id	sid	tid	token	lemma	pos
fcis-1624	1	1	frontiers	frontier	NOUN
fcis-1624	1	2	in	in	ADP
fcis-1624	1	3	computing	computing	NOUN
fcis-1624	1	4	and	and	CCONJ
fcis-1624	1	5	intelligent	intelligent	ADJ
fcis-1624	1	6	systems	system	NOUN
fcis-1624	1	7	issn	issn	VERB
fcis-1624	1	8	:	:	PUNCT
fcis-1624	1	9	2832	2832	NUM
fcis-1624	1	10	-	-	SYM
fcis-1624	1	11	6024	6024	NUM
fcis-1624	1	12	|	|	NOUN
fcis-1624	1	13	vol	vol	NOUN
fcis-1624	1	14	.	.	PROPN
fcis-1624	2	1	1	1	NUM
fcis-1624	2	2	,	,	PUNCT
fcis-1624	2	3	no	no	INTJ
fcis-1624	2	4	.	.	NOUN
fcis-1624	2	5	2	2	NUM
fcis-1624	2	6	,	,	PUNCT
fcis-1624	2	7	2022	2022	NUM
fcis-1624	2	8	1	1	NUM
fcis-1624	2	9	yolov5	yolov5	NOUN
fcis-1624	2	10	-	-	PUNCT
fcis-1624	2	11	based	base	VERB
fcis-1624	2	12	fall	fall	NOUN
fcis-1624	2	13	detection	detection	NOUN
fcis-1624	2	14	algorithm	algorithm	NOUN
fcis-1624	2	15	for	for	ADP
fcis-1624	2	16	homebound	homebound	ADJ
fcis-1624	2	17	people	people	NOUN
fcis-1624	2	18	fumin	fumin	PROPN
fcis-1624	2	19	wang	wang	PROPN
fcis-1624	2	20	*	*	PROPN
fcis-1624	2	21	queen	queen	PROPN
fcis-1624	2	22	mary	mary	PROPN
fcis-1624	2	23	university	university	PROPN
fcis-1624	2	24	of	of	ADP
fcis-1624	2	25	london	london	PROPN
fcis-1624	2	26	engineering	engineering	NOUN
fcis-1624	2	27	school	school	NOUN
fcis-1624	2	28	,	,	PUNCT
fcis-1624	2	29	npu	npu	ADJ
fcis-1624	2	30	,	,	PUNCT
fcis-1624	2	31	northwestern	northwestern	ADJ
fcis-1624	2	32	polytechnical	polytechnical	ADJ
fcis-1624	2	33	university	university	PROPN
fcis-1624	2	34	,	,	PUNCT
fcis-1624	2	35	xian	xian	PROPN
fcis-1624	2	36	,	,	PUNCT
fcis-1624	2	37	shaanxi	shaanxi	PROPN
fcis-1624	2	38	,	,	PUNCT
fcis-1624	2	39	710072	710072	NUM
fcis-1624	2	40	,	,	PUNCT
fcis-1624	2	41	china	china	PROPN
fcis-1624	2	42	*	*	PUNCT
fcis-1624	2	43	corresponding	correspond	VERB
fcis-1624	2	44	author	author	NOUN
fcis-1624	2	45	:	:	PUNCT
fcis-1624	2	46	email	email	NOUN
fcis-1624	2	47	:	:	PUNCT
fcis-1624	2	48	664316139@mail.nwpu.edu.cn	664316139@mail.nwpu.edu.cn	NUM
fcis-1624	2	49	)	)	PUNCT
fcis-1624	2	50	abstract	abstract	NOUN
fcis-1624	2	51	:	:	PUNCT
fcis-1624	2	52	according	accord	VERB
fcis-1624	2	53	to	to	ADP
fcis-1624	2	54	the	the	DET
fcis-1624	2	55	statistics	statistic	NOUN
fcis-1624	2	56	from	from	ADP
fcis-1624	2	57	the	the	DET
fcis-1624	2	58	official	official	ADJ
fcis-1624	2	59	website	website	NOUN
fcis-1624	2	60	of	of	ADP
fcis-1624	2	61	the	the	DET
fcis-1624	2	62	national	national	PROPN
fcis-1624	2	63	bureau	bureau	PROPN
fcis-1624	2	64	of	of	ADP
fcis-1624	2	65	statistics	statistic	NOUN
fcis-1624	2	66	,	,	PUNCT
fcis-1624	2	67	the	the	DET
fcis-1624	2	68	population	population	NOUN
fcis-1624	2	69	over	over	ADP
fcis-1624	2	70	60	60	NUM
fcis-1624	2	71	years	year	NOUN
fcis-1624	2	72	old	old	ADJ
fcis-1624	2	73	in	in	ADP
fcis-1624	2	74	china	china	PROPN
fcis-1624	2	75	is	be	AUX
fcis-1624	2	76	close	close	ADJ
fcis-1624	2	77	to	to	ADP
fcis-1624	2	78	260	260	NUM
fcis-1624	2	79	million	million	NUM
fcis-1624	2	80	,	,	PUNCT
fcis-1624	2	81	accounting	account	VERB
fcis-1624	2	82	for	for	ADP
fcis-1624	2	83	about	about	ADV
fcis-1624	2	84	18.7	18.7	NUM
fcis-1624	2	85	%	%	NOUN
fcis-1624	2	86	of	of	ADP
fcis-1624	2	87	the	the	DET
fcis-1624	2	88	total	total	ADJ
fcis-1624	2	89	population	population	NOUN
fcis-1624	2	90	.	.	PUNCT
fcis-1624	3	1	as	as	SCONJ
fcis-1624	3	2	society	society	NOUN
fcis-1624	3	3	moves	move	VERB
fcis-1624	3	4	towards	towards	ADP
fcis-1624	3	5	aging	aging	NOUN
fcis-1624	3	6	,	,	PUNCT
fcis-1624	3	7	the	the	DET
fcis-1624	3	8	safety	safety	NOUN
fcis-1624	3	9	problem	problem	NOUN
fcis-1624	3	10	of	of	ADP
fcis-1624	3	11	the	the	DET
fcis-1624	3	12	elder	elder	ADJ
fcis-1624	3	13	people	people	NOUN
fcis-1624	3	14	alone	alone	ADV
fcis-1624	3	15	is	be	AUX
fcis-1624	3	16	gradually	gradually	ADV
fcis-1624	3	17	highlighted	highlight	VERB
fcis-1624	3	18	,	,	PUNCT
fcis-1624	3	19	and	and	CCONJ
fcis-1624	3	20	the	the	DET
fcis-1624	3	21	accidental	accidental	ADJ
fcis-1624	3	22	fall	fall	NOUN
fcis-1624	3	23	of	of	ADP
fcis-1624	3	24	the	the	DET
fcis-1624	3	25	elderly	elderly	ADJ
fcis-1624	3	26	at	at	ADP
fcis-1624	3	27	home	home	NOUN
fcis-1624	3	28	has	have	AUX
fcis-1624	3	29	become	become	VERB
fcis-1624	3	30	a	a	DET
fcis-1624	3	31	problem	problem	NOUN
fcis-1624	3	32	that	that	PRON
fcis-1624	3	33	can	can	AUX
fcis-1624	3	34	not	not	PART
fcis-1624	3	35	be	be	AUX
fcis-1624	3	36	ignored	ignore	VERB
fcis-1624	3	37	,	,	PUNCT
fcis-1624	3	38	and	and	CCONJ
fcis-1624	3	39	developing	develop	VERB
fcis-1624	3	40	fall	fall	NOUN
fcis-1624	3	41	detection	detection	NOUN
fcis-1624	3	42	technology	technology	NOUN
fcis-1624	3	43	to	to	PART
fcis-1624	3	44	mitigate	mitigate	VERB
fcis-1624	3	45	the	the	DET
fcis-1624	3	46	danger	danger	NOUN
fcis-1624	3	47	of	of	ADP
fcis-1624	3	48	falls	fall	NOUN
fcis-1624	3	49	of	of	ADP
fcis-1624	3	50	the	the	DET
fcis-1624	3	51	elderly	elderly	ADJ
fcis-1624	3	52	is	be	AUX
fcis-1624	3	53	imperative	imperative	ADJ
fcis-1624	3	54	.	.	PUNCT
fcis-1624	4	1	so	so	ADV
fcis-1624	4	2	,	,	PUNCT
fcis-1624	4	3	this	this	DET
fcis-1624	4	4	paper	paper	NOUN
fcis-1624	4	5	proposes	propose	VERB
fcis-1624	4	6	a	a	DET
fcis-1624	4	7	fall	fall	NOUN
fcis-1624	4	8	detection	detection	NOUN
fcis-1624	4	9	method	method	NOUN
fcis-1624	4	10	which	which	PRON
fcis-1624	4	11	is	be	AUX
fcis-1624	4	12	based	base	VERB
fcis-1624	4	13	on	on	ADP
fcis-1624	4	14	deep	deep	ADJ
fcis-1624	4	15	learning	learning	NOUN
fcis-1624	4	16	.	.	PUNCT
fcis-1624	5	1	specifically	specifically	ADV
fcis-1624	5	2	,	,	PUNCT
fcis-1624	5	3	the	the	DET
fcis-1624	5	4	method	method	NOUN
fcis-1624	5	5	is	be	AUX
fcis-1624	5	6	based	base	VERB
fcis-1624	5	7	on	on	ADP
fcis-1624	5	8	the	the	DET
fcis-1624	5	9	yolov3	yolov3	PROPN
fcis-1624	5	10	method	method	NOUN
fcis-1624	5	11	,	,	PUNCT
fcis-1624	5	12	and	and	CCONJ
fcis-1624	5	13	to	to	PART
fcis-1624	5	14	enhance	enhance	VERB
fcis-1624	5	15	the	the	DET
fcis-1624	5	16	accuracy	accuracy	NOUN
fcis-1624	5	17	and	and	CCONJ
fcis-1624	5	18	speed	speed	NOUN
fcis-1624	5	19	,	,	PUNCT
fcis-1624	5	20	the	the	DET
fcis-1624	5	21	yolov3	yolov3	PROPN
fcis-1624	5	22	algorithm	algorithm	PROPN
fcis-1624	5	23	is	be	AUX
fcis-1624	5	24	updated	update	VERB
fcis-1624	5	25	to	to	ADP
fcis-1624	5	26	yolov5s	yolov5s	PROPN
fcis-1624	5	27	.	.	PUNCT
fcis-1624	6	1	the	the	DET
fcis-1624	6	2	final	final	ADJ
fcis-1624	6	3	experimental	experimental	ADJ
fcis-1624	6	4	results	result	NOUN
fcis-1624	6	5	also	also	ADV
fcis-1624	6	6	confirm	confirm	VERB
fcis-1624	6	7	that	that	SCONJ
fcis-1624	6	8	the	the	DET
fcis-1624	6	9	above	above	ADJ
fcis-1624	6	10	method	method	NOUN
fcis-1624	6	11	has	have	AUX
fcis-1624	6	12	achieved	achieve	VERB
fcis-1624	6	13	the	the	DET
fcis-1624	6	14	purpose	purpose	NOUN
fcis-1624	6	15	.	.	PUNCT
fcis-1624	7	1	keywords	keyword	NOUN
fcis-1624	7	2	:	:	PUNCT
fcis-1624	7	3	yolov5s	yolov5s	NOUN
fcis-1624	7	4	;	;	PUNCT
fcis-1624	7	5	fall	fall	NOUN
fcis-1624	7	6	detection	detection	NOUN
fcis-1624	7	7	;	;	PUNCT
fcis-1624	7	8	deep	deep	ADJ
fcis-1624	7	9	learning	learning	NOUN
fcis-1624	7	10	.	.	PUNCT
fcis-1624	8	1	1	1	X
fcis-1624	8	2	.	.	X
fcis-1624	8	3	introduction	introduction	NOUN
fcis-1624	8	4	at	at	ADP
fcis-1624	8	5	present	present	ADJ
fcis-1624	8	6	,	,	PUNCT
fcis-1624	8	7	society	society	NOUN
fcis-1624	8	8	is	be	AUX
fcis-1624	8	9	gradually	gradually	ADV
fcis-1624	8	10	aging	age	VERB
fcis-1624	8	11	,	,	PUNCT
fcis-1624	8	12	and	and	CCONJ
fcis-1624	8	13	the	the	DET
fcis-1624	8	14	proportion	proportion	NOUN
fcis-1624	8	15	of	of	ADP
fcis-1624	8	16	elderly	elderly	ADJ
fcis-1624	8	17	people	people	NOUN
fcis-1624	8	18	in	in	ADP
fcis-1624	8	19	the	the	DET
fcis-1624	8	20	socio	socio	NOUN
fcis-1624	8	21	-	-	ADJ
fcis-1624	8	22	demographic	demographic	ADJ
fcis-1624	8	23	class	class	NOUN
fcis-1624	8	24	is	be	AUX
fcis-1624	8	25	increasing	increase	VERB
fcis-1624	8	26	,	,	PUNCT
fcis-1624	8	27	and	and	CCONJ
fcis-1624	8	28	the	the	DET
fcis-1624	8	29	health	health	NOUN
fcis-1624	8	30	and	and	CCONJ
fcis-1624	8	31	safety	safety	NOUN
fcis-1624	8	32	issues	issue	NOUN
fcis-1624	8	33	of	of	ADP
fcis-1624	8	34	the	the	DET
fcis-1624	8	35	elderly	elderly	ADJ
fcis-1624	8	36	are	be	AUX
fcis-1624	8	37	becoming	become	VERB
fcis-1624	8	38	more	more	ADV
fcis-1624	8	39	and	and	CCONJ
fcis-1624	8	40	more	more	ADV
fcis-1624	8	41	prominent	prominent	ADJ
fcis-1624	8	42	.	.	PUNCT
fcis-1624	9	1	[	[	X
fcis-1624	9	2	1	1	X
fcis-1624	9	3	]	]	X
fcis-1624	9	4	studies	study	NOUN
fcis-1624	9	5	have	have	AUX
fcis-1624	9	6	shown	show	VERB
fcis-1624	9	7	that	that	SCONJ
fcis-1624	9	8	in	in	ADP
fcis-1624	9	9	china	china	PROPN
fcis-1624	9	10	,	,	PUNCT
fcis-1624	9	11	nearly	nearly	ADV
fcis-1624	9	12	20	20	NUM
fcis-1624	9	13	%	%	NOUN
fcis-1624	9	14	of	of	ADP
fcis-1624	9	15	elderly	elderly	ADJ
fcis-1624	9	16	people	people	NOUN
fcis-1624	9	17	are	be	AUX
fcis-1624	9	18	seriously	seriously	ADV
fcis-1624	9	19	injured	injure	VERB
fcis-1624	9	20	after	after	ADP
fcis-1624	9	21	a	a	DET
fcis-1624	9	22	fall	fall	NOUN
fcis-1624	9	23	,	,	PUNCT
fcis-1624	9	24	even	even	ADV
fcis-1624	9	25	if	if	SCONJ
fcis-1624	9	26	they	they	PRON
fcis-1624	9	27	were	be	AUX
fcis-1624	9	28	very	very	ADV
fcis-1624	9	29	healthy	healthy	ADJ
fcis-1624	9	30	before	before	ADP
fcis-1624	9	31	the	the	DET
fcis-1624	9	32	fall	fall	NOUN
fcis-1624	9	33	,	,	PUNCT
fcis-1624	9	34	and	and	CCONJ
fcis-1624	9	35	most	most	ADJ
fcis-1624	9	36	of	of	ADP
fcis-1624	9	37	them	they	PRON
fcis-1624	9	38	suffer	suffer	VERB
fcis-1624	9	39	secondary	secondary	ADJ
fcis-1624	9	40	injuries	injury	NOUN
fcis-1624	9	41	because	because	SCONJ
fcis-1624	9	42	no	no	DET
fcis-1624	9	43	one	one	NOUN
fcis-1624	9	44	finds	find	VERB
fcis-1624	9	45	and	and	CCONJ
fcis-1624	9	46	sends	send	VERB
fcis-1624	9	47	them	they	PRON
fcis-1624	9	48	to	to	ADP
fcis-1624	9	49	the	the	DET
fcis-1624	9	50	doctor	doctor	NOUN
fcis-1624	9	51	in	in	ADP
fcis-1624	9	52	time	time	NOUN
fcis-1624	9	53	after	after	ADP
fcis-1624	9	54	the	the	DET
fcis-1624	9	55	fall	fall	NOUN
fcis-1624	9	56	,	,	PUNCT
fcis-1624	9	57	which	which	PRON
fcis-1624	9	58	is	be	AUX
fcis-1624	9	59	very	very	ADV
fcis-1624	9	60	common	common	ADJ
fcis-1624	9	61	in	in	ADP
fcis-1624	9	62	elderly	elderly	ADJ
fcis-1624	9	63	families	family	NOUN
fcis-1624	9	64	living	live	VERB
fcis-1624	9	65	alone	alone	ADV
fcis-1624	9	66	.	.	PUNCT
fcis-1624	10	1	[	[	X
fcis-1624	10	2	2	2	X
fcis-1624	10	3	]	]	PUNCT
fcis-1624	10	4	according	accord	VERB
fcis-1624	10	5	to	to	ADP
fcis-1624	10	6	statistics	statistic	NOUN
fcis-1624	10	7	,	,	PUNCT
fcis-1624	10	8	the	the	DET
fcis-1624	10	9	number	number	NOUN
fcis-1624	10	10	of	of	ADP
fcis-1624	10	11	elderly	elderly	ADJ
fcis-1624	10	12	people	people	NOUN
fcis-1624	10	13	living	live	VERB
fcis-1624	10	14	alone	alone	ADJ
fcis-1624	10	15	in	in	ADP
fcis-1624	10	16	china	china	PROPN
fcis-1624	10	17	has	have	AUX
fcis-1624	10	18	exceeded	exceed	VERB
fcis-1624	10	19	30	30	NUM
fcis-1624	10	20	million	million	NUM
fcis-1624	10	21	,	,	PUNCT
fcis-1624	10	22	and	and	CCONJ
fcis-1624	10	23	it	it	PRON
fcis-1624	10	24	is	be	AUX
fcis-1624	10	25	a	a	DET
fcis-1624	10	26	practical	practical	ADJ
fcis-1624	10	27	problem	problem	NOUN
fcis-1624	10	28	to	to	PART
fcis-1624	10	29	detect	detect	VERB
fcis-1624	10	30	falls	fall	NOUN
fcis-1624	10	31	and	and	CCONJ
fcis-1624	10	32	prevent	prevent	VERB
fcis-1624	10	33	secondary	secondary	ADJ
fcis-1624	10	34	injuries	injury	NOUN
fcis-1624	10	35	caused	cause	VERB
fcis-1624	10	36	by	by	ADP
fcis-1624	10	37	falls	fall	NOUN
fcis-1624	10	38	.	.	PUNCT
fcis-1624	11	1	therefore	therefore	ADV
fcis-1624	11	2	,	,	PUNCT
fcis-1624	11	3	it	it	PRON
fcis-1624	11	4	is	be	AUX
fcis-1624	11	5	valuable	valuable	ADJ
fcis-1624	11	6	to	to	PART
fcis-1624	11	7	research	research	VERB
fcis-1624	11	8	and	and	CCONJ
fcis-1624	11	9	develop	develop	VERB
fcis-1624	11	10	an	an	DET
fcis-1624	11	11	accurate	accurate	ADJ
fcis-1624	11	12	fall	fall	NOUN
fcis-1624	11	13	detection	detection	NOUN
fcis-1624	11	14	algorithm	algorithm	NOUN
fcis-1624	11	15	for	for	ADP
fcis-1624	11	16	the	the	DET
fcis-1624	11	17	human	human	ADJ
fcis-1624	11	18	body	body	NOUN
fcis-1624	11	19	.	.	PUNCT
fcis-1624	12	1	it	it	PRON
fcis-1624	12	2	can	can	AUX
fcis-1624	12	3	detect	detect	VERB
fcis-1624	12	4	the	the	DET
fcis-1624	12	5	fall	fall	NOUN
fcis-1624	12	6	in	in	ADP
fcis-1624	12	7	time	time	NOUN
fcis-1624	12	8	so	so	SCONJ
fcis-1624	12	9	that	that	SCONJ
fcis-1624	12	10	the	the	DET
fcis-1624	12	11	elderly	elderly	ADJ
fcis-1624	12	12	can	can	AUX
fcis-1624	12	13	get	get	VERB
fcis-1624	12	14	timely	timely	ADJ
fcis-1624	12	15	treatment	treatment	NOUN
fcis-1624	12	16	and	and	CCONJ
fcis-1624	12	17	avoid	avoid	VERB
fcis-1624	12	18	causing	cause	VERB
fcis-1624	12	19	secondary	secondary	ADJ
fcis-1624	12	20	injuries	injury	NOUN
fcis-1624	12	21	.	.	PUNCT
fcis-1624	13	1	according	accord	VERB
fcis-1624	13	2	to	to	ADP
fcis-1624	13	3	the	the	DET
fcis-1624	13	4	current	current	ADJ
fcis-1624	13	5	state	state	NOUN
fcis-1624	13	6	of	of	ADP
fcis-1624	13	7	research	research	NOUN
fcis-1624	13	8	at	at	ADP
fcis-1624	13	9	home	home	NOUN
fcis-1624	13	10	and	and	CCONJ
fcis-1624	13	11	abroad	abroad	ADV
fcis-1624	13	12	[	[	X
fcis-1624	13	13	3,4	3,4	NUM
fcis-1624	13	14	]	]	PUNCT
fcis-1624	13	15	,	,	PUNCT
fcis-1624	13	16	there	there	PRON
fcis-1624	13	17	are	be	VERB
fcis-1624	13	18	two	two	NUM
fcis-1624	13	19	kinds	kind	NOUN
fcis-1624	13	20	of	of	ADP
fcis-1624	13	21	the	the	DET
fcis-1624	13	22	current	current	ADJ
fcis-1624	13	23	mainstream	mainstream	NOUN
fcis-1624	13	24	algorithms	algorithm	NOUN
fcis-1624	13	25	on	on	ADP
fcis-1624	13	26	target	target	NOUN
fcis-1624	13	27	detection	detection	NOUN
fcis-1624	13	28	.	.	PUNCT
fcis-1624	14	1	the	the	DET
fcis-1624	14	2	first	first	ADJ
fcis-1624	14	3	class	class	NOUN
fcis-1624	14	4	of	of	ADP
fcis-1624	14	5	algorithms	algorithm	NOUN
fcis-1624	14	6	is	be	AUX
fcis-1624	14	7	to	to	PART
fcis-1624	14	8	forecast	forecast	VERB
fcis-1624	14	9	the	the	DET
fcis-1624	14	10	class	class	NOUN
fcis-1624	14	11	and	and	CCONJ
fcis-1624	14	12	location	location	NOUN
fcis-1624	14	13	of	of	ADP
fcis-1624	14	14	many	many	ADJ
fcis-1624	14	15	targets	target	NOUN
fcis-1624	14	16	directly	directly	ADV
fcis-1624	14	17	and	and	CCONJ
fcis-1624	14	18	use	use	VERB
fcis-1624	14	19	only	only	ADV
fcis-1624	14	20	one	one	NUM
fcis-1624	14	21	convolutional	convolutional	ADJ
fcis-1624	14	22	neural	neural	ADJ
fcis-1624	14	23	network	network	NOUN
fcis-1624	14	24	,	,	PUNCT
fcis-1624	14	25	which	which	PRON
fcis-1624	14	26	is	be	AUX
fcis-1624	14	27	represented	represent	VERB
fcis-1624	14	28	by	by	ADP
fcis-1624	14	29	one	one	NUM
fcis-1624	14	30	-	-	PUNCT
fcis-1624	14	31	stage	stage	NOUN
fcis-1624	14	32	algorithms	algorithm	NOUN
fcis-1624	14	33	,	,	PUNCT
fcis-1624	14	34	yolo	yolo	NOUN
fcis-1624	14	35	and	and	CCONJ
fcis-1624	14	36	ssd	ssd	NOUN
fcis-1624	14	37	are	be	AUX
fcis-1624	14	38	the	the	DET
fcis-1624	14	39	representative	representative	ADJ
fcis-1624	14	40	algorithms	algorithm	NOUN
fcis-1624	14	41	of	of	ADP
fcis-1624	14	42	which	which	PRON
fcis-1624	14	43	.	.	PUNCT
fcis-1624	15	1	the	the	DET
fcis-1624	15	2	second	second	ADJ
fcis-1624	15	3	class	class	NOUN
fcis-1624	15	4	,	,	PUNCT
fcis-1624	15	5	such	such	ADJ
fcis-1624	15	6	as	as	ADP
fcis-1624	15	7	r	r	NOUN
fcis-1624	15	8	-	-	PUNCT
fcis-1624	15	9	cnn	cnn	PROPN
fcis-1624	15	10	,	,	PUNCT
fcis-1624	15	11	fast	fast	ADJ
fcis-1624	15	12	r	r	NOUN
fcis-1624	15	13	-	-	PUNCT
fcis-1624	15	14	cnn	cnn	PROPN
fcis-1624	15	15	,	,	PUNCT
fcis-1624	15	16	faster	fast	ADV
fcis-1624	15	17	r	r	NOUN
fcis-1624	15	18	-	-	PUNCT
fcis-1624	15	19	cnn	cnn	PROPN
fcis-1624	15	20	and	and	CCONJ
fcis-1624	15	21	other	other	ADJ
fcis-1624	15	22	two	two	NUM
fcis-1624	15	23	-	-	PUNCT
fcis-1624	15	24	stage	stage	NOUN
fcis-1624	15	25	r	r	NOUN
fcis-1624	15	26	-	-	PUNCT
fcis-1624	15	27	cn	cn	ADJ
fcis-1624	15	28	algorithms	algorithm	NOUN
fcis-1624	15	29	which	which	PRON
fcis-1624	15	30	are	be	AUX
fcis-1624	15	31	derived	derive	VERB
fcis-1624	15	32	from	from	ADP
fcis-1624	15	33	region	region	NOUN
fcis-1624	15	34	proposal	proposal	NOUN
fcis-1624	15	35	.	.	PUNCT
fcis-1624	16	1	in	in	ADP
fcis-1624	16	2	other	other	ADJ
fcis-1624	16	3	words	word	NOUN
fcis-1624	16	4	,	,	PUNCT
fcis-1624	16	5	the	the	DET
fcis-1624	16	6	target	target	NOUN
fcis-1624	16	7	candidate	candidate	NOUN
fcis-1624	16	8	frame	frame	NOUN
fcis-1624	16	9	,	,	PUNCT
fcis-1624	16	10	i.e.	i.e.	X
fcis-1624	16	11	,	,	PUNCT
fcis-1624	16	12	the	the	DET
fcis-1624	16	13	target	target	NOUN
fcis-1624	16	14	location	location	NOUN
fcis-1624	16	15	,	,	PUNCT
fcis-1624	16	16	needs	need	VERB
fcis-1624	16	17	to	to	PART
fcis-1624	16	18	be	be	AUX
fcis-1624	16	19	generated	generate	VERB
fcis-1624	16	20	before	before	SCONJ
fcis-1624	16	21	the	the	DET
fcis-1624	16	22	candidate	candidate	NOUN
fcis-1624	16	23	frame	frame	NOUN
fcis-1624	16	24	can	can	AUX
fcis-1624	16	25	be	be	AUX
fcis-1624	16	26	classified	classify	VERB
fcis-1624	16	27	and	and	CCONJ
fcis-1624	16	28	regressed	regress	VERB
fcis-1624	16	29	.	.	PUNCT
fcis-1624	17	1	in	in	ADP
fcis-1624	17	2	comparison	comparison	NOUN
fcis-1624	17	3	,	,	PUNCT
fcis-1624	17	4	the	the	DET
fcis-1624	17	5	second	second	ADJ
fcis-1624	17	6	type	type	NOUN
fcis-1624	17	7	of	of	ADP
fcis-1624	17	8	algorithm	algorithm	NOUN
fcis-1624	17	9	is	be	AUX
fcis-1624	17	10	slower	slow	ADJ
fcis-1624	17	11	but	but	CCONJ
fcis-1624	17	12	more	more	ADV
fcis-1624	17	13	accurate	accurate	ADJ
fcis-1624	17	14	,	,	PUNCT
fcis-1624	17	15	while	while	SCONJ
fcis-1624	17	16	the	the	DET
fcis-1624	17	17	first	first	ADJ
fcis-1624	17	18	type	type	NOUN
fcis-1624	17	19	is	be	AUX
fcis-1624	17	20	slightly	slightly	ADV
fcis-1624	17	21	less	less	ADV
fcis-1624	17	22	accurate	accurate	ADJ
fcis-1624	17	23	but	but	CCONJ
fcis-1624	17	24	faster	fast	ADV
fcis-1624	17	25	.	.	PUNCT
fcis-1624	18	1	specifically	specifically	ADV
fcis-1624	18	2	,	,	PUNCT
fcis-1624	18	3	for	for	ADP
fcis-1624	18	4	fall	fall	NOUN
fcis-1624	18	5	detection	detection	NOUN
fcis-1624	18	6	,	,	PUNCT
fcis-1624	18	7	most	most	ADJ
fcis-1624	18	8	of	of	ADP
fcis-1624	18	9	the	the	DET
fcis-1624	18	10	existing	exist	VERB
fcis-1624	18	11	algorithms	algorithm	NOUN
fcis-1624	18	12	are	be	AUX
fcis-1624	18	13	based	base	VERB
fcis-1624	18	14	on	on	ADP
fcis-1624	18	15	wearable	wearable	ADJ
fcis-1624	18	16	devices	device	NOUN
fcis-1624	19	1	[	[	X
fcis-1624	19	2	5	5	NUM
fcis-1624	19	3	]	]	PUNCT
fcis-1624	19	4	,	,	PUNCT
fcis-1624	19	5	which	which	PRON
fcis-1624	19	6	are	be	AUX
fcis-1624	19	7	not	not	PART
fcis-1624	19	8	ideal	ideal	ADJ
fcis-1624	19	9	in	in	ADP
fcis-1624	19	10	terms	term	NOUN
fcis-1624	19	11	of	of	ADP
fcis-1624	19	12	complexity	complexity	NOUN
fcis-1624	19	13	and	and	CCONJ
fcis-1624	19	14	recognition	recognition	NOUN
fcis-1624	19	15	capability	capability	NOUN
fcis-1624	19	16	.	.	PUNCT
fcis-1624	20	1	therefore	therefore	ADV
fcis-1624	20	2	,	,	PUNCT
fcis-1624	20	3	studying	study	VERB
fcis-1624	20	4	the	the	DET
fcis-1624	20	5	computer	computer	NOUN
fcis-1624	20	6	vision	vision	NOUN
fcis-1624	20	7	[	[	X
fcis-1624	20	8	6	6	NUM
fcis-1624	20	9	]	]	PUNCT
fcis-1624	20	10	in	in	ADP
fcis-1624	20	11	the	the	DET
fcis-1624	20	12	field	field	NOUN
fcis-1624	20	13	of	of	ADP
fcis-1624	20	14	fall	fall	NOUN
fcis-1624	20	15	detection	detection	NOUN
fcis-1624	20	16	may	may	AUX
fcis-1624	20	17	solve	solve	VERB
fcis-1624	20	18	the	the	DET
fcis-1624	20	19	above	above	ADJ
fcis-1624	20	20	problems	problem	NOUN
fcis-1624	20	21	.	.	PUNCT
fcis-1624	21	1	2	2	X
fcis-1624	21	2	.	.	NUM
fcis-1624	21	3	yolo	yolo	ADJ
fcis-1624	21	4	target	target	NOUN
fcis-1624	21	5	detection	detection	NOUN
fcis-1624	21	6	algorithm	algorithm	NOUN
fcis-1624	21	7	family	family	NOUN
fcis-1624	21	8	2.1	2.1	NUM
fcis-1624	21	9	.	.	PUNCT
fcis-1624	22	1	comparison	comparison	NOUN
fcis-1624	22	2	of	of	ADP
fcis-1624	22	3	different	different	ADJ
fcis-1624	22	4	algorithm	algorithm	NOUN
fcis-1624	22	5	series	series	NOUN
fcis-1624	22	6	table	table	NOUN
fcis-1624	22	7	1	1	NUM
fcis-1624	22	8	.	.	PUNCT
fcis-1624	22	9	yolo	yolo	ADJ
fcis-1624	22	10	target	target	NOUN
fcis-1624	22	11	detection	detection	NOUN
fcis-1624	22	12	algorithm	algorithm	NOUN
fcis-1624	22	13	series	series	PROPN
fcis-1624	22	14	comparison	comparison	PROPN
fcis-1624	22	15	yolo	yolo	PROPN
fcis-1624	22	16	v1	v1	PROPN
fcis-1624	22	17	yolo	yolo	PROPN
fcis-1624	22	18	v2	v2	PROPN
fcis-1624	22	19	yolo	yolo	PROPN
fcis-1624	22	20	v3	v3	PROPN
fcis-1624	22	21	yolo	yolo	PROPN
fcis-1624	22	22	v4	v4	PROPN
fcis-1624	23	1	yolo	yolo	ADV
fcis-1624	23	2	v5	v5	PROPN
fcis-1624	23	3	advantage	advantage	NOUN
fcis-1624	23	4	s	s	X
fcis-1624	23	5	detectio	detectio	NOUN
fcis-1624	23	6	n	n	PRON
fcis-1624	23	7	speed	speed	NOUN
fcis-1624	23	8	is	be	AUX
fcis-1624	23	9	faster	fast	ADJ
fcis-1624	23	10	detection	detection	NOUN
fcis-1624	23	11	speed	speed	NOUN
fcis-1624	23	12	is	be	AUX
fcis-1624	23	13	better	well	ADJ
fcis-1624	23	14	than	than	ADP
fcis-1624	23	15	fasterrcnn	fasterrcnn	PROPN
fcis-1624	23	16	,	,	PUNCT
fcis-1624	23	17	ssd	ssd	NOUN
fcis-1624	23	18	,	,	PUNCT
fcis-1624	23	19	etc	etc	X
fcis-1624	23	20	.	.	X
fcis-1624	23	21	fast	fast	ADJ
fcis-1624	23	22	detectio	detectio	NOUN
fcis-1624	23	23	n	n	PRON
fcis-1624	23	24	speed	speed	VERB
fcis-1624	23	25	efficie	efficie	NOUN
fcis-1624	23	26	nt	not	PART
fcis-1624	23	27	and	and	CCONJ
fcis-1624	23	28	power	power	NOUN
fcis-1624	23	29	ful	ful	NOUN
fcis-1624	23	30	model	model	NOUN
fcis-1624	23	31	s	s	PART
fcis-1624	23	32	high	high	ADJ
fcis-1624	23	33	flexibility	flexibility	NOUN
fcis-1624	23	34	disadvanta	disadvanta	PROPN
fcis-1624	23	35	ges	ge	NOUN
fcis-1624	23	36	the	the	DET
fcis-1624	23	37	model	model	NOUN
fcis-1624	23	38	relies	rely	VERB
fcis-1624	23	39	on	on	ADP
fcis-1624	23	40	object	object	NOUN
fcis-1624	23	41	recogniti	recogniti	ADV
fcis-1624	23	42	on	on	ADP
fcis-1624	23	43	labeling	label	VERB
fcis-1624	23	44	data	data	NOUN
fcis-1624	23	45	accuracy	accuracy	NOUN
fcis-1624	23	46	performa	performa	PROPN
fcis-1624	23	47	nce	nce	NOUN
fcis-1624	23	48	is	be	AUX
fcis-1624	23	49	not	not	PART
fcis-1624	23	50	very	very	ADV
fcis-1624	23	51	good	good	ADJ
fcis-1624	23	52	the	the	DET
fcis-1624	23	53	model	model	NOUN
fcis-1624	23	54	complex	complex	NOUN
fcis-1624	23	55	ity	ity	PROPN
fcis-1624	23	56	is	be	AUX
fcis-1624	23	57	high	high	ADJ
fcis-1624	23	58	slightly	slightly	ADV
fcis-1624	23	59	inferior	inferior	ADJ
fcis-1624	23	60	performa	performa	X
fcis-1624	23	61	nce	nce	NOUN
fcis-1624	23	62	to	to	ADP
fcis-1624	23	63	yolo	yolo	PROPN
fcis-1624	23	64	v4	v4	NOUN
fcis-1624	23	65	in	in	ADP
fcis-1624	23	66	the	the	DET
fcis-1624	23	67	target	target	NOUN
fcis-1624	23	68	detection	detection	NOUN
fcis-1624	23	69	field	field	NOUN
fcis-1624	23	70	,	,	PUNCT
fcis-1624	23	71	yolo	yolo	PROPN
fcis-1624	23	72	v3	v3	PROPN
fcis-1624	23	73	has	have	AUX
fcis-1624	23	74	been	be	AUX
fcis-1624	23	75	applied	apply	VERB
fcis-1624	23	76	to	to	ADP
fcis-1624	23	77	this	this	DET
fcis-1624	23	78	field	field	NOUN
fcis-1624	23	79	and	and	CCONJ
fcis-1624	23	80	achieved	achieve	VERB
fcis-1624	23	81	good	good	ADJ
fcis-1624	23	82	results	result	NOUN
fcis-1624	23	83	.	.	PUNCT
fcis-1624	24	1	this	this	DET
fcis-1624	24	2	paper	paper	NOUN
fcis-1624	24	3	aims	aim	VERB
fcis-1624	24	4	to	to	PART
fcis-1624	24	5	update	update	VERB
fcis-1624	24	6	the	the	DET
fcis-1624	24	7	existing	exist	VERB
fcis-1624	24	8	yolo	yolo	ADJ
fcis-1624	24	9	v3	v3	PROPN
fcis-1624	24	10	fall	fall	PROPN
fcis-1624	24	11	detection	detection	PROPN
fcis-1624	24	12	research	research	NOUN
fcis-1624	24	13	project	project	NOUN
fcis-1624	24	14	to	to	ADP
fcis-1624	24	15	the	the	DET
fcis-1624	24	16	yolo	yolo	ADJ
fcis-1624	24	17	v5	v5	PROPN
fcis-1624	24	18	algorithm	algorithm	NOUN
fcis-1624	24	19	,	,	PUNCT
fcis-1624	24	20	which	which	PRON
fcis-1624	24	21	is	be	AUX
fcis-1624	24	22	more	more	ADV
fcis-1624	24	23	flexible	flexible	ADJ
fcis-1624	24	24	and	and	CCONJ
fcis-1624	24	25	faster	fast	ADV
fcis-1624	24	26	.	.	PUNCT
fcis-1624	25	1	in	in	ADP
fcis-1624	25	2	the	the	DET
fcis-1624	25	3	next	next	ADJ
fcis-1624	25	4	sections	section	NOUN
fcis-1624	25	5	,	,	PUNCT
fcis-1624	25	6	i	i	PRON
fcis-1624	25	7	will	will	AUX
fcis-1624	25	8	present	present	VERB
fcis-1624	25	9	the	the	DET
fcis-1624	25	10	innovations	innovation	NOUN
fcis-1624	25	11	made	make	VERB
fcis-1624	25	12	in	in	ADP
fcis-1624	25	13	the	the	DET
fcis-1624	25	14	network	network	NOUN
fcis-1624	25	15	structure	structure	NOUN
fcis-1624	25	16	of	of	ADP
fcis-1624	25	17	yolo	yolo	ADJ
fcis-1624	25	18	v5	v5	PROPN
fcis-1624	25	19	compared	compare	VERB
fcis-1624	25	20	to	to	ADP
fcis-1624	25	21	yolo	yolo	PROPN
fcis-1624	25	22	v3	v3	PROPN
fcis-1624	25	23	,	,	PUNCT
fcis-1624	25	24	the	the	DET
fcis-1624	25	25	creation	creation	NOUN
fcis-1624	25	26	of	of	ADP
fcis-1624	25	27	the	the	DET
fcis-1624	25	28	dataset	dataset	NOUN
fcis-1624	25	29	,	,	PUNCT
fcis-1624	25	30	the	the	DET
fcis-1624	25	31	training	training	NOUN
fcis-1624	25	32	of	of	ADP
fcis-1624	25	33	the	the	DET
fcis-1624	25	34	model	model	NOUN
fcis-1624	25	35	and	and	CCONJ
fcis-1624	25	36	the	the	DET
fcis-1624	25	37	detection	detection	NOUN
fcis-1624	25	38	of	of	ADP
fcis-1624	25	39	the	the	DET
fcis-1624	25	40	dataset	dataset	NOUN
fcis-1624	25	41	.	.	PUNCT
fcis-1624	26	1	based	base	VERB
fcis-1624	26	2	on	on	ADP
fcis-1624	26	3	the	the	DET
fcis-1624	26	4	yolo	yolo	ADJ
fcis-1624	26	5	v5	v5	PROPN
fcis-1624	26	6	algorithm	algorithm	NOUN
fcis-1624	26	7	,	,	PUNCT
fcis-1624	26	8	it	it	PRON
fcis-1624	26	9	makes	make	VERB
fcis-1624	26	10	the	the	DET
fcis-1624	26	11	detection	detection	NOUN
fcis-1624	26	12	more	more	ADV
fcis-1624	26	13	accurate	accurate	ADJ
fcis-1624	26	14	and	and	CCONJ
fcis-1624	26	15	faster	fast	ADV
fcis-1624	26	16	.	.	PUNCT
fcis-1624	27	1	2.2	2.2	NUM
fcis-1624	27	2	.	.	PUNCT
fcis-1624	28	1	network	network	NOUN
fcis-1624	28	2	structure	structure	NOUN
fcis-1624	28	3	diagram	diagram	NOUN
fcis-1624	28	4	of	of	ADP
fcis-1624	28	5	yolo	yolo	PROPN
fcis-1624	28	6	v5	v5	VERB
fcis-1624	28	7	the	the	DET
fcis-1624	28	8	structure	structure	NOUN
fcis-1624	28	9	of	of	ADP
fcis-1624	28	10	yolo	yolo	ADJ
fcis-1624	28	11	v5	v5	PROPN
fcis-1624	28	12	is	be	AUX
fcis-1624	28	13	similar	similar	ADJ
fcis-1624	28	14	to	to	ADP
fcis-1624	28	15	that	that	PRON
fcis-1624	28	16	of	of	ADP
fcis-1624	28	17	yolo	yolo	ADJ
fcis-1624	28	18	v4	v4	NOUN
fcis-1624	28	19	,	,	PUNCT
fcis-1624	28	20	with	with	ADP
fcis-1624	28	21	more	more	ADJ
fcis-1624	28	22	innovations	innovation	NOUN
fcis-1624	28	23	compared	compare	VERB
fcis-1624	28	24	to	to	ADP
fcis-1624	28	25	yolo	yolo	PROPN
fcis-1624	28	26	v3	v3	PROPN
fcis-1624	28	27	.	.	PUNCT
fcis-1624	29	1	there	there	PRON
fcis-1624	29	2	are	be	VERB
fcis-1624	29	3	four	four	NUM
fcis-1624	29	4	detection	detection	NOUN
fcis-1624	29	5	networks	network	NOUN
fcis-1624	29	6	in	in	ADP
fcis-1624	29	7	yolo	yolo	ADJ
fcis-1624	29	8	v5	v5	PROPN
fcis-1624	29	9	,	,	PUNCT
fcis-1624	29	10	namely	namely	ADV
fcis-1624	29	11	yolo	yolo	ADJ
fcis-1624	29	12	v5s	v5s	PROPN
fcis-1624	29	13	,	,	PUNCT
fcis-1624	29	14	yolo	yolo	PROPN
fcis-1624	29	15	v5	v5	PROPN
fcis-1624	29	16	m	m	PROPN
fcis-1624	29	17	,	,	PUNCT
fcis-1624	29	18	yolo	yolo	PROPN
fcis-1624	29	19	v5l	v5l	PROPN
fcis-1624	29	20	,	,	PUNCT
fcis-1624	29	21	and	and	CCONJ
fcis-1624	29	22	yolo	yolo	ADJ
fcis-1624	29	23	v5x	v5x	NOUN
fcis-1624	29	24	.	.	PUNCT
fcis-1624	30	1	the	the	DET
fcis-1624	30	2	depth	depth	NOUN
fcis-1624	30	3	and	and	CCONJ
fcis-1624	30	4	the	the	DET
fcis-1624	30	5	feature	feature	NOUN
fcis-1624	30	6	map	map	NOUN
fcis-1624	30	7	width	width	NOUN
fcis-1624	30	8	of	of	ADP
fcis-1624	30	9	yolo	yolo	ADJ
fcis-1624	30	10	v5s	v5s	PROPN
fcis-1624	30	11	network	network	NOUN
fcis-1624	30	12	is	be	AUX
fcis-1624	30	13	the	the	DET
fcis-1624	30	14	smallest	small	ADJ
fcis-1624	30	15	in	in	ADP
fcis-1624	30	16	the	the	DET
fcis-1624	30	17	yolov5	yolov5	NOUN
fcis-1624	30	18	series	series	NOUN
fcis-1624	30	19	,	,	PUNCT
fcis-1624	30	20	and	and	CCONJ
fcis-1624	30	21	its	its	PRON
fcis-1624	30	22	detection	detection	NOUN
fcis-1624	30	23	speed	speed	NOUN
fcis-1624	30	24	and	and	CCONJ
fcis-1624	30	25	ap	ap	PROPN
fcis-1624	30	26	accuracy	accuracy	NOUN
fcis-1624	30	27	are	be	AUX
fcis-1624	30	28	not	not	PART
fcis-1624	30	29	excellent	excellent	ADJ
fcis-1624	30	30	compared	compare	VERB
fcis-1624	30	31	with	with	ADP
fcis-1624	30	32	the	the	DET
fcis-1624	30	33	last	last	ADJ
fcis-1624	30	34	three	three	NUM
fcis-1624	30	35	models	model	NOUN
fcis-1624	30	36	,	,	PUNCT
fcis-1624	30	37	but	but	CCONJ
fcis-1624	30	38	it	it	PRON
fcis-1624	30	39	is	be	AUX
fcis-1624	30	40	a	a	DET
fcis-1624	30	41	good	good	ADJ
fcis-1624	30	42	choice	choice	NOUN
fcis-1624	30	43	if	if	SCONJ
fcis-1624	30	44	the	the	DET
fcis-1624	30	45	monitoring	monitoring	NOUN
fcis-1624	30	46	target	target	NOUN
fcis-1624	30	47	is	be	AUX
fcis-1624	30	48	mainly	mainly	ADV
fcis-1624	30	49	large	large	ADJ
fcis-1624	30	50	targets	target	NOUN
fcis-1624	30	51	and	and	CCONJ
fcis-1624	30	52	the	the	DET
fcis-1624	30	53	speed	speed	NOUN
fcis-1624	30	54	is	be	AUX
fcis-1624	30	55	pursued	pursue	VERB
fcis-1624	30	56	.	.	PUNCT
fcis-1624	31	1	the	the	DET
fcis-1624	31	2	model	model	NOUN
fcis-1624	31	3	used	use	VERB
fcis-1624	31	4	in	in	ADP
fcis-1624	31	5	this	this	DET
fcis-1624	31	6	paper	paper	NOUN
fcis-1624	31	7	is	be	AUX
fcis-1624	31	8	the	the	DET
fcis-1624	31	9	yolov5s	yolov5s	PROPN
fcis-1624	31	10	2	2	NUM
fcis-1624	31	11	algorithm	algorithm	NOUN
fcis-1624	31	12	,	,	PUNCT
fcis-1624	31	13	and	and	CCONJ
fcis-1624	31	14	the	the	DET
fcis-1624	31	15	following	follow	VERB
fcis-1624	31	16	figure	figure	NOUN
fcis-1624	31	17	gives	give	VERB
fcis-1624	31	18	a	a	DET
fcis-1624	31	19	view	view	NOUN
fcis-1624	31	20	of	of	ADP
fcis-1624	31	21	the	the	DET
fcis-1624	31	22	structure	structure	NOUN
fcis-1624	31	23	of	of	ADP
fcis-1624	31	24	yolov5s	yolov5s	PROPN
fcis-1624	31	25	network	network	NOUN
fcis-1624	31	26	.	.	PUNCT
fcis-1624	32	1	figure	figure	NOUN
fcis-1624	32	2	1	1	NUM
fcis-1624	32	3	.	.	PUNCT
fcis-1624	33	1	yolo	yolo	PROPN
fcis-1624	33	2	v5	v5	PROPN
fcis-1624	33	3	network	network	NOUN
fcis-1624	33	4	structure	structure	NOUN
fcis-1624	33	5	diagram	diagram	NOUN
fcis-1624	33	6	there	there	PRON
fcis-1624	33	7	are	be	VERB
fcis-1624	33	8	four	four	NUM
fcis-1624	33	9	aspects	aspect	NOUN
fcis-1624	33	10	for	for	ADP
fcis-1624	33	11	the	the	DET
fcis-1624	33	12	structure	structure	NOUN
fcis-1624	33	13	diagram	diagram	NOUN
fcis-1624	33	14	of	of	ADP
fcis-1624	33	15	yolo	yolo	PROPN
fcis-1624	33	16	v5s	v5s	PROPN
fcis-1624	33	17	:	:	PUNCT
fcis-1624	33	18	input	input	NOUN
fcis-1624	33	19	side	side	NOUN
fcis-1624	33	20	,	,	PUNCT
fcis-1624	33	21	backbone	backbone	NOUN
fcis-1624	33	22	,	,	PUNCT
fcis-1624	33	23	neck	neck	NOUN
fcis-1624	33	24	,	,	PUNCT
fcis-1624	33	25	and	and	CCONJ
fcis-1624	33	26	prediction	prediction	NOUN
fcis-1624	33	27	.	.	PUNCT
fcis-1624	34	1	compared	compare	VERB
fcis-1624	34	2	with	with	ADP
fcis-1624	34	3	the	the	DET
fcis-1624	34	4	yolo	yolo	ADJ
fcis-1624	34	5	v3	v3	PROPN
fcis-1624	34	6	algorithm	algorithm	PROPN
fcis-1624	34	7	,	,	PUNCT
fcis-1624	34	8	it	it	PRON
fcis-1624	34	9	has	have	AUX
fcis-1624	34	10	made	make	VERB
fcis-1624	34	11	innovations	innovation	NOUN
fcis-1624	34	12	in	in	ADP
fcis-1624	34	13	all	all	DET
fcis-1624	34	14	four	four	NUM
fcis-1624	34	15	parts	part	NOUN
fcis-1624	34	16	：	：	PUNCT
fcis-1624	34	17	(	(	PUNCT
fcis-1624	34	18	1	1	X
fcis-1624	34	19	)	)	PUNCT
fcis-1624	34	20	input	input	NOUN
fcis-1624	34	21	:	:	PUNCT
fcis-1624	34	22	enhance	enhance	VERB
fcis-1624	34	23	mosaic	mosaic	ADJ
fcis-1624	34	24	data	datum	NOUN
fcis-1624	34	25	and	and	CCONJ
fcis-1624	34	26	adaptive	adaptive	ADJ
fcis-1624	34	27	anchor	anchor	NOUN
fcis-1624	34	28	frame	frame	NOUN
fcis-1624	34	29	calculation	calculation	NOUN
fcis-1624	34	30	(	(	PUNCT
fcis-1624	34	31	2	2	X
fcis-1624	34	32	)	)	PUNCT
fcis-1624	34	33	backbone	backbone	NOUN
fcis-1624	34	34	:	:	PUNCT
fcis-1624	34	35	focus	focus	NOUN
fcis-1624	34	36	structure	structure	NOUN
fcis-1624	34	37	and	and	CCONJ
fcis-1624	34	38	csp	csp	PROPN
fcis-1624	34	39	structure	structure	NOUN
fcis-1624	34	40	(	(	PUNCT
fcis-1624	34	41	3	3	NUM
fcis-1624	34	42	)	)	PUNCT
fcis-1624	34	43	neck	neck	NOUN
fcis-1624	34	44	:	:	PUNCT
fcis-1624	34	45	fpn	fpn	VERB
fcis-1624	34	46	structure	structure	NOUN
fcis-1624	34	47	and	and	CCONJ
fcis-1624	34	48	pan	pan	NOUN
fcis-1624	34	49	structure	structure	NOUN
fcis-1624	34	50	(	(	PUNCT
fcis-1624	34	51	4	4	NUM
fcis-1624	34	52	)	)	PUNCT
fcis-1624	34	53	prediction	prediction	NOUN
fcis-1624	34	54	:	:	PUNCT
fcis-1624	34	55	giou_loss	giou_loss	PROPN
fcis-1624	34	56	2.2.1	2.2.1	NUM
fcis-1624	34	57	.	.	PUNCT
fcis-1624	34	58	inputs	input	NOUN
fcis-1624	34	59	mosaic	mosaic	ADJ
fcis-1624	34	60	data	datum	NOUN
fcis-1624	34	61	enhancement	enhancement	NOUN
fcis-1624	34	62	yolo	yolo	ADJ
fcis-1624	34	63	v5	v5	PROPN
fcis-1624	34	64	carries	carry	VERB
fcis-1624	34	65	out	out	ADP
fcis-1624	34	66	mosaic	mosaic	ADJ
fcis-1624	34	67	data	data	NOUN
fcis-1624	34	68	enhancement	enhancement	NOUN
fcis-1624	34	69	by	by	ADP
fcis-1624	34	70	randomly	randomly	ADV
fcis-1624	34	71	using	use	VERB
fcis-1624	34	72	4	4	NUM
fcis-1624	34	73	images	image	NOUN
fcis-1624	34	74	,	,	PUNCT
fcis-1624	34	75	scaling	scale	VERB
fcis-1624	34	76	randomly	randomly	ADV
fcis-1624	34	77	,	,	PUNCT
fcis-1624	34	78	and	and	CCONJ
fcis-1624	34	79	then	then	ADV
fcis-1624	34	80	randomly	randomly	ADV
fcis-1624	34	81	distributing	distribute	VERB
fcis-1624	34	82	them	they	PRON
fcis-1624	34	83	for	for	ADP
fcis-1624	34	84	stitching	stitching	NOUN
fcis-1624	34	85	,	,	PUNCT
fcis-1624	34	86	the	the	DET
fcis-1624	34	87	detection	detection	NOUN
fcis-1624	34	88	dataset	dataset	VERB
fcis-1624	34	89	was	be	AUX
fcis-1624	34	90	greatly	greatly	ADV
fcis-1624	34	91	supplemented	supplement	VERB
fcis-1624	34	92	,	,	PUNCT
fcis-1624	34	93	especially	especially	ADV
fcis-1624	34	94	the	the	DET
fcis-1624	34	95	random	random	ADJ
fcis-1624	34	96	scaling	scaling	NOUN
fcis-1624	34	97	added	add	VERB
fcis-1624	34	98	many	many	ADJ
fcis-1624	34	99	small	small	ADJ
fcis-1624	34	100	targets	target	NOUN
fcis-1624	34	101	,	,	PUNCT
fcis-1624	34	102	which	which	PRON
fcis-1624	34	103	made	make	VERB
fcis-1624	34	104	the	the	DET
fcis-1624	34	105	network	network	NOUN
fcis-1624	34	106	richer	rich	ADJ
fcis-1624	34	107	.	.	PUNCT
fcis-1624	35	1	this	this	DET
fcis-1624	35	2	improvement	improvement	NOUN
fcis-1624	35	3	is	be	AUX
fcis-1624	35	4	very	very	ADV
fcis-1624	35	5	beneficial	beneficial	ADJ
fcis-1624	35	6	for	for	ADP
fcis-1624	35	7	enhancing	enhance	VERB
fcis-1624	35	8	the	the	DET
fcis-1624	35	9	detection	detection	NOUN
fcis-1624	35	10	of	of	ADP
fcis-1624	35	11	small	small	ADJ
fcis-1624	35	12	objects	object	NOUN
fcis-1624	35	13	.	.	PUNCT
fcis-1624	36	1	adaptive	adaptive	ADJ
fcis-1624	36	2	anchor	anchor	NOUN
fcis-1624	36	3	frame	frame	NOUN
fcis-1624	36	4	calculation	calculation	NOUN
fcis-1624	36	5	yolov5	yolov5	NOUN
fcis-1624	36	6	embeds	embed	VERB
fcis-1624	36	7	calculation	calculation	NOUN
fcis-1624	36	8	of	of	ADP
fcis-1624	36	9	initial	initial	ADJ
fcis-1624	36	10	anchor	anchor	NOUN
fcis-1624	36	11	frame	frame	NOUN
fcis-1624	36	12	value	value	NOUN
fcis-1624	36	13	into	into	ADP
fcis-1624	36	14	the	the	DET
fcis-1624	36	15	code	code	NOUN
fcis-1624	36	16	,	,	PUNCT
fcis-1624	36	17	the	the	DET
fcis-1624	36	18	optimal	optimal	ADJ
fcis-1624	36	19	anchor	anchor	NOUN
fcis-1624	36	20	frame	frame	NOUN
fcis-1624	36	21	values	value	NOUN
fcis-1624	36	22	in	in	ADP
fcis-1624	36	23	different	different	ADJ
fcis-1624	36	24	training	training	NOUN
fcis-1624	36	25	sets	set	NOUN
fcis-1624	36	26	are	be	AUX
fcis-1624	36	27	updated	update	VERB
fcis-1624	36	28	in	in	ADP
fcis-1624	36	29	each	each	DET
fcis-1624	36	30	training	training	NOUN
fcis-1624	36	31	session	session	NOUN
fcis-1624	36	32	with	with	ADP
fcis-1624	36	33	adaptive	adaptive	ADJ
fcis-1624	36	34	calculations	calculation	NOUN
fcis-1624	36	35	.	.	PUNCT
fcis-1624	37	1	2.2.2	2.2.2	X
fcis-1624	37	2	.	.	PUNCT
fcis-1624	37	3	backbone	backbone	NOUN
fcis-1624	37	4	focus	focus	NOUN
fcis-1624	37	5	structure	structure	NOUN
fcis-1624	37	6	figure	figure	NOUN
fcis-1624	37	7	2	2	NUM
fcis-1624	37	8	.	.	PUNCT
fcis-1624	38	1	the	the	DET
fcis-1624	38	2	focus	focus	NOUN
fcis-1624	38	3	structure	structure	NOUN
fcis-1624	38	4	figure	figure	NOUN
fcis-1624	38	5	3	3	NUM
fcis-1624	38	6	.	.	PUNCT
fcis-1624	38	7	slicing	slice	VERB
fcis-1624	38	8	operation	operation	NOUN
fcis-1624	38	9	in	in	ADP
fcis-1624	38	10	the	the	DET
fcis-1624	38	11	focus	focus	NOUN
fcis-1624	38	12	structure	structure	NOUN
fcis-1624	38	13	,	,	PUNCT
fcis-1624	38	14	input	input	VERB
fcis-1624	38	15	the	the	DET
fcis-1624	38	16	original	original	ADJ
fcis-1624	38	17	608×608×3	608×608×3	NUM
fcis-1624	38	18	image	image	NOUN
fcis-1624	38	19	to	to	ADP
fcis-1624	38	20	the	the	DET
fcis-1624	38	21	focus	focus	NOUN
fcis-1624	38	22	structure	structure	NOUN
fcis-1624	38	23	,	,	PUNCT
fcis-1624	38	24	use	use	VERB
fcis-1624	38	25	slicing	slice	VERB
fcis-1624	38	26	operation	operation	NOUN
fcis-1624	38	27	to	to	PART
fcis-1624	38	28	turn	turn	VERB
fcis-1624	38	29	it	it	PRON
fcis-1624	38	30	into	into	ADP
fcis-1624	38	31	a	a	DET
fcis-1624	38	32	304×304×12	304×304×12	NUM
fcis-1624	38	33	feature	feature	NOUN
fcis-1624	38	34	map	map	NOUN
fcis-1624	38	35	first	first	ADV
fcis-1624	38	36	,	,	PUNCT
fcis-1624	38	37	and	and	CCONJ
fcis-1624	38	38	then	then	ADV
fcis-1624	38	39	a	a	DET
fcis-1624	38	40	convolutional	convolutional	ADJ
fcis-1624	38	41	operation	operation	NOUN
fcis-1624	38	42	is	be	AUX
fcis-1624	38	43	then	then	ADV
fcis-1624	38	44	performed	perform	VERB
fcis-1624	38	45	,	,	PUNCT
fcis-1624	38	46	which	which	PRON
fcis-1624	38	47	has	have	VERB
fcis-1624	38	48	32	32	NUM
fcis-1624	38	49	convolutional	convolutional	ADJ
fcis-1624	38	50	kernels	kernel	NOUN
fcis-1624	38	51	,	,	PUNCT
fcis-1624	38	52	which	which	PRON
fcis-1624	38	53	are	be	AUX
fcis-1624	38	54	finally	finally	ADV
fcis-1624	38	55	converted	convert	VERB
fcis-1624	38	56	into	into	ADP
fcis-1624	38	57	304×304×32	304×304×32	NUM
fcis-1624	38	58	feature	feature	NOUN
fcis-1624	38	59	maps	map	NOUN
fcis-1624	38	60	.	.	PUNCT
fcis-1624	39	1	the	the	DET
fcis-1624	39	2	tile	tile	NOUN
fcis-1624	39	3	diagram	diagram	NOUN
fcis-1624	39	4	on	on	ADP
fcis-1624	39	5	the	the	DET
fcis-1624	39	6	right	right	NOUN
fcis-1624	39	7	shows	show	VERB
fcis-1624	39	8	a	a	DET
fcis-1624	39	9	4×4×3	4×4×3	NUM
fcis-1624	39	10	image	image	NOUN
fcis-1624	39	11	with	with	ADP
fcis-1624	39	12	tiles	tile	NOUN
fcis-1624	39	13	as	as	ADP
fcis-1624	39	14	2×2×12	2×2×12	NUM
fcis-1624	39	15	feature	feature	NOUN
fcis-1624	39	16	maps	map	NOUN
fcis-1624	39	17	.	.	PUNCT
fcis-1624	40	1	2.2.3	2.2.3	X
fcis-1624	40	2	.	.	X
fcis-1624	40	3	neck	neck	NOUN
fcis-1624	40	4	yolov5	yolov5	PROPN
fcis-1624	40	5	now	now	ADV
fcis-1624	40	6	uses	use	VERB
fcis-1624	40	7	the	the	DET
fcis-1624	40	8	same	same	ADJ
fcis-1624	40	9	structure	structure	NOUN
fcis-1624	40	10	of	of	ADP
fcis-1624	40	11	fpn+pan	fpn+pan	PROPN
fcis-1624	40	12	for	for	ADP
fcis-1624	40	13	neck	neck	NOUN
fcis-1624	40	14	as	as	ADP
fcis-1624	40	15	in	in	ADP
fcis-1624	40	16	yolov4	yolov4	PROPN
fcis-1624	40	17	.	.	PUNCT
fcis-1624	41	1	the	the	DET
fcis-1624	41	2	fpn	fpn	PROPN
fcis-1624	41	3	(	(	PUNCT
fcis-1624	41	4	feature	feature	NOUN
fcis-1624	41	5	pyramid	pyramid	NOUN
fcis-1624	41	6	network	network	NOUN
fcis-1624	41	7	)	)	PUNCT
fcis-1624	42	1	[	[	X
fcis-1624	42	2	7	7	X
fcis-1624	42	3	]	]	PUNCT
fcis-1624	42	4	is	be	AUX
fcis-1624	42	5	used	use	VERB
fcis-1624	42	6	as	as	ADP
fcis-1624	42	7	a	a	DET
fcis-1624	42	8	reference	reference	NOUN
fcis-1624	42	9	,	,	PUNCT
fcis-1624	42	10	and	and	CCONJ
fcis-1624	42	11	the	the	DET
fcis-1624	42	12	final	final	ADJ
fcis-1624	42	13	output	output	NOUN
fcis-1624	42	14	three	three	NUM
fcis-1624	42	15	feature	feature	NOUN
fcis-1624	42	16	maps	map	NOUN
fcis-1624	42	17	(	(	PUNCT
fcis-1624	42	18	19∗19	19∗19	NUM
fcis-1624	42	19	,	,	PUNCT
fcis-1624	42	20	38∗38	38∗38	NUM
fcis-1624	42	21	,	,	PUNCT
fcis-1624	42	22	76∗76	76∗76	NOUN
fcis-1624	42	23	)	)	PUNCT
fcis-1624	42	24	have	have	VERB
fcis-1624	42	25	different	different	ADJ
fcis-1624	42	26	scales	scale	NOUN
fcis-1624	42	27	by	by	ADP
fcis-1624	42	28	up	up	ADV
fcis-1624	42	29	-	-	PUNCT
fcis-1624	42	30	sampling	sample	VERB
fcis-1624	42	31	and	and	CCONJ
fcis-1624	42	32	feature	feature	NOUN
fcis-1624	42	33	fusion	fusion	NOUN
fcis-1624	42	34	.	.	PUNCT
fcis-1624	43	1	in	in	ADP
fcis-1624	43	2	order	order	NOUN
fcis-1624	43	3	to	to	PART
fcis-1624	43	4	improve	improve	VERB
fcis-1624	43	5	the	the	DET
fcis-1624	43	6	detection	detection	NOUN
fcis-1624	43	7	of	of	ADP
fcis-1624	43	8	large	large	ADJ
fcis-1624	43	9	and	and	CCONJ
fcis-1624	43	10	small	small	ADJ
fcis-1624	43	11	targets	target	NOUN
fcis-1624	43	12	,	,	PUNCT
fcis-1624	43	13	the	the	DET
fcis-1624	43	14	feature	feature	NOUN
fcis-1624	43	15	richness	richness	NOUN
fcis-1624	43	16	can	can	AUX
fcis-1624	43	17	be	be	AUX
fcis-1624	43	18	increased	increase	VERB
fcis-1624	43	19	through	through	ADP
fcis-1624	43	20	multi	multi	ADJ
fcis-1624	43	21	-	-	ADJ
fcis-1624	43	22	scale	scale	ADJ
fcis-1624	43	23	feature	feature	NOUN
fcis-1624	43	24	detection	detection	NOUN
fcis-1624	43	25	to	to	PART
fcis-1624	43	26	achieve	achieve	VERB
fcis-1624	43	27	the	the	DET
fcis-1624	43	28	purpose	purpose	NOUN
fcis-1624	43	29	of	of	ADP
fcis-1624	43	30	detection	detection	NOUN
fcis-1624	43	31	[	[	X
fcis-1624	43	32	8	8	NUM
fcis-1624	43	33	]	]	PUNCT
fcis-1624	43	34	.	.	PUNCT
fcis-1624	44	1	figure	figure	NOUN
fcis-1624	44	2	4	4	NUM
fcis-1624	44	3	.	.	PUNCT
fcis-1624	44	4	neck	neck	NOUN
fcis-1624	44	5	structure	structure	NOUN
fcis-1624	44	6	figure	figure	NOUN
fcis-1624	44	7	5	5	NUM
fcis-1624	44	8	.	.	PUNCT
fcis-1624	44	9	yolov5_neck	yolov5_neck	PUNCT
fcis-1624	45	1	yolov5	yolov5	PROPN
fcis-1624	45	2	's	's	PART
fcis-1624	45	3	neck	neck	NOUN
fcis-1624	45	4	structure	structure	NOUN
fcis-1624	45	5	is	be	AUX
fcis-1624	45	6	enhanced	enhance	VERB
fcis-1624	45	7	with	with	ADP
fcis-1624	45	8	a	a	DET
fcis-1624	45	9	csp2	csp2	NOUN
fcis-1624	45	10	structure	structure	NOUN
fcis-1624	45	11	borrowed	borrow	VERB
fcis-1624	45	12	from	from	ADP
fcis-1624	45	13	the	the	DET
fcis-1624	45	14	cspnet	cspnet	NOUN
fcis-1624	45	15	design	design	NOUN
fcis-1624	45	16	for	for	ADP
fcis-1624	45	17	network	network	NOUN
fcis-1624	45	18	feature	feature	NOUN
fcis-1624	45	19	fusion	fusion	NOUN
fcis-1624	45	20	.	.	PUNCT
fcis-1624	46	1	2.2.4	2.2.4	NUM
fcis-1624	46	2	.	.	X
fcis-1624	46	3	outputs	output	NOUN
fcis-1624	46	4	bounding	bound	VERB
fcis-1624	46	5	box	box	NOUN
fcis-1624	46	6	loss	loss	NOUN
fcis-1624	46	7	function	function	NOUN
fcis-1624	46	8	ciou_loss	ciou_loss	PROPN
fcis-1624	46	9	is	be	AUX
fcis-1624	46	10	used	use	VERB
fcis-1624	46	11	as	as	ADP
fcis-1624	46	12	the	the	DET
fcis-1624	46	13	loss	loss	NOUN
fcis-1624	46	14	function	function	NOUN
fcis-1624	46	15	of	of	ADP
fcis-1624	46	16	the	the	DET
fcis-1624	46	17	bounding	bounding	NOUN
fcis-1624	46	18	box	box	NOUN
fcis-1624	46	19	in	in	ADP
fcis-1624	46	20	yolov5	yolov5	NOUN
fcis-1624	46	21	:	:	PUNCT
fcis-1624	46	22	2	2	NUM
fcis-1624	46	23	2	2	NUM
fcis-1624	46	24	2	2	NUM
fcis-1624	46	25	tan	tan	NOUN
fcis-1624	46	26	_	_	NOUN
fcis-1624	47	1	2	2	NUM
fcis-1624	47	2	_	_	SYM
fcis-1624	47	3	1	1	NUM
fcis-1624	47	4	1	1	NUM
fcis-1624	47	5	(	(	PUNCT
fcis-1624	47	6	)	)	PUNCT
fcis-1624	47	7	tan	tan	PROPN
fcis-1624	47	8	_	_	PUNCT
fcis-1624	47	9	(	(	PUNCT
fcis-1624	47	10	1	1	X
fcis-1624	47	11	)	)	PUNCT
fcis-1624	47	12	dis	dis	PROPN
fcis-1624	47	13	ce	ce	PROPN
fcis-1624	47	14	v	v	NUM
fcis-1624	47	15	ciou	ciou	NOUN
fcis-1624	47	16	loss	loss	NOUN
fcis-1624	47	17	ciou	ciou	NOUN
fcis-1624	47	18	iou	iou	PROPN
fcis-1624	47	19	dis	dis	PROPN
fcis-1624	47	20	ce	ce	PROPN
fcis-1624	47	21	c	c	PROPN
fcis-1624	47	22	iou	iou	PROPN
fcis-1624	47	23	v	v	NOUN
fcis-1624	47	24	=	=	SYM
fcis-1624	47	25	=	=	PUNCT
fcis-1624	47	26	−	−	PROPN
fcis-1624	47	27	−	−	NOUN
fcis-1624	47	28	−	−	PROPN
fcis-1624	47	29	−	−	PROPN
fcis-1624	48	1	+	+	CCONJ
fcis-1624	48	2	where	where	SCONJ
fcis-1624	48	3	v	v	NOUN
fcis-1624	48	4	is	be	AUX
fcis-1624	48	5	a	a	DET
fcis-1624	48	6	parameter	parameter	NOUN
fcis-1624	48	7	measuring	measure	VERB
fcis-1624	48	8	the	the	DET
fcis-1624	48	9	consistency	consistency	NOUN
fcis-1624	48	10	of	of	ADP
fcis-1624	48	11	the	the	DET
fcis-1624	48	12	aspect	aspect	NOUN
fcis-1624	48	13	ratio	ratio	NOUN
fcis-1624	48	14	,	,	PUNCT
fcis-1624	48	15	which	which	PRON
fcis-1624	48	16	we	we	PRON
fcis-1624	48	17	can	can	AUX
fcis-1624	48	18	also	also	ADV
fcis-1624	48	19	define	define	VERB
fcis-1624	48	20	as	as	ADP
fcis-1624	48	21	:	:	SYM
fcis-1624	48	22	2	2	NUM
fcis-1624	48	23	2	2	NUM
fcis-1624	48	24	4	4	NUM
fcis-1624	48	25	(	(	PUNCT
fcis-1624	48	26	arctan	arctan	PROPN
fcis-1624	48	27	arctan	arctan	PROPN
fcis-1624	48	28	)	)	PUNCT
fcis-1624	49	1	gt	gt	PROPN
fcis-1624	50	1	p	p	NOUN
fcis-1624	50	2	gt	gt	PROPN
fcis-1624	50	3	p	p	X
fcis-1624	50	4	w	w	PROPN
fcis-1624	50	5	w	w	PROPN
fcis-1624	50	6	v	v	PROPN
fcis-1624	50	7	h	h	NOUN
fcis-1624	50	8	h	h	PROPN
fcis-1624	50	9	=	=	PUNCT
fcis-1624	51	1	−	−	PROPN
fcis-1624	51	2	the	the	DET
fcis-1624	51	3	introduction	introduction	NOUN
fcis-1624	51	4	of	of	ADP
fcis-1624	51	5	this	this	DET
fcis-1624	51	6	influence	influence	NOUN
fcis-1624	51	7	factor	factor	NOUN
fcis-1624	51	8	v	v	NOUN
fcis-1624	51	9	takes	take	VERB
fcis-1624	51	10	the	the	DET
fcis-1624	51	11	aspect	aspect	NOUN
fcis-1624	51	12	ratio	ratio	NOUN
fcis-1624	51	13	of	of	ADP
fcis-1624	51	14	both	both	CCONJ
fcis-1624	51	15	the	the	DET
fcis-1624	51	16	prediction	prediction	NOUN
fcis-1624	51	17	box	box	NOUN
fcis-1624	51	18	and	and	CCONJ
fcis-1624	51	19	the	the	DET
fcis-1624	51	20	target	target	NOUN
fcis-1624	51	21	box	box	NOUN
fcis-1624	51	22	into	into	ADP
fcis-1624	51	23	account	account	NOUN
fcis-1624	51	24	.	.	PUNCT
fcis-1624	52	1	therefore	therefore	ADV
fcis-1624	52	2	,	,	PUNCT
fcis-1624	52	3	ciou	ciou	NOUN
fcis-1624	52	4	_	_	NOUN
fcis-1624	52	5	loss	loss	NOUN
fcis-1624	52	6	takes	take	VERB
fcis-1624	52	7	into	into	ADP
fcis-1624	52	8	account	account	NOUN
fcis-1624	52	9	three	three	NUM
fcis-1624	52	10	important	important	ADJ
fcis-1624	52	11	geometric	geometric	ADJ
fcis-1624	52	12	factors	factor	NOUN
fcis-1624	52	13	that	that	PRON
fcis-1624	52	14	should	should	AUX
fcis-1624	52	15	be	be	AUX
fcis-1624	52	16	considered	consider	VERB
fcis-1624	52	17	in	in	ADP
fcis-1624	52	18	the	the	DET
fcis-1624	52	19	target	target	NOUN
fcis-1624	52	20	frame	frame	NOUN
fcis-1624	52	21	regression	regression	NOUN
fcis-1624	52	22	function	function	NOUN
fcis-1624	52	23	:	:	PUNCT
fcis-1624	52	24	overlap	overlap	NOUN
fcis-1624	52	25	area	area	NOUN
fcis-1624	52	26	,	,	PUNCT
fcis-1624	52	27	centroid	centroid	NOUN
fcis-1624	52	28	distance	distance	NOUN
fcis-1624	52	29	,	,	PUNCT
fcis-1624	52	30	and	and	CCONJ
fcis-1624	52	31	aspect	aspect	NOUN
fcis-1624	52	32	ratio	ratio	NOUN
fcis-1624	52	33	.	.	PUNCT
fcis-1624	53	1	this	this	PRON
fcis-1624	53	2	makes	make	VERB
fcis-1624	53	3	the	the	DET
fcis-1624	53	4	regression	regression	NOUN
fcis-1624	53	5	of	of	ADP
fcis-1624	53	6	the	the	DET
fcis-1624	53	7	prediction	prediction	NOUN
fcis-1624	53	8	frame	frame	NOUN
fcis-1624	53	9	faster	fast	ADV
fcis-1624	53	10	and	and	CCONJ
fcis-1624	53	11	more	more	ADV
fcis-1624	53	12	accurate	accurate	ADJ
fcis-1624	53	13	.	.	PUNCT
fcis-1624	54	1	3	3	X
fcis-1624	54	2	.	.	X
fcis-1624	54	3	experimental	experimental	ADJ
fcis-1624	54	4	results	result	NOUN
fcis-1624	54	5	and	and	CCONJ
fcis-1624	54	6	analysis	analysis	NOUN
fcis-1624	54	7	3.1	3.1	NUM
fcis-1624	54	8	.	.	PUNCT
fcis-1624	54	9	dataset	dataset	VERB
fcis-1624	54	10	for	for	ADP
fcis-1624	54	11	deep	deep	ADJ
fcis-1624	54	12	learning	learning	NOUN
fcis-1624	54	13	,	,	PUNCT
fcis-1624	54	14	the	the	DET
fcis-1624	54	15	final	final	ADJ
fcis-1624	54	16	detection	detection	NOUN
fcis-1624	54	17	effect	effect	NOUN
fcis-1624	54	18	is	be	AUX
fcis-1624	54	19	closely	closely	ADV
fcis-1624	54	20	related	relate	VERB
fcis-1624	54	21	to	to	ADP
fcis-1624	54	22	the	the	DET
fcis-1624	54	23	quality	quality	NOUN
fcis-1624	54	24	of	of	ADP
fcis-1624	54	25	the	the	DET
fcis-1624	54	26	dataset	dataset	NOUN
fcis-1624	54	27	.	.	PUNCT
fcis-1624	55	1	a	a	DET
fcis-1624	55	2	prerequisite	prerequisite	NOUN
fcis-1624	55	3	for	for	SCONJ
fcis-1624	55	4	the	the	DET
fcis-1624	55	5	network	network	NOUN
fcis-1624	55	6	to	to	PART
fcis-1624	55	7	be	be	AUX
fcis-1624	55	8	able	able	ADJ
fcis-1624	55	9	to	to	PART
fcis-1624	55	10	fully	fully	ADV
fcis-1624	55	11	understand	understand	VERB
fcis-1624	55	12	the	the	DET
fcis-1624	55	13	features	feature	NOUN
fcis-1624	55	14	of	of	ADP
fcis-1624	55	15	the	the	DET
fcis-1624	55	16	detection	detection	NOUN
fcis-1624	55	17	target	target	NOUN
fcis-1624	55	18	is	be	AUX
fcis-1624	55	19	to	to	PART
fcis-1624	55	20	have	have	VERB
fcis-1624	55	21	a	a	DET
fcis-1624	55	22	sufficient	sufficient	ADJ
fcis-1624	55	23	number	number	NOUN
fcis-1624	55	24	of	of	ADP
fcis-1624	55	25	samples	sample	NOUN
fcis-1624	55	26	for	for	SCONJ
fcis-1624	55	27	the	the	DET
fcis-1624	55	28	network	network	NOUN
fcis-1624	55	29	to	to	PART
fcis-1624	55	30	learn	learn	VERB
fcis-1624	55	31	.	.	PUNCT
fcis-1624	56	1	the	the	DET
fcis-1624	56	2	dataset	dataset	NOUN
fcis-1624	56	3	used	use	VERB
fcis-1624	56	4	in	in	ADP
fcis-1624	56	5	this	this	DET
fcis-1624	56	6	experiment	experiment	NOUN
fcis-1624	56	7	is	be	AUX
fcis-1624	56	8	made	make	VERB
fcis-1624	56	9	from	from	ADP
fcis-1624	56	10	open	open	ADJ
fcis-1624	56	11	datasets	dataset	NOUN
fcis-1624	56	12	available	available	ADJ
fcis-1624	56	13	on	on	ADP
fcis-1624	56	14	the	the	DET
fcis-1624	56	15	internet	internet	NOUN
fcis-1624	56	16	.	.	PUNCT
fcis-1624	57	1	in	in	ADP
fcis-1624	57	2	the	the	DET
fcis-1624	57	3	model	model	NOUN
fcis-1624	57	4	training	training	NOUN
fcis-1624	57	5	for	for	ADP
fcis-1624	57	6	this	this	DET
fcis-1624	57	7	experiment	experiment	NOUN
fcis-1624	57	8	,	,	PUNCT
fcis-1624	57	9	it	it	PRON
fcis-1624	57	10	is	be	AUX
fcis-1624	57	11	planned	plan	VERB
fcis-1624	57	12	to	to	PART
fcis-1624	57	13	divide	divide	VERB
fcis-1624	57	14	the	the	DET
fcis-1624	57	15	dataset	dataset	NOUN
fcis-1624	57	16	into	into	ADP
fcis-1624	57	17	training	training	NOUN
fcis-1624	57	18	sets	set	NOUN
fcis-1624	57	19	and	and	CCONJ
fcis-1624	57	20	validation	validation	NOUN
fcis-1624	57	21	sets	set	NOUN
fcis-1624	57	22	in	in	ADP
fcis-1624	57	23	an	an	DET
fcis-1624	57	24	8:2	8:2	NUM
fcis-1624	57	25	ratio	ratio	NOUN
fcis-1624	57	26	.	.	PUNCT
fcis-1624	58	1	3	3	NUM
fcis-1624	58	2	the	the	DET
fcis-1624	58	3	xml	xml	NOUN
fcis-1624	58	4	-	-	PUNCT
fcis-1624	58	5	formatted	format	VERB
fcis-1624	58	6	label	label	NOUN
fcis-1624	58	7	file	file	NOUN
fcis-1624	58	8	is	be	AUX
fcis-1624	58	9	converted	convert	VERB
fcis-1624	58	10	to	to	ADP
fcis-1624	58	11	a	a	DET
fcis-1624	58	12	txt	txt	NOUN
fcis-1624	58	13	-	-	PUNCT
fcis-1624	58	14	formatted	format	VERB
fcis-1624	58	15	label	label	NOUN
fcis-1624	58	16	file	file	NOUN
fcis-1624	58	17	.	.	PUNCT
fcis-1624	59	1	3.2	3.2	NUM
fcis-1624	59	2	.	.	PUNCT
fcis-1624	59	3	model	model	NOUN
fcis-1624	59	4	training	training	NOUN
fcis-1624	59	5	in	in	ADP
fcis-1624	59	6	this	this	DET
fcis-1624	59	7	experiment	experiment	NOUN
fcis-1624	59	8	,	,	PUNCT
fcis-1624	59	9	the	the	DET
fcis-1624	59	10	training	training	NOUN
fcis-1624	59	11	environment	environment	NOUN
fcis-1624	59	12	settings	setting	NOUN
fcis-1624	59	13	are	be	AUX
fcis-1624	59	14	:	:	PUNCT
fcis-1624	59	15	python	python	PROPN
fcis-1624	59	16	programming	programming	NOUN
fcis-1624	59	17	,	,	PUNCT
fcis-1624	59	18	tensorflow	tensorflow	ADJ
fcis-1624	59	19	open	open	ADJ
fcis-1624	59	20	-	-	PUNCT
fcis-1624	59	21	source	source	NOUN
fcis-1624	59	22	framework	framework	NOUN
fcis-1624	59	23	1.14	1.14	NUM
fcis-1624	59	24	,	,	PUNCT
fcis-1624	59	25	keras	keras	PROPN
fcis-1624	59	26	open	open	ADJ
fcis-1624	59	27	-	-	PUNCT
fcis-1624	59	28	source	source	NOUN
fcis-1624	59	29	framework	framework	NOUN
fcis-1624	59	30	2.2.4	2.2.4	NUM
fcis-1624	59	31	,	,	PUNCT
fcis-1624	59	32	cuda10.1	cuda10.1	PROPN
fcis-1624	59	33	cudnn7.4	cudnn7.4	NUM
fcis-1624	59	34	,	,	PUNCT
fcis-1624	59	35	video	video	NOUN
fcis-1624	59	36	card	card	NOUN
fcis-1624	59	37	2070s	2070s	NUM
fcis-1624	59	38	.	.	PUNCT
fcis-1624	60	1	the	the	DET
fcis-1624	60	2	training	training	NOUN
fcis-1624	60	3	of	of	ADP
fcis-1624	60	4	model	model	NOUN
fcis-1624	60	5	:	:	PUNCT
fcis-1624	61	1	that	that	PRON
fcis-1624	61	2	is	be	AUX
fcis-1624	61	3	,	,	PUNCT
fcis-1624	61	4	fitting	fit	VERB
fcis-1624	61	5	the	the	DET
fcis-1624	61	6	network	network	NOUN
fcis-1624	61	7	parameters	parameter	NOUN
fcis-1624	61	8	in	in	ADP
fcis-1624	61	9	the	the	DET
fcis-1624	61	10	model	model	NOUN
fcis-1624	61	11	to	to	PART
fcis-1624	61	12	make	make	VERB
fcis-1624	61	13	the	the	DET
fcis-1624	61	14	predicted	predict	VERB
fcis-1624	61	15	values	value	NOUN
fcis-1624	61	16	more	more	ADV
fcis-1624	61	17	and	and	CCONJ
fcis-1624	61	18	more	more	ADV
fcis-1624	61	19	accurate	accurate	ADJ
fcis-1624	61	20	.	.	PUNCT
fcis-1624	62	1	this	this	DET
fcis-1624	62	2	experiment	experiment	NOUN
fcis-1624	62	3	uses	use	VERB
fcis-1624	62	4	the	the	DET
fcis-1624	62	5	loss	loss	NOUN
fcis-1624	62	6	function	function	NOUN
fcis-1624	62	7	to	to	PART
fcis-1624	62	8	calculate	calculate	VERB
fcis-1624	62	9	the	the	DET
fcis-1624	62	10	error	error	NOUN
fcis-1624	62	11	of	of	ADP
fcis-1624	62	12	the	the	DET
fcis-1624	62	13	network	network	NOUN
fcis-1624	62	14	.	.	PUNCT
fcis-1624	63	1	then	then	ADV
fcis-1624	63	2	the	the	DET
fcis-1624	63	3	obtained	obtain	VERB
fcis-1624	63	4	error	error	NOUN
fcis-1624	63	5	inversion	inversion	NOUN
fcis-1624	63	6	can	can	AUX
fcis-1624	63	7	adjust	adjust	VERB
fcis-1624	63	8	the	the	DET
fcis-1624	63	9	internal	internal	ADJ
fcis-1624	63	10	network	network	NOUN
fcis-1624	63	11	parameters	parameter	NOUN
fcis-1624	63	12	,	,	PUNCT
fcis-1624	63	13	and	and	CCONJ
fcis-1624	63	14	the	the	DET
fcis-1624	63	15	adjusted	adjust	VERB
fcis-1624	63	16	internal	internal	ADJ
fcis-1624	63	17	network	network	NOUN
fcis-1624	63	18	parameters	parameter	NOUN
fcis-1624	63	19	make	make	VERB
fcis-1624	63	20	the	the	DET
fcis-1624	63	21	error	error	NOUN
fcis-1624	63	22	smaller	small	ADJ
fcis-1624	63	23	and	and	CCONJ
fcis-1624	63	24	smaller	small	ADJ
fcis-1624	63	25	.	.	PUNCT
fcis-1624	64	1	the	the	DET
fcis-1624	64	2	above	above	ADJ
fcis-1624	64	3	process	process	NOUN
fcis-1624	64	4	uses	use	VERB
fcis-1624	64	5	pre	pre	ADJ
fcis-1624	64	6	-	-	VERB
fcis-1624	64	7	trained	trained	ADJ
fcis-1624	64	8	weights	weight	NOUN
fcis-1624	64	9	downloaded	download	VERB
fcis-1624	64	10	from	from	ADP
fcis-1624	64	11	the	the	DET
fcis-1624	64	12	internet	internet	NOUN
fcis-1624	64	13	yolov5s.pt	yolov5s.pt	NOUN
fcis-1624	64	14	with	with	ADP
fcis-1624	64	15	training	training	NOUN
fcis-1624	64	16	epochs	epoch	NOUN
fcis-1624	64	17	of	of	ADP
fcis-1624	64	18	50	50	NUM
fcis-1624	64	19	.	.	PUNCT
fcis-1624	65	1	the	the	DET
fcis-1624	65	2	training	training	NOUN
fcis-1624	65	3	process	process	NOUN
fcis-1624	65	4	and	and	CCONJ
fcis-1624	65	5	internal	internal	ADJ
fcis-1624	65	6	parameters	parameter	NOUN
fcis-1624	65	7	can	can	AUX
fcis-1624	65	8	be	be	AUX
fcis-1624	65	9	viewed	view	VERB
fcis-1624	65	10	through	through	ADP
fcis-1624	65	11	the	the	DET
fcis-1624	65	12	visualization	visualization	NOUN
fcis-1624	65	13	tool	tool	NOUN
fcis-1624	65	14	tensor	tensor	NOUN
fcis-1624	65	15	board	board	NOUN
fcis-1624	65	16	,	,	PUNCT
fcis-1624	65	17	as	as	SCONJ
fcis-1624	65	18	shown	show	VERB
fcis-1624	65	19	below	below	ADV
fcis-1624	65	20	.	.	PUNCT
fcis-1624	66	1	figure	figure	VERB
fcis-1624	66	2	6	6	NUM
fcis-1624	66	3	.	.	PUNCT
fcis-1624	67	1	training	train	VERB
fcis-1624	67	2	loss	loss	NOUN
fcis-1624	67	3	curves	curve	NOUN
fcis-1624	67	4	3.3	3.3	NUM
fcis-1624	67	5	.	.	PUNCT
fcis-1624	68	1	evaluation	evaluation	NOUN
fcis-1624	68	2	metrics	metric	NOUN
fcis-1624	68	3	the	the	DET
fcis-1624	68	4	evaluation	evaluation	NOUN
fcis-1624	68	5	metric	metric	NOUN
fcis-1624	68	6	chosen	choose	VERB
fcis-1624	68	7	for	for	ADP
fcis-1624	68	8	this	this	DET
fcis-1624	68	9	experiment	experiment	NOUN
fcis-1624	68	10	is	be	AUX
fcis-1624	68	11	the	the	DET
fcis-1624	68	12	average	average	ADJ
fcis-1624	68	13	accuracy	accuracy	NOUN
fcis-1624	68	14	,	,	PUNCT
fcis-1624	68	15	that	that	ADV
fcis-1624	68	16	is	is	ADV
fcis-1624	68	17	,	,	PUNCT
fcis-1624	68	18	the	the	DET
fcis-1624	68	19	map	map	NOUN
fcis-1624	68	20	index	index	NOUN
fcis-1624	68	21	.	.	PUNCT
fcis-1624	69	1	the	the	DET
fcis-1624	69	2	average	average	ADJ
fcis-1624	69	3	(	(	PUNCT
fcis-1624	69	4	average	average	ADJ
fcis-1624	69	5	accuracy	accuracy	NOUN
fcis-1624	69	6	)	)	PUNCT
fcis-1624	69	7	of	of	ADP
fcis-1624	69	8	the	the	DET
fcis-1624	69	9	ap	ap	PROPN
fcis-1624	69	10	for	for	ADP
fcis-1624	69	11	all	all	DET
fcis-1624	69	12	object	object	NOUN
fcis-1624	69	13	classes	class	NOUN
fcis-1624	69	14	is	be	AUX
fcis-1624	69	15	called	call	VERB
fcis-1624	69	16	the	the	DET
fcis-1624	69	17	map	map	NOUN
fcis-1624	69	18	metric	metric	ADJ
fcis-1624	69	19	,	,	PUNCT
fcis-1624	69	20	and	and	CCONJ
fcis-1624	69	21	different	different	ADJ
fcis-1624	69	22	recall	recall	NOUN
fcis-1624	69	23	values	value	NOUN
fcis-1624	69	24	correspond	correspond	VERB
fcis-1624	69	25	to	to	ADP
fcis-1624	69	26	the	the	DET
fcis-1624	69	27	aps	ap	NOUN
fcis-1624	69	28	of	of	ADP
fcis-1624	69	29	different	different	ADJ
fcis-1624	69	30	object	object	NOUN
fcis-1624	69	31	classes	class	NOUN
fcis-1624	69	32	.	.	PUNCT
fcis-1624	70	1	these	these	DET
fcis-1624	70	2	recall	recall	NOUN
fcis-1624	70	3	values	value	NOUN
fcis-1624	70	4	are	be	AUX
fcis-1624	70	5	equal	equal	ADJ
fcis-1624	70	6	to	to	ADP
fcis-1624	70	7	or	or	CCONJ
fcis-1624	70	8	less	less	ADJ
fcis-1624	70	9	than	than	ADP
fcis-1624	70	10	the	the	DET
fcis-1624	70	11	values	value	NOUN
fcis-1624	70	12	of	of	ADP
fcis-1624	70	13	maximum	maximum	ADJ
fcis-1624	70	14	precision	precision	NOUN
fcis-1624	70	15	,	,	PUNCT
fcis-1624	70	16	whereby	whereby	SCONJ
fcis-1624	70	17	the	the	DET
fcis-1624	70	18	value	value	NOUN
fcis-1624	70	19	of	of	ADP
fcis-1624	70	20	maximum	maximum	ADJ
fcis-1624	70	21	precision	precision	NOUN
fcis-1624	70	22	is	be	AUX
fcis-1624	70	23	obtained	obtain	VERB
fcis-1624	70	24	and	and	CCONJ
fcis-1624	70	25	used	use	VERB
fcis-1624	70	26	as	as	ADP
fcis-1624	70	27	an	an	DET
fcis-1624	70	28	independent	independent	ADJ
fcis-1624	70	29	variable	variable	NOUN
fcis-1624	70	30	,	,	PUNCT
fcis-1624	70	31	and	and	CCONJ
fcis-1624	70	32	then	then	ADV
fcis-1624	70	33	the	the	DET
fcis-1624	70	34	area	area	NOUN
fcis-1624	70	35	under	under	ADP
fcis-1624	70	36	the	the	DET
fcis-1624	70	37	curve	curve	NOUN
fcis-1624	70	38	is	be	AUX
fcis-1624	70	39	plotted	plot	VERB
fcis-1624	70	40	as	as	ADP
fcis-1624	70	41	the	the	DET
fcis-1624	70	42	ap	ap	PROPN
fcis-1624	70	43	for	for	ADP
fcis-1624	70	44	that	that	DET
fcis-1624	70	45	target	target	NOUN
fcis-1624	70	46	,	,	PUNCT
fcis-1624	70	47	as	as	SCONJ
fcis-1624	70	48	shown	show	VERB
fcis-1624	70	49	in	in	ADP
fcis-1624	70	50	figure	figure	NOUN
fcis-1624	70	51	7	7	NUM
fcis-1624	70	52	.	.	PUNCT
fcis-1624	71	1	[	[	X
fcis-1624	71	2	9	9	NUM
fcis-1624	71	3	]	]	PUNCT
fcis-1624	71	4	the	the	DET
fcis-1624	71	5	miss	miss	ADJ
fcis-1624	71	6	rate	rate	NOUN
fcis-1624	71	7	of	of	ADP
fcis-1624	71	8	identified	identify	VERB
fcis-1624	71	9	objects	object	NOUN
fcis-1624	71	10	is	be	AUX
fcis-1624	71	11	called	call	VERB
fcis-1624	71	12	recall	recall	NOUN
fcis-1624	71	13	and	and	CCONJ
fcis-1624	71	14	the	the	DET
fcis-1624	71	15	accuracy	accuracy	NOUN
fcis-1624	71	16	rate	rate	NOUN
fcis-1624	71	17	of	of	ADP
fcis-1624	71	18	identified	identify	VERB
fcis-1624	71	19	objects	object	NOUN
fcis-1624	71	20	is	be	AUX
fcis-1624	71	21	called	call	VERB
fcis-1624	71	22	accuracy	accuracy	NOUN
fcis-1624	71	23	.	.	PUNCT
fcis-1624	72	1	the	the	DET
fcis-1624	72	2	calculation	calculation	NOUN
fcis-1624	72	3	formula	formula	NOUN
fcis-1624	72	4	is	be	AUX
fcis-1624	72	5	shown	show	VERB
fcis-1624	72	6	below	below	ADP
fcis-1624	72	7	(	(	PUNCT
fcis-1624	72	8	1	1	NUM
fcis-1624	72	9	)	)	PUNCT
fcis-1624	72	10	and	and	CCONJ
fcis-1624	72	11	(	(	PUNCT
fcis-1624	72	12	2	2	NUM
fcis-1624	72	13	)	)	PUNCT
fcis-1624	72	14	.	.	PUNCT
fcis-1624	73	1	number	number	NOUN
fcis-1624	73	2	of	of	ADP
fcis-1624	73	3	properly	properly	ADV
fcis-1624	73	4	identified	identify	VERB
fcis-1624	73	5	object	object	NOUN
fcis-1624	73	6	classes	class	NOUN
fcis-1624	73	7	–	–	PUNCT
fcis-1624	73	8	true	true	ADJ
fcis-1624	73	9	positives	positive	NOUN
fcis-1624	73	10	,	,	PUNCT
fcis-1624	73	11	number	number	NOUN
fcis-1624	73	12	of	of	ADP
fcis-1624	73	13	misidentified	misidentifie	VERB
fcis-1624	73	14	object	object	NOUN
fcis-1624	73	15	classes	class	NOUN
fcis-1624	73	16	–	–	PUNCT
fcis-1624	73	17	false	false	ADJ
fcis-1624	73	18	positives	positive	NOUN
fcis-1624	73	19	,	,	PUNCT
fcis-1624	73	20	number	number	NOUN
fcis-1624	73	21	of	of	ADP
fcis-1624	73	22	misidentified	misidentified	ADJ
fcis-1624	73	23	or	or	CCONJ
fcis-1624	73	24	identified	identify	VERB
fcis-1624	73	25	target	target	NOUN
fcis-1624	73	26	classes	class	NOUN
fcis-1624	73	27	–	–	PUNCT
fcis-1624	73	28	false	false	ADJ
fcis-1624	73	29	negatives	negative	NOUN
fcis-1624	73	30	.	.	PUNCT
fcis-1624	74	1	(	(	PUNCT
fcis-1624	74	2	1	1	X
fcis-1624	74	3	)	)	PUNCT
fcis-1624	74	4	(	(	PUNCT
fcis-1624	74	5	2	2	X
fcis-1624	74	6	)	)	PUNCT
fcis-1624	74	7	truepositives	truepositive	NOUN
fcis-1624	74	8	recall	recall	VERB
fcis-1624	74	9	truepositives	truepositive	NOUN
fcis-1624	74	10	falsenegatives	falsenegative	VERB
fcis-1624	74	11	truepositives	truepositive	NOUN
fcis-1624	74	12	precision	precision	NOUN
fcis-1624	74	13	truepositives	truepositive	NOUN
fcis-1624	74	14	falsepositives	falsepositive	VERB
fcis-1624	74	15	=	=	PUNCT
fcis-1624	75	1	+	+	PUNCT
fcis-1624	75	2	=	=	SYM
fcis-1624	76	1	+	+	NUM
fcis-1624	76	2	figure	figure	NOUN
fcis-1624	76	3	7	7	NUM
fcis-1624	76	4	.	.	PUNCT
fcis-1624	76	5	ap	ap	PROPN
fcis-1624	76	6	calculation	calculation	PROPN
fcis-1624	76	7	diagram	diagram	NOUN
fcis-1624	76	8	figure	figure	NOUN
fcis-1624	76	9	8	8	NUM
fcis-1624	76	10	.	.	PUNCT
fcis-1624	76	11	comparison	comparison	NOUN
fcis-1624	76	12	of	of	ADP
fcis-1624	76	13	common	common	ADJ
fcis-1624	76	14	detection	detection	NOUN
fcis-1624	76	15	algorithms	algorithm	VERB
fcis-1624	76	16	3.4	3.4	NUM
fcis-1624	76	17	.	.	PUNCT
fcis-1624	77	1	detection	detection	NOUN
fcis-1624	77	2	results	result	NOUN
fcis-1624	77	3	currently	currently	ADV
fcis-1624	77	4	,	,	PUNCT
fcis-1624	77	5	a	a	DET
fcis-1624	77	6	comparison	comparison	NOUN
fcis-1624	77	7	of	of	ADP
fcis-1624	77	8	commonly	commonly	ADV
fcis-1624	77	9	used	use	VERB
fcis-1624	77	10	object	object	NOUN
fcis-1624	77	11	detection	detection	NOUN
fcis-1624	77	12	algorithms	algorithm	NOUN
fcis-1624	77	13	is	be	AUX
fcis-1624	77	14	shown	show	VERB
fcis-1624	77	15	in	in	ADP
fcis-1624	77	16	figure	figure	NOUN
fcis-1624	77	17	8	8	NUM
fcis-1624	78	1	[	[	X
fcis-1624	78	2	10	10	NUM
fcis-1624	78	3	]	]	PUNCT
fcis-1624	78	4	.	.	PUNCT
fcis-1624	79	1	the	the	DET
fcis-1624	79	2	detection	detection	NOUN
fcis-1624	79	3	results	result	NOUN
fcis-1624	79	4	of	of	ADP
fcis-1624	79	5	the	the	DET
fcis-1624	79	6	dataset	dataset	NOUN
fcis-1624	79	7	show	show	NOUN
fcis-1624	79	8	that	that	SCONJ
fcis-1624	79	9	the	the	DET
fcis-1624	79	10	map-50	map-50	ADJ
fcis-1624	79	11	value	value	NOUN
fcis-1624	79	12	of	of	ADP
fcis-1624	79	13	yolov3	yolov3	PROPN
fcis-1624	79	14	algorithm	algorithm	PROPN
fcis-1624	79	15	is	be	AUX
fcis-1624	79	16	about	about	ADV
fcis-1624	79	17	55	55	NUM
fcis-1624	79	18	,	,	PUNCT
fcis-1624	79	19	while	while	SCONJ
fcis-1624	79	20	the	the	DET
fcis-1624	79	21	yolov5	yolov5	NOUN
fcis-1624	79	22	algorithm	algorithm	NOUN
fcis-1624	79	23	has	have	VERB
fcis-1624	79	24	a	a	DET
fcis-1624	79	25	value	value	NOUN
fcis-1624	79	26	of	of	ADP
fcis-1624	79	27	map-50	map-50	NOUN
fcis-1624	79	28	around	around	ADV
fcis-1624	79	29	60	60	NUM
fcis-1624	79	30	as	as	SCONJ
fcis-1624	79	31	obtained	obtain	VERB
fcis-1624	79	32	by	by	ADP
fcis-1624	79	33	figure	figure	NOUN
fcis-1624	79	34	5	5	NUM
fcis-1624	79	35	above	above	ADV
fcis-1624	79	36	.	.	PUNCT
fcis-1624	80	1	this	this	PRON
fcis-1624	80	2	indicates	indicate	VERB
fcis-1624	80	3	that	that	SCONJ
fcis-1624	80	4	the	the	DET
fcis-1624	80	5	yolov5	yolov5	NOUN
fcis-1624	80	6	algorithm	algorithm	NOUN
fcis-1624	80	7	has	have	VERB
fcis-1624	80	8	a	a	DET
fcis-1624	80	9	higher	high	ADJ
fcis-1624	80	10	map	map	NOUN
fcis-1624	80	11	.	.	PUNCT
fcis-1624	81	1	there	there	PRON
fcis-1624	81	2	are	be	VERB
fcis-1624	81	3	two	two	NUM
fcis-1624	81	4	detection	detection	NOUN
fcis-1624	81	5	results	result	NOUN
fcis-1624	81	6	set	set	VERB
fcis-1624	81	7	up	up	ADP
fcis-1624	81	8	here	here	ADV
fcis-1624	81	9	:	:	PUNCT
fcis-1624	81	10	the	the	DET
fcis-1624	81	11	up	up	ADJ
fcis-1624	81	12	class	class	NOUN
fcis-1624	81	13	and	and	CCONJ
fcis-1624	81	14	the	the	DET
fcis-1624	81	15	down	down	ADJ
fcis-1624	81	16	class	class	NOUN
fcis-1624	81	17	.	.	PUNCT
fcis-1624	82	1	figures	figure	NOUN
fcis-1624	82	2	9	9	NUM
fcis-1624	82	3	and	and	CCONJ
fcis-1624	82	4	10	10	NUM
fcis-1624	82	5	are	be	AUX
fcis-1624	82	6	the	the	DET
fcis-1624	82	7	results	result	NOUN
fcis-1624	82	8	of	of	ADP
fcis-1624	82	9	the	the	DET
fcis-1624	82	10	corresponding	corresponding	ADJ
fcis-1624	82	11	test	test	NOUN
fcis-1624	82	12	dataset	dataset	NOUN
fcis-1624	82	13	.	.	PUNCT
fcis-1624	83	1	figure	figure	NOUN
fcis-1624	83	2	9	9	NUM
fcis-1624	83	3	.	.	PUNCT
fcis-1624	84	1	no	no	DET
fcis-1624	84	2	drop	drop	NOUN
fcis-1624	84	3	test	test	NOUN
fcis-1624	84	4	results	result	NOUN
fcis-1624	84	5	figure	figure	VERB
fcis-1624	84	6	10	10	NUM
fcis-1624	84	7	.	.	PUNCT
fcis-1624	85	1	drop	drop	NOUN
fcis-1624	85	2	test	test	NOUN
fcis-1624	85	3	results	result	VERB
fcis-1624	85	4	the	the	DET
fcis-1624	85	5	ap	ap	PROPN
fcis-1624	85	6	for	for	ADP
fcis-1624	85	7	the	the	DET
fcis-1624	85	8	upward	upward	ADJ
fcis-1624	85	9	class	class	NOUN
fcis-1624	85	10	is	be	AUX
fcis-1624	85	11	0.980	0.980	NUM
fcis-1624	85	12	,	,	PUNCT
fcis-1624	85	13	the	the	DET
fcis-1624	85	14	downward	downward	ADJ
fcis-1624	85	15	class	class	NOUN
fcis-1624	85	16	of	of	ADP
fcis-1624	85	17	ap	ap	PROPN
fcis-1624	85	18	is	be	AUX
fcis-1624	85	19	0.969	0.969	NUM
fcis-1624	85	20	,	,	PUNCT
fcis-1624	85	21	while	while	SCONJ
fcis-1624	85	22	the	the	DET
fcis-1624	85	23	generated	generate	VERB
fcis-1624	85	24	map	map	NOUN
fcis-1624	85	25	is	be	AUX
fcis-1624	85	26	0.0.975	0.0.975	PROPN
fcis-1624	85	27	.	.	PUNCT
fcis-1624	86	1	the	the	DET
fcis-1624	86	2	detection	detection	NOUN
fcis-1624	86	3	results	result	NOUN
fcis-1624	86	4	obtained	obtain	VERB
fcis-1624	86	5	from	from	ADP
fcis-1624	86	6	this	this	DET
fcis-1624	86	7	network	network	NOUN
fcis-1624	86	8	model	model	NOUN
fcis-1624	86	9	are	be	AUX
fcis-1624	86	10	compared	compare	VERB
fcis-1624	86	11	with	with	ADP
fcis-1624	86	12	the	the	DET
fcis-1624	86	13	detection	detection	NOUN
fcis-1624	86	14	results	result	NOUN
fcis-1624	86	15	of	of	ADP
fcis-1624	86	16	other	other	ADJ
fcis-1624	86	17	algorithms	algorithm	NOUN
fcis-1624	86	18	on	on	ADP
fcis-1624	86	19	the	the	DET
fcis-1624	86	20	voc	voc	NOUN
fcis-1624	86	21	dataset	dataset	NOUN
fcis-1624	86	22	,	,	PUNCT
fcis-1624	86	23	it	it	PRON
fcis-1624	86	24	can	can	AUX
fcis-1624	86	25	be	be	AUX
fcis-1624	86	26	seen	see	VERB
fcis-1624	86	27	that	that	SCONJ
fcis-1624	86	28	our	our	PRON
fcis-1624	86	29	algorithm	algorithm	NOUN
fcis-1624	86	30	has	have	AUX
fcis-1624	86	31	further	far	ADV
fcis-1624	86	32	improved	improve	VERB
fcis-1624	86	33	in	in	ADP
fcis-1624	86	34	terms	term	NOUN
fcis-1624	86	35	of	of	ADP
fcis-1624	86	36	accuracy	accuracy	NOUN
fcis-1624	86	37	.	.	PUNCT
fcis-1624	87	1	4	4	X
fcis-1624	87	2	.	.	X
fcis-1624	87	3	conclusion	conclusion	NOUN
fcis-1624	87	4	this	this	DET
fcis-1624	87	5	experiment	experiment	NOUN
fcis-1624	87	6	introduced	introduce	VERB
fcis-1624	87	7	a	a	DET
fcis-1624	87	8	fall	fall	NOUN
fcis-1624	87	9	detection	detection	NOUN
fcis-1624	87	10	method	method	NOUN
fcis-1624	87	11	which	which	PRON
fcis-1624	87	12	4	4	NUM
fcis-1624	87	13	is	be	AUX
fcis-1624	87	14	derived	derive	VERB
fcis-1624	87	15	from	from	ADP
fcis-1624	87	16	deep	deep	ADJ
fcis-1624	87	17	learning	learning	NOUN
fcis-1624	87	18	,	,	PUNCT
fcis-1624	87	19	and	and	CCONJ
fcis-1624	87	20	this	this	DET
fcis-1624	87	21	experiment	experiment	NOUN
fcis-1624	87	22	updates	update	VERB
fcis-1624	87	23	the	the	DET
fcis-1624	87	24	algorithm	algorithm	NOUN
fcis-1624	87	25	in	in	ADP
fcis-1624	87	26	the	the	DET
fcis-1624	87	27	previous	previous	ADJ
fcis-1624	87	28	yolov3	yolov3	PROPN
fcis-1624	87	29	fall	fall	NOUN
fcis-1624	87	30	detection	detection	NOUN
fcis-1624	87	31	method	method	NOUN
fcis-1624	87	32	to	to	ADP
fcis-1624	87	33	yolov5s	yolov5s	PROPN
fcis-1624	87	34	model	model	NOUN
fcis-1624	87	35	,	,	PUNCT
fcis-1624	87	36	and	and	CCONJ
fcis-1624	87	37	the	the	DET
fcis-1624	87	38	fall	fall	NOUN
fcis-1624	87	39	detection	detection	NOUN
fcis-1624	87	40	proceed	proceed	VERB
fcis-1624	87	41	with	with	ADP
fcis-1624	87	42	the	the	DET
fcis-1624	87	43	latest	late	ADJ
fcis-1624	87	44	yolov5	yolov5	NOUN
fcis-1624	87	45	network	network	NOUN
fcis-1624	87	46	.	.	PUNCT
fcis-1624	88	1	the	the	DET
fcis-1624	88	2	final	final	ADJ
fcis-1624	88	3	results	result	NOUN
fcis-1624	88	4	can	can	AUX
fcis-1624	88	5	confirm	confirm	VERB
fcis-1624	88	6	that	that	SCONJ
fcis-1624	88	7	the	the	DET
fcis-1624	88	8	method	method	NOUN
fcis-1624	88	9	introduced	introduce	VERB
fcis-1624	88	10	in	in	ADP
fcis-1624	88	11	this	this	DET
fcis-1624	88	12	experiment	experiment	NOUN
fcis-1624	88	13	makes	make	VERB
fcis-1624	88	14	the	the	DET
fcis-1624	88	15	fall	fall	NOUN
fcis-1624	88	16	detection	detection	NOUN
fcis-1624	88	17	more	more	ADV
fcis-1624	88	18	accurate	accurate	ADJ
fcis-1624	88	19	.	.	PUNCT
fcis-1624	89	1	in	in	ADP
fcis-1624	89	2	the	the	DET
fcis-1624	89	3	next	next	ADJ
fcis-1624	89	4	study	study	NOUN
fcis-1624	89	5	research	research	NOUN
fcis-1624	89	6	,	,	PUNCT
fcis-1624	89	7	i	i	PRON
fcis-1624	89	8	will	will	AUX
fcis-1624	89	9	work	work	VERB
fcis-1624	89	10	on	on	ADP
fcis-1624	89	11	applying	apply	VERB
fcis-1624	89	12	semi	semi	ADJ
fcis-1624	89	13	-	-	ADJ
fcis-1624	89	14	supervised	supervised	ADJ
fcis-1624	89	15	deep	deep	ADJ
fcis-1624	89	16	learning	learning	NOUN
fcis-1624	89	17	networks	network	NOUN
fcis-1624	89	18	to	to	ADP
fcis-1624	89	19	the	the	DET
fcis-1624	89	20	study	study	NOUN
fcis-1624	89	21	of	of	ADP
fcis-1624	89	22	fall	fall	NOUN
fcis-1624	89	23	detection	detection	NOUN
fcis-1624	89	24	in	in	ADP
fcis-1624	89	25	order	order	NOUN
fcis-1624	89	26	to	to	PART
fcis-1624	89	27	use	use	VERB
fcis-1624	89	28	unlabeled	unlabeled	ADJ
fcis-1624	89	29	images	image	NOUN
fcis-1624	89	30	and	and	CCONJ
fcis-1624	89	31	videos	video	NOUN
fcis-1624	89	32	comprehensively	comprehensively	ADV
fcis-1624	89	33	.	.	PUNCT
fcis-1624	90	1	references	reference	NOUN
fcis-1624	90	2	[	[	X
fcis-1624	90	3	1	1	X
fcis-1624	90	4	]	]	PUNCT
fcis-1624	90	5	t.	t.	PROPN
fcis-1624	90	6	xu	xu	PROPN
fcis-1624	90	7	,	,	PUNCT
fcis-1624	90	8	j.	j.	PROPN
fcis-1624	90	9	chen	chen	PROPN
fcis-1624	90	10	,	,	PUNCT
fcis-1624	90	11	z.	z.	PROPN
fcis-1624	90	12	li	li	PROPN
fcis-1624	90	13	and	and	CCONJ
fcis-1624	90	14	y.	y.	PROPN
fcis-1624	90	15	cai	cai	PROPN
fcis-1624	90	16	(	(	PUNCT
fcis-1624	90	17	2021	2021	NUM
fcis-1624	90	18	)	)	PUNCT
fcis-1624	90	19	fall	fall	NOUN
fcis-1624	90	20	detection	detection	NOUN
fcis-1624	90	21	based	base	VERB
fcis-1624	90	22	on	on	ADP
fcis-1624	90	23	person	person	NOUN
fcis-1624	90	24	detection	detection	NOUN
fcis-1624	90	25	and	and	CCONJ
fcis-1624	90	26	multi	multi	ADJ
fcis-1624	90	27	-	-	NOUN
fcis-1624	90	28	target	target	ADJ
fcis-1624	90	29	tracking	tracking	NOUN
fcis-1624	90	30	in	in	ADP
fcis-1624	90	31	:	:	PUNCT
fcis-1624	90	32	2021	2021	NUM
fcis-1624	90	33	11th	11th	ADJ
fcis-1624	90	34	international	international	ADJ
fcis-1624	90	35	conference	conference	NOUN
fcis-1624	90	36	on	on	ADP
fcis-1624	90	37	information	information	NOUN
fcis-1624	90	38	technology	technology	NOUN
fcis-1624	90	39	in	in	ADP
fcis-1624	90	40	medicine	medicine	NOUN
fcis-1624	90	41	and	and	CCONJ
fcis-1624	90	42	education	education	NOUN
fcis-1624	90	43	(	(	PUNCT
fcis-1624	90	44	itme	itme	NOUN
fcis-1624	90	45	)	)	PUNCT
fcis-1624	90	46	,	,	PUNCT
fcis-1624	91	1	pp	pp	ADJ
fcis-1624	91	2	.	.	PUNCT
fcis-1624	92	1	60	60	NUM
fcis-1624	92	2	-	-	SYM
fcis-1624	92	3	65	65	NUM
fcis-1624	93	1	[	[	X
fcis-1624	93	2	2	2	NUM
fcis-1624	93	3	]	]	PUNCT
fcis-1624	93	4	w	w	PROPN
fcis-1624	93	5	j	j	PROPN
fcis-1624	93	6	choi	choi	NOUN
fcis-1624	93	7	,	,	PUNCT
fcis-1624	93	8	k	k	PROPN
fcis-1624	93	9	lim	lim	PROPN
fcis-1624	93	10	,	,	PUNCT
fcis-1624	93	11	s	s	PROPN
fcis-1624	93	12	kim	kim	PROPN
fcis-1624	93	13	et	et	PROPN
fcis-1624	93	14	al	al	PROPN
fcis-1624	93	15	.	.	PROPN
fcis-1624	94	1	(	(	PUNCT
fcis-1624	94	2	2021	2021	NUM
fcis-1624	94	3	)	)	PUNCT
fcis-1624	94	4	science	science	NOUN
fcis-1624	94	5	of	of	ADP
fcis-1624	94	6	falling	fall	VERB
fcis-1624	94	7	and	and	CCONJ
fcis-1624	94	8	injury	injury	NOUN
fcis-1624	94	9	in	in	ADP
fcis-1624	94	10	older	old	ADJ
fcis-1624	94	11	adults	adult	NOUN
fcis-1624	94	12	-	-	PUNCT
fcis-1624	94	13	do	do	AUX
fcis-1624	94	14	all	all	PRON
fcis-1624	94	15	falls	fall	VERB
fcis-1624	94	16	lead	lead	VERB
fcis-1624	94	17	to	to	ADP
fcis-1624	94	18	death	death	NOUN
fcis-1624	94	19	?	?	PUNCT
fcis-1624	95	1	literature	literature	PROPN
fcis-1624	95	2	review	review	PROPN
fcis-1624	95	3	.	.	PUNCT
fcis-1624	96	1	j.	j.	PROPN
fcis-1624	96	2	physical	physical	PROPN
fcis-1624	96	3	therapy	therapy	PROPN
fcis-1624	96	4	korea	korea	PROPN
fcis-1624	96	5	,	,	PUNCT
fcis-1624	96	6	vol	vol	NOUN
fcis-1624	96	7	.	.	PROPN
fcis-1624	96	8	28	28	NUM
fcis-1624	96	9	,	,	PUNCT
fcis-1624	96	10	no	no	INTJ
fcis-1624	96	11	.	.	NOUN
fcis-1624	96	12	3	3	NUM
fcis-1624	96	13	,	,	PUNCT
fcis-1624	96	14	pp	pp	ADJ
fcis-1624	96	15	.	.	PUNCT
fcis-1624	97	1	161	161	NUM
fcis-1624	97	2	-	-	SYM
fcis-1624	97	3	167	167	NUM
fcis-1624	97	4	.	.	PUNCT
fcis-1624	98	1	[	[	X
fcis-1624	98	2	3	3	NUM
fcis-1624	98	3	]	]	X
fcis-1624	98	4	h	h	NOUN
fcis-1624	98	5	ramirez	ramirez	PROPN
fcis-1624	98	6	,	,	PUNCT
fcis-1624	98	7	s	s	VERB
fcis-1624	98	8	a	a	DET
fcis-1624	98	9	velastin	velastin	ADJ
fcis-1624	98	10	,	,	PUNCT
fcis-1624	98	11	i	i	PRON
fcis-1624	98	12	meza	meza	VERB
fcis-1624	98	13	et	et	PROPN
fcis-1624	98	14	al	al	PROPN
fcis-1624	98	15	.	.	PROPN
fcis-1624	99	1	(	(	PUNCT
fcis-1624	99	2	2021	2021	NUM
fcis-1624	99	3	)	)	PUNCT
fcis-1624	99	4	fall	fall	NOUN
fcis-1624	99	5	detection	detection	NOUN
fcis-1624	99	6	and	and	CCONJ
fcis-1624	99	7	activity	activity	NOUN
fcis-1624	99	8	recognition	recognition	NOUN
fcis-1624	99	9	using	use	VERB
fcis-1624	99	10	human	human	ADJ
fcis-1624	99	11	skeleton	skeleton	NOUN
fcis-1624	99	12	features	feature	NOUN
fcis-1624	99	13	.	.	PUNCT
fcis-1624	100	1	j.	j.	PROPN
fcis-1624	100	2	ieee	ieee	PROPN
fcis-1624	100	3	access	access	PROPN
fcis-1624	100	4	,	,	PUNCT
fcis-1624	100	5	vol	vol	NOUN
fcis-1624	100	6	.	.	NOUN
fcis-1624	100	7	9	9	NUM
fcis-1624	100	8	,	,	PUNCT
fcis-1624	100	9	pp	pp	ADJ
fcis-1624	100	10	.	.	PUNCT
fcis-1624	100	11	33532	33532	NUM
fcis-1624	100	12	-	-	SYM
fcis-1624	100	13	33542	33542	NUM
fcis-1624	100	14	.	.	PUNCT
fcis-1624	101	1	[	[	X
fcis-1624	101	2	4	4	X
fcis-1624	101	3	]	]	X
fcis-1624	101	4	m	m	PROPN
fcis-1624	101	5	s	s	PROPN
fcis-1624	101	6	islam	islam	PROPN
fcis-1624	101	7	,	,	PUNCT
fcis-1624	101	8	h	h	PROPN
fcis-1624	101	9	shahriar	shahriar	PROPN
fcis-1624	101	10	,	,	PUNCT
fcis-1624	101	11	s	s	PART
fcis-1624	101	12	sneha	sneha	PROPN
fcis-1624	101	13	et	et	PROPN
fcis-1624	101	14	al	al	PROPN
fcis-1624	101	15	.	.	PROPN
fcis-1624	102	1	(	(	PUNCT
fcis-1624	102	2	2020	2020	NUM
fcis-1624	102	3	)	)	PUNCT
fcis-1624	102	4	mobile	mobile	NOUN
fcis-1624	102	5	sensorbased	sensorbase	VERB
fcis-1624	102	6	fall	fall	NOUN
fcis-1624	102	7	detection	detection	NOUN
fcis-1624	102	8	framework	framework	NOUN
fcis-1624	102	9	.	.	PUNCT
fcis-1624	103	1	in	in	ADP
fcis-1624	103	2	:	:	PUNCT
fcis-1624	103	3	proceedings	proceeding	NOUN
fcis-1624	103	4	of	of	ADP
fcis-1624	103	5	2020	2020	NUM
fcis-1624	103	6	ieee	ieee	PROPN
fcis-1624	103	7	44th	44th	ADJ
fcis-1624	103	8	annual	annual	ADJ
fcis-1624	103	9	computers	computer	NOUN
fcis-1624	103	10	software	software	NOUN
fcis-1624	103	11	and	and	CCONJ
fcis-1624	103	12	applications	application	NOUN
fcis-1624	103	13	conference	conference	NOUN
fcis-1624	103	14	(	(	PUNCT
fcis-1624	103	15	compsac	compsac	NOUN
fcis-1624	103	16	)	)	PUNCT
fcis-1624	103	17	,	,	PUNCT
fcis-1624	103	18	pp	pp	PROPN
fcis-1624	103	19	.	.	PUNCT
fcis-1624	103	20	693	693	NUM
fcis-1624	103	21	-	-	NUM
fcis-1624	103	22	698	698	NUM
fcis-1624	103	23	.	.	PUNCT
fcis-1624	104	1	[	[	X
fcis-1624	104	2	5	5	NUM
fcis-1624	104	3	]	]	PUNCT
fcis-1624	104	4	a	a	DET
fcis-1624	104	5	núñez	núñez	PROPN
fcis-1624	104	6	-	-	PUNCT
fcis-1624	104	7	marcos	marcos	PROPN
fcis-1624	104	8	,	,	PUNCT
fcis-1624	104	9	g	g	PROPN
fcis-1624	104	10	azkune	azkune	NOUN
fcis-1624	104	11	and	and	CCONJ
fcis-1624	104	12	i.	i.	PROPN
fcis-1624	104	13	arganda	arganda	PROPN
fcis-1624	104	14	-	-	PUNCT
fcis-1624	104	15	carreras	carreras	PROPN
fcis-1624	104	16	(	(	PUNCT
fcis-1624	104	17	2017	2017	NUM
fcis-1624	104	18	)	)	PUNCT
fcis-1624	104	19	vision	vision	NOUN
fcis-1624	104	20	-	-	PUNCT
fcis-1624	104	21	based	base	VERB
fcis-1624	104	22	fall	fall	NOUN
fcis-1624	104	23	detection	detection	NOUN
fcis-1624	104	24	with	with	ADP
fcis-1624	104	25	convolutional	convolutional	ADJ
fcis-1624	104	26	neural	neural	ADJ
fcis-1624	104	27	networks	network	NOUN
fcis-1624	104	28	.	.	PUNCT
fcis-1624	105	1	j.	j.	PROPN
fcis-1624	105	2	wireless	wireless	PROPN
fcis-1624	105	3	communications	communication	NOUN
fcis-1624	105	4	and	and	CCONJ
fcis-1624	105	5	mobile	mobile	NOUN
fcis-1624	105	6	computing	computing	NOUN
fcis-1624	105	7	,	,	PUNCT
fcis-1624	105	8	vol	vol	NOUN
fcis-1624	105	9	.	.	PROPN
fcis-1624	105	10	2017	2017	NUM
fcis-1624	105	11	,	,	PUNCT
fcis-1624	105	12	pp	pp	ADJ
fcis-1624	105	13	.	.	PUNCT
fcis-1624	106	1	1	1	NUM
fcis-1624	106	2	-	-	SYM
fcis-1624	106	3	16	16	NUM
fcis-1624	106	4	.	.	PUNCT
fcis-1624	107	1	[	[	X
fcis-1624	107	2	6	6	NUM
fcis-1624	107	3	]	]	X
fcis-1624	107	4	k	k	PROPN
fcis-1624	107	5	chaccour	chaccour	PROPN
fcis-1624	107	6	,	,	PUNCT
fcis-1624	107	7	r	r	NOUN
fcis-1624	107	8	darazi	darazi	NOUN
fcis-1624	107	9	,	,	PUNCT
fcis-1624	107	10	a	a	DET
fcis-1624	107	11	h	h	NOUN
fcis-1624	107	12	el	el	PROPN
fcis-1624	107	13	hassani	hassani	PROPN
fcis-1624	107	14	et	et	PROPN
fcis-1624	107	15	al	al	PROPN
fcis-1624	107	16	.	.	PROPN
fcis-1624	107	17	(	(	PUNCT
fcis-1624	107	18	2016	2016	NUM
fcis-1624	107	19	)	)	PUNCT
fcis-1624	107	20	from	from	ADP
fcis-1624	107	21	fall	fall	NOUN
fcis-1624	107	22	detection	detection	NOUN
fcis-1624	107	23	to	to	PART
fcis-1624	107	24	fall	fall	VERB
fcis-1624	107	25	prevention	prevention	NOUN
fcis-1624	107	26	:	:	PUNCT
fcis-1624	107	27	a	a	DET
fcis-1624	107	28	generic	generic	ADJ
fcis-1624	107	29	classification	classification	NOUN
fcis-1624	107	30	of	of	ADP
fcis-1624	107	31	fallrelated	fallrelated	ADJ
fcis-1624	107	32	systems	system	NOUN
fcis-1624	107	33	.	.	PUNCT
fcis-1624	108	1	j.	j.	PROPN
fcis-1624	108	2	ieee	ieee	PROPN
fcis-1624	108	3	sensors	sensors	PROPN
fcis-1624	108	4	journal	journal	PROPN
fcis-1624	108	5	,	,	PUNCT
fcis-1624	108	6	vol	vol	NOUN
fcis-1624	108	7	.	.	PROPN
fcis-1624	108	8	17	17	NUM
fcis-1624	108	9	,	,	PUNCT
fcis-1624	108	10	no	no	INTJ
fcis-1624	108	11	.	.	NOUN
fcis-1624	108	12	3	3	NUM
fcis-1624	108	13	,	,	PUNCT
fcis-1624	108	14	pp	pp	ADJ
fcis-1624	108	15	.	.	PUNCT
fcis-1624	109	1	812	812	NUM
fcis-1624	109	2	-	-	PUNCT
fcis-1624	109	3	822	822	NUM
fcis-1624	109	4	.	.	PUNCT
fcis-1624	110	1	[	[	X
fcis-1624	110	2	7	7	NUM
fcis-1624	110	3	]	]	X
fcis-1624	110	4	tsung	tsung	PROPN
fcis-1624	110	5	-	-	PUNCT
fcis-1624	110	6	yi	yi	PROPN
fcis-1624	110	7	lin	lin	PROPN
fcis-1624	110	8	,	,	PUNCT
fcis-1624	110	9	piotr	piotr	PROPN
fcis-1624	110	10	dollár	dollár	PROPN
fcis-1624	110	11	,	,	PUNCT
fcis-1624	110	12	ross	ross	PROPN
fcis-1624	110	13	girshick	girshick	PROPN
fcis-1624	110	14	,	,	PUNCT
fcis-1624	110	15	kaiming	kaime	VERB
fcis-1624	110	16	he	he	PRON
fcis-1624	110	17	and	and	CCONJ
fcis-1624	110	18	bharath	bharath	PROPN
fcis-1624	110	19	hariharan	hariharan	PROPN
fcis-1624	110	20	(	(	PUNCT
fcis-1624	110	21	2017	2017	NUM
fcis-1624	110	22	)	)	PUNCT
fcis-1624	110	23	feature	feature	NOUN
fcis-1624	110	24	pyramid	pyramid	NOUN
fcis-1624	110	25	networks	network	NOUN
fcis-1624	110	26	for	for	ADP
fcis-1624	110	27	object	object	NOUN
fcis-1624	110	28	detection	detection	NOUN
fcis-1624	110	29	.	.	PUNCT
fcis-1624	111	1	in	in	ADP
fcis-1624	111	2	:	:	PUNCT
fcis-1624	111	3	proceedings	proceeding	NOUN
fcis-1624	111	4	of	of	ADP
fcis-1624	111	5	the	the	DET
fcis-1624	111	6	ieee	ieee	NOUN
fcis-1624	111	7	conference	conference	NOUN
fcis-1624	111	8	on	on	ADP
fcis-1624	111	9	computer	computer	NOUN
fcis-1624	111	10	vision	vision	NOUN
fcis-1624	111	11	and	and	CCONJ
fcis-1624	111	12	pattern	pattern	NOUN
fcis-1624	111	13	recognition	recognition	NOUN
fcis-1624	111	14	,	,	PUNCT
fcis-1624	111	15	pp	pp	PROPN
fcis-1624	111	16	.	.	PUNCT
fcis-1624	112	1	936	936	NUM
fcis-1624	112	2	-	-	SYM
fcis-1624	112	3	944	944	NUM
fcis-1624	112	4	.	.	PUNCT
fcis-1624	113	1	[	[	X
fcis-1624	113	2	8	8	NUM
fcis-1624	113	3	]	]	PUNCT
fcis-1624	113	4	zhihua	zhihua	PROPN
fcis-1624	113	5	zhou	zhou	PROPN
fcis-1624	113	6	(	(	PUNCT
fcis-1624	113	7	2016	2016	NUM
fcis-1624	113	8	)	)	PUNCT
fcis-1624	113	9	machine	machine	NOUN
fcis-1624	113	10	learning	learning	NOUN
fcis-1624	113	11	.	.	PUNCT
fcis-1624	114	1	in	in	ADP
fcis-1624	114	2	:	:	PUNCT
fcis-1624	114	3	tsinghua	tsinghua	PROPN
fcis-1624	114	4	university	university	PROPN
fcis-1624	114	5	press	press	NOUN
fcis-1624	114	6	.	.	PUNCT
fcis-1624	115	1	beijing	beijing	PROPN
fcis-1624	115	2	.	.	PUNCT
fcis-1624	116	1	[	[	X
fcis-1624	116	2	9	9	NUM
fcis-1624	116	3	]	]	PUNCT
fcis-1624	116	4	x.	x.	NOUN
fcis-1624	116	5	wang	wang	PROPN
fcis-1624	116	6	and	and	CCONJ
fcis-1624	116	7	k.	k.	PROPN
fcis-1624	116	8	jia	jia	PROPN
fcis-1624	116	9	(	(	PUNCT
fcis-1624	116	10	2020	2020	NUM
fcis-1624	116	11	)	)	PUNCT
fcis-1624	116	12	human	human	ADJ
fcis-1624	116	13	fall	fall	NOUN
fcis-1624	116	14	detection	detection	NOUN
fcis-1624	116	15	algorithm	algorithm	NOUN
fcis-1624	116	16	based	base	VERB
fcis-1624	116	17	on	on	ADP
fcis-1624	116	18	yolov3	yolov3	PROPN
fcis-1624	116	19	.	.	PUNCT
fcis-1624	117	1	in	in	ADP
fcis-1624	117	2	:	:	PUNCT
fcis-1624	117	3	2020	2020	NUM
fcis-1624	117	4	ieee	ieee	NOUN
fcis-1624	117	5	5th	5th	ADJ
fcis-1624	117	6	international	international	ADJ
fcis-1624	117	7	conference	conference	NOUN
fcis-1624	117	8	on	on	ADP
fcis-1624	117	9	image	image	NOUN
fcis-1624	117	10	,	,	PUNCT
fcis-1624	117	11	vision	vision	NOUN
fcis-1624	117	12	and	and	CCONJ
fcis-1624	117	13	computing	computing	NOUN
fcis-1624	117	14	(	(	PUNCT
fcis-1624	117	15	icivc	icivc	NOUN
fcis-1624	117	16	)	)	PUNCT
fcis-1624	117	17	.	.	PUNCT
fcis-1624	118	1	pp	pp	ADV
fcis-1624	118	2	.	.	PUNCT
fcis-1624	119	1	5054	5054	NUM
fcis-1624	119	2	.	.	PUNCT
fcis-1624	120	1	[	[	X
fcis-1624	120	2	10	10	NUM
fcis-1624	120	3	]	]	X
fcis-1624	120	4	e	e	X
fcis-1624	120	5	casilari	casilari	NOUN
fcis-1624	120	6	,	,	PUNCT
fcis-1624	120	7	r	r	NOUN
fcis-1624	120	8	lora	lora	PROPN
fcis-1624	120	9	-	-	PUNCT
fcis-1624	120	10	rivera	rivera	PROPN
fcis-1624	120	11	and	and	CCONJ
fcis-1624	120	12	f.	f.	PROPN
fcis-1624	120	13	garcía	garcía	PROPN
fcis-1624	120	14	-	-	PUNCT
fcis-1624	120	15	lagos	lago	NOUN
fcis-1624	120	16	(	(	PUNCT
fcis-1624	120	17	2020	2020	NUM
fcis-1624	120	18	)	)	PUNCT
fcis-1624	120	19	a	a	DET
fcis-1624	120	20	study	study	NOUN
fcis-1624	120	21	on	on	ADP
fcis-1624	120	22	the	the	DET
fcis-1624	120	23	application	application	NOUN
fcis-1624	120	24	of	of	ADP
fcis-1624	120	25	convolutional	convolutional	ADJ
fcis-1624	120	26	neural	neural	ADJ
fcis-1624	120	27	networks	network	NOUN
fcis-1624	120	28	to	to	PART
fcis-1624	120	29	fall	fall	VERB
fcis-1624	120	30	detection	detection	NOUN
fcis-1624	120	31	evaluated	evaluate	VERB
fcis-1624	120	32	with	with	ADP
fcis-1624	120	33	multiple	multiple	ADJ
fcis-1624	120	34	public	public	ADJ
fcis-1624	120	35	datasets	dataset	NOUN
fcis-1624	120	36	.	.	PUNCT
fcis-1624	121	1	j.	j.	PROPN
fcis-1624	121	2	sensors	sensors	PROPN
fcis-1624	121	3	,	,	PUNCT
fcis-1624	121	4	vol	vol	NOUN
fcis-1624	121	5	.	.	PROPN
fcis-1624	121	6	20	20	NUM
fcis-1624	121	7	,	,	PUNCT
fcis-1624	121	8	no	no	INTJ
fcis-1624	121	9	.	.	NOUN
fcis-1624	121	10	5	5	NUM
fcis-1624	121	11	,	,	PUNCT
fcis-1624	121	12	pp	pp	ADJ
fcis-1624	121	13	.	.	PUNCT
fcis-1624	121	14	1466	1466	NUM
fcis-1624	121	15	.	.	PUNCT