	on	on	ADP
bracis-28383	255	30	convolutional	convolutional	ADJ
bracis-28383	255	31	neural	neural	ADJ
bracis-28383	255	32	network	network	NOUN
bracis-28383	255	33	.	.	PUNCT
bracis-28383	256	1	int	int	NOUN
bracis-28383	256	2	.	.	PUNCT
bracis-28383	257	1	j.	j.	PROPN
bracis-28383	257	2	distrib	distrib	PROPN
bracis-28383	257	3	.	.	PUNCT
bracis-28383	257	4	sens	sens	PROPN
bracis-28383	257	5	.	.	PROPN
bracis-28383	257	6	netw	netw	PROPN
bracis-28383	257	7	.	.	PUNCT
bracis-28383	258	1	15(5	15(5	NUM
bracis-28383	258	2	)	)	PUNCT
bracis-28383	258	3	,	,	PUNCT
bracis-28383	258	4	1550147719829675	1550147719829675	NUM
bracis-28383	258	5	(	(	PUNCT
bracis-28383	258	6	2019	2019	NUM
bracis-28383	258	7	)	)	PUNCT
bracis-28383	258	8	article	article	NOUN
bracis-28383	258	9	  	  	SPACE
bracis-28383	258	10	google	google	PROPN
bracis-28383	258	11	scholar	scholar	NOUN
bracis-28383	258	12	  	  	SPACE
bracis-28383	258	13	simonyan	simonyan	ADJ
bracis-28383	258	14	,	,	PUNCT
bracis-28383	258	15	k.	k.	PROPN
bracis-28383	258	16	,	,	PUNCT
bracis-28383	258	17	zisserman	zisserman	PROPN
bracis-28383	258	18	,	,	PUNCT
bracis-28383	258	19	a.	a.	NOUN
bracis-28383	258	20	:	:	PUNCT
bracis-28383	258	21	very	very	ADV
bracis-28383	258	22	deep	deep	ADJ
bracis-28383	258	23	convolutional	convolutional	ADJ
bracis-28383	258	24	networks	network	NOUN
bracis-28383	258	25	for	for	ADP
bracis-28383	258	26	large	large	ADJ
bracis-28383	258	27	-	-	PUNCT
bracis-28383	258	28	scale	scale	NOUN
bracis-28383	258	29	image	image	NOUN
bracis-28383	258	30	recognition	recognition	NOUN
bracis-28383	258	31	.	.	PUNCT
bracis-28383	259	1	arxiv	arxiv	PROPN
bracis-28383	259	2	preprint	preprint	PROPN
bracis-28383	259	3	arxiv:1409.1556	arxiv:1409.1556	NOUN
bracis-28383	259	4	(	(	PUNCT
bracis-28383	259	5	2014	2014	NUM
bracis-28383	259	6	)	)	PUNCT
bracis-28383	259	7	sundaram	sundaram	PROPN
bracis-28383	259	8	,	,	PUNCT
bracis-28383	259	9	d.m	d.m	PROPN
bracis-28383	259	10	.	.	PROPN
bracis-28383	259	11	,	,	PUNCT
bracis-28383	259	12	loganathan	loganathan	PROPN
bracis-28383	259	13	,	,	PUNCT
bracis-28383	259	14	a.	a.	NOUN
bracis-28383	259	15	:	:	PUNCT
bracis-28383	259	16	a	a	DET
bracis-28383	259	17	new	new	ADJ
bracis-28383	259	18	supervised	supervised	ADJ
bracis-28383	259	19	clustering	clustering	NOUN
bracis-28383	259	20	framework	framework	NOUN
bracis-28383	259	21	using	use	VERB
bracis-28383	259	22	multi	multi	ADJ
bracis-28383	259	23	discriminative	discriminative	NOUN
bracis-28383	259	24	parts	part	NOUN
bracis-28383	259	25	and	and	CCONJ
bracis-28383	259	26	expectation	expectation	NOUN
bracis-28383	259	27	-	-	PUNCT
bracis-28383	259	28	maximization	maximization	NOUN
bracis-28383	259	29	approach	approach	NOUN
bracis-28383	259	30	for	for	ADP
bracis-28383	259	31	a	a	DET
bracis-28383	259	32	fine	fine	ADV
bracis-28383	259	33	-	-	PUNCT
bracis-28383	259	34	grained	grain	VERB
bracis-28383	259	35	animal	animal	NOUN
bracis-28383	259	36	breed	breed	NOUN
bracis-28383	259	37	classification	classification	NOUN
bracis-28383	259	38	(	(	PUNCT
bracis-28383	259	39	sc	sc	NOUN
bracis-28383	259	40	-	-	PUNCT
bracis-28383	259	41	mpem	mpem	NOUN
bracis-28383	259	42	)	)	PUNCT
bracis-28383	259	43	.	.	PUNCT
bracis-28383	260	1	neural	neural	ADJ
bracis-28383	260	2	process	process	NOUN
bracis-28383	260	3	.	.	PUNCT
bracis-28383	261	1	lett	lett	PROPN
bracis-28383	261	2	.	.	PUNCT
bracis-28383	262	1	52(1	52(1	NOUN
bracis-28383	262	2	)	)	PUNCT
bracis-28383	262	3	,	,	PUNCT
bracis-28383	262	4	727–766	727–766	NUM
bracis-28383	262	5	(	(	PUNCT
bracis-28383	262	6	2020	2020	NUM
bracis-28383	262	7	)	)	PUNCT
bracis-28383	262	8	article	article	NOUN
bracis-28383	262	9	  	  	SPACE
bracis-28383	262	10	google	google	PROPN
bracis-28383	262	11	scholar	scholar	NOUN
bracis-28383	262	12	  	  	SPACE
bracis-28383	262	13	swanson	swanson	PROPN
bracis-28383	262	14	,	,	PUNCT
bracis-28383	262	15	a.	a.	PROPN
bracis-28383	262	16	,	,	PUNCT
bracis-28383	262	17	kosmala	kosmala	PROPN
bracis-28383	262	18	,	,	PUNCT
bracis-28383	262	19	m.	m.	NOUN
bracis-28383	262	20	,	,	PUNCT
bracis-28383	262	21	lintott	lintott	PROPN
bracis-28383	262	22	,	,	PUNCT
bracis-28383	262	23	c.	c.	PROPN
bracis-28383	262	24	,	,	PUNCT
bracis-28383	262	25	simpson	simpson	PROPN
bracis-28383	262	26	,	,	PUNCT
bracis-28383	262	27	r.	r.	PROPN
bracis-28383	262	28	,	,	PUNCT
bracis-28383	262	29	smith	smith	PROPN
bracis-28383	262	30	,	,	PUNCT
bracis-28383	262	31	a.	a.	PROPN
bracis-28383	262	32	,	,	PUNCT
bracis-28383	262	33	packer	packer	NOUN
bracis-28383	262	34	,	,	PUNCT
bracis-28383	262	35	c.	c.	PROPN
bracis-28383	262	36	:	:	PUNCT
bracis-28383	262	37	data	datum	NOUN
bracis-28383	262	38	from	from	ADP
bracis-28383	262	39	:	:	PUNCT
bracis-28383	262	40	snapshot	snapshot	PROPN
bracis-28383	262	41	serengeti	serengeti	PROPN
bracis-28383	262	42	,	,	PUNCT
bracis-28383	262	43	high	high	ADJ
bracis-28383	262	44	-	-	PUNCT
bracis-28383	262	45	frequency	frequency	NOUN
bracis-28383	262	46	annotated	annotate	VERB
bracis-28383	262	47	camera	camera	NOUN
bracis-28383	262	48	trap	trap	NOUN
bracis-28383	262	49	images	image	NOUN
bracis-28383	262	50	of	of	ADP
bracis-28383	262	51	40	40	NUM
bracis-28383	262	52	mammalian	mammalian	ADJ
bracis-28383	262	53	species	specie	NOUN
bracis-28383	262	54	in	in	ADP
bracis-28383	262	55	an	an	DET
bracis-28383	262	56	african	african	ADJ
bracis-28383	262	57	savanna	savanna	NOUN
bracis-28383	262	58	(	(	PUNCT
bracis-28383	262	59	2015	2015	NUM
bracis-28383	262	60	)	)	PUNCT
bracis-28383	262	61	.	.	PUNCT
bracis-28383	263	1	https://doi.org/10.5061/dryad.5pt92	https://doi.org/10.5061/dryad.5pt92	PROPN
bracis-28383	263	2	tangsripairoj	tangsripairoj	PROPN
bracis-28383	263	3	,	,	PUNCT
bracis-28383	263	4	s.	s.	PROPN
bracis-28383	263	5	,	,	PUNCT
bracis-28383	263	6	kittirattanaviwat	kittirattanaviwat	VERB
bracis-28383	263	7	,	,	PUNCT
bracis-28383	263	8	p.	p.	PROPN
bracis-28383	263	9	,	,	PUNCT
bracis-28383	263	10	koophiran	koophiran	PROPN
bracis-28383	263	11	,	,	PUNCT
bracis-28383	263	12	k.	k.	PROPN
bracis-28383	263	13	,	,	PUNCT
bracis-28383	263	14	raksaithong	raksaithong	PROPN
bracis-28383	263	15	,	,	PUNCT
bracis-28383	263	16	l.	l.	PROPN
bracis-28383	263	17	:	:	PUNCT
bracis-28383	263	18	bokk	bokk	PROPN
bracis-28383	264	1	meow	meow	ADV
bracis-28383	264	2	:	:	PUNCT
bracis-28383	264	3	a	a	DET
bracis-28383	264	4	mobile	mobile	ADJ
bracis-28383	264	5	application	application	NOUN
bracis-28383	264	6	for	for	ADP
bracis-28383	264	7	finding	finding	NOUN
bracis-28383	264	8	and	and	CCONJ
bracis-28383	264	9	tracking	tracking	NOUN
bracis-28383	264	10	pets	pet	NOUN
bracis-28383	264	11	.	.	PUNCT
bracis-28383	265	1	in	in	ADP
bracis-28383	265	2	:	:	PUNCT
bracis-28383	265	3	2018	2018	NUM
bracis-28383	265	4	15th	15th	ADJ
bracis-28383	265	5	international	international	ADJ
bracis-28383	265	6	joint	joint	ADJ
bracis-28383	265	7	conference	conference	NOUN
bracis-28383	265	8	on	on	ADP
bracis-28383	265	9	computer	computer	NOUN
bracis-28383	265	10	science	science	NOUN
bracis-28383	265	11	and	and	CCONJ
bracis-28383	265	12	software	software	NOUN
bracis-28383	265	13	engineering	engineering	NOUN
bracis-28383	265	14	(	(	PUNCT
bracis-28383	265	15	jcsse	jcsse	NOUN
bracis-28383	265	16	)	)	PUNCT
bracis-28383	265	17	,	,	PUNCT
bracis-28383	265	18	pp	pp	ADJ
bracis-28383	265	19	.	.	PUNCT
bracis-28383	266	1	1–6	1–6	X
bracis-28383	266	2	.	.	X
bracis-28383	267	1	ieee	ieee	NOUN
bracis-28383	267	2	(	(	PUNCT
bracis-28383	267	3	2018	2018	NUM
bracis-28383	267	4	)	)	PUNCT
bracis-28383	267	5	google	google	PROPN
bracis-28383	267	6	scholar	scholar	NOUN
bracis-28383	267	7	  	  	SPACE
bracis-28383	267	8	turečková	turečková	NOUN
bracis-28383	267	9	,	,	PUNCT
bracis-28383	267	10	a.	a.	NOUN
bracis-28383	267	11	,	,	PUNCT
bracis-28383	267	12	holík	holík	ADV
bracis-28383	267	13	,	,	PUNCT
bracis-28383	267	14	t.	t.	PROPN
bracis-28383	267	15	,	,	PUNCT
bracis-28383	267	16	komínková	komínková	NOUN
bracis-28383	267	17	oplatková	oplatková	PROPN
bracis-28383	267	18	,	,	PUNCT
bracis-28383	267	19	z.	z.	PROPN
bracis-28383	267	20	:	:	PUNCT
bracis-28383	267	21	dog	dog	NOUN
bracis-28383	267	22	face	face	NOUN
bracis-28383	267	23	detection	detection	NOUN
bracis-28383	267	24	using	use	VERB
bracis-28383	267	25	yolo	yolo	ADJ
bracis-28383	267	26	network	network	NOUN
bracis-28383	267	27	.	.	PUNCT
bracis-28383	268	1	in	in	ADP
bracis-28383	268	2	:	:	PUNCT
bracis-28383	268	3	mendel	mendel	PROPN
bracis-28383	268	4	.	.	PROPN
bracis-28383	268	5	brno	brno	PROPN
bracis-28383	268	6	university	university	PROPN
bracis-28383	268	7	of	of	ADP
bracis-28383	268	8	technology	technology	NOUN
bracis-28383	268	9	(	(	PUNCT
bracis-28383	268	10	2020	2020	NUM
bracis-28383	268	11	)	)	PUNCT
bracis-28383	268	12	google	google	PROPN
bracis-28383	268	13	scholar	scholar	NOUN
bracis-28383	268	14	  	  	SPACE
bracis-28383	268	15	weerasekara	weerasekara	NOUN
bracis-28383	268	16	,	,	PUNCT
bracis-28383	268	17	d.	d.	PROPN
bracis-28383	268	18	,	,	PUNCT
bracis-28383	268	19	gamage	gamage	NOUN
bracis-28383	268	20	,	,	PUNCT
bracis-28383	268	21	m.	m.	NOUN
bracis-28383	268	22	,	,	PUNCT
bracis-28383	268	23	kulasooriya	kulasooriya	ADV
bracis-28383	268	24	,	,	PUNCT
bracis-28383	268	25	k.	k.	PROPN
bracis-28383	268	26	:	:	PUNCT
bracis-28383	268	27	combined	combined	ADJ
bracis-28383	268	28	approach	approach	NOUN
bracis-28383	268	29	of	of	ADP
bracis-28383	268	30	supervised	supervised	ADJ
bracis-28383	268	31	and	and	CCONJ
bracis-28383	268	32	unsupervised	unsupervised	ADJ
bracis-28383	268	33	learning	learning	NOUN
bracis-28383	268	34	for	for	ADP
bracis-28383	268	35	dog	dog	NOUN
bracis-28383	268	36	face	face	NOUN
bracis-28383	268	37	recognition	recognition	NOUN
bracis-28383	268	38	.	.	PUNCT
bracis-28383	269	1	in	in	ADP
bracis-28383	269	2	:	:	PUNCT
bracis-28383	269	3	2021	2021	NUM
bracis-28383	269	4	6th	6th	ADJ
bracis-28383	269	5	international	international	ADJ
bracis-28383	269	6	conference	conference	NOUN
bracis-28383	269	7	for	for	ADP
bracis-28383	269	8	convergence	convergence	NOUN
bracis-28383	269	9	in	in	ADP
bracis-28383	269	10	technology	technology	NOUN
bracis-28383	269	11	(	(	PUNCT
bracis-28383	269	12	i2ct	i2ct	NOUN
bracis-28383	269	13	)	)	PUNCT
bracis-28383	269	14	,	,	PUNCT
bracis-28383	269	15	pp	pp	ADP
bracis-28383	269	16	.	.	PUNCT
bracis-28383	270	1	1–5	1–5	X
bracis-28383	270	2	.	.	X
bracis-28383	270	3	ieee	ieee	NOUN
bracis-28383	270	4	(	(	PUNCT
bracis-28383	270	5	2021	2021	NUM
bracis-28383	270	6	)	)	PUNCT
bracis-28383	270	7	google	google	PROPN
bracis-28383	270	8	scholar	scholar	NOUN
bracis-28383	270	9	  	  	SPACE
bracis-28383	270	10	yoon	yoon	PROPN
bracis-28383	270	11	,	,	PUNCT
bracis-28383	270	12	b.	b.	PROPN
bracis-28383	270	13	,	,	PUNCT
bracis-28383	270	14	so	so	ADV
bracis-28383	270	15	,	,	PUNCT
bracis-28383	270	16	h.	h.	PROPN
bracis-28383	270	17	,	,	PUNCT
bracis-28383	270	18	rhee	rhee	PROPN
bracis-28383	270	19	,	,	PUNCT
bracis-28383	270	20	j.	j.	PROPN
bracis-28383	270	21	:	:	PUNCT
bracis-28383	270	22	a	a	DET
bracis-28383	270	23	methodology	methodology	NOUN
bracis-28383	270	24	for	for	ADP
bracis-28383	270	25	utilizing	utilize	VERB
bracis-28383	270	26	vector	vector	NOUN
bracis-28383	270	27	space	space	NOUN
bracis-28383	270	28	to	to	PART
bracis-28383	270	29	improve	improve	VERB
bracis-28383	270	30	the	the	DET
bracis-28383	270	31	performance	performance	NOUN
bracis-28383	270	32	of	of	ADP
bracis-28383	270	33	a	a	DET
bracis-28383	270	34	dog	dog	NOUN
bracis-28383	270	35	face	face	NOUN
bracis-28383	270	36	identification	identification	NOUN
bracis-28383	270	37	model	model	NOUN
bracis-28383	270	38	.	.	PUNCT
bracis-28383	271	1	appl	appl	PROPN
bracis-28383	271	2	.	.	PUNCT
bracis-28383	272	1	sci	sci	PROPN
bracis-28383	272	2	.	.	PROPN
bracis-28383	272	3	11(5	11(5	NUM
bracis-28383	272	4	)	)	PUNCT
bracis-28383	272	5	,	,	PUNCT
bracis-28383	272	6	2074	2074	NUM
bracis-28383	272	7	(	(	PUNCT
bracis-28383	272	8	2021	2021	NUM
bracis-28383	272	9	)	)	PUNCT
bracis-28383	272	10	.	.	PUNCT
bracis-28383	273	1	https://doi.org/10.3390/app11052074	https://doi.org/10.3390/app11052074	PROPN
bracis-28383	273	2	article	article	PROPN
bracis-28383	273	3	  	  	SPACE
bracis-28383	273	4	google	google	PROPN
bracis-28383	273	5	scholar	scholar	NOUN
bracis-28383	273	6	  	  	SPACE
bracis-28383	273	7	download	download	NOUN
bracis-28383	273	8	references	reference	NOUN
bracis-28383	273	9	acknowledgements	acknowledgement	NOUN
bracis-28383	273	10	the	the	DET
bracis-28383	273	11	authors	author	NOUN
bracis-28383	273	12	would	would	AUX
bracis-28383	273	13	like	like	VERB
bracis-28383	273	14	to	to	PART
bracis-28383	273	15	thank	thank	VERB
bracis-28383	273	16	the	the	DET
bracis-28383	273	17	ceará	ceará	NOUN
bracis-28383	273	18	state	state	PROPN
bracis-28383	273	19	foundation	foundation	PROPN
bracis-28383	273	20	for	for	ADP
bracis-28383	273	21	the	the	DET
bracis-28383	273	22	support	support	NOUN
bracis-28383	273	23	of	of	ADP
bracis-28383	273	24	scientific	scientific	ADJ
bracis-28383	273	25	and	and	CCONJ
bracis-28383	273	26	technological	technological	ADJ
bracis-28383	273	27	development	development	NOUN
bracis-28383	273	28	(	(	PUNCT
bracis-28383	273	29	funcap	funcap	NOUN
bracis-28383	273	30	)	)	PUNCT
bracis-28383	273	31	for	for	ADP
bracis-28383	273	32	the	the	DET
bracis-28383	273	33	financial	financial	ADJ
bracis-28383	273	34	support	support	NOUN
bracis-28383	273	35	(	(	PUNCT
bracis-28383	273	36	6945087/2019	6945087/2019	NUM
bracis-28383	273	37	)	)	PUNCT
bracis-28383	273	38	.	.	PUNCT
bracis-28383	274	1	author	author	NOUN
bracis-28383	274	2	information	information	NOUN
bracis-28383	274	3	authors	author	NOUN
bracis-28383	274	4	and	and	CCONJ
bracis-28383	274	5	affiliations	affiliation	NOUN
bracis-28383	274	6	mdcc	mdcc	NOUN
bracis-28383	274	7	,	,	PUNCT
bracis-28383	274	8	universidade	universidade	PROPN
bracis-28383	274	9	federal	federal	PROPN
bracis-28383	274	10	do	do	AUX
bracis-28383	274	11	ceará	ceará	NOUN
bracis-28383	274	12	,	,	PUNCT
bracis-28383	274	13	campus	campus	NOUN
bracis-28383	274	14	do	do	VERB
bracis-28383	274	15	pici	pici	PROPN
bracis-28383	274	16	bloco	bloco	PROPN
bracis-28383	274	17	910	910	PROPN
bracis-28383	274	18	,	,	PUNCT
bracis-28383	274	19	fortaleza	fortaleza	PROPN
bracis-28383	274	20	,	,	PUNCT
bracis-28383	274	21	ce	ce	PROPN
bracis-28383	274	22	,	,	PUNCT
bracis-28383	274	23	brazil	brazil	PROPN
bracis-28383	274	24	joão	joão	PROPN
bracis-28383	274	25	p.	p.	PROPN
bracis-28383	274	26	b.	b.	PROPN
bracis-28383	274	27	andrade	andrade	PROPN
bracis-28383	274	28	,	,	PUNCT
bracis-28383	274	29	 	 	SPACE
bracis-28383	274	30	leonardo	leonardo	PROPN
bracis-28383	274	31	f.	f.	PROPN
bracis-28383	274	32	costa	costa	PROPN
bracis-28383	274	33	,	,	PUNCT
bracis-28383	274	34	 	 	SPACE
bracis-28383	274	35	lucas	lucas	PROPN
bracis-28383	274	36	s.	s.	PROPN
bracis-28383	274	37	fernandes	fernandes	PROPN
bracis-28383	274	38	 	 	SPACE
bracis-28383	274	39	&	&	CCONJ
bracis-28383	274	40	 	 	SPACE
bracis-28383	274	41	paulo	paulo	PROPN
bracis-28383	274	42	a.	a.	PROPN
bracis-28383	274	43	l.	l.	PROPN
bracis-28383	274	44	rego	rego	PROPN
bracis-28383	274	45	instituto	instituto	PROPN
bracis-28383	274	46	ufc	ufc	PROPN
bracis-28383	274	47	virtual	virtual	PROPN
bracis-28383	274	48	universidade	universidade	PROPN
bracis-28383	274	49	federal	federal	PROPN
bracis-28383	274	50	do	do	AUX
bracis-28383	274	51	ceará	ceará	NOUN
bracis-28383	274	52	,	,	PUNCT
bracis-28383	274	53	campus	campus	NOUN
bracis-28383	274	54	do	do	AUX
bracis-28383	274	55	pici	pici	PROPN
bracis-28383	274	56	bloco	bloco	PROPN
bracis-28383	274	57	901	901	NUM
bracis-28383	274	58	,	,	PUNCT
bracis-28383	274	59	fortaleza	fortaleza	PROPN
bracis-28383	274	60	,	,	PUNCT
bracis-28383	274	61	ce	ce	PROPN
bracis-28383	274	62	,	,	PUNCT
bracis-28383	274	63	brazil	brazil	PROPN
bracis-28383	274	64	josé	josé	PROPN
bracis-28383	274	65	g.	g.	PROPN
bracis-28383	274	66	r.	r.	PROPN
bracis-28383	274	67	maia	maia	PROPN
bracis-28383	274	68	authors	author	NOUN
bracis-28383	274	69	joão	joão	PROPN
bracis-28383	274	70	p.	p.	PROPN
bracis-28383	274	71	b.	b.	PROPN
bracis-28383	275	1	andradeview	andradeview	ADJ
bracis-28383	275	2	author	author	NOUN
bracis-28383	275	3	publications	publication	NOUN
bracis-28383	275	4	search	search	NOUN
bracis-28383	275	5	author	author	NOUN
bracis-28383	275	6	on	on	ADP
bracis-28383	275	7	:	:	PUNCT
bracis-28383	275	8	pubmed	pubmed	PROPN
bracis-28383	275	9	 	 	SPACE
bracis-28383	275	10	google	google	PROPN
bracis-28383	275	11	scholar	scholar	PROPN
bracis-28383	275	12	leonardo	leonardo	PROPN
bracis-28383	275	13	f.	f.	PROPN
bracis-28383	275	14	costaview	costaview	PROPN
bracis-28383	275	15	author	author	NOUN
bracis-28383	275	16	publications	publication	NOUN
bracis-28383	275	17	search	search	NOUN
bracis-28383	275	18	author	author	NOUN
bracis-28383	275	19	on	on	ADP
bracis-28383	275	20	:	:	PUNCT
bracis-28383	275	21	pubmed	pubmed	PROPN
bracis-28383	275	22	 	 	SPACE
bracis-28383	275	23	google	google	PROPN
bracis-28383	275	24	scholar	scholar	PROPN
bracis-28383	275	25	lucas	lucas	PROPN
bracis-28383	275	26	s.	s.	PROPN
bracis-28383	275	27	fernandesview	fernandesview	PROPN
bracis-28383	275	28	author	author	NOUN
bracis-28383	275	29	publications	publication	NOUN
bracis-28383	275	30	search	search	NOUN
bracis-28383	275	31	author	author	NOUN
bracis-28383	275	32	on	on	ADP
bracis-28383	275	33	:	:	PUNCT
bracis-28383	275	34	pubmed	pubmed	PROPN
bracis-28383	275	35	 	 	SPACE
bracis-28383	275	36	google	google	PROPN
bracis-28383	275	37	scholar	scholar	PROPN
bracis-28383	275	38	paulo	paulo	PROPN
bracis-28383	275	39	a.	a.	PROPN
bracis-28383	275	40	l.	l.	PROPN
bracis-28383	275	41	regoview	regoview	PROPN
bracis-28383	275	42	author	author	NOUN
bracis-28383	275	43	publications	publication	VERB
bracis-28383	275	44	search	search	NOUN
bracis-28383	275	45	author	author	NOUN
bracis-28383	275	46	on	on	ADP
bracis-28383	275	47	:	:	PUNCT
bracis-28383	275	48	pubmed	pubmed	PROPN
bracis-28383	275	49	 	 	SPACE
bracis-28383	275	50	google	google	PROPN
bracis-28383	275	51	scholar	scholar	PROPN
bracis-28383	275	52	josé	josé	PROPN
bracis-28383	275	53	g.	g.	PROPN
bracis-28383	275	54	r.	r.	PROPN
bracis-28383	275	55	maiaview	maiaview	PROPN
bracis-28383	275	56	author	author	NOUN
bracis-28383	275	57	publications	publication	NOUN
bracis-28383	275	58	search	search	NOUN
bracis-28383	275	59	author	author	NOUN
bracis-28383	275	60	on	on	ADP
bracis-28383	275	61	:	:	PUNCT
bracis-28383	275	62	pubmed	pubmed	PROPN
bracis-28383	275	63	 	 	SPACE
bracis-28383	275	64	google	google	PROPN
bracis-28383	275	65	scholar	scholar	NOUN
bracis-28383	275	66	corresponding	correspond	VERB
bracis-28383	275	67	author	author	NOUN
bracis-28383	275	68	correspondence	correspondence	NOUN
bracis-28383	275	69	to	to	ADP
bracis-28383	275	70	joão	joão	PROPN
bracis-28383	275	71	p.	p.	PROPN
bracis-28383	275	72	b.	b.	PROPN
bracis-28383	275	73	andrade	andrade	PROPN
bracis-28383	275	74	.	.	PUNCT
bracis-28383	276	1	editor	editor	NOUN
bracis-28383	276	2	information	information	NOUN
bracis-28383	276	3	editors	editor	NOUN
bracis-28383	276	4	and	and	CCONJ
bracis-28383	276	5	affiliations	affiliation	NOUN
bracis-28383	276	6	federal	federal	PROPN
bracis-28383	276	7	university	university	PROPN
bracis-28383	276	8	of	of	ADP
bracis-28383	276	9	são	são	PROPN
bracis-28383	276	10	carlos	carlos	PROPN
bracis-28383	276	11	,	,	PUNCT
bracis-28383	276	12	são	são	PROPN
bracis-28383	276	13	carlos	carlos	PROPN
bracis-28383	276	14	,	,	PUNCT
bracis-28383	276	15	brazil	brazil	PROPN
bracis-28383	276	16	murilo	murilo	PROPN
bracis-28383	276	17	c.	c.	PROPN
bracis-28383	276	18	naldi	naldi	PROPN
bracis-28383	276	19	centro	centro	PROPN
bracis-28383	276	20	universitario	universitario	PROPN
bracis-28383	276	21	da	da	PROPN
bracis-28383	276	22	fei	fei	PROPN
bracis-28383	276	23	,	,	PUNCT
bracis-28383	276	24	são	são	PROPN
bracis-28383	276	25	bernardo	bernardo	PROPN
bracis-28383	276	26	do	do	AUX
bracis-28383	276	27	campo	campo	PROPN
bracis-28383	276	28	,	,	PUNCT
bracis-28383	276	29	brazil	brazil	PROPN
bracis-28383	276	30	reinaldo	reinaldo	PROPN
bracis-28383	276	31	a.	a.	PROPN
bracis-28383	276	32	c.	c.	PROPN
bracis-28383	276	33	bianchi	bianchi	PROPN
bracis-28383	276	34	rights	right	NOUN
bracis-28383	276	35	and	and	CCONJ
bracis-28383	276	36	permissions	permission	NOUN
bracis-28383	276	37	reprints	reprint	NOUN
bracis-28383	276	38	and	and	CCONJ
bracis-28383	276	39	permissions	permission	VERB
bracis-28383	276	40	copyright	copyright	NOUN
bracis-28383	276	41	information	information	NOUN
bracis-28383	276	42	©	©	ADP
bracis-28383	276	43	2023	2023	NUM
bracis-28383	276	44	the	the	DET
bracis-28383	276	45	author(s	author(s	NOUN
bracis-28383	276	46	)	)	PUNCT
bracis-28383	276	47	,	,	PUNCT
bracis-28383	276	48	under	under	ADP
bracis-28383	276	49	exclusive	exclusive	ADJ
bracis-28383	276	50	license	license	NOUN
bracis-28383	276	51	to	to	ADP
bracis-28383	276	52	springer	springer	NOUN
bracis-28383	276	53	nature	nature	PROPN
bracis-28383	276	54	switzerland	switzerland	PROPN
bracis-28383	276	55	ag	ag	PROPN
bracis-28383	276	56	about	about	ADP
bracis-28383	276	57	this	this	DET
bracis-28383	276	58	paper	paper	NOUN
bracis-28383	276	59	cite	cite	VERB
bracis-28383	276	60	this	this	DET
bracis-28383	276	61	paper	paper	NOUN
bracis-28383	276	62	andrade	andrade	PROPN
bracis-28383	276	63	,	,	PUNCT
bracis-28383	276	64	j.p.b	j.p.b	PROPN
bracis-28383	276	65	.	.	PROPN
bracis-28383	276	66	,	,	PUNCT
bracis-28383	276	67	costa	costa	PROPN
bracis-28383	276	68	,	,	PUNCT
bracis-28383	276	69	l.f	l.f	PROPN
bracis-28383	276	70	.	.	PROPN
bracis-28383	276	71	,	,	PUNCT
bracis-28383	276	72	fernandes	fernandes	PROPN
bracis-28383	276	73	,	,	PUNCT
bracis-28383	276	74	l.s	l.s	PROPN
bracis-28383	276	75	.	.	PROPN
bracis-28383	276	76	,	,	PUNCT
bracis-28383	276	77	rego	rego	NOUN
bracis-28383	276	78	,	,	PUNCT
bracis-28383	276	79	p.a.l	p.a.l	ADP
bracis-28383	276	80	.	.	PROPN
bracis-28383	276	81	,	,	PUNCT
bracis-28383	276	82	maia	maia	PROPN
bracis-28383	276	83	,	,	PUNCT
bracis-28383	276	84	j.g.r	j.g.r	PROPN
bracis-28383	276	85	.	.	PUNCT
bracis-28383	277	1	(	(	PUNCT
bracis-28383	277	2	2023	2023	NUM
bracis-28383	277	3	)	)	PUNCT
bracis-28383	277	4	.	.	PUNCT
bracis-28383	278	1	dog	dog	NOUN
bracis-28383	278	2	face	face	NOUN
bracis-28383	278	3	recognition	recognition	NOUN
bracis-28383	278	4	using	use	VERB
bracis-28383	278	5	deep	deep	ADJ
bracis-28383	278	6	features	feature	NOUN
bracis-28383	278	7	embeddings	embedding	NOUN
bracis-28383	278	8	.	.	PUNCT
bracis-28383	279	1	in	in	ADP
bracis-28383	279	2	:	:	PUNCT
bracis-28383	279	3	naldi	naldi	PROPN
bracis-28383	279	4	,	,	PUNCT
bracis-28383	279	5	m.c	m.c	PROPN
bracis-28383	279	6	.	.	PROPN
bracis-28383	279	7	,	,	PUNCT
bracis-28383	279	8	bianchi	bianchi	PROPN
bracis-28383	279	9	,	,	PUNCT
bracis-28383	279	10	r.a.c	r.a.c	ADP
bracis-28383	279	11	.	.	PUNCT
bracis-28383	279	12	(	(	PUNCT
bracis-28383	279	13	eds	ed	NOUN
bracis-28383	279	14	)	)	PUNCT
bracis-28383	279	15	intelligent	intelligent	ADJ
bracis-28383	279	16	systems	system	NOUN
bracis-28383	279	17	.	.	PUNCT
bracis-28383	280	1	bracis	bracis	PROPN
bracis-28383	280	2	2023	2023	NUM
bracis-28383	280	3	.	.	PUNCT
bracis-28383	281	1	lecture	lecture	NOUN
bracis-28383	281	2	notes	note	NOUN
bracis-28383	281	3	in	in	ADP
bracis-28383	281	4	computer	computer	NOUN
bracis-28383	281	5	science	science	NOUN
bracis-28383	281	6	(	(	PUNCT
bracis-28383	281	7	)	)	PUNCT
bracis-28383	281	8	,	,	PUNCT
bracis-28383	281	9	vol	vol	NOUN
bracis-28383	281	10	14196	14196	NUM
bracis-28383	281	11	.	.	PUNCT
bracis-28383	282	1	springer	springer	NOUN
bracis-28383	282	2	,	,	PUNCT
bracis-28383	282	3	cham	cham	PROPN
bracis-28383	282	4	.	.	PUNCT
bracis-28383	283	1	https://doi.org/10.1007/978-3-031-45389-2_9	https://doi.org/10.1007/978-3-031-45389-2_9	ADJ
bracis-28383	283	2	download	download	NOUN
bracis-28383	283	3	citation	citation	NOUN
bracis-28383	283	4	.ris	.ris	PUNCT
bracis-28383	283	5	.enw	.enw	PROPN
bracis-28383	283	6	.bib	.bib	PUNCT
bracis-28383	284	1	doi	doi	PROPN
bracis-28383	284	2	:	:	PUNCT
bracis-28383	284	3	https://doi.org/10.1007/978-3-031-45389-2_9	https://doi.org/10.1007/978-3-031-45389-2_9	NUM
bracis-28383	284	4	published	publish	VERB
bracis-28383	284	5	:	:	PUNCT
bracis-28383	284	6	12	12	NUM
bracis-28383	284	7	october	october	PROPN
bracis-28383	284	8	2023	2023	NUM
bracis-28383	284	9	publisher	publisher	NOUN
bracis-28383	284	10	name	name	NOUN
bracis-28383	284	11	:	:	PUNCT
bracis-28383	284	12	springer	springer	NOUN
bracis-28383	284	13	,	,	PUNCT
bracis-28383	284	14	cham	cham	PROPN
bracis-28383	284	15	print	print	PROPN
bracis-28383	284	16	isbn	isbn	PROPN
bracis-28383	284	17	:	:	PUNCT
bracis-28383	284	18	978	978	NUM
bracis-28383	284	19	-	-	SYM
bracis-28383	284	20	3	3	NUM
bracis-28383	284	21	-	-	PUNCT
bracis-28383	284	22	031	031	NUM
bracis-28383	284	23	-	-	PUNCT
bracis-28383	284	24	45388	45388	NUM
bracis-28383	284	25	-	-	SYM
bracis-28383	284	26	5	5	NUM
bracis-28383	284	27	online	online	ADJ
bracis-28383	284	28	isbn	isbn	NOUN
bracis-28383	284	29	:	:	PUNCT
bracis-28383	284	30	978	978	NUM
bracis-28383	284	31	-	-	SYM
bracis-28383	284	32	3	3	NUM
bracis-28383	284	33	-	-	PUNCT
bracis-28383	284	34	031	031	NUM
bracis-28383	284	35	-	-	PUNCT
bracis-28383	284	36	45389	45389	NUM
bracis-28383	284	37	-	-	SYM
bracis-28383	284	38	2	2	NUM
bracis-28383	284	39	ebook	ebook	NOUN
bracis-28383	284	40	packages	package	NOUN
bracis-28383	284	41	:	:	PUNCT
bracis-28383	284	42	computer	computer	NOUN
bracis-28383	284	43	sciencecomputer	sciencecomputer	NOUN
bracis-28383	284	44	science	science	NOUN
bracis-28383	284	45	(	(	PUNCT
bracis-28383	284	46	r0	r0	NOUN
bracis-28383	284	47	)	)	PUNCT
bracis-28383	284	48	share	share	VERB
bracis-28383	284	49	this	this	DET
bracis-28383	284	50	paper	paper	NOUN
bracis-28383	284	51	anyone	anyone	PRON
bracis-28383	284	52	you	you	PRON
bracis-28383	284	53	share	share	VERB
bracis-28383	284	54	the	the	DET
bracis-28383	284	55	following	follow	VERB
bracis-28383	284	56	link	link	NOUN
bracis-28383	284	57	with	with	ADP
bracis-28383	284	58	will	will	AUX
bracis-28383	284	59	be	be	AUX
bracis-28383	284	60	able	able	ADJ
bracis-28383	284	61	to	to	PART
bracis-28383	284	62	read	read	VERB
bracis-28383	284	63	this	this	DET
bracis-28383	284	64	content	content	NOUN
bracis-28383	284	65	:	:	PUNCT
bracis-28383	284	66	get	get	VERB
bracis-28383	284	67	shareable	shareable	ADJ
bracis-28383	284	68	linksorry	linksorry	NOUN
bracis-28383	284	69	,	,	PUNCT
bracis-28383	284	70	a	a	DET
bracis-28383	284	71	shareable	shareable	ADJ
bracis-28383	284	72	link	link	NOUN
bracis-28383	284	73	is	be	AUX
bracis-28383	284	74	not	not	PART
bracis-28383	284	75	currently	currently	ADV
bracis-28383	284	76	available	available	ADJ
bracis-28383	284	77	for	for	ADP
bracis-28383	284	78	this	this	DET
bracis-28383	284	79	article	article	NOUN
bracis-28383	284	80	.	.	PUNCT
bracis-28383	285	1	copy	copy	VERB
bracis-28383	285	2	shareable	shareable	ADJ
bracis-28383	285	3	link	link	NOUN
bracis-28383	285	4	to	to	PART
bracis-28383	285	5	clipboard	clipboard	NOUN
bracis-28383	285	6	provided	provide	VERB
bracis-28383	285	7	by	by	ADP
bracis-28383	285	8	the	the	DET
bracis-28383	285	9	springer	springer	NOUN
bracis-28383	285	10	nature	nature	PROPN
bracis-28383	285	11	sharedit	sharedit	PROPN
bracis-28383	285	12	content	content	NOUN
bracis-28383	285	13	-	-	PUNCT
bracis-28383	285	14	sharing	share	VERB
bracis-28383	285	15	initiative	initiative	NOUN
bracis-28383	285	16	keywords	keyword	VERB
bracis-28383	285	17	residual	residual	ADJ
bracis-28383	285	18	neural	neural	ADJ
bracis-28383	285	19	networks	network	NOUN
bracis-28383	285	20	deep	deep	ADV
bracis-28383	285	21	features	feature	VERB
bracis-28383	285	22	facial	facial	ADJ
bracis-28383	285	23	identification	identification	NOUN
bracis-28383	285	24	dogs	dog	NOUN
bracis-28383	285	25	publish	publish	VERB
bracis-28383	285	26	with	with	ADP
bracis-28383	285	27	us	us	PROPN
bracis-28383	285	28	policies	policy	NOUN
bracis-28383	285	29	and	and	CCONJ
bracis-28383	285	30	ethics	ethic	NOUN
bracis-28383	285	31	search	search	NOUN
bracis-28383	285	32	search	search	NOUN
bracis-28383	285	33	by	by	ADP
bracis-28383	285	34	keyword	keyword	NOUN
bracis-28383	285	35	or	or	CCONJ
bracis-28383	285	36	author	author	NOUN
bracis-28383	285	37	search	search	NOUN
bracis-28383	285	38	navigation	navigation	NOUN
bracis-28383	285	39	find	find	VERB
bracis-28383	285	40	a	a	DET
bracis-28383	285	41	journal	journal	NOUN
bracis-28383	285	42	publish	publish	VERB
bracis-28383	285	43	with	with	ADP
bracis-28383	285	44	us	we	PRON
bracis-28383	285	45	track	track	VERB
bracis-28383	285	46	your	your	PRON
bracis-28383	285	47	research	research	NOUN
bracis-28383	285	48	discover	discover	VERB
bracis-28383	285	49	content	content	NOUN
bracis-28383	285	50	journals	journal	NOUN
bracis-28383	285	51	a	a	DET
bracis-28383	285	52	-	-	PUNCT
bracis-28383	285	53	z	z	NOUN
bracis-28383	285	54	books	book	NOUN
bracis-28383	285	55	a	a	DET
bracis-28383	285	56	-	-	PUNCT
bracis-28383	285	57	z	z	NOUN
bracis-28383	285	58	publish	publish	NOUN
bracis-28383	285	59	with	with	ADP
bracis-28383	285	60	us	us	PROPN
bracis-28383	285	61	journal	journal	PROPN
bracis-28383	285	62	finder	finder	PROPN
bracis-28383	285	63	publish	publish	VERB
bracis-28383	285	64	your	your	PRON
bracis-28383	285	65	research	research	NOUN
bracis-28383	285	66	language	language	NOUN
bracis-28383	285	67	editing	edit	VERB
bracis-28383	285	68	open	open	ADJ
bracis-28383	285	69	access	access	NOUN
bracis-28383	285	70	publishing	publishing	NOUN
bracis-28383	285	71	products	product	NOUN
bracis-28383	285	72	and	and	CCONJ
bracis-28383	285	73	services	service	NOUN
bracis-28383	285	74	our	our	PRON
bracis-28383	285	75	products	product	NOUN
bracis-28383	285	76	librarians	librarian	VERB
bracis-28383	285	77	societies	society	NOUN
bracis-28383	285	78	partners	partner	NOUN
bracis-28383	285	79	and	and	CCONJ
bracis-28383	285	80	advertisers	advertiser	NOUN
bracis-28383	285	81	our	our	PRON
bracis-28383	285	82	brands	brand	NOUN
bracis-28383	285	83	springer	springer	NOUN
bracis-28383	285	84	nature	nature	PROPN
bracis-28383	285	85	portfolio	portfolio	PROPN
bracis-28383	285	86	bmc	bmc	PROPN
bracis-28383	285	87	palgrave	palgrave	PROPN
bracis-28383	285	88	macmillan	macmillan	PROPN
bracis-28383	285	89	apress	apress	PROPN
bracis-28383	285	90	discover	discover	VERB
bracis-28383	285	91	your	your	PRON
bracis-28383	285	92	privacy	privacy	NOUN
bracis-28383	285	93	choices	choice	NOUN
bracis-28383	285	94	/	/	SYM
bracis-28383	285	95	manage	manage	NOUN
bracis-28383	285	96	cookies	cookie	NOUN
bracis-28383	285	97	your	your	PRON
bracis-28383	285	98	us	us	PROPN
bracis-28383	286	1	state	state	NOUN
bracis-28383	286	2	privacy	privacy	NOUN
bracis-28383	286	3	rights	right	NOUN
bracis-28383	286	4	accessibility	accessibility	NOUN
bracis-28383	286	5	statement	statement	NOUN
bracis-28383	286	6	terms	term	NOUN
bracis-28383	286	7	and	and	CCONJ
bracis-28383	286	8	conditions	condition	NOUN
bracis-28383	286	9	privacy	privacy	NOUN
bracis-28383	286	10	policy	policy	NOUN
bracis-28383	286	11	help	help	NOUN
bracis-28383	286	12	and	and	CCONJ
bracis-28383	286	13	support	support	VERB
bracis-28383	286	14	legal	legal	ADJ
bracis-28383	286	15	notice	notice	NOUN
bracis-28383	286	16	cancel	cancel	VERB
bracis-28383	286	17	contracts	contract	NOUN
bracis-28383	286	18	here	here	ADV
bracis-28383	286	19	129.74.145.123	129.74.145.123	NUM
bracis-28383	286	20	hesburgh	hesburgh	PROPN
bracis-28383	286	21	library	library	PROPN
bracis-28383	286	22	er	er	INTJ
bracis-28383	286	23	unit	unit	NOUN
bracis-28383	286	24	(	(	PUNCT
bracis-28383	286	25	3005732405	3005732405	NUM
bracis-28383	286	26	)	)	PUNCT
bracis-28383	286	27	northeast	northeast	ADJ
bracis-28383	286	28	research	research	NOUN
bracis-28383	286	29	libraries	library	NOUN
bracis-28383	286	30	(	(	PUNCT
bracis-28383	286	31	nerl	nerl	PROPN
bracis-28383	286	32	)	)	PUNCT
bracis-28383	286	33	(	(	PUNCT
bracis-28383	286	34	8200828607	8200828607	NUM
bracis-28383	286	35	)	)	PUNCT
bracis-28383	286	36	nerl	nerl	VERB
bracis-28383	286	37	ta	ta	X
bracis-28383	286	38	account	account	NOUN
bracis-28383	286	39	(	(	PUNCT
bracis-28383	286	40	3006206169	3006206169	NUM
bracis-28383	286	41	)	)	PUNCT
bracis-28383	286	42	university	university	NOUN
bracis-28383	286	43	of	of	ADP
bracis-28383	286	44	notre	notre	PROPN
bracis-28383	286	45	dame	dame	PROPN
bracis-28383	286	46	hesburgh	hesburgh	PROPN
bracis-28383	286	47	library	library	NOUN
bracis-28383	286	48	(	(	PUNCT
bracis-28383	286	49	3000184373	3000184373	NUM
bracis-28383	286	50	)	)	PUNCT
bracis-28383	287	1	©	©	ADP
bracis-28383	287	2	2025	2025	NUM
bracis-28383	287	3	springer	springer	NOUN
bracis-28383	287	4	nature	nature	NOUN