id	sid	tid	token	lemma	pos
cuesj-1010	1	1	tx_1	tx_1	PROPN
cuesj-1010	1	2	~	~	X
cuesj-1010	1	3	abs	ab	NOUN
cuesj-1010	1	4	:	:	PUNCT
cuesj-1010	1	5	at	at	ADP
cuesj-1010	1	6	/	/	SYM
cuesj-1010	1	7	add	add	VERB
cuesj-1010	1	8	:	:	PUNCT
cuesj-1010	1	9	tx_2	tx_2	PROPN
cuesj-1010	1	10	~	~	NOUN
cuesj-1010	1	11	abs	ab	NOUN
cuesj-1010	1	12	:	:	PUNCT
cuesj-1010	1	13	at	at	ADP
cuesj-1010	1	14	1	1	NUM
cuesj-1010	1	15	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	1	16	cuesj	cuesj	NOUN
cuesj-1010	1	17	2024	2024	NUM
cuesj-1010	1	18	,	,	PUNCT
cuesj-1010	2	1	8	8	NUM
cuesj-1010	2	2	(	(	PUNCT
cuesj-1010	2	3	1	1	NUM
cuesj-1010	2	4	):	):	PUNCT
cuesj-1010	2	5	1	1	NUM
cuesj-1010	2	6	-	-	SYM
cuesj-1010	2	7	7	7	NUM
cuesj-1010	2	8	research	research	NOUN
cuesj-1010	2	9	article	article	NOUN
cuesj-1010	2	10	real	real	ADJ
cuesj-1010	2	11	-	-	PUNCT
cuesj-1010	2	12	time	time	NOUN
cuesj-1010	2	13	vehicles	vehicle	NOUN
cuesj-1010	2	14	detection	detection	NOUN
cuesj-1010	2	15	using	use	VERB
cuesj-1010	2	16	a	a	DET
cuesj-1010	2	17	tello	tello	NOUN
cuesj-1010	2	18	drone	drone	NOUN
cuesj-1010	2	19	with	with	ADP
cuesj-1010	2	20	yolov5	yolov5	NOUN
cuesj-1010	2	21	algorithm	algorithm	NOUN
cuesj-1010	2	22	shalaw	shalaw	NOUN
cuesj-1010	2	23	m.	m.	PROPN
cuesj-1010	2	24	abdallah	abdallah	PROPN
cuesj-1010	2	25	department	department	PROPN
cuesj-1010	2	26	of	of	ADP
cuesj-1010	2	27	software	software	PROPN
cuesj-1010	2	28	engineering	engineering	PROPN
cuesj-1010	2	29	,	,	PUNCT
cuesj-1010	2	30	koya	koya	PROPN
cuesj-1010	2	31	university	university	PROPN
cuesj-1010	2	32	,	,	PUNCT
cuesj-1010	2	33	kurdistan	kurdistan	PROPN
cuesj-1010	2	34	region	region	NOUN
cuesj-1010	2	35	,	,	PUNCT
cuesj-1010	2	36	iraq	iraq	PROPN
cuesj-1010	2	37	abstract	abstract	PROPN
cuesj-1010	2	38	incorporating	incorporate	VERB
cuesj-1010	2	39	unmanned	unmanned	ADJ
cuesj-1010	2	40	aerial	aerial	ADJ
cuesj-1010	2	41	vehicles	vehicle	NOUN
cuesj-1010	2	42	(	(	PUNCT
cuesj-1010	2	43	uavs	uavs	NOUN
cuesj-1010	2	44	)	)	PUNCT
cuesj-1010	2	45	within	within	ADP
cuesj-1010	2	46	artificial	artificial	ADJ
cuesj-1010	2	47	intelligence	intelligence	NOUN
cuesj-1010	2	48	systems	system	NOUN
cuesj-1010	2	49	has	have	AUX
cuesj-1010	2	50	given	give	VERB
cuesj-1010	2	51	rise	rise	NOUN
cuesj-1010	2	52	to	to	ADP
cuesj-1010	2	53	an	an	DET
cuesj-1010	2	54	essential	essential	ADJ
cuesj-1010	2	55	and	and	CCONJ
cuesj-1010	2	56	academically	academically	ADV
cuesj-1010	2	57	significant	significant	ADJ
cuesj-1010	2	58	approach	approach	NOUN
cuesj-1010	2	59	in	in	ADP
cuesj-1010	2	60	the	the	DET
cuesj-1010	2	61	domain	domain	NOUN
cuesj-1010	2	62	of	of	ADP
cuesj-1010	2	63	vehicle	vehicle	NOUN
cuesj-1010	2	64	detection	detection	NOUN
cuesj-1010	2	65	.	.	PUNCT
cuesj-1010	3	1	this	this	DET
cuesj-1010	3	2	study	study	NOUN
cuesj-1010	3	3	introduces	introduce	VERB
cuesj-1010	3	4	a	a	DET
cuesj-1010	3	5	real	real	ADJ
cuesj-1010	3	6	-	-	PUNCT
cuesj-1010	3	7	time	time	NOUN
cuesj-1010	3	8	vehicle	vehicle	NOUN
cuesj-1010	3	9	detection	detection	NOUN
cuesj-1010	3	10	framework	framework	NOUN
cuesj-1010	3	11	leveraging	leverage	VERB
cuesj-1010	3	12	the	the	PRON
cuesj-1010	3	13	you	you	PRON
cuesj-1010	3	14	only	only	ADV
cuesj-1010	3	15	look	look	VERB
cuesj-1010	3	16	once	once	ADV
cuesj-1010	3	17	(	(	PUNCT
cuesj-1010	3	18	yolo	yolo	ADJ
cuesj-1010	3	19	)	)	PUNCT
cuesj-1010	3	20	algorithm	algorithm	NOUN
cuesj-1010	3	21	for	for	ADP
cuesj-1010	3	22	precise	precise	ADJ
cuesj-1010	3	23	identification	identification	NOUN
cuesj-1010	3	24	of	of	ADP
cuesj-1010	3	25	vehicles	vehicle	NOUN
cuesj-1010	3	26	,	,	PUNCT
cuesj-1010	3	27	employing	employ	VERB
cuesj-1010	3	28	the	the	DET
cuesj-1010	3	29	camera	camera	NOUN
cuesj-1010	3	30	on	on	ADP
cuesj-1010	3	31	the	the	DET
cuesj-1010	3	32	dji	dji	PROPN
cuesj-1010	3	33	tello	tello	PROPN
cuesj-1010	3	34	drone	drone	PROPN
cuesj-1010	3	35	.	.	PUNCT
cuesj-1010	4	1	the	the	DET
cuesj-1010	4	2	research	research	NOUN
cuesj-1010	4	3	is	be	AUX
cuesj-1010	4	4	underpinned	underpin	VERB
cuesj-1010	4	5	by	by	ADP
cuesj-1010	4	6	a	a	DET
cuesj-1010	4	7	rich	rich	ADJ
cuesj-1010	4	8	dataset	dataset	NOUN
cuesj-1010	4	9	encompassing	encompass	VERB
cuesj-1010	4	10	approximately	approximately	ADV
cuesj-1010	4	11	2000	2000	NUM
cuesj-1010	4	12	images	image	NOUN
cuesj-1010	4	13	,	,	PUNCT
cuesj-1010	4	14	meticulously	meticulously	ADV
cuesj-1010	4	15	annotated	annotate	VERB
cuesj-1010	4	16	with	with	ADP
cuesj-1010	4	17	the	the	DET
cuesj-1010	4	18	respective	respective	ADJ
cuesj-1010	4	19	vehicle	vehicle	NOUN
cuesj-1010	4	20	angles	angle	NOUN
cuesj-1010	4	21	.	.	PUNCT
cuesj-1010	5	1	the	the	DET
cuesj-1010	5	2	framework	framework	NOUN
cuesj-1010	5	3	’s	’s	PART
cuesj-1010	5	4	innovation	innovation	NOUN
cuesj-1010	5	5	lies	lie	VERB
cuesj-1010	5	6	in	in	ADP
cuesj-1010	5	7	its	its	PRON
cuesj-1010	5	8	comprehensive	comprehensive	ADJ
cuesj-1010	5	9	training	training	NOUN
cuesj-1010	5	10	regimen	regiman	NOUN
cuesj-1010	5	11	,	,	PUNCT
cuesj-1010	5	12	encompassing	encompass	VERB
cuesj-1010	5	13	all	all	DET
cuesj-1010	5	14	angles	angle	NOUN
cuesj-1010	5	15	of	of	ADP
cuesj-1010	5	16	vehicles	vehicle	NOUN
cuesj-1010	5	17	:	:	PUNCT
cuesj-1010	5	18	vehicle	vehicle	NOUN
cuesj-1010	5	19	-	-	PUNCT
cuesj-1010	5	20	front	front	NOUN
cuesj-1010	5	21	,	,	PUNCT
cuesj-1010	5	22	vehicle	vehicle	NOUN
cuesj-1010	5	23	-	-	PUNCT
cuesj-1010	5	24	rear	rear	NOUN
cuesj-1010	5	25	,	,	PUNCT
cuesj-1010	5	26	vehicle	vehicle	NOUN
cuesj-1010	5	27	-	-	PUNCT
cuesj-1010	5	28	above	above	NOUN
cuesj-1010	5	29	,	,	PUNCT
cuesj-1010	5	30	and	and	CCONJ
cuesj-1010	5	31	vehicle	vehicle	NOUN
cuesj-1010	5	32	-	-	PUNCT
cuesj-1010	5	33	sides	side	NOUN
cuesj-1010	5	34	.	.	PUNCT
cuesj-1010	6	1	this	this	DET
cuesj-1010	6	2	holistic	holistic	ADJ
cuesj-1010	6	3	approach	approach	NOUN
cuesj-1010	6	4	aims	aim	VERB
cuesj-1010	6	5	to	to	PART
cuesj-1010	6	6	yield	yield	VERB
cuesj-1010	6	7	a	a	DET
cuesj-1010	6	8	model	model	NOUN
cuesj-1010	6	9	capable	capable	ADJ
cuesj-1010	6	10	of	of	ADP
cuesj-1010	6	11	accurately	accurately	ADV
cuesj-1010	6	12	identifying	identify	VERB
cuesj-1010	6	13	and	and	CCONJ
cuesj-1010	6	14	tracking	tracking	NOUN
cuesj-1010	6	15	vehicles	vehicle	NOUN
cuesj-1010	6	16	from	from	ADP
cuesj-1010	6	17	a	a	DET
cuesj-1010	6	18	multitude	multitude	NOUN
cuesj-1010	6	19	of	of	ADP
cuesj-1010	6	20	viewing	view	VERB
cuesj-1010	6	21	angles	angle	NOUN
cuesj-1010	6	22	.	.	PUNCT
cuesj-1010	7	1	the	the	DET
cuesj-1010	7	2	choice	choice	NOUN
cuesj-1010	7	3	of	of	ADP
cuesj-1010	7	4	the	the	DET
cuesj-1010	7	5	yolo	yolo	ADJ
cuesj-1010	7	6	algorithm	algorithm	NOUN
cuesj-1010	7	7	,	,	PUNCT
cuesj-1010	7	8	further	far	ADV
cuesj-1010	7	9	enhanced	enhance	VERB
cuesj-1010	7	10	by	by	ADP
cuesj-1010	7	11	ultralytics	ultralytic	NOUN
cuesj-1010	7	12	hub	hub	PROPN
cuesj-1010	7	13	,	,	PUNCT
cuesj-1010	7	14	ensures	ensure	VERB
cuesj-1010	7	15	the	the	DET
cuesj-1010	7	16	robustness	robustness	NOUN
cuesj-1010	7	17	and	and	CCONJ
cuesj-1010	7	18	accuracy	accuracy	NOUN
cuesj-1010	7	19	required	require	VERB
cuesj-1010	7	20	for	for	ADP
cuesj-1010	7	21	the	the	DET
cuesj-1010	7	22	detection	detection	NOUN
cuesj-1010	7	23	of	of	ADP
cuesj-1010	7	24	moving	move	VERB
cuesj-1010	7	25	objects	object	NOUN
cuesj-1010	7	26	.	.	PUNCT
cuesj-1010	8	1	the	the	DET
cuesj-1010	8	2	model	model	NOUN
cuesj-1010	8	3	’s	’s	PART
cuesj-1010	8	4	capability	capability	NOUN
cuesj-1010	8	5	to	to	PART
cuesj-1010	8	6	effectively	effectively	ADV
cuesj-1010	8	7	track	track	VERB
cuesj-1010	8	8	objects	object	NOUN
cuesj-1010	8	9	is	be	AUX
cuesj-1010	8	10	a	a	DET
cuesj-1010	8	11	testament	testament	NOUN
cuesj-1010	8	12	to	to	ADP
cuesj-1010	8	13	the	the	DET
cuesj-1010	8	14	algorithm	algorithm	NOUN
cuesj-1010	8	15	’s	’s	PART
cuesj-1010	8	16	efficacy	efficacy	NOUN
cuesj-1010	8	17	.	.	PUNCT
cuesj-1010	9	1	in	in	ADP
cuesj-1010	9	2	assessing	assess	VERB
cuesj-1010	9	3	the	the	DET
cuesj-1010	9	4	framework	framework	NOUN
cuesj-1010	9	5	’s	’s	PART
cuesj-1010	9	6	performance	performance	NOUN
cuesj-1010	9	7	,	,	PUNCT
cuesj-1010	9	8	we	we	PRON
cuesj-1010	9	9	employed	employ	VERB
cuesj-1010	9	10	a	a	DET
cuesj-1010	9	11	comprehensive	comprehensive	ADJ
cuesj-1010	9	12	set	set	NOUN
cuesj-1010	9	13	of	of	ADP
cuesj-1010	9	14	evaluation	evaluation	NOUN
cuesj-1010	9	15	parameters	parameter	NOUN
cuesj-1010	9	16	,	,	PUNCT
cuesj-1010	9	17	including	include	VERB
cuesj-1010	9	18	mean	mean	ADJ
cuesj-1010	9	19	average	average	ADJ
cuesj-1010	9	20	precision	precision	NOUN
cuesj-1010	9	21	,	,	PUNCT
cuesj-1010	9	22	precision	precision	NOUN
cuesj-1010	9	23	,	,	PUNCT
cuesj-1010	9	24	recall	recall	NOUN
cuesj-1010	9	25	,	,	PUNCT
cuesj-1010	9	26	and	and	CCONJ
cuesj-1010	9	27	f1	f1	PROPN
cuesj-1010	9	28	score	score	NOUN
cuesj-1010	9	29	.	.	PUNCT
cuesj-1010	10	1	this	this	DET
cuesj-1010	10	2	research	research	NOUN
cuesj-1010	10	3	not	not	PART
cuesj-1010	10	4	only	only	ADV
cuesj-1010	10	5	underscores	underscore	VERB
cuesj-1010	10	6	the	the	DET
cuesj-1010	10	7	practicality	practicality	NOUN
cuesj-1010	10	8	of	of	ADP
cuesj-1010	10	9	uavs	uavs	NOUN
cuesj-1010	10	10	in	in	ADP
cuesj-1010	10	11	the	the	DET
cuesj-1010	10	12	field	field	NOUN
cuesj-1010	10	13	of	of	ADP
cuesj-1010	10	14	artificial	artificial	ADJ
cuesj-1010	10	15	intelligence	intelligence	NOUN
cuesj-1010	10	16	but	but	CCONJ
cuesj-1010	10	17	also	also	ADV
cuesj-1010	10	18	highlights	highlight	VERB
cuesj-1010	10	19	the	the	DET
cuesj-1010	10	20	excellence	excellence	NOUN
cuesj-1010	10	21	achieved	achieve	VERB
cuesj-1010	10	22	in	in	ADP
cuesj-1010	10	23	real	real	ADJ
cuesj-1010	10	24	-	-	PUNCT
cuesj-1010	10	25	time	time	NOUN
cuesj-1010	10	26	vehicle	vehicle	NOUN
cuesj-1010	10	27	detection	detection	NOUN
cuesj-1010	10	28	.	.	PUNCT
cuesj-1010	11	1	keywords	keyword	NOUN
cuesj-1010	11	2	:	:	PUNCT
cuesj-1010	11	3	tello	tello	PROPN
cuesj-1010	11	4	drone	drone	PROPN
cuesj-1010	11	5	,	,	PUNCT
cuesj-1010	11	6	you	you	PRON
cuesj-1010	11	7	only	only	ADV
cuesj-1010	11	8	look	look	VERB
cuesj-1010	11	9	once	once	ADV
cuesj-1010	11	10	v5	v5	NOUN
cuesj-1010	11	11	,	,	PUNCT
cuesj-1010	11	12	roboflow	roboflow	ADJ
cuesj-1010	11	13	,	,	PUNCT
cuesj-1010	11	14	intersection	intersection	NOUN
cuesj-1010	11	15	over	over	ADP
cuesj-1010	11	16	union	union	NOUN
cuesj-1010	11	17	,	,	PUNCT
cuesj-1010	11	18	ultralytics	ultralytic	NOUN
cuesj-1010	11	19	introduction	introduction	VERB
cuesj-1010	11	20	the	the	DET
cuesj-1010	11	21	use	use	NOUN
cuesj-1010	11	22	of	of	ADP
cuesj-1010	11	23	drones	drone	NOUN
cuesj-1010	11	24	has	have	AUX
cuesj-1010	11	25	become	become	VERB
cuesj-1010	11	26	increasingly	increasingly	ADV
cuesj-1010	11	27	popular	popular	ADJ
cuesj-1010	11	28	,	,	PUNCT
cuesj-1010	11	29	especially	especially	ADV
cuesj-1010	11	30	in	in	ADP
cuesj-1010	11	31	the	the	DET
cuesj-1010	11	32	military	military	ADJ
cuesj-1010	11	33	and	and	CCONJ
cuesj-1010	11	34	transportation	transportation	NOUN
cuesj-1010	11	35	sectors	sector	NOUN
cuesj-1010	11	36	.	.	PUNCT
cuesj-1010	12	1	we	we	PRON
cuesj-1010	12	2	can	can	AUX
cuesj-1010	12	3	teach	teach	VERB
cuesj-1010	12	4	drones	drone	NOUN
cuesj-1010	12	5	how	how	SCONJ
cuesj-1010	12	6	to	to	PART
cuesj-1010	12	7	locate	locate	VERB
cuesj-1010	12	8	and	and	CCONJ
cuesj-1010	12	9	recognize	recognize	VERB
cuesj-1010	12	10	targets	target	NOUN
cuesj-1010	12	11	and	and	CCONJ
cuesj-1010	12	12	gather	gather	VERB
cuesj-1010	12	13	the	the	DET
cuesj-1010	12	14	necessary	necessary	ADJ
cuesj-1010	12	15	information	information	NOUN
cuesj-1010	12	16	about	about	ADP
cuesj-1010	12	17	them	they	PRON
cuesj-1010	12	18	.	.	PUNCT
cuesj-1010	13	1	we	we	PRON
cuesj-1010	13	2	can	can	AUX
cuesj-1010	13	3	make	make	VERB
cuesj-1010	13	4	drones	drone	NOUN
cuesj-1010	13	5	more	more	ADV
cuesj-1010	13	6	advanced	advanced	ADJ
cuesj-1010	13	7	and	and	CCONJ
cuesj-1010	13	8	smarter	smarter	ADV
cuesj-1010	13	9	using	use	VERB
cuesj-1010	13	10	artificial	artificial	ADJ
cuesj-1010	13	11	intelligence	intelligence	NOUN
cuesj-1010	13	12	-	-	PUNCT
cuesj-1010	13	13	trained	train	VERB
cuesj-1010	13	14	models	model	NOUN
cuesj-1010	13	15	such	such	ADJ
cuesj-1010	13	16	as	as	SCONJ
cuesj-1010	13	17	you	you	PRON
cuesj-1010	13	18	only	only	ADV
cuesj-1010	13	19	look	look	VERB
cuesj-1010	13	20	once	once	ADV
cuesj-1010	13	21	(	(	PUNCT
cuesj-1010	13	22	yolo	yolo	ADJ
cuesj-1010	13	23	)	)	PUNCT
cuesj-1010	13	24	and	and	CCONJ
cuesj-1010	13	25	tensorflow	tensorflow	NOUN
cuesj-1010	13	26	.	.	PUNCT
cuesj-1010	14	1	the	the	DET
cuesj-1010	14	2	purpose	purpose	NOUN
cuesj-1010	14	3	of	of	ADP
cuesj-1010	14	4	this	this	DET
cuesj-1010	14	5	article	article	NOUN
cuesj-1010	14	6	is	be	AUX
cuesj-1010	14	7	to	to	PART
cuesj-1010	14	8	train	train	VERB
cuesj-1010	14	9	a	a	DET
cuesj-1010	14	10	yolo	yolo	ADJ
cuesj-1010	14	11	model	model	NOUN
cuesj-1010	14	12	that	that	PRON
cuesj-1010	14	13	can	can	AUX
cuesj-1010	14	14	recognize	recognize	VERB
cuesj-1010	14	15	all	all	DET
cuesj-1010	14	16	angles	angle	NOUN
cuesj-1010	14	17	of	of	ADP
cuesj-1010	14	18	a	a	DET
cuesj-1010	14	19	vehicle	vehicle	NOUN
cuesj-1010	14	20	for	for	ADP
cuesj-1010	14	21	military	military	ADJ
cuesj-1010	14	22	and	and	CCONJ
cuesj-1010	14	23	traffic	traffic	NOUN
cuesj-1010	14	24	use	use	NOUN
cuesj-1010	14	25	.	.	PUNCT
cuesj-1010	15	1	by	by	ADP
cuesj-1010	15	2	training	train	VERB
cuesj-1010	15	3	drones	drone	NOUN
cuesj-1010	15	4	to	to	PART
cuesj-1010	15	5	recognize	recognize	VERB
cuesj-1010	15	6	all	all	DET
cuesj-1010	15	7	angles	angle	NOUN
cuesj-1010	15	8	of	of	ADP
cuesj-1010	15	9	a	a	DET
cuesj-1010	15	10	vehicle	vehicle	NOUN
cuesj-1010	15	11	,	,	PUNCT
cuesj-1010	15	12	we	we	PRON
cuesj-1010	15	13	can	can	AUX
cuesj-1010	15	14	enhance	enhance	VERB
cuesj-1010	15	15	their	their	PRON
cuesj-1010	15	16	surveillance	surveillance	NOUN
cuesj-1010	15	17	capabilities	capability	NOUN
cuesj-1010	15	18	and	and	CCONJ
cuesj-1010	15	19	improve	improve	VERB
cuesj-1010	15	20	situational	situational	ADJ
cuesj-1010	15	21	awareness	awareness	NOUN
cuesj-1010	15	22	in	in	ADP
cuesj-1010	15	23	both	both	PRON
cuesj-1010	15	24	military	military	ADJ
cuesj-1010	15	25	operations	operation	NOUN
cuesj-1010	15	26	and	and	CCONJ
cuesj-1010	15	27	traffic	traffic	NOUN
cuesj-1010	15	28	management	management	NOUN
cuesj-1010	15	29	.	.	PUNCT
cuesj-1010	16	1	this	this	DET
cuesj-1010	16	2	advanced	advanced	ADJ
cuesj-1010	16	3	technology	technology	NOUN
cuesj-1010	16	4	has	have	VERB
cuesj-1010	16	5	the	the	DET
cuesj-1010	16	6	potential	potential	NOUN
cuesj-1010	16	7	to	to	PART
cuesj-1010	16	8	revolutionize	revolutionize	VERB
cuesj-1010	16	9	the	the	DET
cuesj-1010	16	10	way	way	NOUN
cuesj-1010	16	11	we	we	PRON
cuesj-1010	16	12	gather	gather	VERB
cuesj-1010	16	13	information	information	NOUN
cuesj-1010	16	14	and	and	CCONJ
cuesj-1010	16	15	make	make	VERB
cuesj-1010	16	16	informed	informed	ADJ
cuesj-1010	16	17	decisions	decision	NOUN
cuesj-1010	16	18	in	in	ADP
cuesj-1010	16	19	these	these	DET
cuesj-1010	16	20	sectors	sector	NOUN
cuesj-1010	16	21	.	.	PUNCT
cuesj-1010	17	1	the	the	DET
cuesj-1010	17	2	yolo	yolo	ADJ
cuesj-1010	17	3	framework	framework	NOUN
cuesj-1010	17	4	has	have	AUX
cuesj-1010	17	5	emerged	emerge	VERB
cuesj-1010	17	6	as	as	ADP
cuesj-1010	17	7	a	a	DET
cuesj-1010	17	8	popular	popular	ADJ
cuesj-1010	17	9	opensource	opensource	NOUN
cuesj-1010	17	10	deep	deep	ADJ
cuesj-1010	17	11	learning	learning	NOUN
cuesj-1010	17	12	tool	tool	NOUN
cuesj-1010	17	13	for	for	ADP
cuesj-1010	17	14	various	various	ADJ
cuesj-1010	17	15	machine	machine	NOUN
cuesj-1010	17	16	learning	learn	VERB
cuesj-1010	17	17	applications	application	NOUN
cuesj-1010	17	18	.	.	PUNCT
cuesj-1010	18	1	initially	initially	ADV
cuesj-1010	18	2	introduced	introduce	VERB
cuesj-1010	18	3	in	in	ADP
cuesj-1010	18	4	2016	2016	NUM
cuesj-1010	18	5	,	,	PUNCT
cuesj-1010	18	6	yolo	yolo	PROPN
cuesj-1010	18	7	employs	employ	VERB
cuesj-1010	18	8	a	a	DET
cuesj-1010	18	9	single	single	ADJ
cuesj-1010	18	10	neural	neural	ADJ
cuesj-1010	18	11	network	network	NOUN
cuesj-1010	18	12	to	to	PART
cuesj-1010	18	13	perform	perform	VERB
cuesj-1010	18	14	object	object	NOUN
cuesj-1010	18	15	identification	identification	NOUN
cuesj-1010	18	16	,	,	PUNCT
cuesj-1010	18	17	categorization	categorization	NOUN
cuesj-1010	18	18	,	,	PUNCT
cuesj-1010	18	19	and	and	CCONJ
cuesj-1010	18	20	localization	localization	NOUN
cuesj-1010	18	21	.	.	PUNCT
cuesj-1010	19	1	over	over	ADP
cuesj-1010	19	2	the	the	DET
cuesj-1010	19	3	years	year	NOUN
cuesj-1010	19	4	,	,	PUNCT
cuesj-1010	19	5	yolo	yolo	PROPN
cuesj-1010	19	6	has	have	AUX
cuesj-1010	19	7	undergone	undergo	VERB
cuesj-1010	19	8	several	several	ADJ
cuesj-1010	19	9	iterations	iteration	NOUN
cuesj-1010	19	10	,	,	PUNCT
cuesj-1010	19	11	with	with	ADP
cuesj-1010	19	12	the	the	DET
cuesj-1010	19	13	most	most	ADV
cuesj-1010	19	14	recent	recent	ADJ
cuesj-1010	19	15	versions	version	NOUN
cuesj-1010	19	16	,	,	PUNCT
cuesj-1010	19	17	yolov5	yolov5	NOUN
cuesj-1010	19	18	and	and	CCONJ
cuesj-1010	19	19	yolov6	yolov6	PROPN
cuesj-1010	19	20	/	/	SYM
cuesj-1010	19	21	v7	v7	VERB
cuesj-1010	19	22	,	,	PUNCT
cuesj-1010	19	23	released	release	VERB
cuesj-1010	19	24	in	in	ADP
cuesj-1010	19	25	2020	2020	NUM
cuesj-1010	19	26	and	and	CCONJ
cuesj-1010	19	27	2022	2022	NUM
cuesj-1010	19	28	,	,	PUNCT
cuesj-1010	19	29	respectively	respectively	ADV
cuesj-1010	19	30	,	,	PUNCT
cuesj-1010	19	31	and	and	CCONJ
cuesj-1010	19	32	offering	offer	VERB
cuesj-1010	19	33	significant	significant	ADJ
cuesj-1010	19	34	enhancements	enhancement	NOUN
cuesj-1010	19	35	.	.	PUNCT
cuesj-1010	20	1	it	it	PRON
cuesj-1010	20	2	’s	’	VERB
cuesj-1010	20	3	important	important	ADJ
cuesj-1010	20	4	to	to	PART
cuesj-1010	20	5	note	note	VERB
cuesj-1010	20	6	that	that	SCONJ
cuesj-1010	20	7	versions	version	NOUN
cuesj-1010	20	8	6	6	NUM
cuesj-1010	20	9	and	and	CCONJ
cuesj-1010	20	10	7	7	NUM
cuesj-1010	20	11	were	be	AUX
cuesj-1010	20	12	independently	independently	ADV
cuesj-1010	20	13	developed	develop	VERB
cuesj-1010	20	14	by	by	ADP
cuesj-1010	20	15	different	different	ADJ
cuesj-1010	20	16	teams	team	NOUN
cuesj-1010	20	17	and	and	CCONJ
cuesj-1010	20	18	do	do	AUX
cuesj-1010	20	19	n’t	not	PART
cuesj-1010	20	20	belong	belong	VERB
cuesj-1010	20	21	to	to	ADP
cuesj-1010	20	22	the	the	DET
cuesj-1010	20	23	core	core	ADJ
cuesj-1010	20	24	yolo	yolo	ADJ
cuesj-1010	20	25	series	series	NOUN
cuesj-1010	20	26	.	.	PUNCT
cuesj-1010	21	1	as	as	ADP
cuesj-1010	21	2	of	of	ADP
cuesj-1010	21	3	now	now	ADV
cuesj-1010	21	4	,	,	PUNCT
cuesj-1010	21	5	yolov5	yolov5	NOUN
cuesj-1010	21	6	remains	remain	VERB
cuesj-1010	21	7	the	the	DET
cuesj-1010	21	8	most	most	ADV
cuesj-1010	21	9	widely	widely	ADV
cuesj-1010	21	10	used	use	VERB
cuesj-1010	21	11	among	among	ADP
cuesj-1010	21	12	these	these	DET
cuesj-1010	21	13	iterations	iteration	NOUN
cuesj-1010	21	14	.	.	PUNCT
cuesj-1010	22	1	yolov5	yolov5	NOUN
cuesj-1010	22	2	stands	stand	VERB
cuesj-1010	22	3	out	out	ADP
cuesj-1010	22	4	for	for	ADP
cuesj-1010	22	5	several	several	ADJ
cuesj-1010	22	6	key	key	ADJ
cuesj-1010	22	7	advantages	advantage	NOUN
cuesj-1010	22	8	over	over	ADP
cuesj-1010	22	9	its	its	PRON
cuesj-1010	22	10	predecessors	predecessor	NOUN
cuesj-1010	22	11	.	.	PUNCT
cuesj-1010	23	1	these	these	PRON
cuesj-1010	23	2	include	include	VERB
cuesj-1010	23	3	a	a	DET
cuesj-1010	23	4	reduced	reduce	VERB
cuesj-1010	23	5	model	model	NOUN
cuesj-1010	23	6	size	size	NOUN
cuesj-1010	23	7	,	,	PUNCT
cuesj-1010	23	8	increased	increase	VERB
cuesj-1010	23	9	processing	processing	NOUN
cuesj-1010	23	10	speed	speed	NOUN
cuesj-1010	23	11	,	,	PUNCT
cuesj-1010	23	12	improved	improved	ADJ
cuesj-1010	23	13	accuracy	accuracy	NOUN
cuesj-1010	23	14	,	,	PUNCT
cuesj-1010	23	15	and	and	CCONJ
cuesj-1010	23	16	seamless	seamless	ADJ
cuesj-1010	23	17	integration	integration	NOUN
cuesj-1010	23	18	with	with	ADP
cuesj-1010	23	19	the	the	DET
cuesj-1010	23	20	popular	popular	ADJ
cuesj-1010	23	21	pytorch	pytorch	NOUN
cuesj-1010	23	22	opensource	opensource	NOUN
cuesj-1010	23	23	machine	machine	NOUN
cuesj-1010	23	24	learning	learning	NOUN
cuesj-1010	23	25	framework	framework	NOUN
cuesj-1010	23	26	.	.	PUNCT
cuesj-1010	24	1	furthermore	furthermore	ADV
cuesj-1010	24	2	,	,	PUNCT
cuesj-1010	24	3	yolov5	yolov5	NOUN
cuesj-1010	24	4	introduces	introduce	VERB
cuesj-1010	24	5	ten	ten	NUM
cuesj-1010	24	6	new	new	ADJ
cuesj-1010	24	7	types	type	NOUN
cuesj-1010	24	8	not	not	PART
cuesj-1010	24	9	found	find	VERB
cuesj-1010	24	10	in	in	ADP
cuesj-1010	24	11	earlier	early	ADJ
cuesj-1010	24	12	versions	version	NOUN
cuesj-1010	24	13	,	,	PUNCT
cuesj-1010	24	14	expanding	expand	VERB
cuesj-1010	24	15	its	its	PRON
cuesj-1010	24	16	versatility.[1	versatility.[1	NOUN
cuesj-1010	24	17	]	]	PUNCT
cuesj-1010	24	18	in	in	ADP
cuesj-1010	24	19	the	the	DET
cuesj-1010	24	20	realm	realm	NOUN
cuesj-1010	24	21	of	of	ADP
cuesj-1010	24	22	data	datum	NOUN
cuesj-1010	24	23	science	science	NOUN
cuesj-1010	24	24	projects	project	NOUN
cuesj-1010	24	25	,	,	PUNCT
cuesj-1010	24	26	the	the	DET
cuesj-1010	24	27	yolov5	yolov5	NOUN
cuesj-1010	24	28	algorithm	algorithm	NOUN
cuesj-1010	24	29	finds	find	VERB
cuesj-1010	24	30	significant	significant	ADJ
cuesj-1010	24	31	utility	utility	NOUN
cuesj-1010	24	32	in	in	ADP
cuesj-1010	24	33	categorizing	categorize	VERB
cuesj-1010	24	34	freely	freely	ADV
cuesj-1010	24	35	available	available	ADJ
cuesj-1010	24	36	datasets	dataset	NOUN
cuesj-1010	24	37	accessible	accessible	ADJ
cuesj-1010	24	38	from	from	ADP
cuesj-1010	24	39	internet	internet	NOUN
cuesj-1010	24	40	repositories	repository	NOUN
cuesj-1010	24	41	like	like	ADP
cuesj-1010	24	42	roboflow.[2	roboflow.[2	NOUN
cuesj-1010	24	43	]	]	PUNCT
cuesj-1010	24	44	roboflow	roboflow	NOUN
cuesj-1010	24	45	,	,	PUNCT
cuesj-1010	24	46	a	a	DET
cuesj-1010	24	47	free	free	ADJ
cuesj-1010	24	48	tool	tool	NOUN
cuesj-1010	24	49	,	,	PUNCT
cuesj-1010	24	50	simplifies	simplify	VERB
cuesj-1010	24	51	the	the	DET
cuesj-1010	24	52	annotation	annotation	NOUN
cuesj-1010	24	53	and	and	CCONJ
cuesj-1010	24	54	labeling	labeling	NOUN
cuesj-1010	24	55	of	of	ADP
cuesj-1010	24	56	images	image	NOUN
cuesj-1010	24	57	,	,	PUNCT
cuesj-1010	24	58	addressing	address	VERB
cuesj-1010	24	59	the	the	DET
cuesj-1010	24	60	complexities	complexity	NOUN
cuesj-1010	24	61	of	of	ADP
cuesj-1010	24	62	managing	manage	VERB
cuesj-1010	24	63	large	large	ADJ
cuesj-1010	24	64	datasets	dataset	NOUN
cuesj-1010	24	65	.	.	PUNCT
cuesj-1010	25	1	beyond	beyond	ADP
cuesj-1010	25	2	labeling	labeling	NOUN
cuesj-1010	25	3	,	,	PUNCT
cuesj-1010	25	4	the	the	DET
cuesj-1010	25	5	platform	platform	NOUN
cuesj-1010	25	6	streamlines	streamline	VERB
cuesj-1010	25	7	the	the	DET
cuesj-1010	25	8	assignment	assignment	NOUN
cuesj-1010	25	9	of	of	ADP
cuesj-1010	25	10	images	image	NOUN
cuesj-1010	25	11	to	to	ADP
cuesj-1010	25	12	appropriate	appropriate	ADJ
cuesj-1010	25	13	classes	class	NOUN
cuesj-1010	25	14	,	,	PUNCT
cuesj-1010	25	15	project	project	NOUN
cuesj-1010	25	16	organization	organization	NOUN
cuesj-1010	25	17	,	,	PUNCT
cuesj-1010	25	18	and	and	CCONJ
cuesj-1010	25	19	numerous	numerous	ADJ
cuesj-1010	25	20	other	other	ADJ
cuesj-1010	25	21	workflow	workflow	NOUN
cuesj-1010	25	22	tasks	task	NOUN
cuesj-1010	25	23	,	,	PUNCT
cuesj-1010	25	24	providing	provide	VERB
cuesj-1010	25	25	comprehensive	comprehensive	ADJ
cuesj-1010	25	26	solutions	solution	NOUN
cuesj-1010	25	27	for	for	ADP
cuesj-1010	25	28	data	datum	NOUN
cuesj-1010	25	29	scientists	scientist	NOUN
cuesj-1010	25	30	.	.	PUNCT
cuesj-1010	26	1	corresponding	correspond	VERB
cuesj-1010	26	2	author	author	NOUN
cuesj-1010	26	3	:	:	PUNCT
cuesj-1010	26	4	shalaw	shalaw	NOUN
cuesj-1010	26	5	m.	m.	PROPN
cuesj-1010	26	6	abdallah	abdallah	PROPN
cuesj-1010	26	7	,	,	PUNCT
cuesj-1010	26	8	department	department	NOUN
cuesj-1010	26	9	of	of	ADP
cuesj-1010	26	10	software	software	PROPN
cuesj-1010	26	11	engineering	engineering	PROPN
cuesj-1010	26	12	,	,	PUNCT
cuesj-1010	26	13	koya	koya	PROPN
cuesj-1010	26	14	university	university	PROPN
cuesj-1010	26	15	,	,	PUNCT
cuesj-1010	26	16	kurdistan	kurdistan	PROPN
cuesj-1010	26	17	region	region	NOUN
cuesj-1010	26	18	,	,	PUNCT
cuesj-1010	26	19	iraq	iraq	PROPN
cuesj-1010	26	20	.	.	PUNCT
cuesj-1010	27	1	e	e	X
cuesj-1010	27	2	-	-	NOUN
cuesj-1010	27	3	mail	mail	NOUN
cuesj-1010	27	4	:	:	PUNCT
cuesj-1010	27	5	shalaw.mshir@gmail.com	shalaw.mshir@gmail.com	X
cuesj-1010	27	6	received	receive	VERB
cuesj-1010	27	7	:	:	PUNCT
cuesj-1010	27	8	august	august	PROPN
cuesj-1010	27	9	09	09	NUM
cuesj-1010	27	10	,	,	PUNCT
cuesj-1010	27	11	2023	2023	NUM
cuesj-1010	27	12	accepted	accept	VERB
cuesj-1010	27	13	:	:	PUNCT
cuesj-1010	27	14	november	november	PROPN
cuesj-1010	27	15	21	21	NUM
cuesj-1010	27	16	,	,	PUNCT
cuesj-1010	27	17	2023	2023	NUM
cuesj-1010	27	18	published	publish	VERB
cuesj-1010	27	19	:	:	PUNCT
cuesj-1010	27	20	january	january	PROPN
cuesj-1010	27	21	20	20	NUM
cuesj-1010	27	22	,	,	PUNCT
cuesj-1010	27	23	2024	2024	NUM
cuesj-1010	27	24	doi	doi	NOUN
cuesj-1010	27	25	:	:	PUNCT
cuesj-1010	27	26	10.24086	10.24086	NUM
cuesj-1010	27	27	/	/	SYM
cuesj-1010	27	28	cuesj.v8n1y2024.pp1	cuesj.v8n1y2024.pp1	PROPN
cuesj-1010	27	29	-	-	PUNCT
cuesj-1010	27	30	7	7	NUM
cuesj-1010	27	31	copyright	copyright	NOUN
cuesj-1010	27	32	©	©	PROPN
cuesj-1010	27	33	2024	2024	NUM
cuesj-1010	27	34	shalaw	shalaw	NOUN
cuesj-1010	27	35	m.	m.	NOUN
cuesj-1010	27	36	abdallah	abdallah	PROPN
cuesj-1010	27	37	.	.	PUNCT
cuesj-1010	28	1	this	this	PRON
cuesj-1010	28	2	is	be	AUX
cuesj-1010	28	3	an	an	DET
cuesj-1010	28	4	open	open	ADJ
cuesj-1010	28	5	-	-	PUNCT
cuesj-1010	28	6	access	access	NOUN
cuesj-1010	28	7	article	article	NOUN
cuesj-1010	28	8	distributed	distribute	VERB
cuesj-1010	28	9	under	under	ADP
cuesj-1010	28	10	the	the	DET
cuesj-1010	28	11	creative	creative	ADJ
cuesj-1010	28	12	commons	common	NOUN
cuesj-1010	28	13	attribution	attribution	NOUN
cuesj-1010	28	14	license	license	NOUN
cuesj-1010	28	15	.	.	PUNCT
cuesj-1010	29	1	cihan	cihan	VERB
cuesj-1010	29	2	university	university	NOUN
cuesj-1010	29	3	-	-	PUNCT
cuesj-1010	29	4	erbil	erbil	PROPN
cuesj-1010	29	5	scientific	scientific	ADJ
cuesj-1010	29	6	journal	journal	NOUN
cuesj-1010	29	7	(	(	PUNCT
cuesj-1010	29	8	cuesj	cuesj	PROPN
cuesj-1010	29	9	)	)	PUNCT
cuesj-1010	29	10	https://creativecommons.org/licenses/by-nc-nd/4.0/	https://creativecommons.org/licenses/by-nc-nd/4.0/	PROPN
cuesj-1010	29	11	abdallah	abdallah	PROPN
cuesj-1010	29	12	:	:	PUNCT
cuesj-1010	29	13	vehicle	vehicle	NOUN
cuesj-1010	29	14	detection	detection	NOUN
cuesj-1010	29	15	with	with	ADP
cuesj-1010	29	16	tello	tello	PROPN
cuesj-1010	29	17	drone	drone	PROPN
cuesj-1010	29	18	and	and	CCONJ
cuesj-1010	29	19	yolov5	yolov5	NOUN
cuesj-1010	29	20	2	2	NUM
cuesj-1010	29	21	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	29	22	cuesj	cuesj	NOUN
cuesj-1010	29	23	2024	2024	NUM
cuesj-1010	29	24	,	,	PUNCT
cuesj-1010	29	25	8	8	NUM
cuesj-1010	29	26	(	(	PUNCT
cuesj-1010	29	27	1	1	NUM
cuesj-1010	29	28	):	):	PUNCT
cuesj-1010	29	29	1	1	NUM
cuesj-1010	29	30	-	-	SYM
cuesj-1010	29	31	7	7	NUM
cuesj-1010	29	32	at	at	ADP
cuesj-1010	29	33	the	the	DET
cuesj-1010	29	34	core	core	NOUN
cuesj-1010	29	35	of	of	ADP
cuesj-1010	29	36	yolo	yolo	PROPN
cuesj-1010	29	37	’s	’s	PART
cuesj-1010	29	38	model	model	NOUN
cuesj-1010	29	39	is	be	AUX
cuesj-1010	29	40	its	its	PRON
cuesj-1010	29	41	employment	employment	NOUN
cuesj-1010	29	42	of	of	ADP
cuesj-1010	29	43	a	a	DET
cuesj-1010	29	44	single	single	ADJ
cuesj-1010	29	45	neural	neural	ADJ
cuesj-1010	29	46	network	network	NOUN
cuesj-1010	29	47	to	to	PART
cuesj-1010	29	48	analyze	analyze	VERB
cuesj-1010	29	49	an	an	DET
cuesj-1010	29	50	entire	entire	ADJ
cuesj-1010	29	51	image	image	NOUN
cuesj-1010	29	52	.	.	PUNCT
cuesj-1010	30	1	this	this	DET
cuesj-1010	30	2	network	network	NOUN
cuesj-1010	30	3	partitions	partition	VERB
cuesj-1010	30	4	the	the	DET
cuesj-1010	30	5	image	image	NOUN
cuesj-1010	30	6	into	into	ADP
cuesj-1010	30	7	grid	grid	NOUN
cuesj-1010	30	8	sections	section	NOUN
cuesj-1010	30	9	,	,	PUNCT
cuesj-1010	30	10	estimating	estimate	VERB
cuesj-1010	30	11	probabilities	probability	NOUN
cuesj-1010	30	12	and	and	CCONJ
cuesj-1010	30	13	bounding	bounding	NOUN
cuesj-1010	30	14	boxes	box	NOUN
cuesj-1010	30	15	for	for	ADP
cuesj-1010	30	16	each	each	DET
cuesj-1010	30	17	section	section	NOUN
cuesj-1010	30	18	.	.	PUNCT
cuesj-1010	31	1	these	these	DET
cuesj-1010	31	2	bounding	bounding	NOUN
cuesj-1010	31	3	boxes	box	NOUN
cuesj-1010	31	4	are	be	AUX
cuesj-1010	31	5	weighted	weight	VERB
cuesj-1010	31	6	according	accord	VERB
cuesj-1010	31	7	to	to	ADP
cuesj-1010	31	8	their	their	PRON
cuesj-1010	31	9	expected	expect	VERB
cuesj-1010	31	10	probability	probability	NOUN
cuesj-1010	31	11	,	,	PUNCT
cuesj-1010	31	12	allowing	allow	VERB
cuesj-1010	31	13	the	the	DET
cuesj-1010	31	14	model	model	NOUN
cuesj-1010	31	15	to	to	PART
cuesj-1010	31	16	consider	consider	VERB
cuesj-1010	31	17	the	the	DET
cuesj-1010	31	18	global	global	ADJ
cuesj-1010	31	19	context	context	NOUN
cuesj-1010	31	20	when	when	SCONJ
cuesj-1010	31	21	evaluating	evaluate	VERB
cuesj-1010	31	22	the	the	DET
cuesj-1010	31	23	entire	entire	ADJ
cuesj-1010	31	24	image	image	NOUN
cuesj-1010	31	25	during	during	ADP
cuesj-1010	31	26	testing.[3	testing.[3	NOUN
cuesj-1010	31	27	]	]	PUNCT
cuesj-1010	31	28	in	in	ADP
cuesj-1010	31	29	addition	addition	NOUN
cuesj-1010	31	30	,	,	PUNCT
cuesj-1010	31	31	yolo	yolo	PROPN
cuesj-1010	31	32	delivers	deliver	VERB
cuesj-1010	31	33	efficient	efficient	ADJ
cuesj-1010	31	34	predictions	prediction	NOUN
cuesj-1010	31	35	through	through	ADP
cuesj-1010	31	36	a	a	DET
cuesj-1010	31	37	single	single	ADJ
cuesj-1010	31	38	pass	pass	NOUN
cuesj-1010	31	39	,	,	PUNCT
cuesj-1010	31	40	as	as	SCONJ
cuesj-1010	31	41	opposed	oppose	VERB
cuesj-1010	31	42	to	to	ADP
cuesj-1010	31	43	repeated	repeat	VERB
cuesj-1010	31	44	analyses	analysis	NOUN
cuesj-1010	31	45	required	require	VERB
cuesj-1010	31	46	in	in	ADP
cuesj-1010	31	47	other	other	ADJ
cuesj-1010	31	48	systems	system	NOUN
cuesj-1010	31	49	.	.	PUNCT
cuesj-1010	32	1	it	it	PRON
cuesj-1010	32	2	employs	employ	VERB
cuesj-1010	32	3	three	three	NUM
cuesj-1010	32	4	methods	method	NOUN
cuesj-1010	32	5	for	for	ADP
cuesj-1010	32	6	detecting	detect	VERB
cuesj-1010	32	7	which	which	PRON
cuesj-1010	32	8	are	be	AUX
cuesj-1010	32	9	as	as	ADP
cuesj-1010	32	10	below	below	ADV
cuesj-1010	32	11	:	:	PUNCT
cuesj-1010	32	12	1	1	X
cuesj-1010	32	13	.	.	X
cuesj-1010	32	14	blocks	block	NOUN
cuesj-1010	32	15	of	of	ADP
cuesj-1010	32	16	residual	residual	ADJ
cuesj-1010	32	17	(	(	PUNCT
cuesj-1010	32	18	making	make	VERB
cuesj-1010	32	19	an	an	DET
cuesj-1010	32	20	s	s	NOUN
cuesj-1010	32	21	*	*	PUNCT
cuesj-1010	32	22	s	s	NOUN
cuesj-1010	32	23	grid	grid	NOUN
cuesj-1010	32	24	from	from	ADP
cuesj-1010	32	25	the	the	DET
cuesj-1010	32	26	image	image	NOUN
cuesj-1010	32	27	)	)	PUNCT
cuesj-1010	33	1	2	2	NUM
cuesj-1010	33	2	.	.	X
cuesj-1010	33	3	regression	regression	NOUN
cuesj-1010	33	4	of	of	ADP
cuesj-1010	33	5	bounding	bounding	NOUN
cuesj-1010	33	6	box	box	NOUN
cuesj-1010	33	7	(	(	PUNCT
cuesj-1010	33	8	predict	predict	VERB
cuesj-1010	33	9	the	the	DET
cuesj-1010	33	10	class	class	NOUN
cuesj-1010	33	11	,	,	PUNCT
cuesj-1010	33	12	center	center	NOUN
cuesj-1010	33	13	,	,	PUNCT
cuesj-1010	33	14	width	width	ADJ
cuesj-1010	33	15	,	,	PUNCT
cuesj-1010	33	16	and	and	CCONJ
cuesj-1010	33	17	height	height	NOUN
cuesj-1010	33	18	of	of	ADP
cuesj-1010	33	19	objects	object	NOUN
cuesj-1010	33	20	)	)	PUNCT
cuesj-1010	33	21	3	3	NUM
cuesj-1010	33	22	.	.	X
cuesj-1010	33	23	intersection	intersection	NOUN
cuesj-1010	33	24	over	over	ADP
cuesj-1010	33	25	union	union	NOUN
cuesj-1010	33	26	(	(	PUNCT
cuesj-1010	33	27	iou	iou	PROPN
cuesj-1010	33	28	)	)	PUNCT
cuesj-1010	33	29	for	for	ADP
cuesj-1010	33	30	checking	check	VERB
cuesj-1010	33	31	the	the	DET
cuesj-1010	33	32	way	way	NOUN
cuesj-1010	33	33	that	that	PRON
cuesj-1010	33	34	bbox	bbox	NOUN
cuesj-1010	33	35	overlaps	overlap	VERB
cuesj-1010	33	36	and	and	CCONJ
cuesj-1010	33	37	achieves	achieve	VERB
cuesj-1010	33	38	the	the	DET
cuesj-1010	33	39	greatest	great	ADJ
cuesj-1010	33	40	fit	fit	NOUN
cuesj-1010	33	41	.	.	PUNCT
cuesj-1010	34	1	a	a	DET
cuesj-1010	34	2	fundamental	fundamental	ADJ
cuesj-1010	34	3	element	element	NOUN
cuesj-1010	34	4	of	of	ADP
cuesj-1010	34	5	yolo	yolo	PROPN
cuesj-1010	34	6	is	be	AUX
cuesj-1010	34	7	its	its	PRON
cuesj-1010	34	8	use	use	NOUN
cuesj-1010	34	9	of	of	ADP
cuesj-1010	34	10	residual	residual	ADJ
cuesj-1010	34	11	blocks	block	NOUN
cuesj-1010	34	12	to	to	PART
cuesj-1010	34	13	construct	construct	VERB
cuesj-1010	34	14	an	an	DET
cuesj-1010	34	15	s	s	NOUN
cuesj-1010	34	16	*	*	PUNCT
cuesj-1010	34	17	s	s	NOUN
cuesj-1010	34	18	grid	grid	NOUN
cuesj-1010	34	19	from	from	ADP
cuesj-1010	34	20	the	the	DET
cuesj-1010	34	21	input	input	NOUN
cuesj-1010	34	22	image	image	NOUN
cuesj-1010	34	23	.	.	PUNCT
cuesj-1010	35	1	this	this	DET
cuesj-1010	35	2	grid	grid	NOUN
cuesj-1010	35	3	allows	allow	VERB
cuesj-1010	35	4	the	the	DET
cuesj-1010	35	5	image	image	NOUN
cuesj-1010	35	6	to	to	PART
cuesj-1010	35	7	be	be	AUX
cuesj-1010	35	8	divided	divide	VERB
cuesj-1010	35	9	into	into	ADP
cuesj-1010	35	10	a	a	DET
cuesj-1010	35	11	set	set	NOUN
cuesj-1010	35	12	of	of	ADP
cuesj-1010	35	13	cells	cell	NOUN
cuesj-1010	35	14	,	,	PUNCT
cuesj-1010	35	15	each	each	PRON
cuesj-1010	35	16	responsible	responsible	ADJ
cuesj-1010	35	17	for	for	ADP
cuesj-1010	35	18	predicting	predict	VERB
cuesj-1010	35	19	specific	specific	ADJ
cuesj-1010	35	20	class	class	NOUN
cuesj-1010	35	21	probabilities	probability	NOUN
cuesj-1010	35	22	and	and	CCONJ
cuesj-1010	35	23	bounding	bounding	NOUN
cuesj-1010	35	24	boxes	box	NOUN
cuesj-1010	35	25	.	.	PUNCT
cuesj-1010	36	1	the	the	DET
cuesj-1010	36	2	adoption	adoption	NOUN
cuesj-1010	36	3	of	of	ADP
cuesj-1010	36	4	residual	residual	ADJ
cuesj-1010	36	5	blocks	block	NOUN
cuesj-1010	36	6	enables	enable	VERB
cuesj-1010	36	7	yolo	yolo	ADV
cuesj-1010	36	8	to	to	PART
cuesj-1010	36	9	analyze	analyze	VERB
cuesj-1010	36	10	the	the	DET
cuesj-1010	36	11	entire	entire	ADJ
cuesj-1010	36	12	image	image	NOUN
cuesj-1010	36	13	in	in	ADP
cuesj-1010	36	14	a	a	DET
cuesj-1010	36	15	single	single	ADJ
cuesj-1010	36	16	pass	pass	NOUN
cuesj-1010	36	17	,	,	PUNCT
cuesj-1010	36	18	making	make	VERB
cuesj-1010	36	19	it	it	PRON
cuesj-1010	36	20	well	well	ADV
cuesj-1010	36	21	-	-	PUNCT
cuesj-1010	36	22	suited	suit	VERB
cuesj-1010	36	23	for	for	ADP
cuesj-1010	36	24	real	real	ADJ
cuesj-1010	36	25	-	-	PUNCT
cuesj-1010	36	26	world	world	NOUN
cuesj-1010	36	27	object	object	NOUN
cuesj-1010	36	28	recognition	recognition	NOUN
cuesj-1010	36	29	applications.[4	applications.[4	PRON
cuesj-1010	36	30	]	]	PUNCT
cuesj-1010	36	31	for	for	ADP
cuesj-1010	36	32	object	object	NOUN
cuesj-1010	36	33	position	position	NOUN
cuesj-1010	36	34	prediction	prediction	NOUN
cuesj-1010	36	35	,	,	PUNCT
cuesj-1010	36	36	yolo	yolo	PROPN
cuesj-1010	36	37	employs	employ	VERB
cuesj-1010	36	38	bounding	bound	VERB
cuesj-1010	36	39	box	box	NOUN
cuesj-1010	36	40	regression	regression	NOUN
cuesj-1010	36	41	.	.	PUNCT
cuesj-1010	37	1	this	this	PRON
cuesj-1010	37	2	involves	involve	VERB
cuesj-1010	37	3	estimating	estimate	VERB
cuesj-1010	37	4	the	the	DET
cuesj-1010	37	5	height	height	NOUN
cuesj-1010	37	6	,	,	PUNCT
cuesj-1010	37	7	width	width	ADJ
cuesj-1010	37	8	,	,	PUNCT
cuesj-1010	37	9	center	center	NOUN
cuesj-1010	37	10	,	,	PUNCT
cuesj-1010	37	11	and	and	CCONJ
cuesj-1010	37	12	class	class	NOUN
cuesj-1010	37	13	of	of	ADP
cuesj-1010	37	14	objects	object	NOUN
cuesj-1010	37	15	.	.	PUNCT
cuesj-1010	38	1	during	during	ADP
cuesj-1010	38	2	training	training	NOUN
cuesj-1010	38	3	,	,	PUNCT
cuesj-1010	38	4	the	the	DET
cuesj-1010	38	5	model	model	NOUN
cuesj-1010	38	6	adjusts	adjust	VERB
cuesj-1010	38	7	the	the	DET
cuesj-1010	38	8	bounding	bounding	NOUN
cuesj-1010	38	9	boxes	box	NOUN
cuesj-1010	38	10	based	base	VERB
cuesj-1010	38	11	on	on	ADP
cuesj-1010	38	12	the	the	DET
cuesj-1010	38	13	training	training	NOUN
cuesj-1010	38	14	data	datum	NOUN
cuesj-1010	38	15	to	to	PART
cuesj-1010	38	16	improve	improve	VERB
cuesj-1010	38	17	predictions	prediction	NOUN
cuesj-1010	38	18	.	.	PUNCT
cuesj-1010	39	1	this	this	DET
cuesj-1010	39	2	fine	fine	ADJ
cuesj-1010	39	3	-	-	PUNCT
cuesj-1010	39	4	tuning	tuning	NOUN
cuesj-1010	39	5	enhances	enhance	VERB
cuesj-1010	39	6	the	the	DET
cuesj-1010	39	7	model	model	NOUN
cuesj-1010	39	8	’s	’s	PART
cuesj-1010	39	9	accuracy	accuracy	NOUN
cuesj-1010	39	10	in	in	ADP
cuesj-1010	39	11	predicting	predict	VERB
cuesj-1010	39	12	object	object	NOUN
cuesj-1010	39	13	positions	position	NOUN
cuesj-1010	39	14	in	in	ADP
cuesj-1010	39	15	new	new	ADJ
cuesj-1010	39	16	images	image	NOUN
cuesj-1010	39	17	.	.	PUNCT
cuesj-1010	40	1	the	the	DET
cuesj-1010	40	2	next	next	ADJ
cuesj-1010	40	3	critical	critical	ADJ
cuesj-1010	40	4	concept	concept	NOUN
cuesj-1010	40	5	is	be	AUX
cuesj-1010	40	6	iou	iou	NOUN
cuesj-1010	40	7	,	,	PUNCT
cuesj-1010	40	8	also	also	ADV
cuesj-1010	40	9	known	know	VERB
cuesj-1010	40	10	as	as	ADP
cuesj-1010	40	11	the	the	DET
cuesj-1010	40	12	jaccard	jaccard	PROPN
cuesj-1010	40	13	index	index	NOUN
cuesj-1010	40	14	.	.	PUNCT
cuesj-1010	41	1	iou	iou	PROPN
cuesj-1010	41	2	is	be	AUX
cuesj-1010	41	3	a	a	DET
cuesj-1010	41	4	prominent	prominent	ADJ
cuesj-1010	41	5	evaluation	evaluation	NOUN
cuesj-1010	41	6	metric	metric	NOUN
cuesj-1010	41	7	for	for	ADP
cuesj-1010	41	8	object	object	NOUN
cuesj-1010	41	9	detection	detection	NOUN
cuesj-1010	41	10	and	and	CCONJ
cuesj-1010	41	11	is	be	AUX
cuesj-1010	41	12	commonly	commonly	ADV
cuesj-1010	41	13	used	use	VERB
cuesj-1010	41	14	to	to	PART
cuesj-1010	41	15	measure	measure	VERB
cuesj-1010	41	16	the	the	DET
cuesj-1010	41	17	similarity	similarity	NOUN
cuesj-1010	41	18	between	between	ADP
cuesj-1010	41	19	two	two	NUM
cuesj-1010	41	20	shapes	shape	NOUN
cuesj-1010	41	21	.	.	PUNCT
cuesj-1010	42	1	it	it	PRON
cuesj-1010	42	2	provides	provide	VERB
cuesj-1010	42	3	a	a	DET
cuesj-1010	42	4	normalized	normalize	VERB
cuesj-1010	42	5	measure	measure	NOUN
cuesj-1010	42	6	that	that	PRON
cuesj-1010	42	7	focuses	focus	VERB
cuesj-1010	42	8	on	on	ADP
cuesj-1010	42	9	the	the	DET
cuesj-1010	42	10	areas	area	NOUN
cuesj-1010	42	11	(	(	PUNCT
cuesj-1010	42	12	or	or	CCONJ
cuesj-1010	42	13	volumes	volume	NOUN
cuesj-1010	42	14	)	)	PUNCT
cuesj-1010	42	15	of	of	ADP
cuesj-1010	42	16	the	the	DET
cuesj-1010	42	17	compared	compare	VERB
cuesj-1010	42	18	objects	object	NOUN
cuesj-1010	42	19	,	,	PUNCT
cuesj-1010	42	20	taking	take	VERB
cuesj-1010	42	21	into	into	ADP
cuesj-1010	42	22	account	account	NOUN
cuesj-1010	42	23	their	their	PRON
cuesj-1010	42	24	parameters	parameter	NOUN
cuesj-1010	42	25	,	,	PUNCT
cuesj-1010	42	26	such	such	ADJ
cuesj-1010	42	27	as	as	ADP
cuesj-1010	42	28	width	width	ADJ
cuesj-1010	42	29	,	,	PUNCT
cuesj-1010	42	30	height	height	NOUN
cuesj-1010	42	31	,	,	PUNCT
cuesj-1010	42	32	and	and	CCONJ
cuesj-1010	42	33	placement	placement	NOUN
cuesj-1010	42	34	of	of	ADP
cuesj-1010	42	35	bounding	bounding	NOUN
cuesj-1010	42	36	boxes	box	NOUN
cuesj-1010	42	37	.	.	PUNCT
cuesj-1010	43	1	iou	iou	NOUN
cuesj-1010	43	2	is	be	AUX
cuesj-1010	43	3	scale	scale	NOUN
cuesj-1010	43	4	-	-	PUNCT
cuesj-1010	43	5	invariant	invariant	ADJ
cuesj-1010	43	6	,	,	PUNCT
cuesj-1010	43	7	making	make	VERB
cuesj-1010	43	8	it	it	PRON
cuesj-1010	43	9	a	a	DET
cuesj-1010	43	10	valuable	valuable	ADJ
cuesj-1010	43	11	metric	metric	NOUN
cuesj-1010	43	12	for	for	ADP
cuesj-1010	43	13	various	various	ADJ
cuesj-1010	43	14	performance	performance	NOUN
cuesj-1010	43	15	assessments	assessment	NOUN
cuesj-1010	43	16	in	in	ADP
cuesj-1010	43	17	segmentation	segmentation	NOUN
cuesj-1010	43	18	,	,	PUNCT
cuesj-1010	43	19	object	object	VERB
cuesj-1010	43	20	detection	detection	NOUN
cuesj-1010	43	21	,	,	PUNCT
cuesj-1010	43	22	and	and	CCONJ
cuesj-1010	43	23	tracking.[5	tracking.[5	NOUN
cuesj-1010	43	24	-	-	X
cuesj-1010	43	25	7	7	NUM
cuesj-1010	43	26	]	]	PUNCT
cuesj-1010	43	27	as	as	SCONJ
cuesj-1010	43	28	previously	previously	ADV
cuesj-1010	43	29	mentioned	mention	VERB
cuesj-1010	43	30	,	,	PUNCT
cuesj-1010	43	31	each	each	DET
cuesj-1010	43	32	bounding	bounding	NOUN
cuesj-1010	43	33	box	box	NOUN
cuesj-1010	43	34	(	(	PUNCT
cuesj-1010	43	35	y	y	NOUN
cuesj-1010	43	36	=	=	PUNCT
cuesj-1010	44	1	[	[	X
cuesj-1010	44	2	pc	pc	NOUN
cuesj-1010	44	3	,	,	PUNCT
cuesj-1010	44	4	bx	bx	PROPN
cuesj-1010	44	5	,	,	PUNCT
cuesj-1010	44	6	by	by	ADP
cuesj-1010	44	7	,	,	PUNCT
cuesj-1010	44	8	bh	bh	NOUN
cuesj-1010	44	9	,	,	PUNCT
cuesj-1010	44	10	bw	bw	NOUN
cuesj-1010	44	11	,	,	PUNCT
cuesj-1010	44	12	c	c	NOUN
cuesj-1010	44	13	]	]	PUNCT
cuesj-1010	44	14	)	)	PUNCT
cuesj-1010	44	15	in	in	ADP
cuesj-1010	44	16	the	the	DET
cuesj-1010	44	17	image	image	NOUN
cuesj-1010	44	18	comprises	comprise	VERB
cuesj-1010	44	19	features	feature	NOUN
cuesj-1010	44	20	such	such	ADJ
cuesj-1010	44	21	as	as	ADP
cuesj-1010	44	22	the	the	DET
cuesj-1010	44	23	bounding	bounding	NOUN
cuesj-1010	44	24	box	box	NOUN
cuesj-1010	44	25	center	center	NOUN
cuesj-1010	44	26	(	(	PUNCT
cuesj-1010	44	27	bx	bx	X
cuesj-1010	44	28	,	,	PUNCT
cuesj-1010	44	29	by	by	ADP
cuesj-1010	44	30	)	)	PUNCT
cuesj-1010	44	31	,	,	PUNCT
cuesj-1010	44	32	class	class	NOUN
cuesj-1010	44	33	label	label	NOUN
cuesj-1010	44	34	(	(	PUNCT
cuesj-1010	44	35	c	c	NOUN
cuesj-1010	44	36	)	)	PUNCT
cuesj-1010	44	37	,	,	PUNCT
cuesj-1010	44	38	height	height	NOUN
cuesj-1010	44	39	(	(	PUNCT
cuesj-1010	44	40	bh	bh	NOUN
cuesj-1010	44	41	)	)	PUNCT
cuesj-1010	44	42	,	,	PUNCT
cuesj-1010	44	43	and	and	CCONJ
cuesj-1010	44	44	width	width	ADJ
cuesj-1010	44	45	(	(	PUNCT
cuesj-1010	44	46	bw	bw	NOUN
cuesj-1010	44	47	)	)	PUNCT
cuesj-1010	44	48	.	.	PUNCT
cuesj-1010	45	1	a	a	DET
cuesj-1010	45	2	regression	regression	NOUN
cuesj-1010	45	3	of	of	ADP
cuesj-1010	45	4	a	a	DET
cuesj-1010	45	5	single	single	ADJ
cuesj-1010	45	6	bounding	bounding	NOUN
cuesj-1010	45	7	box	box	NOUN
cuesj-1010	45	8	is	be	AUX
cuesj-1010	45	9	employed	employ	VERB
cuesj-1010	45	10	to	to	PART
cuesj-1010	45	11	predict	predict	VERB
cuesj-1010	45	12	the	the	DET
cuesj-1010	45	13	class	class	NOUN
cuesj-1010	45	14	,	,	PUNCT
cuesj-1010	45	15	center	center	NOUN
cuesj-1010	45	16	,	,	PUNCT
cuesj-1010	45	17	width	width	ADJ
cuesj-1010	45	18	,	,	PUNCT
cuesj-1010	45	19	and	and	CCONJ
cuesj-1010	45	20	height	height	NOUN
cuesj-1010	45	21	of	of	ADP
cuesj-1010	45	22	objects	object	NOUN
cuesj-1010	45	23	figure	figure	NOUN
cuesj-1010	45	24	1	1	NUM
cuesj-1010	45	25	.	.	PUNCT
cuesj-1010	46	1	iou	iou	PROPN
cuesj-1010	46	2	attends	attend	VERB
cuesj-1010	46	3	by	by	ADP
cuesj-1010	46	4	below	below	ADV
cuesj-1010	46	5	(	(	PUNCT
cuesj-1010	46	6	1	1	NUM
cuesj-1010	46	7	):	):	PUNCT
cuesj-1010	46	8	=	=	PUNCT
cuesj-1010	46	9	sec	sec	PROPN
cuesj-1010	46	10	_	_	PUNCT
cuesj-1010	46	11	(	(	PUNCT
cuesj-1010	46	12	)	)	PUNCT
cuesj-1010	47	1	_	_	PUNCT
cuesj-1010	47	2	(	(	PUNCT
cuesj-1010	47	3	)	)	PUNCT
cuesj-1010	47	4	inter	inter	PROPN
cuesj-1010	47	5	tion	tion	PROPN
cuesj-1010	47	6	arear	arear	PROPN
cuesj-1010	47	7	ia	ia	PROPN
cuesj-1010	47	8	iou	iou	PROPN
cuesj-1010	47	9	uion	uion	PROPN
cuesj-1010	47	10	area	area	PROPN
cuesj-1010	47	11	ua	ua	PROPN
cuesj-1010	47	12	(	(	PUNCT
cuesj-1010	47	13	1	1	X
cuesj-1010	47	14	)	)	PUNCT
cuesj-1010	47	15	ua	ua	NOUN
cuesj-1010	47	16	=	=	SYM
cuesj-1010	47	17	(	(	PUNCT
cuesj-1010	47	18	x22	x22	PROPN
cuesj-1010	47	19	–	–	PUNCT
cuesj-1010	47	20	x21	x21	NUM
cuesj-1010	47	21	)	)	PUNCT
cuesj-1010	47	22	*	*	PUNCT
cuesj-1010	48	1	(	(	PUNCT
cuesj-1010	48	2	y22	y22	PROPN
cuesj-1010	48	3	–	–	PUNCT
cuesj-1010	48	4	y21	y21	NOUN
cuesj-1010	48	5	)	)	PUNCT
cuesj-1010	49	1	+	+	CCONJ
cuesj-1010	49	2	(	(	PUNCT
cuesj-1010	49	3	x21	x21	NUM
cuesj-1010	49	4	–	–	PUNCT
cuesj-1010	49	5	x11	x11	NUM
cuesj-1010	49	6	)	)	PUNCT
cuesj-1010	49	7	*	*	PUNCT
cuesj-1010	50	1	(	(	PUNCT
cuesj-1010	50	2	y12	y12	NUM
cuesj-1010	50	3	–	–	PUNCT
cuesj-1010	50	4	y11	y11	NOUN
cuesj-1010	50	5	)	)	PUNCT
cuesj-1010	50	6	–	–	PUNCT
cuesj-1010	50	7	ia	ia	PROPN
cuesj-1010	50	8	inter	inter	PROPN
cuesj-1010	50	9	sec	sec	PROPN
cuesj-1010	50	10	tion_height	tion_height	PROPN
cuesj-1010	50	11	(	(	PUNCT
cuesj-1010	50	12	ih	ih	NOUN
cuesj-1010	50	13	)	)	PUNCT
cuesj-1010	50	14	=	=	SYM
cuesj-1010	50	15	min	min	PROPN
cuesj-1010	50	16	(	(	PUNCT
cuesj-1010	50	17	y12	y12	PROPN
cuesj-1010	50	18	,	,	PUNCT
cuesj-1010	50	19	y22	y22	PROPN
cuesj-1010	50	20	)	)	PUNCT
cuesj-1010	50	21	–	–	PUNCT
cuesj-1010	50	22	max	max	PROPN
cuesj-1010	50	23	(	(	PUNCT
cuesj-1010	50	24	y11	y11	NOUN
cuesj-1010	50	25	,	,	PUNCT
cuesj-1010	50	26	y21	y21	PROPN
cuesj-1010	50	27	)	)	PUNCT
cuesj-1010	50	28	inter	inter	PROPN
cuesj-1010	50	29	sec	sec	PROPN
cuesj-1010	50	30	tion_width	tion_width	NOUN
cuesj-1010	50	31	(	(	PUNCT
cuesj-1010	50	32	iw	iw	NOUN
cuesj-1010	50	33	)	)	PUNCT
cuesj-1010	50	34	=	=	SYM
cuesj-1010	50	35	min	min	PROPN
cuesj-1010	50	36	(	(	PUNCT
cuesj-1010	50	37	x12	x12	NUM
cuesj-1010	50	38	,	,	PUNCT
cuesj-1010	50	39	x22	x22	NUM
cuesj-1010	50	40	)	)	PUNCT
cuesj-1010	50	41	–	–	PUNCT
cuesj-1010	50	42	max	max	PROPN
cuesj-1010	50	43	(	(	PUNCT
cuesj-1010	50	44	x11	x11	PROPN
cuesj-1010	50	45	,	,	PUNCT
cuesj-1010	50	46	x21	x21	PROPN
cuesj-1010	50	47	)	)	PUNCT
cuesj-1010	50	48	ia	ia	PROPN
cuesj-1010	50	49	=	=	SYM
cuesj-1010	50	50	ih	ih	PROPN
cuesj-1010	50	51	*	*	PUNCT
cuesj-1010	50	52	iw	iw	INTJ
cuesj-1010	50	53	the	the	DET
cuesj-1010	50	54	calculation	calculation	NOUN
cuesj-1010	50	55	of	of	ADP
cuesj-1010	50	56	iou	iou	NOUN
cuesj-1010	50	57	is	be	AUX
cuesj-1010	50	58	represented	represent	VERB
cuesj-1010	50	59	by	by	ADP
cuesj-1010	50	60	(	(	PUNCT
cuesj-1010	50	61	1	1	NUM
cuesj-1010	50	62	)	)	PUNCT
cuesj-1010	50	63	,	,	PUNCT
cuesj-1010	50	64	where	where	SCONJ
cuesj-1010	50	65	we	we	PRON
cuesj-1010	50	66	determine	determine	VERB
cuesj-1010	50	67	the	the	DET
cuesj-1010	50	68	overlap	overlap	NOUN
cuesj-1010	50	69	area	area	NOUN
cuesj-1010	50	70	between	between	ADP
cuesj-1010	50	71	the	the	DET
cuesj-1010	50	72	ground	ground	NOUN
cuesj-1010	50	73	-	-	PUNCT
cuesj-1010	50	74	truth	truth	NOUN
cuesj-1010	50	75	bounding	bounding	NOUN
cuesj-1010	50	76	box	box	NOUN
cuesj-1010	50	77	and	and	CCONJ
cuesj-1010	50	78	the	the	DET
cuesj-1010	50	79	predicted	predict	VERB
cuesj-1010	50	80	bounding	bounding	NOUN
cuesj-1010	50	81	box	box	NOUN
cuesj-1010	50	82	in	in	ADP
cuesj-1010	50	83	the	the	DET
cuesj-1010	50	84	intersection	intersection	NOUN
cuesj-1010	50	85	area	area	NOUN
cuesj-1010	50	86	.	.	PUNCT
cuesj-1010	51	1	the	the	DET
cuesj-1010	51	2	union	union	PROPN
cuesj-1010	51	3	area	area	NOUN
cuesj-1010	51	4	covers	cover	VERB
cuesj-1010	51	5	both	both	DET
cuesj-1010	51	6	the	the	DET
cuesj-1010	51	7	predicted	predict	VERB
cuesj-1010	51	8	and	and	CCONJ
cuesj-1010	51	9	ground	ground	NOUN
cuesj-1010	51	10	-	-	PUNCT
cuesj-1010	51	11	truth	truth	NOUN
cuesj-1010	51	12	bounding	bounding	NOUN
cuesj-1010	51	13	boxes	box	NOUN
cuesj-1010	51	14	.	.	PUNCT
cuesj-1010	52	1	the	the	DET
cuesj-1010	52	2	resulting	result	VERB
cuesj-1010	52	3	iou	iou	NOUN
cuesj-1010	52	4	score	score	NOUN
cuesj-1010	52	5	,	,	PUNCT
cuesj-1010	52	6	when	when	SCONJ
cuesj-1010	52	7	>	>	X
cuesj-1010	52	8	0.5	0.5	NUM
cuesj-1010	52	9	,	,	PUNCT
cuesj-1010	52	10	signifies	signify	VERB
cuesj-1010	52	11	a	a	DET
cuesj-1010	52	12	successful	successful	ADJ
cuesj-1010	52	13	object	object	NOUN
cuesj-1010	52	14	detection	detection	NOUN
cuesj-1010	52	15	,	,	PUNCT
cuesj-1010	52	16	while	while	SCONJ
cuesj-1010	52	17	values	value	NOUN
cuesj-1010	52	18	below	below	ADV
cuesj-1010	52	19	indicate	indicate	VERB
cuesj-1010	52	20	detection	detection	NOUN
cuesj-1010	52	21	failure.[8	failure.[8	PUNCT
cuesj-1010	52	22	]	]	X
cuesj-1010	52	23	the	the	DET
cuesj-1010	52	24	architecture	architecture	NOUN
cuesj-1010	52	25	of	of	ADP
cuesj-1010	52	26	yolov5	yolov5	NOUN
cuesj-1010	52	27	the	the	DET
cuesj-1010	52	28	network	network	NOUN
cuesj-1010	52	29	architecture	architecture	NOUN
cuesj-1010	52	30	of	of	ADP
cuesj-1010	52	31	yolov5	yolov5	NOUN
cuesj-1010	52	32	is	be	AUX
cuesj-1010	52	33	presented	present	VERB
cuesj-1010	52	34	in	in	ADP
cuesj-1010	52	35	figure	figure	NOUN
cuesj-1010	52	36	2	2	NUM
cuesj-1010	52	37	.	.	PUNCT
cuesj-1010	52	38	choosing	choose	VERB
cuesj-1010	52	39	yolov5	yolov5	NOUN
cuesj-1010	52	40	as	as	SCONJ
cuesj-1010	52	41	our	our	PRON
cuesj-1010	52	42	initial	initial	ADJ
cuesj-1010	52	43	learner	learner	NOUN
cuesj-1010	52	44	has	have	VERB
cuesj-1010	52	45	three	three	NUM
cuesj-1010	52	46	causes	cause	NOUN
cuesj-1010	52	47	.	.	PUNCT
cuesj-1010	53	1	first	first	ADV
cuesj-1010	53	2	,	,	PUNCT
cuesj-1010	53	3	yolov5	yolov5	NOUN
cuesj-1010	53	4	,	,	PUNCT
cuesj-1010	53	5	much	much	ADJ
cuesj-1010	53	6	like	like	ADP
cuesj-1010	53	7	yolov4	yolov4	PROPN
cuesj-1010	53	8	,	,	PUNCT
cuesj-1010	53	9	introduces	introduce	VERB
cuesj-1010	53	10	the	the	DET
cuesj-1010	53	11	crossstage	crossstage	NOUN
cuesj-1010	53	12	partial	partial	ADJ
cuesj-1010	53	13	network	network	NOUN
cuesj-1010	53	14	(	(	PUNCT
cuesj-1010	53	15	cspnet	cspnet	NOUN
cuesj-1010	53	16	)	)	PUNCT
cuesj-1010	53	17	into	into	ADP
cuesj-1010	53	18	the	the	DET
cuesj-1010	53	19	darknet	darknet	NOUN
cuesj-1010	53	20	framework	framework	NOUN
cuesj-1010	53	21	,	,	PUNCT
cuesj-1010	53	22	forming	form	VERB
cuesj-1010	53	23	the	the	DET
cuesj-1010	53	24	foundation	foundation	NOUN
cuesj-1010	53	25	known	know	VERB
cuesj-1010	53	26	as	as	ADP
cuesj-1010	53	27	cspdarknet	cspdarknet	NOUN
cuesj-1010	53	28	.	.	PUNCT
cuesj-1010	54	1	the	the	DET
cuesj-1010	54	2	incorporation	incorporation	NOUN
cuesj-1010	54	3	of	of	ADP
cuesj-1010	54	4	cspnet	cspnet	NOUN
cuesj-1010	54	5	is	be	AUX
cuesj-1010	54	6	primarily	primarily	ADV
cuesj-1010	54	7	designed	design	VERB
cuesj-1010	54	8	to	to	PART
cuesj-1010	54	9	enhance	enhance	VERB
cuesj-1010	54	10	the	the	DET
cuesj-1010	54	11	network	network	NOUN
cuesj-1010	54	12	’s	’s	PART
cuesj-1010	54	13	gradient	gradient	NOUN
cuesj-1010	54	14	combinations	combination	NOUN
cuesj-1010	54	15	,	,	PUNCT
cuesj-1010	54	16	all	all	PRON
cuesj-1010	54	17	while	while	SCONJ
cuesj-1010	54	18	demanding	demand	VERB
cuesj-1010	54	19	fewer	few	ADJ
cuesj-1010	54	20	computational	computational	ADJ
cuesj-1010	54	21	resources	resource	NOUN
cuesj-1010	54	22	.	.	PUNCT
cuesj-1010	55	1	this	this	PRON
cuesj-1010	55	2	is	be	AUX
cuesj-1010	55	3	achieved	achieve	VERB
cuesj-1010	55	4	by	by	ADP
cuesj-1010	55	5	dividing	divide	VERB
cuesj-1010	55	6	the	the	DET
cuesj-1010	55	7	base	base	NOUN
cuesj-1010	55	8	layer	layer	NOUN
cuesj-1010	55	9	’s	’s	PART
cuesj-1010	55	10	feature	feature	NOUN
cuesj-1010	55	11	map	map	NOUN
cuesj-1010	55	12	into	into	ADP
cuesj-1010	55	13	two	two	NUM
cuesj-1010	55	14	distinct	distinct	ADJ
cuesj-1010	55	15	parts	part	NOUN
cuesj-1010	55	16	and	and	CCONJ
cuesj-1010	55	17	skillfully	skillfully	ADV
cuesj-1010	55	18	integrating	integrate	VERB
cuesj-1010	55	19	them	they	PRON
cuesj-1010	55	20	using	use	VERB
cuesj-1010	55	21	a	a	DET
cuesj-1010	55	22	suggested	suggest	VERB
cuesj-1010	55	23	cross	cross	ADJ
cuesj-1010	55	24	-	-	ADJ
cuesj-1010	55	25	stage	stage	ADJ
cuesj-1010	55	26	hierarchy	hierarchy	NOUN
cuesj-1010	55	27	.	.	PUNCT
cuesj-1010	56	1	notably	notably	ADV
cuesj-1010	56	2	,	,	PUNCT
cuesj-1010	56	3	cspnet	cspnet	NOUN
cuesj-1010	56	4	significantly	significantly	ADV
cuesj-1010	56	5	reduces	reduce	VERB
cuesj-1010	56	6	computational	computational	ADJ
cuesj-1010	56	7	overhead	overhead	NOUN
cuesj-1010	56	8	,	,	PUNCT
cuesj-1010	56	9	leading	lead	VERB
cuesj-1010	56	10	to	to	ADP
cuesj-1010	56	11	improved	improved	ADJ
cuesj-1010	56	12	speed	speed	NOUN
cuesj-1010	56	13	and	and	CCONJ
cuesj-1010	56	14	accuracy	accuracy	NOUN
cuesj-1010	56	15	during	during	ADP
cuesj-1010	56	16	inference.[9,10	inference.[9,10	NOUN
cuesj-1010	56	17	]	]	X
cuesj-1010	56	18	second	second	ADJ
cuesj-1010	56	19	,	,	PUNCT
cuesj-1010	56	20	yolov5	yolov5	NOUN
cuesj-1010	56	21	leverages	leverage	VERB
cuesj-1010	56	22	the	the	DET
cuesj-1010	56	23	path	path	NOUN
cuesj-1010	56	24	aggregation	aggregation	NOUN
cuesj-1010	56	25	network	network	NOUN
cuesj-1010	56	26	(	(	PUNCT
cuesj-1010	56	27	panet	panet	PROPN
cuesj-1010	56	28	)	)	PUNCT
cuesj-1010	56	29	as	as	ADP
cuesj-1010	56	30	its	its	PRON
cuesj-1010	56	31	“	"	PUNCT
cuesj-1010	56	32	neck	neck	NOUN
cuesj-1010	56	33	”	"	PUNCT
cuesj-1010	56	34	to	to	PART
cuesj-1010	56	35	enhance	enhance	VERB
cuesj-1010	56	36	information	information	NOUN
cuesj-1010	56	37	flow	flow	NOUN
cuesj-1010	56	38	.	.	PUNCT
cuesj-1010	57	1	panet	panet	PROPN
cuesj-1010	57	2	adopts	adopt	VERB
cuesj-1010	57	3	a	a	DET
cuesj-1010	57	4	novel	novel	ADJ
cuesj-1010	57	5	feature	feature	NOUN
cuesj-1010	57	6	pyramid	pyramid	NOUN
cuesj-1010	57	7	network	network	NOUN
cuesj-1010	57	8	(	(	PUNCT
cuesj-1010	57	9	fpn	fpn	PROPN
cuesj-1010	57	10	)	)	PUNCT
cuesj-1010	57	11	topology	topology	NOUN
cuesj-1010	57	12	with	with	ADP
cuesj-1010	57	13	an	an	DET
cuesj-1010	57	14	augmented	augmented	ADJ
cuesj-1010	57	15	bottom	bottom	ADJ
cuesj-1010	57	16	-	-	PUNCT
cuesj-1010	57	17	up	up	ADP
cuesj-1010	57	18	pathway	pathway	NOUN
cuesj-1010	57	19	,	,	PUNCT
cuesj-1010	57	20	facilitating	facilitate	VERB
cuesj-1010	57	21	the	the	DET
cuesj-1010	57	22	propagation	propagation	NOUN
cuesj-1010	57	23	of	of	ADP
cuesj-1010	57	24	low	low	ADJ
cuesj-1010	57	25	-	-	PUNCT
cuesj-1010	57	26	level	level	NOUN
cuesj-1010	57	27	features	feature	NOUN
cuesj-1010	57	28	.	.	PUNCT
cuesj-1010	58	1	it	it	PRON
cuesj-1010	58	2	further	far	ADV
cuesj-1010	58	3	employs	employ	VERB
cuesj-1010	58	4	adaptive	adaptive	ADJ
cuesj-1010	58	5	feature	feature	NOUN
cuesj-1010	58	6	pooling	pooling	NOUN
cuesj-1010	58	7	,	,	PUNCT
cuesj-1010	58	8	effectively	effectively	ADV
cuesj-1010	58	9	linking	link	VERB
cuesj-1010	58	10	feature	feature	NOUN
cuesj-1010	58	11	grids	grid	NOUN
cuesj-1010	58	12	with	with	ADP
cuesj-1010	58	13	various	various	ADJ
cuesj-1010	58	14	levels	level	NOUN
cuesj-1010	58	15	of	of	ADP
cuesj-1010	58	16	features	feature	NOUN
cuesj-1010	58	17	.	.	PUNCT
cuesj-1010	59	1	this	this	DET
cuesj-1010	59	2	design	design	NOUN
cuesj-1010	59	3	choice	choice	NOUN
cuesj-1010	59	4	enables	enable	VERB
cuesj-1010	59	5	vital	vital	ADJ
cuesj-1010	59	6	information	information	NOUN
cuesj-1010	59	7	from	from	ADP
cuesj-1010	59	8	every	every	DET
cuesj-1010	59	9	level	level	NOUN
cuesj-1010	59	10	of	of	ADP
cuesj-1010	59	11	features	feature	NOUN
cuesj-1010	59	12	to	to	PART
cuesj-1010	59	13	promptly	promptly	ADV
cuesj-1010	59	14	propagate	propagate	VERB
cuesj-1010	59	15	to	to	ADP
cuesj-1010	59	16	the	the	DET
cuesj-1010	59	17	subsequent	subsequent	ADJ
cuesj-1010	59	18	subnetwork	subnetwork	NOUN
cuesj-1010	59	19	.	.	PUNCT
cuesj-1010	60	1	panet	panet	PROPN
cuesj-1010	60	2	amplifies	amplify	VERB
cuesj-1010	60	3	the	the	DET
cuesj-1010	60	4	utilization	utilization	NOUN
cuesj-1010	60	5	of	of	ADP
cuesj-1010	60	6	precise	precise	ADJ
cuesj-1010	60	7	localization	localization	NOUN
cuesj-1010	60	8	cues	cue	NOUN
cuesj-1010	60	9	in	in	ADP
cuesj-1010	60	10	lower	low	ADJ
cuesj-1010	60	11	layers	layer	NOUN
cuesj-1010	60	12	,	,	PUNCT
cuesj-1010	60	13	thereby	thereby	ADV
cuesj-1010	60	14	enhancing	enhance	VERB
cuesj-1010	60	15	the	the	DET
cuesj-1010	60	16	accuracy	accuracy	NOUN
cuesj-1010	60	17	of	of	ADP
cuesj-1010	60	18	object	object	NOUN
cuesj-1010	60	19	localization.[11,12	localization.[11,12	NOUN
cuesj-1010	60	20	]	]	X
cuesj-1010	60	21	third	third	ADV
cuesj-1010	60	22	,	,	PUNCT
cuesj-1010	60	23	the	the	DET
cuesj-1010	60	24	yolov5	yolov5	NOUN
cuesj-1010	60	25	“	"	PUNCT
cuesj-1010	60	26	head	head	NOUN
cuesj-1010	60	27	,	,	PUNCT
cuesj-1010	60	28	”	"	PUNCT
cuesj-1010	60	29	encompassing	encompass	VERB
cuesj-1010	60	30	the	the	DET
cuesj-1010	60	31	yolo	yolo	ADJ
cuesj-1010	60	32	layer	layer	NOUN
cuesj-1010	60	33	,	,	PUNCT
cuesj-1010	60	34	is	be	AUX
cuesj-1010	60	35	responsible	responsible	ADJ
cuesj-1010	60	36	for	for	ADP
cuesj-1010	60	37	producing	produce	VERB
cuesj-1010	60	38	multi	multi	ADJ
cuesj-1010	60	39	-	-	ADJ
cuesj-1010	60	40	scale	scale	ADJ
cuesj-1010	60	41	predictions	prediction	NOUN
cuesj-1010	60	42	.	.	PUNCT
cuesj-1010	61	1	it	it	PRON
cuesj-1010	61	2	generates	generate	VERB
cuesj-1010	61	3	three	three	NUM
cuesj-1010	61	4	distinct	distinct	ADJ
cuesj-1010	61	5	feature	feature	NOUN
cuesj-1010	61	6	map	map	NOUN
cuesj-1010	61	7	sizes	size	NOUN
cuesj-1010	61	8	,	,	PUNCT
cuesj-1010	61	9	namely	namely	ADV
cuesj-1010	61	10	72	72	NUM
cuesj-1010	61	11	×	×	NOUN
cuesj-1010	61	12	72	72	NUM
cuesj-1010	61	13	,	,	PUNCT
cuesj-1010	61	14	36	36	NUM
cuesj-1010	61	15	×	×	NOUN
cuesj-1010	61	16	36	36	NUM
cuesj-1010	61	17	,	,	PUNCT
cuesj-1010	61	18	and	and	CCONJ
cuesj-1010	61	19	18	18	NUM
cuesj-1010	61	20	×	×	NOUN
cuesj-1010	61	21	18	18	NUM
cuesj-1010	61	22	,	,	PUNCT
cuesj-1010	61	23	enabling	enable	VERB
cuesj-1010	61	24	the	the	DET
cuesj-1010	61	25	model	model	NOUN
cuesj-1010	61	26	to	to	PART
cuesj-1010	61	27	effectively	effectively	ADV
cuesj-1010	61	28	handle	handle	VERB
cuesj-1010	61	29	objects	object	NOUN
cuesj-1010	61	30	of	of	ADP
cuesj-1010	61	31	various	various	ADJ
cuesj-1010	61	32	scales	scale	NOUN
cuesj-1010	61	33	,	,	PUNCT
cuesj-1010	61	34	encompassing	encompass	VERB
cuesj-1010	61	35	those	those	PRON
cuesj-1010	61	36	that	that	PRON
cuesj-1010	61	37	are	be	AUX
cuesj-1010	61	38	large	large	ADJ
cuesj-1010	61	39	,	,	PUNCT
cuesj-1010	61	40	medium	medium	ADJ
cuesj-1010	61	41	,	,	PUNCT
cuesj-1010	61	42	and	and	CCONJ
cuesj-1010	61	43	small.[12	small.[12	NOUN
cuesj-1010	61	44	]	]	X
cuesj-1010	61	45	moreover	moreover	ADV
cuesj-1010	61	46	,	,	PUNCT
cuesj-1010	61	47	one	one	NUM
cuesj-1010	61	48	of	of	ADP
cuesj-1010	61	49	the	the	DET
cuesj-1010	61	50	most	most	ADV
cuesj-1010	61	51	captivating	captivating	ADJ
cuesj-1010	61	52	features	feature	NOUN
cuesj-1010	61	53	of	of	ADP
cuesj-1010	61	54	yolov5	yolov5	NOUN
cuesj-1010	61	55	is	be	AUX
cuesj-1010	61	56	its	its	PRON
cuesj-1010	61	57	versatility	versatility	NOUN
cuesj-1010	61	58	in	in	ADP
cuesj-1010	61	59	offering	offer	VERB
cuesj-1010	61	60	ten	ten	NUM
cuesj-1010	61	61	different	different	ADJ
cuesj-1010	61	62	types	type	NOUN
cuesj-1010	61	63	,	,	PUNCT
cuesj-1010	61	64	with	with	ADP
cuesj-1010	61	65	the	the	DET
cuesj-1010	61	66	most	most	ADV
cuesj-1010	61	67	common	common	ADJ
cuesj-1010	61	68	ones	one	NOUN
cuesj-1010	61	69	being	be	AUX
cuesj-1010	61	70	categorized	categorize	VERB
cuesj-1010	61	71	into	into	ADP
cuesj-1010	61	72	small	small	ADJ
cuesj-1010	61	73	,	,	PUNCT
cuesj-1010	61	74	medium	medium	ADJ
cuesj-1010	61	75	,	,	PUNCT
cuesj-1010	61	76	large	large	ADJ
cuesj-1010	61	77	,	,	PUNCT
cuesj-1010	61	78	and	and	CCONJ
cuesj-1010	61	79	extra	extra	ADJ
cuesj-1010	61	80	-	-	ADJ
cuesj-1010	61	81	large	large	ADJ
cuesj-1010	61	82	variations	variation	NOUN
cuesj-1010	61	83	.	.	PUNCT
cuesj-1010	62	1	these	these	DET
cuesj-1010	62	2	classifications	classification	NOUN
cuesj-1010	62	3	are	be	AUX
cuesj-1010	62	4	distinguished	distinguish	VERB
cuesj-1010	62	5	primarily	primarily	ADV
cuesj-1010	62	6	by	by	ADP
cuesj-1010	62	7	the	the	DET
cuesj-1010	62	8	number	number	NOUN
cuesj-1010	62	9	of	of	ADP
cuesj-1010	62	10	feature	feature	NOUN
cuesj-1010	62	11	extraction	extraction	NOUN
cuesj-1010	62	12	modules	module	NOUN
cuesj-1010	62	13	and	and	CCONJ
cuesj-1010	62	14	convolution	convolution	NOUN
cuesj-1010	62	15	kernels	kernel	NOUN
cuesj-1010	62	16	they	they	PRON
cuesj-1010	62	17	incorporate	incorporate	VERB
cuesj-1010	62	18	.	.	PUNCT
cuesj-1010	63	1	this	this	DET
cuesj-1010	63	2	distinction	distinction	NOUN
cuesj-1010	63	3	carries	carry	VERB
cuesj-1010	63	4	substantial	substantial	ADJ
cuesj-1010	63	5	practical	practical	ADJ
cuesj-1010	63	6	implications	implication	NOUN
cuesj-1010	63	7	when	when	SCONJ
cuesj-1010	63	8	working	work	VERB
cuesj-1010	63	9	with	with	ADP
cuesj-1010	63	10	yolo	yolo	ADJ
cuesj-1010	63	11	models	model	NOUN
cuesj-1010	63	12	,	,	PUNCT
cuesj-1010	63	13	as	as	SCONJ
cuesj-1010	63	14	it	it	PRON
cuesj-1010	63	15	directly	directly	ADV
cuesj-1010	63	16	impacts	impact	VERB
cuesj-1010	63	17	the	the	DET
cuesj-1010	63	18	model	model	NOUN
cuesj-1010	63	19	’s	’s	PART
cuesj-1010	63	20	capabilities	capability	NOUN
cuesj-1010	63	21	and	and	CCONJ
cuesj-1010	63	22	resource	resource	NOUN
cuesj-1010	63	23	requirements	requirement	NOUN
cuesj-1010	63	24	.	.	PUNCT
cuesj-1010	64	1	moreover	moreover	ADV
cuesj-1010	64	2	,	,	PUNCT
cuesj-1010	64	3	these	these	DET
cuesj-1010	64	4	variations	variation	NOUN
cuesj-1010	64	5	in	in	ADP
cuesj-1010	64	6	yolov5	yolov5	NOUN
cuesj-1010	64	7	also	also	ADV
cuesj-1010	64	8	translate	translate	VERB
cuesj-1010	64	9	into	into	ADP
cuesj-1010	64	10	variations	variation	NOUN
cuesj-1010	64	11	in	in	ADP
cuesj-1010	64	12	the	the	DET
cuesj-1010	64	13	quantity	quantity	NOUN
cuesj-1010	64	14	of	of	ADP
cuesj-1010	64	15	neural	neural	ADJ
cuesj-1010	64	16	network	network	NOUN
cuesj-1010	64	17	parameters	parameter	NOUN
cuesj-1010	64	18	.	.	PUNCT
cuesj-1010	65	1	figure	figure	VERB
cuesj-1010	65	2	1	1	NUM
cuesj-1010	65	3	:	:	PUNCT
cuesj-1010	65	4	example	example	NOUN
cuesj-1010	65	5	of	of	ADP
cuesj-1010	65	6	applying	apply	VERB
cuesj-1010	65	7	bbox	bbox	NOUN
cuesj-1010	65	8	to	to	ADP
cuesj-1010	65	9	a	a	DET
cuesj-1010	65	10	grid	grid	NOUN
cuesj-1010	65	11	class	class	NOUN
cuesj-1010	65	12	abdallah	abdallah	PROPN
cuesj-1010	65	13	:	:	PUNCT
cuesj-1010	65	14	vehicle	vehicle	NOUN
cuesj-1010	65	15	detection	detection	NOUN
cuesj-1010	65	16	with	with	ADP
cuesj-1010	65	17	tello	tello	PROPN
cuesj-1010	65	18	drone	drone	PROPN
cuesj-1010	65	19	and	and	CCONJ
cuesj-1010	65	20	yolov5	yolov5	NOUN
cuesj-1010	65	21	3	3	NUM
cuesj-1010	65	22	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	65	23	cuesj	cuesj	NOUN
cuesj-1010	65	24	2024	2024	NUM
cuesj-1010	65	25	,	,	PUNCT
cuesj-1010	65	26	8	8	NUM
cuesj-1010	65	27	(	(	PUNCT
cuesj-1010	65	28	1	1	NUM
cuesj-1010	65	29	):	):	PUNCT
cuesj-1010	65	30	1	1	NUM
cuesj-1010	65	31	-	-	SYM
cuesj-1010	65	32	7	7	NUM
cuesj-1010	65	33	figure	figure	NOUN
cuesj-1010	65	34	2	2	NUM
cuesj-1010	65	35	:	:	PUNCT
cuesj-1010	65	36	the	the	DET
cuesj-1010	65	37	network	network	NOUN
cuesj-1010	65	38	structure	structure	NOUN
cuesj-1010	65	39	of	of	ADP
cuesj-1010	65	40	you	you	PRON
cuesj-1010	65	41	only	only	ADV
cuesj-1010	65	42	looks	look	VERB
cuesj-1010	65	43	once	once	ADV
cuesj-1010	65	44	(	(	PUNCT
cuesj-1010	65	45	yolo	yolo	ADJ
cuesj-1010	65	46	)	)	PUNCT
cuesj-1010	65	47	.	.	PUNCT
cuesj-1010	66	1	(	(	PUNCT
cuesj-1010	66	2	a	a	X
cuesj-1010	66	3	)	)	PUNCT
cuesj-1010	66	4	cspdarknet	cspdarknet	NOUN
cuesj-1010	66	5	.	.	PUNCT
cuesj-1010	67	1	(	(	PUNCT
cuesj-1010	67	2	b	b	X
cuesj-1010	67	3	)	)	PUNCT
cuesj-1010	67	4	panet	panet	ADJ
cuesj-1010	67	5	neck	neck	NOUN
cuesj-1010	67	6	.	.	PUNCT
cuesj-1010	68	1	(	(	PUNCT
cuesj-1010	68	2	c	c	X
cuesj-1010	68	3	)	)	PUNCT
cuesj-1010	68	4	yolo	yolo	ADJ
cuesj-1010	68	5	head	head	NOUN
cuesj-1010	68	6	notably	notably	ADV
cuesj-1010	68	7	,	,	PUNCT
cuesj-1010	68	8	the	the	DET
cuesj-1010	68	9	size	size	NOUN
cuesj-1010	68	10	of	of	ADP
cuesj-1010	68	11	these	these	DET
cuesj-1010	68	12	parameters	parameter	NOUN
cuesj-1010	68	13	varies	vary	VERB
cuesj-1010	68	14	across	across	ADP
cuesj-1010	68	15	yolov5s	yolov5s	PROPN
cuesj-1010	68	16	,	,	PUNCT
cuesj-1010	68	17	yolov5	yolov5	PROPN
cuesj-1010	68	18	m	m	PROPN
cuesj-1010	68	19	,	,	PUNCT
cuesj-1010	68	20	yolov5l	yolov5l	NOUN
cuesj-1010	68	21	,	,	PUNCT
cuesj-1010	68	22	and	and	CCONJ
cuesj-1010	68	23	yolov5x	yolov5x	PROPN
cuesj-1010	68	24	.	.	PUNCT
cuesj-1010	69	1	the	the	DET
cuesj-1010	69	2	file	file	NOUN
cuesj-1010	69	3	sizes	size	NOUN
cuesj-1010	69	4	for	for	ADP
cuesj-1010	69	5	these	these	DET
cuesj-1010	69	6	respective	respective	ADJ
cuesj-1010	69	7	versions	version	NOUN
cuesj-1010	69	8	are	be	AUX
cuesj-1010	69	9	approximately	approximately	ADV
cuesj-1010	69	10	14	14	NUM
cuesj-1010	69	11	mb	mb	NOUN
cuesj-1010	69	12	,	,	PUNCT
cuesj-1010	69	13	41	41	NUM
cuesj-1010	69	14	mb	mb	NOUN
cuesj-1010	69	15	,	,	PUNCT
cuesj-1010	69	16	89	89	NUM
cuesj-1010	69	17	mb	mb	NOUN
cuesj-1010	69	18	,	,	PUNCT
cuesj-1010	69	19	and	and	CCONJ
cuesj-1010	69	20	can	can	AUX
cuesj-1010	69	21	extend	extend	VERB
cuesj-1010	69	22	up	up	ADP
cuesj-1010	69	23	to	to	PART
cuesj-1010	69	24	166	166	NUM
cuesj-1010	69	25	mb	mb	NOUN
cuesj-1010	69	26	.	.	PUNCT
cuesj-1010	70	1	these	these	DET
cuesj-1010	70	2	differences	difference	NOUN
cuesj-1010	70	3	in	in	ADP
cuesj-1010	70	4	size	size	NOUN
cuesj-1010	70	5	and	and	CCONJ
cuesj-1010	70	6	complexity	complexity	NOUN
cuesj-1010	70	7	provide	provide	VERB
cuesj-1010	70	8	users	user	NOUN
cuesj-1010	70	9	with	with	ADP
cuesj-1010	70	10	a	a	DET
cuesj-1010	70	11	range	range	NOUN
cuesj-1010	70	12	of	of	ADP
cuesj-1010	70	13	options	option	NOUN
cuesj-1010	70	14	to	to	PART
cuesj-1010	70	15	suit	suit	VERB
cuesj-1010	70	16	their	their	PRON
cuesj-1010	70	17	specific	specific	ADJ
cuesj-1010	70	18	project	project	NOUN
cuesj-1010	70	19	requirements	requirement	NOUN
cuesj-1010	70	20	and	and	CCONJ
cuesj-1010	70	21	computational	computational	ADJ
cuesj-1010	70	22	resources.[12,13	resources.[12,13	NOUN
cuesj-1010	70	23	]	]	PUNCT
cuesj-1010	70	24	materials	material	NOUN
cuesj-1010	70	25	and	and	CCONJ
cuesj-1010	70	26	methods	method	NOUN
cuesj-1010	70	27	the	the	DET
cuesj-1010	70	28	methodologies	methodology	NOUN
cuesj-1010	70	29	and	and	CCONJ
cuesj-1010	70	30	approaches	approach	NOUN
cuesj-1010	70	31	used	use	VERB
cuesj-1010	70	32	in	in	ADP
cuesj-1010	70	33	this	this	DET
cuesj-1010	70	34	study	study	NOUN
cuesj-1010	70	35	to	to	PART
cuesj-1010	70	36	locate	locate	VERB
cuesj-1010	70	37	the	the	DET
cuesj-1010	70	38	top	top	ADJ
cuesj-1010	70	39	outcome	outcome	NOUN
cuesj-1010	70	40	in	in	ADP
cuesj-1010	70	41	the	the	DET
cuesj-1010	70	42	selected	select	VERB
cuesj-1010	70	43	dataset	dataset	NOUN
cuesj-1010	70	44	are	be	AUX
cuesj-1010	70	45	presented	present	VERB
cuesj-1010	70	46	in	in	ADP
cuesj-1010	70	47	this	this	DET
cuesj-1010	70	48	section	section	NOUN
cuesj-1010	70	49	.	.	PUNCT
cuesj-1010	71	1	the	the	DET
cuesj-1010	71	2	datasets	dataset	NOUN
cuesj-1010	71	3	are	be	AUX
cuesj-1010	71	4	additionally	additionally	ADV
cuesj-1010	71	5	described	describe	VERB
cuesj-1010	71	6	in	in	ADP
cuesj-1010	71	7	segment	segment	NOUN
cuesj-1010	71	8	3.1	3.1	NUM
cuesj-1010	71	9	.	.	PUNCT
cuesj-1010	72	1	in	in	ADP
cuesj-1010	72	2	segment	segment	NOUN
cuesj-1010	72	3	3.2	3.2	NUM
cuesj-1010	72	4	,	,	PUNCT
cuesj-1010	72	5	experimenting	experiment	VERB
cuesj-1010	72	6	with	with	ADP
cuesj-1010	72	7	the	the	DET
cuesj-1010	72	8	yolov5	yolov5	NOUN
cuesj-1010	72	9	on	on	ADP
cuesj-1010	72	10	the	the	DET
cuesj-1010	72	11	dataset	dataset	NOUN
cuesj-1010	72	12	is	be	AUX
cuesj-1010	72	13	given	give	VERB
cuesj-1010	72	14	.	.	PUNCT
cuesj-1010	73	1	in	in	ADP
cuesj-1010	73	2	segment	segment	NOUN
cuesj-1010	73	3	3.3	3.3	NUM
cuesj-1010	73	4	,	,	PUNCT
cuesj-1010	73	5	the	the	DET
cuesj-1010	73	6	yolo	yolo	ADV
cuesj-1010	73	7	-	-	PUNCT
cuesj-1010	73	8	trained	train	VERB
cuesj-1010	73	9	model	model	NOUN
cuesj-1010	73	10	is	be	AUX
cuesj-1010	73	11	fed	feed	VERB
cuesj-1010	73	12	to	to	ADP
cuesj-1010	73	13	the	the	DET
cuesj-1010	73	14	tello	tello	PROPN
cuesj-1010	73	15	drone	drone	PROPN
cuesj-1010	73	16	.	.	PUNCT
cuesj-1010	74	1	dataset	dataset	NOUN
cuesj-1010	74	2	of	of	ADP
cuesj-1010	74	3	experimental	experimental	ADJ
cuesj-1010	74	4	in	in	ADP
cuesj-1010	74	5	the	the	DET
cuesj-1010	74	6	dataset	dataset	NOUN
cuesj-1010	74	7	,	,	PUNCT
cuesj-1010	74	8	we	we	PRON
cuesj-1010	74	9	have	have	AUX
cuesj-1010	74	10	meticulously	meticulously	ADV
cuesj-1010	74	11	curated	curate	VERB
cuesj-1010	74	12	four	four	NUM
cuesj-1010	74	13	distinct	distinct	ADJ
cuesj-1010	74	14	vehicle	vehicle	NOUN
cuesj-1010	74	15	classes	class	NOUN
cuesj-1010	74	16	:	:	PUNCT
cuesj-1010	74	17	“	"	PUNCT
cuesj-1010	74	18	vehicle_above	vehicle_above	PROPN
cuesj-1010	74	19	,	,	PUNCT
cuesj-1010	74	20	”	"	PUNCT
cuesj-1010	74	21	“	"	PUNCT
cuesj-1010	74	22	vehicle_rear	vehicle_rear	PROPN
cuesj-1010	74	23	,	,	PUNCT
cuesj-1010	74	24	”	"	PUNCT
cuesj-1010	74	25	“	"	PUNCT
cuesj-1010	74	26	vehicle	vehicle	NOUN
cuesj-1010	74	27	_	_	PUNCT
cuesj-1010	74	28	front	front	NOUN
cuesj-1010	74	29	,	,	PUNCT
cuesj-1010	74	30	”	"	PUNCT
cuesj-1010	74	31	and	and	CCONJ
cuesj-1010	74	32	“	"	PUNCT
cuesj-1010	74	33	vehicle_sides	vehicle_side	NOUN
cuesj-1010	74	34	.	.	PUNCT
cuesj-1010	74	35	”	"	PUNCT
cuesj-1010	75	1	our	our	PRON
cuesj-1010	75	2	source	source	NOUN
cuesj-1010	75	3	of	of	ADP
cuesj-1010	75	4	data	data	NOUN
cuesj-1010	75	5	is	be	AUX
cuesj-1010	75	6	the	the	DET
cuesj-1010	75	7	official	official	ADJ
cuesj-1010	75	8	roboflow	roboflow	ADJ
cuesj-1010	75	9	universe	universe	NOUN
cuesj-1010	75	10	website	website	NOUN
cuesj-1010	75	11	,	,	PUNCT
cuesj-1010	75	12	ensuring	ensure	VERB
cuesj-1010	75	13	data	datum	NOUN
cuesj-1010	75	14	reliability	reliability	NOUN
cuesj-1010	75	15	.	.	PUNCT
cuesj-1010	76	1	comprising	comprise	VERB
cuesj-1010	76	2	a	a	DET
cuesj-1010	76	3	total	total	NOUN
cuesj-1010	76	4	of	of	ADP
cuesj-1010	76	5	2000	2000	NUM
cuesj-1010	76	6	images	image	NOUN
cuesj-1010	76	7	,	,	PUNCT
cuesj-1010	76	8	each	each	DET
cuesj-1010	76	9	class	class	NOUN
cuesj-1010	76	10	is	be	AUX
cuesj-1010	76	11	thoughtfully	thoughtfully	ADV
cuesj-1010	76	12	represented	represent	VERB
cuesj-1010	76	13	with	with	ADP
cuesj-1010	76	14	an	an	DET
cuesj-1010	76	15	equal	equal	ADJ
cuesj-1010	76	16	share	share	NOUN
cuesj-1010	76	17	of	of	ADP
cuesj-1010	76	18	500	500	NUM
cuesj-1010	76	19	images	image	NOUN
cuesj-1010	76	20	,	,	PUNCT
cuesj-1010	76	21	contributing	contribute	VERB
cuesj-1010	76	22	to	to	ADP
cuesj-1010	76	23	a	a	DET
cuesj-1010	76	24	well	well	ADV
cuesj-1010	76	25	-	-	PUNCT
cuesj-1010	76	26	balanced	balance	VERB
cuesj-1010	76	27	dataset	dataset	NOUN
cuesj-1010	76	28	.	.	PUNCT
cuesj-1010	77	1	all	all	DET
cuesj-1010	77	2	images	image	NOUN
cuesj-1010	77	3	have	have	AUX
cuesj-1010	77	4	been	be	AUX
cuesj-1010	77	5	thoughtfully	thoughtfully	ADV
cuesj-1010	77	6	curated	curate	VERB
cuesj-1010	77	7	and	and	CCONJ
cuesj-1010	77	8	collected	collect	VERB
cuesj-1010	77	9	from	from	ADP
cuesj-1010	77	10	the	the	DET
cuesj-1010	77	11	roboflow	roboflow	ADJ
cuesj-1010	77	12	universe.[14	universe.[14	PROPN
cuesj-1010	77	13	]	]	PUNCT
cuesj-1010	77	14	to	to	PART
cuesj-1010	77	15	enhance	enhance	VERB
cuesj-1010	77	16	the	the	DET
cuesj-1010	77	17	dataset	dataset	ADJ
cuesj-1010	77	18	quality	quality	NOUN
cuesj-1010	77	19	and	and	CCONJ
cuesj-1010	77	20	uniformity	uniformity	NOUN
cuesj-1010	77	21	,	,	PUNCT
cuesj-1010	77	22	we	we	PRON
cuesj-1010	77	23	’ve	’ve	AUX
cuesj-1010	77	24	harnessed	harness	VERB
cuesj-1010	77	25	the	the	DET
cuesj-1010	77	26	power	power	NOUN
cuesj-1010	77	27	of	of	ADP
cuesj-1010	77	28	roboflow	roboflow	PROPN
cuesj-1010	77	29	’s	’s	PART
cuesj-1010	77	30	preprocessing	preprocesse	VERB
cuesj-1010	77	31	capabilities	capability	NOUN
cuesj-1010	77	32	.	.	PUNCT
cuesj-1010	78	1	this	this	DET
cuesj-1010	78	2	crucial	crucial	ADJ
cuesj-1010	78	3	step	step	NOUN
cuesj-1010	78	4	guarantees	guarantee	VERB
cuesj-1010	78	5	consistency	consistency	NOUN
cuesj-1010	78	6	across	across	ADP
cuesj-1010	78	7	the	the	DET
cuesj-1010	78	8	dataset	dataset	NOUN
cuesj-1010	78	9	,	,	PUNCT
cuesj-1010	78	10	especially	especially	ADV
cuesj-1010	78	11	in	in	ADP
cuesj-1010	78	12	image	image	NOUN
cuesj-1010	78	13	dimensions	dimension	NOUN
cuesj-1010	78	14	,	,	PUNCT
cuesj-1010	78	15	making	make	VERB
cuesj-1010	78	16	it	it	PRON
cuesj-1010	78	17	ideal	ideal	ADJ
cuesj-1010	78	18	for	for	ADP
cuesj-1010	78	19	model	model	NOUN
cuesj-1010	78	20	training	training	NOUN
cuesj-1010	78	21	,	,	PUNCT
cuesj-1010	78	22	testing	testing	NOUN
cuesj-1010	78	23	,	,	PUNCT
cuesj-1010	78	24	and	and	CCONJ
cuesj-1010	78	25	validation	validation	NOUN
cuesj-1010	78	26	.	.	PUNCT
cuesj-1010	79	1	the	the	DET
cuesj-1010	79	2	class	class	NOUN
cuesj-1010	79	3	annotations	annotation	NOUN
cuesj-1010	79	4	are	be	AUX
cuesj-1010	79	5	meticulously	meticulously	ADV
cuesj-1010	79	6	represented	represent	VERB
cuesj-1010	79	7	as	as	ADP
cuesj-1010	79	8	bounding	bounding	NOUN
cuesj-1010	79	9	boxes	box	NOUN
cuesj-1010	79	10	,	,	PUNCT
cuesj-1010	79	11	a	a	DET
cuesj-1010	79	12	common	common	ADJ
cuesj-1010	79	13	practice	practice	NOUN
cuesj-1010	79	14	in	in	ADP
cuesj-1010	79	15	object	object	NOUN
cuesj-1010	79	16	detection	detection	NOUN
cuesj-1010	79	17	tasks	task	NOUN
cuesj-1010	79	18	.	.	PUNCT
cuesj-1010	80	1	these	these	DET
cuesj-1010	80	2	bounding	bounding	NOUN
cuesj-1010	80	3	boxes	box	NOUN
cuesj-1010	80	4	have	have	AUX
cuesj-1010	80	5	been	be	AUX
cuesj-1010	80	6	skillfully	skillfully	ADV
cuesj-1010	80	7	converted	convert	VERB
cuesj-1010	80	8	to	to	ADP
cuesj-1010	80	9	the	the	DET
cuesj-1010	80	10	yolov5	yolov5	NOUN
cuesj-1010	80	11	format	format	NOUN
cuesj-1010	80	12	,	,	PUNCT
cuesj-1010	80	13	all	all	PRON
cuesj-1010	80	14	within	within	ADP
cuesj-1010	80	15	the	the	DET
cuesj-1010	80	16	roboflow	roboflow	ADJ
cuesj-1010	80	17	universe	universe	NOUN
cuesj-1010	80	18	.	.	PUNCT
cuesj-1010	81	1	furthermore	furthermore	ADV
cuesj-1010	81	2	,	,	PUNCT
cuesj-1010	81	3	each	each	DET
cuesj-1010	81	4	image	image	NOUN
cuesj-1010	81	5	in	in	ADP
cuesj-1010	81	6	the	the	DET
cuesj-1010	81	7	dataset	dataset	NOUN
cuesj-1010	81	8	has	have	AUX
cuesj-1010	81	9	been	be	AUX
cuesj-1010	81	10	enriched	enrich	VERB
cuesj-1010	81	11	with	with	ADP
cuesj-1010	81	12	bounding	bounding	NOUN
cuesj-1010	81	13	box	box	NOUN
cuesj-1010	81	14	annotations	annotation	NOUN
cuesj-1010	81	15	using	use	VERB
cuesj-1010	81	16	roboflow	roboflow	ADJ
cuesj-1010	81	17	annotate	annotate	NOUN
cuesj-1010	81	18	.	.	PUNCT
cuesj-1010	82	1	these	these	DET
cuesj-1010	82	2	annotations	annotation	NOUN
cuesj-1010	82	3	act	act	VERB
cuesj-1010	82	4	as	as	ADP
cuesj-1010	82	5	the	the	DET
cuesj-1010	82	6	“	"	PUNCT
cuesj-1010	82	7	answer	answer	NOUN
cuesj-1010	82	8	key	key	NOUN
cuesj-1010	82	9	”	"	PUNCT
cuesj-1010	82	10	for	for	ADP
cuesj-1010	82	11	our	our	PRON
cuesj-1010	82	12	computer	computer	NOUN
cuesj-1010	82	13	vision	vision	NOUN
cuesj-1010	82	14	model	model	NOUN
cuesj-1010	82	15	during	during	ADP
cuesj-1010	82	16	training	training	NOUN
cuesj-1010	82	17	,	,	PUNCT
cuesj-1010	82	18	enabling	enable	VERB
cuesj-1010	82	19	it	it	PRON
cuesj-1010	82	20	to	to	PART
cuesj-1010	82	21	learn	learn	VERB
cuesj-1010	82	22	and	and	CCONJ
cuesj-1010	82	23	understand	understand	VERB
cuesj-1010	82	24	each	each	DET
cuesj-1010	82	25	class	class	NOUN
cuesj-1010	82	26	with	with	ADP
cuesj-1010	82	27	precision	precision	NOUN
cuesj-1010	82	28	.	.	PUNCT
cuesj-1010	83	1	as	as	SCONJ
cuesj-1010	83	2	we	we	PRON
cuesj-1010	83	3	continue	continue	VERB
cuesj-1010	83	4	to	to	PART
cuesj-1010	83	5	add	add	VERB
cuesj-1010	83	6	more	more	ADJ
cuesj-1010	83	7	annotated	annotate	VERB
cuesj-1010	83	8	images	image	NOUN
cuesj-1010	83	9	to	to	ADP
cuesj-1010	83	10	the	the	DET
cuesj-1010	83	11	dataset	dataset	NOUN
cuesj-1010	83	12	,	,	PUNCT
cuesj-1010	83	13	our	our	PRON
cuesj-1010	83	14	model	model	NOUN
cuesj-1010	83	15	gains	gain	VERB
cuesj-1010	83	16	access	access	NOUN
cuesj-1010	83	17	to	to	ADP
cuesj-1010	83	18	richer	rich	ADJ
cuesj-1010	83	19	and	and	CCONJ
cuesj-1010	83	20	more	more	ADV
cuesj-1010	83	21	diverse	diverse	ADJ
cuesj-1010	83	22	data	datum	NOUN
cuesj-1010	83	23	,	,	PUNCT
cuesj-1010	83	24	further	far	ADV
cuesj-1010	83	25	improving	improve	VERB
cuesj-1010	83	26	its	its	PRON
cuesj-1010	83	27	performance	performance	NOUN
cuesj-1010	83	28	.	.	PUNCT
cuesj-1010	84	1	figure	figure	NOUN
cuesj-1010	84	2	3	3	NUM
cuesj-1010	84	3	provides	provide	VERB
cuesj-1010	84	4	a	a	DET
cuesj-1010	84	5	visual	visual	ADJ
cuesj-1010	84	6	example	example	NOUN
cuesj-1010	84	7	,	,	PUNCT
cuesj-1010	84	8	shedding	shed	VERB
cuesj-1010	84	9	light	light	NOUN
cuesj-1010	84	10	on	on	ADP
cuesj-1010	84	11	the	the	DET
cuesj-1010	84	12	dataset	dataset	NOUN
cuesj-1010	84	13	’s	’s	PART
cuesj-1010	84	14	composition	composition	NOUN
cuesj-1010	84	15	and	and	CCONJ
cuesj-1010	84	16	structure	structure	NOUN
cuesj-1010	84	17	.	.	PUNCT
cuesj-1010	85	1	experimenting	experiment	VERB
cuesj-1010	85	2	the	the	DET
cuesj-1010	85	3	yolov5	yolov5	NOUN
cuesj-1010	85	4	on	on	ADP
cuesj-1010	85	5	the	the	DET
cuesj-1010	85	6	dataset	dataset	NOUN
cuesj-1010	85	7	in	in	ADP
cuesj-1010	85	8	this	this	DET
cuesj-1010	85	9	experimental	experimental	ADJ
cuesj-1010	85	10	setup	setup	NOUN
cuesj-1010	85	11	,	,	PUNCT
cuesj-1010	85	12	each	each	DET
cuesj-1010	85	13	yolov5s	yolov5s	NOUN
cuesj-1010	85	14	model	model	NOUN
cuesj-1010	85	15	architecture	architecture	NOUN
cuesj-1010	85	16	was	be	AUX
cuesj-1010	85	17	diligently	diligently	ADV
cuesj-1010	85	18	trained	train	VERB
cuesj-1010	85	19	using	use	VERB
cuesj-1010	85	20	the	the	DET
cuesj-1010	85	21	t4	t4	PROPN
cuesj-1010	85	22	gpu	gpu	PROPN
cuesj-1010	85	23	accelerator	accelerator	NOUN
cuesj-1010	85	24	,	,	PUNCT
cuesj-1010	85	25	a	a	DET
cuesj-1010	85	26	resource	resource	NOUN
cuesj-1010	85	27	generously	generously	ADV
cuesj-1010	85	28	provided	provide	VERB
cuesj-1010	85	29	by	by	ADP
cuesj-1010	85	30	the	the	DET
cuesj-1010	85	31	google	google	PROPN
cuesj-1010	85	32	colab	colab	PROPN
cuesj-1010	85	33	platform	platform	NOUN
cuesj-1010	85	34	,	,	PUNCT
cuesj-1010	85	35	readily	readily	ADV
cuesj-1010	85	36	accessible	accessible	ADJ
cuesj-1010	85	37	online	online	ADV
cuesj-1010	85	38	.	.	PUNCT
cuesj-1010	86	1	for	for	ADP
cuesj-1010	86	2	those	those	PRON
cuesj-1010	86	3	interested	interested	ADJ
cuesj-1010	86	4	in	in	ADP
cuesj-1010	86	5	replicating	replicate	VERB
cuesj-1010	86	6	this	this	DET
cuesj-1010	86	7	experiment	experiment	NOUN
cuesj-1010	86	8	,	,	PUNCT
cuesj-1010	86	9	the	the	DET
cuesj-1010	86	10	entire	entire	ADJ
cuesj-1010	86	11	process	process	NOUN
cuesj-1010	86	12	is	be	AUX
cuesj-1010	86	13	made	make	VERB
cuesj-1010	86	14	accessible	accessible	ADJ
cuesj-1010	86	15	and	and	CCONJ
cuesj-1010	86	16	straightforward	straightforward	ADJ
cuesj-1010	86	17	.	.	PUNCT
cuesj-1010	87	1	you	you	PRON
cuesj-1010	87	2	can	can	AUX
cuesj-1010	87	3	access	access	VERB
cuesj-1010	87	4	the	the	DET
cuesj-1010	87	5	notebook	notebook	NOUN
cuesj-1010	87	6	code	code	NOUN
cuesj-1010	87	7	for	for	ADP
cuesj-1010	87	8	yolov5	yolov5	NOUN
cuesj-1010	87	9	with	with	ADP
cuesj-1010	87	10	a	a	DET
cuesj-1010	87	11	free	free	ADJ
cuesj-1010	87	12	gpu	gpu	NOUN
cuesj-1010	87	13	on	on	ADP
cuesj-1010	87	14	this	this	DET
cuesj-1010	87	15	link	link	NOUN
cuesj-1010	87	16	:	:	PUNCT
cuesj-1010	87	17	google	google	PROPN
cuesj-1010	87	18	colab	colab	PROPN
cuesj-1010	87	19	yolov5	yolov5	PROPN
cuesj-1010	87	20	notebook	notebook	NOUN
cuesj-1010	87	21	.	.	PUNCT
cuesj-1010	88	1	here	here	ADV
cuesj-1010	88	2	are	be	AUX
cuesj-1010	88	3	the	the	DET
cuesj-1010	88	4	key	key	ADJ
cuesj-1010	88	5	steps	step	NOUN
cuesj-1010	88	6	to	to	PART
cuesj-1010	88	7	train	train	VERB
cuesj-1010	88	8	your	your	PRON
cuesj-1010	88	9	custom	custom	NOUN
cuesj-1010	88	10	model	model	NOUN
cuesj-1010	88	11	:	:	PUNCT
cuesj-1010	88	12	1	1	X
cuesj-1010	88	13	.	.	X
cuesj-1010	88	14	cloning	cloning	NOUN
cuesj-1010	88	15	yolov5	yolov5	NOUN
cuesj-1010	88	16	:	:	PUNCT
cuesj-1010	88	17	begin	begin	VERB
cuesj-1010	88	18	by	by	ADP
cuesj-1010	88	19	cloning	clone	VERB
cuesj-1010	88	20	the	the	DET
cuesj-1010	88	21	yolov5	yolov5	NOUN
cuesj-1010	88	22	repository	repository	NOUN
cuesj-1010	88	23	from	from	ADP
cuesj-1010	88	24	ultralytics	ultralytic	NOUN
cuesj-1010	88	25	on	on	ADP
cuesj-1010	88	26	github	github	NOUN
cuesj-1010	88	27	.	.	PUNCT
cuesj-1010	89	1	2	2	X
cuesj-1010	89	2	.	.	X
cuesj-1010	89	3	cloning	cloning	NOUN
cuesj-1010	89	4	requirements	requirement	NOUN
cuesj-1010	89	5	:	:	PUNCT
cuesj-1010	89	6	clone	clone	NOUN
cuesj-1010	89	7	all	all	DET
cuesj-1010	89	8	the	the	DET
cuesj-1010	89	9	necessary	necessary	ADJ
cuesj-1010	89	10	dependencies	dependency	NOUN
cuesj-1010	89	11	and	and	CCONJ
cuesj-1010	89	12	libraries	library	NOUN
cuesj-1010	89	13	,	,	PUNCT
cuesj-1010	89	14	such	such	ADJ
cuesj-1010	89	15	as	as	ADP
cuesj-1010	89	16	opencv	opencv	PROPN
cuesj-1010	89	17	-	-	PUNCT
cuesj-1010	89	18	python	python	PROPN
cuesj-1010	89	19	and	and	CCONJ
cuesj-1010	89	20	matplotlib	matplotlib	PROPN
cuesj-1010	89	21	.	.	PUNCT
cuesj-1010	90	1	3	3	X
cuesj-1010	90	2	.	.	X
cuesj-1010	90	3	dataset	dataset	NOUN
cuesj-1010	90	4	upload	upload	NOUN
cuesj-1010	90	5	:	:	PUNCT
cuesj-1010	90	6	upload	upload	VERB
cuesj-1010	90	7	your	your	PRON
cuesj-1010	90	8	dataset	dataset	NOUN
cuesj-1010	90	9	using	use	VERB
cuesj-1010	90	10	the	the	DET
cuesj-1010	90	11	code	code	NOUN
cuesj-1010	90	12	provided	provide	VERB
cuesj-1010	90	13	by	by	ADP
cuesj-1010	90	14	roboflow	roboflow	NOUN
cuesj-1010	90	15	.	.	PUNCT
cuesj-1010	91	1	this	this	DET
cuesj-1010	91	2	code	code	NOUN
cuesj-1010	91	3	facilitates	facilitate	VERB
cuesj-1010	91	4	the	the	DET
cuesj-1010	91	5	seamless	seamless	ADJ
cuesj-1010	91	6	retrieval	retrieval	NOUN
cuesj-1010	91	7	and	and	CCONJ
cuesj-1010	91	8	organization	organization	NOUN
cuesj-1010	91	9	of	of	ADP
cuesj-1010	91	10	the	the	DET
cuesj-1010	91	11	dataset	dataset	NOUN
cuesj-1010	91	12	from	from	ADP
cuesj-1010	91	13	the	the	DET
cuesj-1010	91	14	roboflow	roboflow	ADJ
cuesj-1010	91	15	account	account	NOUN
cuesj-1010	91	16	where	where	SCONJ
cuesj-1010	91	17	it	it	PRON
cuesj-1010	91	18	was	be	AUX
cuesj-1010	91	19	created	create	VERB
cuesj-1010	91	20	.	.	PUNCT
cuesj-1010	92	1	4	4	X
cuesj-1010	92	2	.	.	X
cuesj-1010	92	3	model	model	NOUN
cuesj-1010	92	4	training	training	NOUN
cuesj-1010	92	5	:	:	PUNCT
cuesj-1010	92	6	with	with	ADP
cuesj-1010	92	7	the	the	DET
cuesj-1010	92	8	dataset	dataset	NOUN
cuesj-1010	92	9	in	in	ADP
cuesj-1010	92	10	place	place	NOUN
cuesj-1010	92	11	,	,	PUNCT
cuesj-1010	92	12	your	your	PRON
cuesj-1010	92	13	model	model	NOUN
cuesj-1010	92	14	is	be	AUX
cuesj-1010	92	15	now	now	ADV
cuesj-1010	92	16	poised	poise	VERB
cuesj-1010	92	17	for	for	ADP
cuesj-1010	92	18	training	training	NOUN
cuesj-1010	92	19	.	.	PUNCT
cuesj-1010	93	1	specify	specify	VERB
cuesj-1010	93	2	the	the	DET
cuesj-1010	93	3	yolov5	yolov5	NOUN
cuesj-1010	93	4	variant	variant	NOUN
cuesj-1010	93	5	and	and	CCONJ
cuesj-1010	93	6	set	set	VERB
cuesj-1010	93	7	the	the	DET
cuesj-1010	93	8	number	number	NOUN
cuesj-1010	93	9	of	of	ADP
cuesj-1010	93	10	epochs	epoch	NOUN
cuesj-1010	93	11	for	for	ADP
cuesj-1010	93	12	training	training	NOUN
cuesj-1010	93	13	.	.	PUNCT
cuesj-1010	94	1	in	in	ADP
cuesj-1010	94	2	this	this	DET
cuesj-1010	94	3	particular	particular	ADJ
cuesj-1010	94	4	experiment	experiment	NOUN
cuesj-1010	94	5	,	,	PUNCT
cuesj-1010	94	6	100	100	NUM
cuesj-1010	94	7	epochs	epoch	NOUN
cuesj-1010	94	8	were	be	AUX
cuesj-1010	94	9	employed	employ	VERB
cuesj-1010	94	10	,	,	PUNCT
cuesj-1010	94	11	encompassing	encompass	VERB
cuesj-1010	94	12	both	both	CCONJ
cuesj-1010	94	13	small	small	ADJ
cuesj-1010	94	14	and	and	CCONJ
cuesj-1010	94	15	extra	extra	ADJ
cuesj-1010	94	16	-	-	ADJ
cuesj-1010	94	17	large	large	ADJ
cuesj-1010	94	18	types	type	NOUN
cuesj-1010	94	19	of	of	ADP
cuesj-1010	94	20	yolov5	yolov5	NOUN
cuesj-1010	94	21	.	.	PUNCT
cuesj-1010	95	1	table	table	NOUN
cuesj-1010	95	2	1	1	NUM
cuesj-1010	95	3	records	record	NOUN
cuesj-1010	95	4	the	the	DET
cuesj-1010	95	5	time	time	NOUN
cuesj-1010	95	6	taken	take	VERB
cuesj-1010	95	7	for	for	ADP
cuesj-1010	95	8	each	each	DET
cuesj-1010	95	9	model	model	NOUN
cuesj-1010	95	10	to	to	PART
cuesj-1010	95	11	complete	complete	VERB
cuesj-1010	95	12	training	training	NOUN
cuesj-1010	95	13	after	after	ADP
cuesj-1010	95	14	100	100	NUM
cuesj-1010	95	15	epochs	epoch	NOUN
cuesj-1010	95	16	:	:	PUNCT
cuesj-1010	95	17	the	the	DET
cuesj-1010	95	18	results	result	NOUN
cuesj-1010	95	19	suggest	suggest	VERB
cuesj-1010	95	20	that	that	SCONJ
cuesj-1010	95	21	the	the	DET
cuesj-1010	95	22	yolov5s	yolov5s	PROPN
cuesj-1010	95	23	model	model	NOUN
cuesj-1010	95	24	achieved	achieve	VERB
cuesj-1010	95	25	the	the	DET
cuesj-1010	95	26	quickest	quick	ADJ
cuesj-1010	95	27	training	training	NOUN
cuesj-1010	95	28	time	time	NOUN
cuesj-1010	95	29	,	,	PUNCT
cuesj-1010	95	30	taking	take	VERB
cuesj-1010	95	31	less	less	ADJ
cuesj-1010	95	32	than	than	ADP
cuesj-1010	95	33	an	an	DET
cuesj-1010	95	34	hour	hour	NOUN
cuesj-1010	95	35	,	,	PUNCT
cuesj-1010	95	36	while	while	SCONJ
cuesj-1010	95	37	yolov5x	yolov5x	PROPN
cuesj-1010	95	38	training	training	NOUN
cuesj-1010	95	39	proved	prove	VERB
cuesj-1010	95	40	to	to	PART
cuesj-1010	95	41	be	be	AUX
cuesj-1010	95	42	the	the	DET
cuesj-1010	95	43	most	most	ADJ
cuesj-1010	95	44	time	time	NOUN
cuesj-1010	95	45	-	-	PUNCT
cuesj-1010	95	46	intensive	intensive	ADJ
cuesj-1010	95	47	,	,	PUNCT
cuesj-1010	95	48	spanning	span	VERB
cuesj-1010	95	49	approximately	approximately	ADV
cuesj-1010	95	50	9	9	NUM
cuesj-1010	95	51	h.	h.	NOUN
cuesj-1010	95	52	after	after	ADP
cuesj-1010	95	53	the	the	DET
cuesj-1010	95	54	successful	successful	ADJ
cuesj-1010	95	55	completion	completion	NOUN
cuesj-1010	95	56	of	of	ADP
cuesj-1010	95	57	training	training	NOUN
cuesj-1010	95	58	,	,	PUNCT
cuesj-1010	95	59	the	the	DET
cuesj-1010	95	60	trained	train	VERB
cuesj-1010	95	61	model	model	NOUN
cuesj-1010	95	62	is	be	AUX
cuesj-1010	95	63	saved	save	VERB
cuesj-1010	95	64	in	in	ADP
cuesj-1010	95	65	the.pt	the.pt	PROPN
cuesj-1010	95	66	format	format	NOUN
cuesj-1010	95	67	,	,	PUNCT
cuesj-1010	95	68	ready	ready	ADJ
cuesj-1010	95	69	to	to	PART
cuesj-1010	95	70	be	be	AUX
cuesj-1010	95	71	loaded	load	VERB
cuesj-1010	95	72	and	and	CCONJ
cuesj-1010	95	73	utilized	utilize	VERB
cuesj-1010	95	74	with	with	ADP
cuesj-1010	95	75	pytorch	pytorch	NOUN
cuesj-1010	95	76	for	for	ADP
cuesj-1010	95	77	further	further	ADJ
cuesj-1010	95	78	applications	application	NOUN
cuesj-1010	95	79	.	.	PUNCT
cuesj-1010	96	1	a	a	DET
cuesj-1010	96	2	b	b	X
cuesj-1010	96	3	c	c	X
cuesj-1010	96	4	abdallah	abdallah	PROPN
cuesj-1010	96	5	:	:	PUNCT
cuesj-1010	96	6	vehicle	vehicle	NOUN
cuesj-1010	96	7	detection	detection	NOUN
cuesj-1010	96	8	with	with	ADP
cuesj-1010	96	9	tello	tello	PROPN
cuesj-1010	96	10	drone	drone	PROPN
cuesj-1010	96	11	and	and	CCONJ
cuesj-1010	96	12	yolov5	yolov5	NOUN
cuesj-1010	96	13	4	4	NUM
cuesj-1010	96	14	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	96	15	cuesj	cuesj	ADJ
cuesj-1010	96	16	2024	2024	NUM
cuesj-1010	96	17	,	,	PUNCT
cuesj-1010	96	18	8	8	NUM
cuesj-1010	96	19	(	(	PUNCT
cuesj-1010	96	20	1	1	NUM
cuesj-1010	96	21	):	):	PUNCT
cuesj-1010	96	22	1	1	NUM
cuesj-1010	96	23	-	-	SYM
cuesj-1010	96	24	7	7	NUM
cuesj-1010	96	25	table	table	NOUN
cuesj-1010	96	26	2	2	NUM
cuesj-1010	96	27	:	:	PUNCT
cuesj-1010	96	28	commands	command	NOUN
cuesj-1010	96	29	used	use	VERB
cuesj-1010	96	30	to	to	ADP
cuesj-1010	96	31	controlling	control	VERB
cuesj-1010	96	32	the	the	DET
cuesj-1010	96	33	drone	drone	NOUN
cuesj-1010	96	34	effect	effect	NOUN
cuesj-1010	96	35	command	command	NOUN
cuesj-1010	96	36	take	take	VERB
cuesj-1010	96	37	off	off	ADP
cuesj-1010	96	38	e	e	PROPN
cuesj-1010	96	39	land	land	NOUN
cuesj-1010	96	40	q	q	PUNCT
cuesj-1010	96	41	move	move	VERB
cuesj-1010	96	42	down	down	ADP
cuesj-1010	96	43	s	s	PRON
cuesj-1010	96	44	move	move	NOUN
cuesj-1010	96	45	up	up	ADP
cuesj-1010	96	46	w	w	NOUN
cuesj-1010	96	47	move	move	NOUN
cuesj-1010	96	48	backward	backward	ADV
cuesj-1010	96	49	down	down	ADP
cuesj-1010	96	50	arrow	arrow	NOUN
cuesj-1010	96	51	move	move	NOUN
cuesj-1010	96	52	forward	forward	ADV
cuesj-1010	96	53	up	up	ADP
cuesj-1010	96	54	arrow	arrow	NOUN
cuesj-1010	96	55	move	move	NOUN
cuesj-1010	96	56	right	right	ADV
cuesj-1010	96	57	right	right	ADV
cuesj-1010	96	58	arrow	arrow	NOUN
cuesj-1010	96	59	move	move	NOUN
cuesj-1010	96	60	left	leave	VERB
cuesj-1010	96	61	left	leave	VERB
cuesj-1010	96	62	arrow	arrow	NOUN
cuesj-1010	96	63	table	table	NOUN
cuesj-1010	96	64	1	1	NUM
cuesj-1010	96	65	:	:	PUNCT
cuesj-1010	96	66	the	the	DET
cuesj-1010	96	67	training	training	NOUN
cuesj-1010	96	68	times	time	NOUN
cuesj-1010	96	69	for	for	ADP
cuesj-1010	96	70	each	each	DET
cuesj-1010	96	71	yolov5	yolov5	NOUN
cuesj-1010	96	72	model	model	NOUN
cuesj-1010	96	73	after	after	ADP
cuesj-1010	96	74	100	100	NUM
cuesj-1010	96	75	epochs	epoch	NOUN
cuesj-1010	96	76	symbol	symbol	NOUN
cuesj-1010	96	77	model	model	NOUN
cuesj-1010	96	78	time	time	NOUN
cuesj-1010	96	79	(	(	PUNCT
cuesj-1010	96	80	h	h	NOUN
cuesj-1010	96	81	)	)	PUNCT
cuesj-1010	96	82	v5x	v5x	NOUN
cuesj-1010	96	83	yolov5x	yolov5x	PROPN
cuesj-1010	96	84	8.6	8.6	NUM
cuesj-1010	97	1	v5s	v5s	NOUN
cuesj-1010	97	2	yolov5s	yolov5s	PROPN
cuesj-1010	97	3	0.73	0.73	NUM
cuesj-1010	97	4	v5	v5	PROPN
cuesj-1010	97	5	yolov5l	yolov5l	PROPN
cuesj-1010	97	6	2.1	2.1	NUM
cuesj-1010	97	7	v5	v5	PROPN
cuesj-1010	97	8	m	m	NOUN
cuesj-1010	97	9	yolov5	yolov5	NOUN
cuesj-1010	97	10	m	m	VERB
cuesj-1010	97	11	1.4	1.4	NUM
cuesj-1010	97	12	tello	tello	PROPN
cuesj-1010	97	13	drone	drone	NOUN
cuesj-1010	97	14	feeding	feed	VERB
cuesj-1010	97	15	the	the	DET
cuesj-1010	97	16	initial	initial	ADJ
cuesj-1010	97	17	step	step	NOUN
cuesj-1010	97	18	in	in	ADP
cuesj-1010	97	19	connecting	connect	VERB
cuesj-1010	97	20	with	with	ADP
cuesj-1010	97	21	the	the	DET
cuesj-1010	97	22	tello	tello	PROPN
cuesj-1010	97	23	drone	drone	NOUN
cuesj-1010	97	24	involves	involve	VERB
cuesj-1010	97	25	the	the	DET
cuesj-1010	97	26	integration	integration	NOUN
cuesj-1010	97	27	of	of	ADP
cuesj-1010	97	28	another	another	DET
cuesj-1010	97	29	library	library	NOUN
cuesj-1010	97	30	called	call	VERB
cuesj-1010	97	31	“	"	PUNCT
cuesj-1010	97	32	djitellopy	djitellopy	ADJ
cuesj-1010	97	33	.	.	PUNCT
cuesj-1010	97	34	”	"	PUNCT
cuesj-1010	98	1	this	this	DET
cuesj-1010	98	2	library	library	NOUN
cuesj-1010	98	3	serves	serve	VERB
cuesj-1010	98	4	as	as	ADP
cuesj-1010	98	5	the	the	DET
cuesj-1010	98	6	gateway	gateway	NOUN
cuesj-1010	98	7	to	to	ADP
cuesj-1010	98	8	interfacing	interface	VERB
cuesj-1010	98	9	with	with	ADP
cuesj-1010	98	10	the	the	DET
cuesj-1010	98	11	drone	drone	NOUN
cuesj-1010	98	12	’s	’s	PART
cuesj-1010	98	13	sensors	sensor	NOUN
cuesj-1010	98	14	and	and	CCONJ
cuesj-1010	98	15	facilitates	facilitate	VERB
cuesj-1010	98	16	control	control	NOUN
cuesj-1010	98	17	over	over	ADP
cuesj-1010	98	18	its	its	PRON
cuesj-1010	98	19	movements.[15	movements.[15	NOUN
cuesj-1010	98	20	]	]	PUNCT
cuesj-1010	98	21	the	the	DET
cuesj-1010	98	22	following	follow	VERB
cuesj-1010	98	23	code	code	NOUN
cuesj-1010	98	24	snippet	snippet	NOUN
cuesj-1010	98	25	demonstrates	demonstrate	VERB
cuesj-1010	98	26	how	how	SCONJ
cuesj-1010	98	27	to	to	PART
cuesj-1010	98	28	establish	establish	VERB
cuesj-1010	98	29	this	this	DET
cuesj-1010	98	30	connection	connection	NOUN
cuesj-1010	98	31	:	:	PUNCT
cuesj-1010	98	32	import	import	NOUN
cuesj-1010	98	33	djitellopy	djitellopy	NOUN
cuesj-1010	98	34	as	as	ADP
cuesj-1010	98	35	tello	tello	PROPN
cuesj-1010	98	36	kp.init	kp.init	PROPN
cuesj-1010	98	37	(	(	PUNCT
cuesj-1010	98	38	)	)	PUNCT
cuesj-1010	98	39	me	i	PRON
cuesj-1010	98	40	=	=	SYM
cuesj-1010	98	41	tello.tello	tello.tello	PROPN
cuesj-1010	98	42	(	(	PUNCT
cuesj-1010	98	43	)	)	PUNCT
cuesj-1010	98	44	me.connect	me.connect	PROPN
cuesj-1010	98	45	(	(	PUNCT
cuesj-1010	98	46	)	)	PUNCT
cuesj-1010	98	47	print(me.get_battery	print(me.get_battery	PROPN
cuesj-1010	98	48	(	(	PUNCT
cuesj-1010	98	49	)	)	PUNCT
cuesj-1010	98	50	)	)	PUNCT
cuesj-1010	98	51	with	with	ADP
cuesj-1010	98	52	this	this	DET
cuesj-1010	98	53	connection	connection	NOUN
cuesj-1010	98	54	in	in	ADP
cuesj-1010	98	55	place	place	NOUN
cuesj-1010	98	56	,	,	PUNCT
cuesj-1010	98	57	the	the	DET
cuesj-1010	98	58	drone	drone	NOUN
cuesj-1010	98	59	is	be	AUX
cuesj-1010	98	60	linked	link	VERB
cuesj-1010	98	61	exclusively	exclusively	ADV
cuesj-1010	98	62	to	to	ADP
cuesj-1010	98	63	the	the	DET
cuesj-1010	98	64	host	host	NOUN
cuesj-1010	98	65	device	device	NOUN
cuesj-1010	98	66	,	,	PUNCT
cuesj-1010	98	67	which	which	PRON
cuesj-1010	98	68	,	,	PUNCT
cuesj-1010	98	69	in	in	ADP
cuesj-1010	98	70	most	most	ADJ
cuesj-1010	98	71	cases	case	NOUN
cuesj-1010	98	72	,	,	PUNCT
cuesj-1010	98	73	is	be	AUX
cuesj-1010	98	74	a	a	DET
cuesj-1010	98	75	laptop	laptop	NOUN
cuesj-1010	98	76	,	,	PUNCT
cuesj-1010	98	77	through	through	ADP
cuesj-1010	98	78	a	a	DET
cuesj-1010	98	79	wi	wi	PROPN
cuesj-1010	98	80	-	-	PUNCT
cuesj-1010	98	81	fi	fi	NOUN
cuesj-1010	98	82	connection	connection	NOUN
cuesj-1010	98	83	.	.	PUNCT
cuesj-1010	99	1	simultaneously	simultaneously	ADV
cuesj-1010	99	2	,	,	PUNCT
cuesj-1010	99	3	to	to	PART
cuesj-1010	99	4	effectively	effectively	ADV
cuesj-1010	99	5	load	load	VERB
cuesj-1010	99	6	and	and	CCONJ
cuesj-1010	99	7	run	run	VERB
cuesj-1010	99	8	the	the	DET
cuesj-1010	99	9	yolo	yolo	ADJ
cuesj-1010	99	10	models	model	NOUN
cuesj-1010	99	11	on	on	ADP
cuesj-1010	99	12	any	any	DET
cuesj-1010	99	13	device	device	NOUN
cuesj-1010	99	14	,	,	PUNCT
cuesj-1010	99	15	we	we	PRON
cuesj-1010	99	16	need	need	VERB
cuesj-1010	99	17	the	the	DET
cuesj-1010	99	18	pytorch	pytorch	NOUN
cuesj-1010	99	19	framework.[15	framework.[15	NOUN
cuesj-1010	99	20	]	]	X
cuesj-1010	99	21	pytorch	pytorch	NOUN
cuesj-1010	99	22	necessitates	necessitate	VERB
cuesj-1010	99	23	access	access	NOUN
cuesj-1010	99	24	to	to	ADP
cuesj-1010	99	25	the	the	DET
cuesj-1010	99	26	internet	internet	NOUN
cuesj-1010	99	27	too	too	ADV
cuesj-1010	99	28	;	;	PUNCT
cuesj-1010	99	29	therefore	therefore	ADV
cuesj-1010	99	30	,	,	PUNCT
cuesj-1010	99	31	there	there	PRON
cuesj-1010	99	32	will	will	AUX
cuesj-1010	99	33	be	be	AUX
cuesj-1010	99	34	confusion	confusion	NOUN
cuesj-1010	99	35	between	between	ADP
cuesj-1010	99	36	loading	loading	NOUN
cuesj-1010	99	37	models	model	NOUN
cuesj-1010	99	38	and	and	CCONJ
cuesj-1010	99	39	feeding	feeding	NOUN
cuesj-1010	99	40	models	model	NOUN
cuesj-1010	99	41	to	to	ADP
cuesj-1010	99	42	the	the	DET
cuesj-1010	99	43	drone	drone	NOUN
cuesj-1010	99	44	.	.	PUNCT
cuesj-1010	100	1	resolving	resolve	VERB
cuesj-1010	100	2	this	this	DET
cuesj-1010	100	3	compatibility	compatibility	NOUN
cuesj-1010	100	4	challenge	challenge	NOUN
cuesj-1010	100	5	involves	involve	VERB
cuesj-1010	100	6	a	a	DET
cuesj-1010	100	7	specific	specific	ADJ
cuesj-1010	100	8	sequence	sequence	NOUN
cuesj-1010	100	9	of	of	ADP
cuesj-1010	100	10	actions	action	NOUN
cuesj-1010	100	11	.	.	PUNCT
cuesj-1010	101	1	the	the	DET
cuesj-1010	101	2	model	model	NOUN
cuesj-1010	101	3	must	must	AUX
cuesj-1010	101	4	be	be	AUX
cuesj-1010	101	5	loaded	load	VERB
cuesj-1010	101	6	while	while	SCONJ
cuesj-1010	101	7	the	the	DET
cuesj-1010	101	8	laptop	laptop	NOUN
cuesj-1010	101	9	is	be	AUX
cuesj-1010	101	10	still	still	ADV
cuesj-1010	101	11	connected	connect	VERB
cuesj-1010	101	12	to	to	ADP
cuesj-1010	101	13	the	the	DET
cuesj-1010	101	14	internet	internet	NOUN
cuesj-1010	101	15	.	.	PUNCT
cuesj-1010	102	1	once	once	SCONJ
cuesj-1010	102	2	the	the	DET
cuesj-1010	102	3	model	model	NOUN
cuesj-1010	102	4	is	be	AUX
cuesj-1010	102	5	operational	operational	ADJ
cuesj-1010	102	6	,	,	PUNCT
cuesj-1010	102	7	the	the	DET
cuesj-1010	102	8	internet	internet	NOUN
cuesj-1010	102	9	connection	connection	NOUN
cuesj-1010	102	10	can	can	AUX
cuesj-1010	102	11	be	be	AUX
cuesj-1010	102	12	interrupted	interrupt	VERB
cuesj-1010	102	13	,	,	PUNCT
cuesj-1010	102	14	and	and	CCONJ
cuesj-1010	102	15	the	the	DET
cuesj-1010	102	16	laptop	laptop	NOUN
cuesj-1010	102	17	is	be	AUX
cuesj-1010	102	18	subsequently	subsequently	ADV
cuesj-1010	102	19	connected	connect	VERB
cuesj-1010	102	20	to	to	ADP
cuesj-1010	102	21	the	the	DET
cuesj-1010	102	22	drone	drone	NOUN
cuesj-1010	102	23	.	.	PUNCT
cuesj-1010	103	1	an	an	DET
cuesj-1010	103	2	intriguing	intriguing	ADJ
cuesj-1010	103	3	aspect	aspect	NOUN
cuesj-1010	103	4	to	to	PART
cuesj-1010	103	5	note	note	VERB
cuesj-1010	103	6	is	be	AUX
cuesj-1010	103	7	that	that	SCONJ
cuesj-1010	103	8	the	the	DET
cuesj-1010	103	9	model	model	NOUN
cuesj-1010	103	10	remains	remains	AUX
cuesj-1010	103	11	cached	cache	VERB
cuesj-1010	103	12	in	in	ADP
cuesj-1010	103	13	memory	memory	NOUN
cuesj-1010	103	14	,	,	PUNCT
cuesj-1010	103	15	serving	serve	VERB
cuesj-1010	103	16	as	as	ADP
cuesj-1010	103	17	temporary	temporary	ADJ
cuesj-1010	103	18	storage	storage	NOUN
cuesj-1010	103	19	and	and	CCONJ
cuesj-1010	103	20	enabling	enable	VERB
cuesj-1010	103	21	us	we	PRON
cuesj-1010	103	22	to	to	PART
cuesj-1010	103	23	retrieve	retrieve	VERB
cuesj-1010	103	24	the	the	DET
cuesj-1010	103	25	model	model	NOUN
cuesj-1010	103	26	without	without	ADP
cuesj-1010	103	27	requiring	require	VERB
cuesj-1010	103	28	internet	internet	NOUN
cuesj-1010	103	29	access	access	NOUN
cuesj-1010	103	30	.	.	PUNCT
cuesj-1010	104	1	this	this	DET
cuesj-1010	104	2	cached	cache	VERB
cuesj-1010	104	3	model	model	NOUN
cuesj-1010	104	4	can	can	AUX
cuesj-1010	104	5	be	be	AUX
cuesj-1010	104	6	effortlessly	effortlessly	ADV
cuesj-1010	104	7	loaded	load	VERB
cuesj-1010	104	8	using	use	VERB
cuesj-1010	104	9	the	the	DET
cuesj-1010	104	10	pytorch	pytorch	NOUN
cuesj-1010	104	11	hub	hub	NOUN
cuesj-1010	104	12	.	.	PUNCT
cuesj-1010	105	1	this	this	DET
cuesj-1010	105	2	code	code	NOUN
cuesj-1010	105	3	below	below	ADV
cuesj-1010	105	4	is	be	AUX
cuesj-1010	105	5	used	use	VERB
cuesj-1010	105	6	to	to	PART
cuesj-1010	105	7	load	load	VERB
cuesj-1010	105	8	the	the	DET
cuesj-1010	105	9	model	model	NOUN
cuesj-1010	105	10	by	by	ADP
cuesj-1010	105	11	torch	torch	NOUN
cuesj-1010	105	12	(	(	PUNCT
cuesj-1010	105	13	torch	torch	NOUN
cuesj-1010	105	14	:	:	PUNCT
cuesj-1010	105	15	model	model	NOUN
cuesj-1010	105	16	=	=	NOUN
cuesj-1010	105	17	torch	torch	NOUN
cuesj-1010	105	18	.	.	PUNCT
cuesj-1010	106	1	hub.load	hub.load	PROPN
cuesj-1010	107	1	(	(	PUNCT
cuesj-1010	107	2	“	"	PUNCT
cuesj-1010	107	3	ultralytics	ultralytic	NOUN
cuesj-1010	107	4	/	/	SYM
cuesj-1010	107	5	yolov5	yolov5	NOUN
cuesj-1010	107	6	,	,	PUNCT
cuesj-1010	107	7	”	"	PUNCT
cuesj-1010	107	8	“	"	PUNCT
cuesj-1010	107	9	custom	custom	ADJ
cuesj-1010	107	10	,	,	PUNCT
cuesj-1010	107	11	”	"	PUNCT
cuesj-1010	107	12	path_or_model=	path_or_model=	PRON
cuesj-1010	107	13	“	"	PUNCT
cuesj-1010	107	14	model.pt	model.pt	PROPN
cuesj-1010	107	15	”	"	PUNCT
cuesj-1010	107	16	)	)	PUNCT
cuesj-1010	107	17	with	with	ADP
cuesj-1010	107	18	the	the	DET
cuesj-1010	107	19	model	model	NOUN
cuesj-1010	107	20	successfully	successfully	ADV
cuesj-1010	107	21	loaded	load	VERB
cuesj-1010	107	22	,	,	PUNCT
cuesj-1010	107	23	the	the	DET
cuesj-1010	107	24	next	next	ADJ
cuesj-1010	107	25	step	step	NOUN
cuesj-1010	107	26	involves	involve	VERB
cuesj-1010	107	27	utilizing	utilize	VERB
cuesj-1010	107	28	specific	specific	ADJ
cuesj-1010	107	29	commands	command	NOUN
cuesj-1010	107	30	to	to	PART
cuesj-1010	107	31	control	control	VERB
cuesj-1010	107	32	the	the	DET
cuesj-1010	107	33	drone	drone	NOUN
cuesj-1010	107	34	’s	’s	PART
cuesj-1010	107	35	activities	activity	NOUN
cuesj-1010	107	36	.	.	PUNCT
cuesj-1010	108	1	the	the	DET
cuesj-1010	108	2	following	follow	VERB
cuesj-1010	108	3	commands	command	NOUN
cuesj-1010	108	4	in	in	ADP
cuesj-1010	108	5	table	table	NOUN
cuesj-1010	108	6	2	2	NUM
cuesj-1010	108	7	could	could	AUX
cuesj-1010	108	8	be	be	AUX
cuesj-1010	108	9	employed	employ	VERB
cuesj-1010	108	10	to	to	ADP
cuesj-1010	108	11	commanding	command	VERB
cuesj-1010	108	12	the	the	DET
cuesj-1010	108	13	drone	drone	NOUN
cuesj-1010	108	14	and	and	CCONJ
cuesj-1010	108	15	controlling	control	VERB
cuesj-1010	108	16	the	the	DET
cuesj-1010	108	17	movements	movement	NOUN
cuesj-1010	108	18	.	.	PUNCT
cuesj-1010	109	1	note	note	NOUN
cuesj-1010	109	2	:	:	PUNCT
cuesj-1010	109	3	in	in	ADP
cuesj-1010	109	4	this	this	DET
cuesj-1010	109	5	step	step	NOUN
cuesj-1010	109	6	,	,	PUNCT
cuesj-1010	109	7	pycharm	pycharm	PROPN
cuesj-1010	109	8	serves	serve	VERB
cuesj-1010	109	9	as	as	ADP
cuesj-1010	109	10	the	the	DET
cuesj-1010	109	11	software	software	NOUN
cuesj-1010	109	12	of	of	ADP
cuesj-1010	109	13	choice	choice	NOUN
cuesj-1010	109	14	for	for	ADP
cuesj-1010	109	15	model	model	NOUN
cuesj-1010	109	16	testing	testing	NOUN
cuesj-1010	109	17	.	.	PUNCT
cuesj-1010	110	1	utilizing	utilize	VERB
cuesj-1010	110	2	pycharm	pycharm	NOUN
cuesj-1010	110	3	,	,	PUNCT
cuesj-1010	110	4	we	we	PRON
cuesj-1010	110	5	can	can	AUX
cuesj-1010	110	6	effectively	effectively	ADV
cuesj-1010	110	7	evaluate	evaluate	VERB
cuesj-1010	110	8	the	the	DET
cuesj-1010	110	9	model	model	NOUN
cuesj-1010	110	10	’s	’s	PART
cuesj-1010	110	11	performance	performance	NOUN
cuesj-1010	110	12	.	.	PUNCT
cuesj-1010	111	1	to	to	PART
cuesj-1010	111	2	initiate	initiate	VERB
cuesj-1010	111	3	the	the	DET
cuesj-1010	111	4	testing	testing	NOUN
cuesj-1010	111	5	process	process	NOUN
cuesj-1010	111	6	,	,	PUNCT
cuesj-1010	111	7	the	the	DET
cuesj-1010	111	8	yolov5	yolov5	NOUN
cuesj-1010	111	9	model	model	NOUN
cuesj-1010	111	10	must	must	AUX
cuesj-1010	111	11	be	be	AUX
cuesj-1010	111	12	obtained	obtain	VERB
cuesj-1010	111	13	by	by	ADP
cuesj-1010	111	14	cloning	clone	VERB
cuesj-1010	111	15	or	or	CCONJ
cuesj-1010	111	16	downloading	download	VERB
cuesj-1010	111	17	it	it	PRON
cuesj-1010	111	18	from	from	ADP
cuesj-1010	111	19	the	the	DET
cuesj-1010	111	20	repository	repository	NOUN
cuesj-1010	111	21	accessible	accessible	ADJ
cuesj-1010	111	22	through	through	ADP
cuesj-1010	111	23	this	this	DET
cuesj-1010	111	24	link	link	NOUN
cuesj-1010	111	25	:	:	PUNCT
cuesj-1010	111	26	yolov5_drone	yolov5_drone	PROPN
cuesj-1010	111	27	repository	repository	NOUN
cuesj-1010	111	28	.	.	PUNCT
cuesj-1010	112	1	once	once	ADV
cuesj-1010	112	2	obtained	obtain	VERB
cuesj-1010	112	3	,	,	PUNCT
cuesj-1010	112	4	the	the	DET
cuesj-1010	112	5	model	model	NOUN
cuesj-1010	112	6	files	file	NOUN
cuesj-1010	112	7	can	can	AUX
cuesj-1010	112	8	be	be	AUX
cuesj-1010	112	9	uploaded	upload	VERB
cuesj-1010	112	10	to	to	PART
cuesj-1010	112	11	pycharm	pycharm	VERB
cuesj-1010	112	12	.	.	PUNCT
cuesj-1010	113	1	particularly	particularly	ADV
cuesj-1010	113	2	,	,	PUNCT
cuesj-1010	113	3	the.pt	the.pt	NOUN
cuesj-1010	113	4	file	file	NOUN
cuesj-1010	113	5	,	,	PUNCT
cuesj-1010	113	6	derived	derive	VERB
cuesj-1010	113	7	from	from	ADP
cuesj-1010	113	8	the	the	DET
cuesj-1010	113	9	final	final	ADJ
cuesj-1010	113	10	stage	stage	NOUN
cuesj-1010	113	11	of	of	ADP
cuesj-1010	113	12	model	model	NOUN
cuesj-1010	113	13	training	training	NOUN
cuesj-1010	113	14	,	,	PUNCT
cuesj-1010	113	15	is	be	AUX
cuesj-1010	113	16	paramount	paramount	ADJ
cuesj-1010	113	17	in	in	ADP
cuesj-1010	113	18	this	this	DET
cuesj-1010	113	19	section	section	NOUN
cuesj-1010	113	20	as	as	SCONJ
cuesj-1010	113	21	it	it	PRON
cuesj-1010	113	22	is	be	AUX
cuesj-1010	113	23	instrumental	instrumental	ADJ
cuesj-1010	113	24	in	in	ADP
cuesj-1010	113	25	conducting	conduct	VERB
cuesj-1010	113	26	comprehensive	comprehensive	ADJ
cuesj-1010	113	27	testing	testing	NOUN
cuesj-1010	113	28	and	and	CCONJ
cuesj-1010	113	29	assessment	assessment	NOUN
cuesj-1010	113	30	.	.	PUNCT
cuesj-1010	114	1	results	result	NOUN
cuesj-1010	114	2	and	and	CCONJ
cuesj-1010	114	3	discussion	discussion	VERB
cuesj-1010	114	4	our	our	PRON
cuesj-1010	114	5	yolov5	yolov5	NOUN
cuesj-1010	114	6	model	model	NOUN
cuesj-1010	114	7	embarked	embark	VERB
cuesj-1010	114	8	on	on	ADP
cuesj-1010	114	9	a	a	DET
cuesj-1010	114	10	rigorous	rigorous	ADJ
cuesj-1010	114	11	training	training	NOUN
cuesj-1010	114	12	regimen	regiman	NOUN
cuesj-1010	114	13	involving	involve	VERB
cuesj-1010	114	14	100	100	NUM
cuesj-1010	114	15	epochs	epoch	NOUN
cuesj-1010	114	16	,	,	PUNCT
cuesj-1010	114	17	fed	feed	VERB
cuesj-1010	114	18	by	by	ADP
cuesj-1010	114	19	a	a	DET
cuesj-1010	114	20	dataset	dataset	NOUN
cuesj-1010	114	21	comprising	comprise	VERB
cuesj-1010	114	22	2000	2000	NUM
cuesj-1010	114	23	images	image	NOUN
cuesj-1010	114	24	.	.	PUNCT
cuesj-1010	115	1	this	this	DET
cuesj-1010	115	2	arduous	arduous	ADJ
cuesj-1010	115	3	training	training	NOUN
cuesj-1010	115	4	journey	journey	NOUN
cuesj-1010	115	5	extended	extend	VERB
cuesj-1010	115	6	over	over	ADP
cuesj-1010	115	7	approximately	approximately	ADV
cuesj-1010	115	8	12	12	NUM
cuesj-1010	115	9	h	h	NOUN
cuesj-1010	115	10	,	,	PUNCT
cuesj-1010	115	11	with	with	ADP
cuesj-1010	115	12	yolov5x	yolov5x	PROPN
cuesj-1010	115	13	,	,	PUNCT
cuesj-1010	115	14	our	our	PRON
cuesj-1010	115	15	heavyweight	heavyweight	ADJ
cuesj-1010	115	16	contender	contender	NOUN
cuesj-1010	115	17	,	,	PUNCT
cuesj-1010	115	18	claiming	claim	VERB
cuesj-1010	115	19	the	the	DET
cuesj-1010	115	20	most	most	ADV
cuesj-1010	115	21	substantial	substantial	ADJ
cuesj-1010	115	22	share	share	NOUN
cuesj-1010	115	23	of	of	ADP
cuesj-1010	115	24	that	that	DET
cuesj-1010	115	25	time	time	NOUN
cuesj-1010	115	26	.	.	PUNCT
cuesj-1010	116	1	the	the	DET
cuesj-1010	116	2	colossal	colossal	PROPN
cuesj-1010	116	3	yolov5x	yolov5x	PROPN
cuesj-1010	116	4	model	model	NOUN
cuesj-1010	116	5	boasts	boast	VERB
cuesj-1010	116	6	a	a	DET
cuesj-1010	116	7	formidable	formidable	ADJ
cuesj-1010	116	8	166	166	NUM
cuesj-1010	116	9	mb	mb	NOUN
cuesj-1010	116	10	of	of	ADP
cuesj-1010	116	11	parameters	parameter	NOUN
cuesj-1010	116	12	.	.	PUNCT
cuesj-1010	117	1	in	in	ADP
cuesj-1010	117	2	our	our	PRON
cuesj-1010	117	3	pursuit	pursuit	NOUN
cuesj-1010	117	4	of	of	ADP
cuesj-1010	117	5	exceptional	exceptional	ADJ
cuesj-1010	117	6	precision	precision	NOUN
cuesj-1010	117	7	,	,	PUNCT
cuesj-1010	117	8	speed	speed	NOUN
cuesj-1010	117	9	was	be	AUX
cuesj-1010	117	10	a	a	DET
cuesj-1010	117	11	secondary	secondary	ADJ
cuesj-1010	117	12	consideration	consideration	NOUN
cuesj-1010	117	13	,	,	PUNCT
cuesj-1010	117	14	ensuring	ensure	VERB
cuesj-1010	117	15	that	that	SCONJ
cuesj-1010	117	16	our	our	PRON
cuesj-1010	117	17	primary	primary	ADJ
cuesj-1010	117	18	focus	focus	NOUN
cuesj-1010	117	19	remains	remain	VERB
cuesj-1010	117	20	on	on	ADP
cuesj-1010	117	21	achieving	achieve	VERB
cuesj-1010	117	22	the	the	DET
cuesj-1010	117	23	highest	high	ADJ
cuesj-1010	117	24	levels	level	NOUN
cuesj-1010	117	25	of	of	ADP
cuesj-1010	117	26	accuracy	accuracy	NOUN
cuesj-1010	117	27	.	.	PUNCT
cuesj-1010	118	1	for	for	ADP
cuesj-1010	118	2	testing	testing	NOUN
cuesj-1010	118	3	purposes	purpose	NOUN
cuesj-1010	118	4	,	,	PUNCT
cuesj-1010	118	5	we	we	PRON
cuesj-1010	118	6	exclusively	exclusively	ADV
cuesj-1010	118	7	scrutinized	scrutinize	VERB
cuesj-1010	118	8	the	the	DET
cuesj-1010	118	9	yolov5x	yolov5x	PROPN
cuesj-1010	118	10	model	model	NOUN
cuesj-1010	118	11	,	,	PUNCT
cuesj-1010	118	12	aligning	align	VERB
cuesj-1010	118	13	with	with	ADP
cuesj-1010	118	14	our	our	PRON
cuesj-1010	118	15	precision	precision	NOUN
cuesj-1010	118	16	-	-	PUNCT
cuesj-1010	118	17	driven	drive	VERB
cuesj-1010	118	18	objectives	objective	NOUN
cuesj-1010	118	19	.	.	PUNCT
cuesj-1010	119	1	however	however	ADV
cuesj-1010	119	2	,	,	PUNCT
cuesj-1010	119	3	for	for	ADP
cuesj-1010	119	4	a	a	DET
cuesj-1010	119	5	comprehensive	comprehensive	ADJ
cuesj-1010	119	6	comparative	comparative	ADJ
cuesj-1010	119	7	analysis	analysis	NOUN
cuesj-1010	119	8	,	,	PUNCT
cuesj-1010	119	9	we	we	PRON
cuesj-1010	119	10	present	present	VERB
cuesj-1010	119	11	the	the	DET
cuesj-1010	119	12	results	result	NOUN
cuesj-1010	119	13	for	for	ADP
cuesj-1010	119	14	all	all	DET
cuesj-1010	119	15	four	four	NUM
cuesj-1010	119	16	model	model	NOUN
cuesj-1010	119	17	variants	variant	NOUN
cuesj-1010	119	18	across	across	ADP
cuesj-1010	119	19	the	the	DET
cuesj-1010	119	20	dataset	dataset	NOUN
cuesj-1010	119	21	.	.	PUNCT
cuesj-1010	120	1	the	the	DET
cuesj-1010	120	2	performance	performance	NOUN
cuesj-1010	120	3	evaluation	evaluation	NOUN
cuesj-1010	120	4	of	of	ADP
cuesj-1010	120	5	our	our	PRON
cuesj-1010	120	6	yolo	yolo	ADJ
cuesj-1010	120	7	model	model	NOUN
cuesj-1010	120	8	extends	extend	VERB
cuesj-1010	120	9	to	to	ADP
cuesj-1010	120	10	an	an	DET
cuesj-1010	120	11	array	array	NOUN
cuesj-1010	120	12	of	of	ADP
cuesj-1010	120	13	critical	critical	ADJ
cuesj-1010	120	14	metrics	metric	NOUN
cuesj-1010	120	15	,	,	PUNCT
cuesj-1010	120	16	including	include	VERB
cuesj-1010	120	17	accuracy	accuracy	NOUN
cuesj-1010	120	18	,	,	PUNCT
cuesj-1010	120	19	mean	mean	ADJ
cuesj-1010	120	20	average	average	ADJ
cuesj-1010	120	21	precision	precision	NOUN
cuesj-1010	120	22	(	(	PUNCT
cuesj-1010	120	23	map	map	NOUN
cuesj-1010	120	24	)	)	PUNCT
cuesj-1010	120	25	,	,	PUNCT
cuesj-1010	120	26	average	average	ADJ
cuesj-1010	120	27	precision	precision	NOUN
cuesj-1010	120	28	(	(	PUNCT
cuesj-1010	120	29	ap	ap	PROPN
cuesj-1010	120	30	)	)	PUNCT
cuesj-1010	120	31	,	,	PUNCT
cuesj-1010	120	32	f1	f1	NOUN
cuesj-1010	120	33	score	score	NOUN
cuesj-1010	120	34	,	,	PUNCT
cuesj-1010	120	35	recall	recall	NOUN
cuesj-1010	120	36	,	,	PUNCT
cuesj-1010	120	37	and	and	CCONJ
cuesj-1010	120	38	precision.[16	precision.[16	PROPN
cuesj-1010	120	39	]	]	PUNCT
cuesj-1010	120	40	precision	precision	NOUN
cuesj-1010	120	41	,	,	PUNCT
cuesj-1010	120	42	one	one	NUM
cuesj-1010	120	43	of	of	ADP
cuesj-1010	120	44	the	the	DET
cuesj-1010	120	45	fundamental	fundamental	ADJ
cuesj-1010	120	46	metrics	metric	NOUN
cuesj-1010	120	47	,	,	PUNCT
cuesj-1010	120	48	is	be	AUX
cuesj-1010	120	49	derived	derive	VERB
cuesj-1010	120	50	by	by	ADP
cuesj-1010	120	51	dividing	divide	VERB
cuesj-1010	120	52	the	the	DET
cuesj-1010	120	53	total	total	ADJ
cuesj-1010	120	54	count	count	NOUN
cuesj-1010	120	55	of	of	ADP
cuesj-1010	120	56	samples	sample	NOUN
cuesj-1010	120	57	classified	classify	VERB
cuesj-1010	120	58	as	as	ADP
cuesj-1010	120	59	positive	positive	ADJ
cuesj-1010	120	60	(	(	PUNCT
cuesj-1010	120	61	regardless	regardless	ADV
cuesj-1010	120	62	of	of	ADP
cuesj-1010	120	63	correctness	correctness	NOUN
cuesj-1010	120	64	)	)	PUNCT
cuesj-1010	120	65	by	by	ADP
cuesj-1010	120	66	the	the	DET
cuesj-1010	120	67	count	count	NOUN
cuesj-1010	120	68	of	of	ADP
cuesj-1010	120	69	positive	positive	ADJ
cuesj-1010	120	70	samples	sample	NOUN
cuesj-1010	120	71	correctly	correctly	ADV
cuesj-1010	120	72	identified	identify	VERB
cuesj-1010	120	73	.	.	PUNCT
cuesj-1010	121	1	precision	precision	NOUN
cuesj-1010	121	2	serves	serve	VERB
cuesj-1010	121	3	as	as	ADP
cuesj-1010	121	4	a	a	DET
cuesj-1010	121	5	valuable	valuable	ADJ
cuesj-1010	121	6	gauge	gauge	NOUN
cuesj-1010	121	7	of	of	ADP
cuesj-1010	121	8	the	the	DET
cuesj-1010	121	9	model	model	NOUN
cuesj-1010	121	10	’s	’s	PART
cuesj-1010	121	11	proficiency	proficiency	NOUN
cuesj-1010	121	12	in	in	ADP
cuesj-1010	121	13	accurately	accurately	ADV
cuesj-1010	121	14	classifying	classify	VERB
cuesj-1010	121	15	samples	sample	NOUN
cuesj-1010	121	16	as	as	ADP
cuesj-1010	121	17	positive.[17	positive.[17	PROPN
cuesj-1010	121	18	]	]	PUNCT
cuesj-1010	121	19	the	the	PRON
cuesj-1010	121	20	below	below	ADV
cuesj-1010	121	21	(	(	PUNCT
cuesj-1010	121	22	2	2	X
cuesj-1010	121	23	)	)	PUNCT
cuesj-1010	121	24	define	define	VERB
cuesj-1010	121	25	the	the	DET
cuesj-1010	121	26	precision	precision	NOUN
cuesj-1010	121	27	:	:	PUNCT
cuesj-1010	122	1	=	=	SYM
cuesj-1010	122	2	+	+	NUM
cuesj-1010	122	3	precision	precision	NOUN
cuesj-1010	122	4	tp	tp	NOUN
cuesj-1010	122	5	tp	tp	ADP
cuesj-1010	122	6	fp	fp	PROPN
cuesj-1010	122	7	(	(	PUNCT
cuesj-1010	122	8	2	2	NUM
cuesj-1010	122	9	)	)	PUNCT
cuesj-1010	122	10	tp	tp	NOUN
cuesj-1010	122	11	is	be	AUX
cuesj-1010	122	12	assigned	assign	VERB
cuesj-1010	122	13	to	to	ADP
cuesj-1010	122	14	true	true	ADJ
cuesj-1010	122	15	positive	positive	ADJ
cuesj-1010	122	16	,	,	PUNCT
cuesj-1010	122	17	when	when	SCONJ
cuesj-1010	122	18	the	the	DET
cuesj-1010	122	19	model	model	NOUN
cuesj-1010	122	20	accurately	accurately	ADV
cuesj-1010	122	21	predicts	predict	VERB
cuesj-1010	122	22	the	the	DET
cuesj-1010	122	23	positive	positive	ADJ
cuesj-1010	122	24	class	class	NOUN
cuesj-1010	122	25	,	,	PUNCT
cuesj-1010	122	26	the	the	DET
cuesj-1010	122	27	result	result	NOUN
cuesj-1010	122	28	is	be	AUX
cuesj-1010	122	29	a	a	DET
cuesj-1010	122	30	true	true	ADJ
cuesj-1010	122	31	positive	positive	ADJ
cuesj-1010	122	32	.	.	PUNCT
cuesj-1010	123	1	fp	fp	NOUN
cuesj-1010	123	2	figure	figure	NOUN
cuesj-1010	123	3	3	3	NUM
cuesj-1010	123	4	:	:	PUNCT
cuesj-1010	123	5	the	the	DET
cuesj-1010	123	6	dataset	dataset	NOUN
cuesj-1010	123	7	example	example	NOUN
cuesj-1010	123	8	of	of	ADP
cuesj-1010	123	9	vehicle	vehicle	NOUN
cuesj-1010	123	10	detection	detection	NOUN
cuesj-1010	123	11	abdallah	abdallah	PROPN
cuesj-1010	123	12	:	:	PUNCT
cuesj-1010	123	13	vehicle	vehicle	NOUN
cuesj-1010	123	14	detection	detection	NOUN
cuesj-1010	123	15	with	with	ADP
cuesj-1010	123	16	tello	tello	PROPN
cuesj-1010	123	17	drone	drone	PROPN
cuesj-1010	123	18	and	and	CCONJ
cuesj-1010	123	19	yolov5	yolov5	NOUN
cuesj-1010	123	20	5	5	NUM
cuesj-1010	123	21	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	123	22	cuesj	cuesj	NOUN
cuesj-1010	123	23	2024	2024	NUM
cuesj-1010	123	24	,	,	PUNCT
cuesj-1010	123	25	8	8	NUM
cuesj-1010	123	26	(	(	PUNCT
cuesj-1010	123	27	1	1	NUM
cuesj-1010	123	28	):	):	PUNCT
cuesj-1010	123	29	1	1	NUM
cuesj-1010	123	30	-	-	SYM
cuesj-1010	123	31	7	7	NUM
cuesj-1010	123	32	assigned	assign	VERB
cuesj-1010	123	33	to	to	ADP
cuesj-1010	123	34	false	false	ADJ
cuesj-1010	123	35	positive	positive	ADJ
cuesj-1010	123	36	,	,	PUNCT
cuesj-1010	123	37	when	when	SCONJ
cuesj-1010	123	38	the	the	DET
cuesj-1010	123	39	model	model	NOUN
cuesj-1010	123	40	forecasts	forecast	VERB
cuesj-1010	123	41	the	the	DET
cuesj-1010	123	42	positive	positive	ADJ
cuesj-1010	123	43	class	class	NOUN
cuesj-1010	123	44	inaccurately	inaccurately	ADV
cuesj-1010	123	45	,	,	PUNCT
cuesj-1010	123	46	the	the	DET
cuesj-1010	123	47	result	result	NOUN
cuesj-1010	123	48	is	be	AUX
cuesj-1010	123	49	called	call	VERB
cuesj-1010	123	50	a	a	DET
cuesj-1010	123	51	false	false	ADJ
cuesj-1010	123	52	positive.[17	positive.[17	NOUN
cuesj-1010	123	53	]	]	PUNCT
cuesj-1010	123	54	the	the	DET
cuesj-1010	123	55	second	second	ADJ
cuesj-1010	123	56	metric	metric	NOUN
cuesj-1010	123	57	is	be	AUX
cuesj-1010	123	58	recalled	recall	VERB
cuesj-1010	123	59	,	,	PUNCT
cuesj-1010	123	60	recall	recall	PROPN
cuesj-1010	123	61	is	be	AUX
cuesj-1010	123	62	determined	determine	VERB
cuesj-1010	123	63	by	by	ADP
cuesj-1010	123	64	dividing	divide	VERB
cuesj-1010	123	65	the	the	DET
cuesj-1010	123	66	total	total	ADJ
cuesj-1010	123	67	number	number	NOUN
cuesj-1010	123	68	of	of	ADP
cuesj-1010	123	69	positive	positive	ADJ
cuesj-1010	123	70	samples	sample	NOUN
cuesj-1010	123	71	by	by	ADP
cuesj-1010	123	72	the	the	DET
cuesj-1010	123	73	number	number	NOUN
cuesj-1010	123	74	of	of	ADP
cuesj-1010	123	75	positive	positive	ADJ
cuesj-1010	123	76	samples	sample	NOUN
cuesj-1010	123	77	that	that	PRON
cuesj-1010	123	78	were	be	AUX
cuesj-1010	123	79	accurately	accurately	ADV
cuesj-1010	123	80	categorized	categorize	VERB
cuesj-1010	123	81	as	as	ADP
cuesj-1010	123	82	positive	positive	ADJ
cuesj-1010	123	83	.	.	PUNCT
cuesj-1010	124	1	the	the	DET
cuesj-1010	124	2	recall	recall	NOUN
cuesj-1010	124	3	gauges	gauge	VERB
cuesj-1010	124	4	how	how	SCONJ
cuesj-1010	124	5	well	well	ADV
cuesj-1010	124	6	the	the	DET
cuesj-1010	124	7	model	model	NOUN
cuesj-1010	124	8	can	can	AUX
cuesj-1010	124	9	identify	identify	VERB
cuesj-1010	124	10	positive	positive	ADJ
cuesj-1010	124	11	samples	sample	NOUN
cuesj-1010	124	12	.	.	PUNCT
cuesj-1010	125	1	positive	positive	ADJ
cuesj-1010	125	2	samples	sample	NOUN
cuesj-1010	125	3	are	be	AUX
cuesj-1010	125	4	found	find	VERB
cuesj-1010	125	5	in	in	ADP
cuesj-1010	125	6	greater	great	ADJ
cuesj-1010	125	7	numbers	number	NOUN
cuesj-1010	125	8	the	the	PRON
cuesj-1010	125	9	higher	high	ADJ
cuesj-1010	125	10	the	the	DET
cuesj-1010	125	11	recall	recall	NOUN
cuesj-1010	125	12	.	.	PUNCT
cuesj-1010	126	1	to	to	ADP
cuesj-1010	126	2	defining	define	VERB
cuesj-1010	126	3	recall	recall	NOUN
cuesj-1010	126	4	(	(	PUNCT
cuesj-1010	126	5	3	3	NUM
cuesj-1010	126	6	)	)	PUNCT
cuesj-1010	126	7	is	be	AUX
cuesj-1010	126	8	used	use	VERB
cuesj-1010	126	9	:	:	PUNCT
cuesj-1010	126	10	=	=	SYM
cuesj-1010	127	1	+	+	CCONJ
cuesj-1010	127	2	re	re	VERB
cuesj-1010	127	3	tp	tp	PART
cuesj-1010	127	4	call	call	VERB
cuesj-1010	127	5	tp	tp	ADP
cuesj-1010	127	6	fn	fn	PROPN
cuesj-1010	127	7	(	(	PUNCT
cuesj-1010	127	8	3	3	NUM
cuesj-1010	127	9	)	)	PUNCT
cuesj-1010	127	10	here	here	ADV
cuesj-1010	127	11	fn	fn	PRON
cuesj-1010	127	12	false	false	ADJ
cuesj-1010	127	13	negative	negative	ADJ
cuesj-1010	127	14	,	,	PUNCT
cuesj-1010	127	15	when	when	SCONJ
cuesj-1010	127	16	the	the	DET
cuesj-1010	127	17	model	model	NOUN
cuesj-1010	127	18	forecasts	forecast	VERB
cuesj-1010	127	19	the	the	DET
cuesj-1010	127	20	negative	negative	ADJ
cuesj-1010	127	21	class	class	NOUN
cuesj-1010	127	22	inaccurately	inaccurately	ADV
cuesj-1010	127	23	,	,	PUNCT
cuesj-1010	127	24	the	the	DET
cuesj-1010	127	25	result	result	NOUN
cuesj-1010	127	26	is	be	AUX
cuesj-1010	127	27	known	know	VERB
cuesj-1010	127	28	as	as	ADP
cuesj-1010	127	29	a	a	DET
cuesj-1010	127	30	false	false	ADJ
cuesj-1010	127	31	negative	negative	NOUN
cuesj-1010	127	32	.	.	PUNCT
cuesj-1010	128	1	the	the	DET
cuesj-1010	128	2	number	number	NOUN
cuesj-1010	128	3	of	of	ADP
cuesj-1010	128	4	f1	f1	ADJ
cuesj-1010	128	5	score	score	NOUN
cuesj-1010	128	6	is	be	AUX
cuesj-1010	128	7	from	from	ADP
cuesj-1010	128	8	0	0	NUM
cuesj-1010	128	9	to	to	ADP
cuesj-1010	128	10	1	1	NUM
cuesj-1010	128	11	;	;	PUNCT
cuesj-1010	128	12	the	the	PRON
cuesj-1010	128	13	greater	great	ADJ
cuesj-1010	128	14	the	the	DET
cuesj-1010	128	15	value	value	NOUN
cuesj-1010	128	16	,	,	PUNCT
cuesj-1010	128	17	the	the	PRON
cuesj-1010	128	18	better	well	ADJ
cuesj-1010	128	19	the	the	DET
cuesj-1010	128	20	accuracy	accuracy	NOUN
cuesj-1010	128	21	of	of	ADP
cuesj-1010	128	22	identifying	identify	VERB
cuesj-1010	128	23	an	an	DET
cuesj-1010	128	24	item	item	NOUN
cuesj-1010	128	25	.	.	PUNCT
cuesj-1010	129	1	the	the	DET
cuesj-1010	129	2	f1	f1	PROPN
cuesj-1010	129	3	score	score	NOUN
cuesj-1010	129	4	indicates	indicate	VERB
cuesj-1010	129	5	how	how	SCONJ
cuesj-1010	129	6	well	well	ADV
cuesj-1010	129	7	the	the	DET
cuesj-1010	129	8	model	model	NOUN
cuesj-1010	129	9	balances	balance	VERB
cuesj-1010	129	10	its	its	PRON
cuesj-1010	129	11	capacity	capacity	NOUN
cuesj-1010	129	12	to	to	PART
cuesj-1010	129	13	capture	capture	VERB
cuesj-1010	129	14	positive	positive	ADJ
cuesj-1010	129	15	cases	case	NOUN
cuesj-1010	129	16	(	(	PUNCT
cuesj-1010	129	17	recall	recall	NOUN
cuesj-1010	129	18	)	)	PUNCT
cuesj-1010	129	19	and	and	CCONJ
cuesj-1010	129	20	be	be	AUX
cuesj-1010	129	21	correct	correct	ADJ
cuesj-1010	129	22	with	with	ADP
cuesj-1010	129	23	the	the	DET
cuesj-1010	129	24	cases	case	NOUN
cuesj-1010	129	25	it	it	PRON
cuesj-1010	129	26	captures	capture	VERB
cuesj-1010	129	27	(	(	PUNCT
cuesj-1010	129	28	precision	precision	NOUN
cuesj-1010	129	29	)	)	PUNCT
cuesj-1010	129	30	.	.	PUNCT
cuesj-1010	130	1	the	the	DET
cuesj-1010	130	2	f1	f1	PROPN
cuesj-1010	130	3	score	score	NOUN
cuesj-1010	130	4	is	be	AUX
cuesj-1010	130	5	calculated	calculate	VERB
cuesj-1010	130	6	as	as	ADP
cuesj-1010	130	7	the	the	DET
cuesj-1010	130	8	harmonic	harmonic	ADJ
cuesj-1010	130	9	mean	mean	NOUN
cuesj-1010	130	10	of	of	ADP
cuesj-1010	130	11	precision	precision	NOUN
cuesj-1010	130	12	and	and	CCONJ
cuesj-1010	130	13	recall	recall	NOUN
cuesj-1010	130	14	.	.	PUNCT
cuesj-1010	131	1	it	it	PRON
cuesj-1010	131	2	is	be	AUX
cuesj-1010	131	3	a	a	DET
cuesj-1010	131	4	widely	widely	ADV
cuesj-1010	131	5	used	use	VERB
cuesj-1010	131	6	statistic	statistic	NOUN
cuesj-1010	131	7	for	for	ADP
cuesj-1010	131	8	classification	classification	NOUN
cuesj-1010	131	9	models	model	NOUN
cuesj-1010	131	10	since	since	SCONJ
cuesj-1010	131	11	it	it	PRON
cuesj-1010	131	12	assesses	assess	VERB
cuesj-1010	131	13	the	the	DET
cuesj-1010	131	14	model	model	NOUN
cuesj-1010	131	15	’s	’s	PART
cuesj-1010	131	16	capacity	capacity	NOUN
cuesj-1010	131	17	for	for	ADP
cuesj-1010	131	18	both	both	CCONJ
cuesj-1010	131	19	recall	recall	NOUN
cuesj-1010	131	20	and	and	CCONJ
cuesj-1010	131	21	precision	precision	NOUN
cuesj-1010	131	22	and	and	CCONJ
cuesj-1010	131	23	yields	yield	NOUN
cuesj-1010	131	24	reliable	reliable	ADJ
cuesj-1010	131	25	findings	finding	NOUN
cuesj-1010	131	26	even	even	ADV
cuesj-1010	131	27	for	for	ADP
cuesj-1010	131	28	unbalanced	unbalanced	ADJ
cuesj-1010	131	29	datasets.[18	datasets.[18	PROPN
cuesj-1010	131	30	]	]	PUNCT
cuesj-1010	131	31	for	for	ADP
cuesj-1010	131	32	this	this	DET
cuesj-1010	131	33	reason	reason	NOUN
cuesj-1010	131	34	,	,	PUNCT
cuesj-1010	131	35	f1	f1	PROPN
cuesj-1010	131	36	can	can	AUX
cuesj-1010	131	37	assess	assess	VERB
cuesj-1010	131	38	a	a	DET
cuesj-1010	131	39	model	model	NOUN
cuesj-1010	131	40	’s	’s	PART
cuesj-1010	131	41	recall	recall	NOUN
cuesj-1010	131	42	and	and	CCONJ
cuesj-1010	131	43	precision	precision	NOUN
cuesj-1010	131	44	because	because	SCONJ
cuesj-1010	131	45	of	of	ADP
cuesj-1010	131	46	the	the	DET
cuesj-1010	131	47	following	follow	VERB
cuesj-1010	131	48	derivation	derivation	NOUN
cuesj-1010	131	49	(	(	PUNCT
cuesj-1010	131	50	4	4	NUM
cuesj-1010	131	51	)	)	PUNCT
cuesj-1010	131	52	method	method	NOUN
cuesj-1010	131	53	:	:	PUNCT
cuesj-1010	132	1	=	=	NOUN
cuesj-1010	132	2	+	+	NUM
cuesj-1010	132	3	2	2	NUM
cuesj-1010	132	4	*	*	NUM
cuesj-1010	132	5	(	(	PUNCT
cuesj-1010	132	6	*	*	PUNCT
cuesj-1010	132	7	re	re	ADP
cuesj-1010	132	8	1	1	NUM
cuesj-1010	132	9	_	_	PRON
cuesj-1010	132	10	(	(	PUNCT
cuesj-1010	132	11	pr	pr	X
cuesj-1010	132	12	re	re	NOUN
cuesj-1010	132	13	)	)	PUNCT
cuesj-1010	132	14	precision	precision	NOUN
cuesj-1010	132	15	call	call	NOUN
cuesj-1010	132	16	f	f	PROPN
cuesj-1010	132	17	score	score	NOUN
cuesj-1010	132	18	ecision	ecision	PROPN
cuesj-1010	132	19	call	call	NOUN
cuesj-1010	132	20	(	(	PUNCT
cuesj-1010	132	21	4	4	NUM
cuesj-1010	132	22	)	)	PUNCT
cuesj-1010	132	23	the	the	DET
cuesj-1010	132	24	final	final	ADJ
cuesj-1010	132	25	metric	metric	NOUN
cuesj-1010	132	26	is	be	AUX
cuesj-1010	132	27	a	a	DET
cuesj-1010	132	28	map	map	NOUN
cuesj-1010	132	29	,	,	PUNCT
cuesj-1010	132	30	computer	computer	NOUN
cuesj-1010	132	31	vision	vision	NOUN
cuesj-1010	132	32	model	model	NOUN
cuesj-1010	132	33	performance	performance	NOUN
cuesj-1010	132	34	is	be	AUX
cuesj-1010	132	35	measured	measure	VERB
cuesj-1010	132	36	using	using	NOUN
cuesj-1010	132	37	map	map	NOUN
cuesj-1010	132	38	.	.	PUNCT
cuesj-1010	133	1	the	the	DET
cuesj-1010	133	2	mean	mean	NOUN
cuesj-1010	133	3	of	of	ADP
cuesj-1010	133	4	the	the	DET
cuesj-1010	133	5	average	average	ADJ
cuesj-1010	133	6	precision	precision	NOUN
cuesj-1010	133	7	metric	metric	ADJ
cuesj-1010	133	8	over	over	ADP
cuesj-1010	133	9	all	all	DET
cuesj-1010	133	10	classes	class	NOUN
cuesj-1010	133	11	in	in	ADP
cuesj-1010	133	12	a	a	DET
cuesj-1010	133	13	model	model	NOUN
cuesj-1010	133	14	is	be	AUX
cuesj-1010	133	15	equal	equal	ADJ
cuesj-1010	133	16	to	to	PART
cuesj-1010	133	17	map	map	VERB
cuesj-1010	133	18	.	.	PUNCT
cuesj-1010	134	1	map	map	NOUN
cuesj-1010	134	2	can	can	AUX
cuesj-1010	134	3	be	be	AUX
cuesj-1010	134	4	used	use	VERB
cuesj-1010	134	5	to	to	PART
cuesj-1010	134	6	compare	compare	VERB
cuesj-1010	134	7	different	different	ADJ
cuesj-1010	134	8	iterations	iteration	NOUN
cuesj-1010	134	9	of	of	ADP
cuesj-1010	134	10	the	the	DET
cuesj-1010	134	11	same	same	ADJ
cuesj-1010	134	12	model	model	NOUN
cuesj-1010	134	13	as	as	ADV
cuesj-1010	134	14	well	well	ADV
cuesj-1010	134	15	as	as	ADP
cuesj-1010	134	16	various	various	ADJ
cuesj-1010	134	17	models	model	NOUN
cuesj-1010	134	18	on	on	ADP
cuesj-1010	134	19	the	the	DET
cuesj-1010	134	20	same	same	ADJ
cuesj-1010	134	21	job	job	NOUN
cuesj-1010	134	22	.	.	PUNCT
cuesj-1010	135	1	the	the	DET
cuesj-1010	135	2	range	range	NOUN
cuesj-1010	135	3	of	of	ADP
cuesj-1010	135	4	map	map	NOUN
cuesj-1010	135	5	measurement	measurement	NOUN
cuesj-1010	135	6	is	be	AUX
cuesj-1010	135	7	0	0	NUM
cuesj-1010	135	8	to	to	PART
cuesj-1010	135	9	1	1	NUM
cuesj-1010	135	10	.	.	PUNCT
cuesj-1010	136	1	before	before	ADP
cuesj-1010	136	2	showing	show	VERB
cuesj-1010	136	3	the	the	DET
cuesj-1010	136	4	equation	equation	NOUN
cuesj-1010	136	5	of	of	ADP
cuesj-1010	136	6	map	map	NOUN
cuesj-1010	136	7	,	,	PUNCT
cuesj-1010	136	8	there	there	PRON
cuesj-1010	136	9	are	be	VERB
cuesj-1010	136	10	specific	specific	ADJ
cuesj-1010	136	11	metrics	metric	NOUN
cuesj-1010	136	12	that	that	PRON
cuesj-1010	136	13	affect	affect	VERB
cuesj-1010	136	14	the	the	DET
cuesj-1010	136	15	calculating	calculating	NOUN
cuesj-1010	136	16	in	in	ADP
cuesj-1010	136	17	map	map	NOUN
cuesj-1010	136	18	.	.	PUNCT
cuesj-1010	137	1	the	the	DET
cuesj-1010	137	2	metrics	metric	NOUN
cuesj-1010	137	3	involve	involve	VERB
cuesj-1010	137	4	iou	iou	NOUN
cuesj-1010	137	5	,	,	PUNCT
cuesj-1010	137	6	recall	recall	NOUN
cuesj-1010	137	7	,	,	PUNCT
cuesj-1010	137	8	precision	precision	NOUN
cuesj-1010	137	9	,	,	PUNCT
cuesj-1010	137	10	and	and	CCONJ
cuesj-1010	137	11	confusion	confusion	NOUN
cuesj-1010	137	12	matrix.[19	matrix.[19	NOUN
cuesj-1010	137	13	]	]	PUNCT
cuesj-1010	137	14	in	in	ADP
cuesj-1010	137	15	the	the	DET
cuesj-1010	137	16	previous	previous	ADJ
cuesj-1010	137	17	we	we	PRON
cuesj-1010	137	18	mention	mention	VERB
cuesj-1010	137	19	iou	iou	PROPN
cuesj-1010	137	20	,	,	PUNCT
cuesj-1010	137	21	recall	recall	NOUN
cuesj-1010	137	22	,	,	PUNCT
cuesj-1010	137	23	and	and	CCONJ
cuesj-1010	137	24	precision	precision	NOUN
cuesj-1010	137	25	.	.	PUNCT
cuesj-1010	138	1	let	let	VERB
cuesj-1010	138	2	us	we	PRON
cuesj-1010	138	3	go	go	VERB
cuesj-1010	138	4	into	into	ADP
cuesj-1010	138	5	the	the	DET
cuesj-1010	138	6	confusion	confusion	NOUN
cuesj-1010	138	7	matrix	matrix	NOUN
cuesj-1010	138	8	,	,	PUNCT
cuesj-1010	138	9	basicly	basicly	ADV
cuesj-1010	138	10	this	this	DET
cuesj-1010	138	11	elements	element	NOUN
cuesj-1010	138	12	are	be	AUX
cuesj-1010	138	13	explained	explain	VERB
cuesj-1010	138	14	in	in	ADP
cuesj-1010	138	15	table	table	NOUN
cuesj-1010	138	16	3	3	NUM
cuesj-1010	138	17	.	.	PUNCT
cuesj-1010	139	1	the	the	DET
cuesj-1010	139	2	following	follow	VERB
cuesj-1010	139	3	four	four	NUM
cuesj-1010	139	4	elements	element	NOUN
cuesj-1010	139	5	are	be	AUX
cuesj-1010	139	6	essential	essential	ADJ
cuesj-1010	139	7	for	for	ADP
cuesj-1010	139	8	the	the	DET
cuesj-1010	139	9	confusion	confusion	NOUN
cuesj-1010	139	10	matrix:[20	matrix:[20	NOUN
cuesj-1010	139	11	]	]	X
cuesj-1010	139	12	true	true	ADJ
cuesj-1010	139	13	positives	positive	NOUN
cuesj-1010	139	14	(	(	PUNCT
cuesj-1010	139	15	tp	tp	NOUN
cuesj-1010	139	16	):	):	PUNCT
cuesj-1010	139	17	the	the	DET
cuesj-1010	139	18	label	label	NOUN
cuesj-1010	139	19	was	be	AUX
cuesj-1010	139	20	correctly	correctly	ADV
cuesj-1010	139	21	predicted	predict	VERB
cuesj-1010	139	22	by	by	ADP
cuesj-1010	139	23	the	the	DET
cuesj-1010	139	24	model	model	NOUN
cuesj-1010	139	25	and	and	CCONJ
cuesj-1010	139	26	it	it	PRON
cuesj-1010	139	27	matched	match	VERB
cuesj-1010	139	28	the	the	DET
cuesj-1010	139	29	ground	ground	NOUN
cuesj-1010	139	30	truth	truth	NOUN
cuesj-1010	139	31	.	.	PUNCT
cuesj-1010	140	1	the	the	DET
cuesj-1010	140	2	appropriate	appropriate	ADJ
cuesj-1010	140	3	object	object	NOUN
cuesj-1010	140	4	type	type	NOUN
cuesj-1010	140	5	and	and	CCONJ
cuesj-1010	140	6	location	location	NOUN
cuesj-1010	140	7	have	have	AUX
cuesj-1010	140	8	been	be	AUX
cuesj-1010	140	9	identified	identify	VERB
cuesj-1010	140	10	by	by	ADP
cuesj-1010	140	11	the	the	DET
cuesj-1010	140	12	model	model	NOUN
cuesj-1010	140	13	.	.	PUNCT
cuesj-1010	141	1	true	true	ADJ
cuesj-1010	141	2	negatives	negative	NOUN
cuesj-1010	141	3	(	(	PUNCT
cuesj-1010	141	4	tn	tn	NOUN
cuesj-1010	141	5	):	):	PUNCT
cuesj-1010	141	6	neither	neither	CCONJ
cuesj-1010	141	7	does	do	AUX
cuesj-1010	141	8	the	the	DET
cuesj-1010	141	9	model	model	NOUN
cuesj-1010	141	10	predict	predict	VERB
cuesj-1010	141	11	the	the	DET
cuesj-1010	141	12	label	label	NOUN
cuesj-1010	141	13	nor	nor	CCONJ
cuesj-1010	141	14	is	be	AUX
cuesj-1010	141	15	it	it	PRON
cuesj-1010	141	16	a	a	DET
cuesj-1010	141	17	part	part	NOUN
cuesj-1010	141	18	of	of	ADP
cuesj-1010	141	19	the	the	DET
cuesj-1010	141	20	ground	ground	NOUN
cuesj-1010	141	21	truth	truth	NOUN
cuesj-1010	141	22	.	.	PUNCT
cuesj-1010	142	1	the	the	DET
cuesj-1010	142	2	incorrect	incorrect	ADJ
cuesj-1010	142	3	class	class	NOUN
cuesj-1010	142	4	was	be	AUX
cuesj-1010	142	5	accurately	accurately	ADV
cuesj-1010	142	6	predicted	predict	VERB
cuesj-1010	142	7	by	by	ADP
cuesj-1010	142	8	the	the	DET
cuesj-1010	142	9	model	model	NOUN
cuesj-1010	142	10	.	.	PUNCT
cuesj-1010	143	1	false	false	ADJ
cuesj-1010	143	2	positives	positive	NOUN
cuesj-1010	143	3	(	(	PUNCT
cuesj-1010	143	4	fp	fp	X
cuesj-1010	143	5	):	):	PUNCT
cuesj-1010	143	6	the	the	DET
cuesj-1010	143	7	model	model	NOUN
cuesj-1010	143	8	predicted	predict	VERB
cuesj-1010	143	9	a	a	DET
cuesj-1010	143	10	label	label	NOUN
cuesj-1010	143	11	,	,	PUNCT
cuesj-1010	143	12	but	but	CCONJ
cuesj-1010	143	13	it	it	PRON
cuesj-1010	143	14	is	be	AUX
cuesj-1010	143	15	not	not	PART
cuesj-1010	143	16	included	include	VERB
cuesj-1010	143	17	in	in	ADP
cuesj-1010	143	18	the	the	DET
cuesj-1010	143	19	ground	ground	NOUN
cuesj-1010	143	20	truth	truth	NOUN
cuesj-1010	143	21	.	.	PUNCT
cuesj-1010	144	1	the	the	DET
cuesj-1010	144	2	model	model	NOUN
cuesj-1010	144	3	has	have	AUX
cuesj-1010	144	4	either	either	CCONJ
cuesj-1010	144	5	given	give	VERB
cuesj-1010	144	6	the	the	DET
cuesj-1010	144	7	incorrect	incorrect	ADJ
cuesj-1010	144	8	label	label	NOUN
cuesj-1010	144	9	or	or	CCONJ
cuesj-1010	144	10	recognized	recognize	VERB
cuesj-1010	144	11	an	an	DET
cuesj-1010	144	12	object	object	NOUN
cuesj-1010	144	13	that	that	PRON
cuesj-1010	144	14	is	be	AUX
cuesj-1010	144	15	not	not	PART
cuesj-1010	144	16	there	there	ADV
cuesj-1010	144	17	.	.	PUNCT
cuesj-1010	145	1	false	false	ADJ
cuesj-1010	145	2	negatives	negative	NOUN
cuesj-1010	145	3	(	(	PUNCT
cuesj-1010	145	4	fn	fn	NOUN
cuesj-1010	145	5	):	):	PUNCT
cuesj-1010	145	6	although	although	SCONJ
cuesj-1010	145	7	a	a	DET
cuesj-1010	145	8	label	label	NOUN
cuesj-1010	145	9	is	be	AUX
cuesj-1010	145	10	a	a	DET
cuesj-1010	145	11	part	part	NOUN
cuesj-1010	145	12	of	of	ADP
cuesj-1010	145	13	the	the	DET
cuesj-1010	145	14	ground	ground	NOUN
cuesj-1010	145	15	truth	truth	NOUN
cuesj-1010	145	16	,	,	PUNCT
cuesj-1010	145	17	it	it	PRON
cuesj-1010	145	18	is	be	AUX
cuesj-1010	145	19	not	not	PART
cuesj-1010	145	20	predicted	predict	VERB
cuesj-1010	145	21	by	by	ADP
cuesj-1010	145	22	the	the	DET
cuesj-1010	145	23	model	model	NOUN
cuesj-1010	145	24	.	.	PUNCT
cuesj-1010	146	1	the	the	DET
cuesj-1010	146	2	model	model	NOUN
cuesj-1010	146	3	has	have	AUX
cuesj-1010	146	4	n’t	not	PART
cuesj-1010	146	5	picked	pick	VERB
cuesj-1010	146	6	up	up	ADP
cuesj-1010	146	7	on	on	ADP
cuesj-1010	146	8	the	the	DET
cuesj-1010	146	9	object	object	NOUN
cuesj-1010	146	10	.	.	PUNCT
cuesj-1010	147	1	table	table	NOUN
cuesj-1010	147	2	3	3	NUM
cuesj-1010	147	3	shows	show	VERB
cuesj-1010	147	4	the	the	DET
cuesj-1010	147	5	confusion	confusion	NOUN
cuesj-1010	147	6	matrix	matrix	NOUN
cuesj-1010	147	7	’s	’s	PART
cuesj-1010	147	8	states	state	NOUN
cuesj-1010	147	9	and	and	CCONJ
cuesj-1010	147	10	labels	label	NOUN
cuesj-1010	147	11	:	:	PUNCT
cuesj-1010	147	12	now	now	ADV
cuesj-1010	147	13	the	the	DET
cuesj-1010	147	14	calculation	calculation	NOUN
cuesj-1010	147	15	of	of	ADP
cuesj-1010	147	16	the	the	DET
cuesj-1010	147	17	map	map	NOUN
cuesj-1010	147	18	will	will	AUX
cuesj-1010	147	19	be	be	AUX
cuesj-1010	147	20	like	like	ADP
cuesj-1010	147	21	(	(	PUNCT
cuesj-1010	147	22	5	5	NUM
cuesj-1010	147	23	):	):	PUNCT
cuesj-1010	147	24	=	=	SYM
cuesj-1010	147	25	=	=	SYM
cuesj-1010	147	26	∑	∑	PROPN
cuesj-1010	148	1	1	1	NUM
cuesj-1010	148	2	1	1	NUM
cuesj-1010	148	3	c	c	NOUN
cuesj-1010	148	4	i	i	PRON
cuesj-1010	148	5	map	map	VERB
cuesj-1010	148	6	api	api	NOUN
cuesj-1010	148	7	c	c	PROPN
cuesj-1010	148	8	(	(	PUNCT
cuesj-1010	148	9	5	5	NUM
cuesj-1010	148	10	)	)	PUNCT
cuesj-1010	148	11	here	here	ADV
cuesj-1010	148	12	the	the	DET
cuesj-1010	148	13	c	c	PROPN
cuesj-1010	148	14	refers	refer	VERB
cuesj-1010	148	15	to	to	ADP
cuesj-1010	148	16	the	the	DET
cuesj-1010	148	17	number	number	NOUN
cuesj-1010	148	18	of	of	ADP
cuesj-1010	148	19	the	the	DET
cuesj-1010	148	20	class	class	NOUN
cuesj-1010	148	21	used	use	VERB
cuesj-1010	148	22	in	in	ADP
cuesj-1010	148	23	the	the	DET
cuesj-1010	148	24	model	model	NOUN
cuesj-1010	148	25	,	,	PUNCT
cuesj-1010	148	26	and	and	CCONJ
cuesj-1010	148	27	the	the	DET
cuesj-1010	148	28	api	api	NOUN
cuesj-1010	148	29	is	be	AUX
cuesj-1010	148	30	the	the	DET
cuesj-1010	148	31	average	average	ADJ
cuesj-1010	148	32	precision	precision	NOUN
cuesj-1010	148	33	of	of	ADP
cuesj-1010	148	34	the	the	DET
cuesj-1010	148	35	certain	certain	ADJ
cuesj-1010	148	36	class	class	NOUN
cuesj-1010	148	37	.	.	PUNCT
cuesj-1010	149	1	ap	ap	PROPN
cuesj-1010	149	2	computes	compute	VERB
cuesj-1010	149	3	the	the	DET
cuesj-1010	149	4	precision	precision	NOUN
cuesj-1010	149	5	at	at	ADP
cuesj-1010	149	6	each	each	DET
cuesj-1010	149	7	threshold	threshold	NOUN
cuesj-1010	149	8	as	as	ADP
cuesj-1010	149	9	a	a	DET
cuesj-1010	149	10	weighted	weighted	ADJ
cuesj-1010	149	11	average	average	NOUN
cuesj-1010	149	12	,	,	PUNCT
cuesj-1010	149	13	where	where	SCONJ
cuesj-1010	149	14	the	the	DET
cuesj-1010	149	15	weight	weight	NOUN
cuesj-1010	149	16	represents	represent	VERB
cuesj-1010	149	17	the	the	DET
cuesj-1010	149	18	increase	increase	NOUN
cuesj-1010	149	19	in	in	ADP
cuesj-1010	149	20	recall	recall	NOUN
cuesj-1010	149	21	from	from	ADP
cuesj-1010	149	22	the	the	DET
cuesj-1010	149	23	previous	previous	ADJ
cuesj-1010	149	24	threshold.[19	threshold.[19	PROPN
cuesj-1010	149	25	]	]	X
cuesj-1010	149	26	the	the	DET
cuesj-1010	149	27	ap	ap	PROPN
cuesj-1010	149	28	will	will	AUX
cuesj-1010	149	29	be	be	AUX
cuesj-1010	149	30	calculated	calculate	VERB
cuesj-1010	149	31	like	like	ADP
cuesj-1010	149	32	(	(	PUNCT
cuesj-1010	149	33	6	6	NUM
cuesj-1010	149	34	)	)	PUNCT
cuesj-1010	149	35	:	:	PUNCT
cuesj-1010	149	36	(	(	PUNCT
cuesj-1010	149	37	)	)	PUNCT
cuesj-1010	149	38	=	=	SYM
cuesj-1010	150	1	−	−	PROPN
cuesj-1010	151	1	+	+	NOUN
cuesj-1010	151	2	∑	∑	PROPN
cuesj-1010	151	3	(	(	PUNCT
cuesj-1010	151	4	1	1	NUM
cuesj-1010	151	5	)	)	PUNCT
cuesj-1010	151	6	*	*	PUNCT
cuesj-1010	151	7	ap	ap	PROPN
cuesj-1010	151	8	rn	rn	PROPN
cuesj-1010	151	9	r	r	PROPN
cuesj-1010	151	10	n	n	PROPN
cuesj-1010	151	11	pn	pn	NOUN
cuesj-1010	151	12	(	(	PUNCT
cuesj-1010	151	13	6	6	NUM
cuesj-1010	151	14	)	)	PUNCT
cuesj-1010	151	15	now	now	ADV
cuesj-1010	151	16	that	that	SCONJ
cuesj-1010	151	17	we	we	PRON
cuesj-1010	151	18	have	have	VERB
cuesj-1010	151	19	a	a	DET
cuesj-1010	151	20	comprehensive	comprehensive	ADJ
cuesj-1010	151	21	understanding	understanding	NOUN
cuesj-1010	151	22	of	of	ADP
cuesj-1010	151	23	how	how	SCONJ
cuesj-1010	151	24	the	the	DET
cuesj-1010	151	25	model	model	NOUN
cuesj-1010	151	26	calculates	calculate	VERB
cuesj-1010	151	27	its	its	PRON
cuesj-1010	151	28	metrics	metric	NOUN
cuesj-1010	151	29	,	,	PUNCT
cuesj-1010	151	30	let	let	VERB
cuesj-1010	151	31	’s	’s	PRON
cuesj-1010	151	32	delve	delve	VERB
cuesj-1010	151	33	into	into	ADP
cuesj-1010	151	34	the	the	DET
cuesj-1010	151	35	results	result	NOUN
cuesj-1010	151	36	for	for	ADP
cuesj-1010	151	37	the	the	DET
cuesj-1010	151	38	various	various	ADJ
cuesj-1010	151	39	yolov5	yolov5	NOUN
cuesj-1010	151	40	types	type	NOUN
cuesj-1010	151	41	.	.	PUNCT
cuesj-1010	152	1	table	table	NOUN
cuesj-1010	152	2	4	4	NUM
cuesj-1010	152	3	shows	show	VERB
cuesj-1010	152	4	the	the	DET
cuesj-1010	152	5	result	result	NOUN
cuesj-1010	152	6	of	of	ADP
cuesj-1010	152	7	yolov5s	yolov5s	PROPN
cuesj-1010	152	8	or	or	CCONJ
cuesj-1010	152	9	the	the	DET
cuesj-1010	152	10	small	small	ADJ
cuesj-1010	152	11	type	type	NOUN
cuesj-1010	152	12	.	.	PUNCT
cuesj-1010	153	1	v5s	v5s	NOUN
cuesj-1010	153	2	get	get	VERB
cuesj-1010	153	3	a	a	DET
cuesj-1010	153	4	good	good	ADJ
cuesj-1010	153	5	result	result	NOUN
cuesj-1010	153	6	;	;	PUNCT
cuesj-1010	153	7	100	100	NUM
cuesj-1010	153	8	epochs	epoch	NOUN
cuesj-1010	153	9	are	be	AUX
cuesj-1010	153	10	used	use	VERB
cuesj-1010	153	11	here	here	ADV
cuesj-1010	153	12	with	with	ADP
cuesj-1010	153	13	416	416	NUM
cuesj-1010	153	14	*	*	SYM
cuesj-1010	153	15	416	416	NUM
cuesj-1010	153	16	image	image	NOUN
cuesj-1010	153	17	sizes	size	NOUN
cuesj-1010	153	18	.	.	PUNCT
cuesj-1010	154	1	note	note	NOUN
cuesj-1010	154	2	:	:	PUNCT
cuesj-1010	154	3	the	the	DET
cuesj-1010	154	4	epochs	epoch	NOUN
cuesj-1010	154	5	and	and	CCONJ
cuesj-1010	154	6	the	the	DET
cuesj-1010	154	7	image	image	NOUN
cuesj-1010	154	8	size	size	NOUN
cuesj-1010	154	9	are	be	AUX
cuesj-1010	154	10	standard	standard	ADJ
cuesj-1010	154	11	but	but	CCONJ
cuesj-1010	154	12	they	they	PRON
cuesj-1010	154	13	can	can	AUX
cuesj-1010	154	14	be	be	AUX
cuesj-1010	154	15	changed	change	VERB
cuesj-1010	154	16	according	accord	VERB
cuesj-1010	154	17	to	to	ADP
cuesj-1010	154	18	the	the	DET
cuesj-1010	154	19	model	model	NOUN
cuesj-1010	154	20	.	.	PUNCT
cuesj-1010	155	1	table	table	NOUN
cuesj-1010	155	2	5	5	NUM
cuesj-1010	155	3	indicates	indicate	VERB
cuesj-1010	155	4	the	the	DET
cuesj-1010	155	5	final	final	ADJ
cuesj-1010	155	6	training	training	NOUN
cuesj-1010	155	7	of	of	ADP
cuesj-1010	155	8	v5	v5	PROPN
cuesj-1010	155	9	m	m	PROPN
cuesj-1010	155	10	,	,	PUNCT
cuesj-1010	155	11	which	which	PRON
cuesj-1010	155	12	scores	score	VERB
cuesj-1010	155	13	the	the	DET
cuesj-1010	155	14	result	result	NOUN
cuesj-1010	155	15	that	that	SCONJ
cuesj-1010	155	16	we	we	PRON
cuesj-1010	155	17	expected	expect	VERB
cuesj-1010	155	18	from	from	ADP
cuesj-1010	155	19	the	the	DET
cuesj-1010	155	20	training	training	NOUN
cuesj-1010	155	21	.	.	PUNCT
cuesj-1010	156	1	the	the	DET
cuesj-1010	156	2	model	model	NOUN
cuesj-1010	156	3	needs	need	VERB
cuesj-1010	156	4	around	around	ADV
cuesj-1010	156	5	1	1	NUM
cuesj-1010	156	6	h	h	NOUN
cuesj-1010	156	7	and	and	CCONJ
cuesj-1010	156	8	40	40	NUM
cuesj-1010	156	9	min	min	NOUN
cuesj-1010	156	10	to	to	PART
cuesj-1010	156	11	finalize	finalize	VERB
cuesj-1010	156	12	the	the	DET
cuesj-1010	156	13	training	training	NOUN
cuesj-1010	156	14	.	.	PUNCT
cuesj-1010	157	1	as	as	ADP
cuesj-1010	157	2	a	a	DET
cuesj-1010	157	3	model	model	NOUN
cuesj-1010	157	4	v5s	v5s	PROPN
cuesj-1010	157	5	,	,	PUNCT
cuesj-1010	157	6	here	here	ADV
cuesj-1010	157	7	we	we	PRON
cuesj-1010	157	8	use	use	VERB
cuesj-1010	157	9	a	a	DET
cuesj-1010	157	10	416	416	NUM
cuesj-1010	157	11	*	*	SYM
cuesj-1010	157	12	416	416	NUM
cuesj-1010	157	13	image	image	NOUN
cuesj-1010	157	14	size	size	NOUN
cuesj-1010	157	15	and	and	CCONJ
cuesj-1010	157	16	100	100	NUM
cuesj-1010	157	17	epochs	epoch	NOUN
cuesj-1010	157	18	.	.	PUNCT
cuesj-1010	157	19	table	table	NOUN
cuesj-1010	157	20	6	6	NUM
cuesj-1010	157	21	represents	represent	VERB
cuesj-1010	157	22	the	the	DET
cuesj-1010	157	23	outcome	outcome	NOUN
cuesj-1010	157	24	of	of	ADP
cuesj-1010	157	25	v5l	v5l	PROPN
cuesj-1010	157	26	or	or	CCONJ
cuesj-1010	157	27	the	the	DET
cuesj-1010	157	28	large	large	ADJ
cuesj-1010	157	29	.	.	PUNCT
cuesj-1010	158	1	here	here	ADV
cuesj-1010	158	2	we	we	PRON
cuesj-1010	158	3	got	get	VERB
cuesj-1010	158	4	an	an	DET
cuesj-1010	158	5	interesting	interesting	ADJ
cuesj-1010	158	6	result	result	NOUN
cuesj-1010	158	7	,	,	PUNCT
cuesj-1010	158	8	but	but	CCONJ
cuesj-1010	158	9	we	we	PRON
cuesj-1010	158	10	still	still	ADV
cuesj-1010	158	11	need	need	VERB
cuesj-1010	158	12	a	a	DET
cuesj-1010	158	13	more	more	ADJ
cuesj-1010	158	14	table	table	NOUN
cuesj-1010	158	15	4	4	NUM
cuesj-1010	158	16	:	:	PUNCT
cuesj-1010	158	17	indicator	indicator	NOUN
cuesj-1010	158	18	of	of	ADP
cuesj-1010	158	19	yolov5s	yolov5s	PROPN
cuesj-1010	158	20	model	model	NOUN
cuesj-1010	158	21	class	class	NOUN
cuesj-1010	158	22	precision	precision	NOUN
cuesj-1010	158	23	recal	recal	ADJ
cuesj-1010	158	24	f1	f1	NOUN
cuesj-1010	158	25	score	score	NOUN
cuesj-1010	158	26	map	map	NOUN
cuesj-1010	158	27	(	(	PUNCT
cuesj-1010	158	28	0.5	0.5	NUM
cuesj-1010	158	29	)	)	PUNCT
cuesj-1010	158	30	vehicle_above	vehicle_above	ADV
cuesj-1010	158	31	0.57	0.57	NUM
cuesj-1010	158	32	0.70	0.70	NUM
cuesj-1010	158	33	0.60	0.60	NUM
cuesj-1010	158	34	0.59	0.59	NUM
cuesj-1010	158	35	yolov5s	yolov5s	NOUN
cuesj-1010	158	36	vehicle_rear	vehicle_rear	ADJ
cuesj-1010	158	37	0.84	0.84	NUM
cuesj-1010	158	38	0.55	0.55	NUM
cuesj-1010	158	39	0.66	0.66	NUM
cuesj-1010	158	40	0.71	0.71	NUM
cuesj-1010	158	41	vehicle_front	vehicle_front	NOUN
cuesj-1010	158	42	0.88	0.88	NUM
cuesj-1010	158	43	0.70	0.70	NUM
cuesj-1010	158	44	0.78	0.78	NUM
cuesj-1010	158	45	0.75	0.75	NUM
cuesj-1010	158	46	vehicle_sides	vehicle_side	NOUN
cuesj-1010	158	47	0.87	0.87	NUM
cuesj-1010	158	48	0.57	0.57	NUM
cuesj-1010	158	49	0.68	0.68	NUM
cuesj-1010	158	50	0.60	0.60	NUM
cuesj-1010	158	51	table	table	NOUN
cuesj-1010	158	52	5	5	NUM
cuesj-1010	158	53	:	:	PUNCT
cuesj-1010	158	54	indicator	indicator	NOUN
cuesj-1010	158	55	of	of	ADP
cuesj-1010	158	56	yolov5	yolov5	PROPN
cuesj-1010	158	57	m	m	PROPN
cuesj-1010	158	58	model	model	NOUN
cuesj-1010	158	59	class	class	NOUN
cuesj-1010	158	60	precision	precision	NOUN
cuesj-1010	158	61	recal	recal	ADJ
cuesj-1010	158	62	f1	f1	NOUN
cuesj-1010	158	63	score	score	NOUN
cuesj-1010	158	64	map	map	NOUN
cuesj-1010	158	65	(	(	PUNCT
cuesj-1010	158	66	0.5	0.5	NUM
cuesj-1010	158	67	)	)	PUNCT
cuesj-1010	158	68	vehicle_above	vehicle_above	ADV
cuesj-1010	158	69	0.91	0.91	NUM
cuesj-1010	158	70	0.70	0.70	NUM
cuesj-1010	158	71	0.85	0.85	NUM
cuesj-1010	158	72	0.88	0.88	NUM
cuesj-1010	158	73	yolov5	yolov5	NOUN
cuesj-1010	158	74	m	m	NOUN
cuesj-1010	158	75	vehicle_rear	vehicle_rear	NOUN
cuesj-1010	158	76	0.66	0.66	NUM
cuesj-1010	158	77	0.55	0.55	NUM
cuesj-1010	158	78	0.76	0.76	NUM
cuesj-1010	158	79	0.87	0.87	NUM
cuesj-1010	158	80	vehicle_front	vehicle_front	NOUN
cuesj-1010	158	81	0.76	0.76	NUM
cuesj-1010	158	82	0.70	0.70	NUM
cuesj-1010	158	83	0.84	0.84	NUM
cuesj-1010	158	84	0.80	0.80	NUM
cuesj-1010	158	85	vehicle_sides	vehicle_side	NOUN
cuesj-1010	158	86	0.72	0.72	NUM
cuesj-1010	158	87	0.57	0.57	NUM
cuesj-1010	158	88	0.70	0.70	NUM
cuesj-1010	158	89	0.70	0.70	NUM
cuesj-1010	158	90	table	table	NOUN
cuesj-1010	158	91	6	6	NUM
cuesj-1010	158	92	:	:	PUNCT
cuesj-1010	158	93	indicator	indicator	NOUN
cuesj-1010	158	94	of	of	ADP
cuesj-1010	158	95	yolov5l	yolov5l	PROPN
cuesj-1010	158	96	model	model	NOUN
cuesj-1010	158	97	class	class	PROPN
cuesj-1010	158	98	precision	precision	NOUN
cuesj-1010	158	99	recal	recal	ADJ
cuesj-1010	158	100	f1	f1	NOUN
cuesj-1010	158	101	score	score	NOUN
cuesj-1010	158	102	map	map	NOUN
cuesj-1010	158	103	(	(	PUNCT
cuesj-1010	158	104	0.5	0.5	NUM
cuesj-1010	158	105	)	)	PUNCT
cuesj-1010	158	106	vehicle_above	vehicle_above	ADV
cuesj-1010	158	107	0.90	0.90	NUM
cuesj-1010	158	108	0.80	0.80	NUM
cuesj-1010	158	109	0.85	0.85	NUM
cuesj-1010	158	110	0.88	0.88	NUM
cuesj-1010	158	111	yolov5l	yolov5l	NOUN
cuesj-1010	158	112	vehicle_rear	vehicle_rear	PROPN
cuesj-1010	158	113	0.94	0.94	NUM
cuesj-1010	158	114	0.87	0.87	NUM
cuesj-1010	158	115	0.90	0.90	NUM
cuesj-1010	158	116	0.91	0.91	NUM
cuesj-1010	158	117	vehicle_front	vehicle_front	NOUN
cuesj-1010	158	118	0.94	0.94	NUM
cuesj-1010	158	119	0.88	0.88	NUM
cuesj-1010	158	120	0.91	0.91	NUM
cuesj-1010	158	121	0.92	0.92	NUM
cuesj-1010	158	122	vehicle_sides	vehicle_side	NOUN
cuesj-1010	158	123	0.93	0.93	NUM
cuesj-1010	158	124	0.88	0.88	NUM
cuesj-1010	158	125	0.90	0.90	NUM
cuesj-1010	158	126	0.93	0.93	NUM
cuesj-1010	158	127	table	table	NOUN
cuesj-1010	158	128	3	3	NUM
cuesj-1010	158	129	:	:	PUNCT
cuesj-1010	158	130	confusion	confusion	NOUN
cuesj-1010	158	131	matrix	matrix	NOUN
cuesj-1010	158	132	elements	element	NOUN
cuesj-1010	158	133	matrix	matrix	NOUN
cuesj-1010	158	134	name	name	NOUN
cuesj-1010	158	135	state	state	NOUN
cuesj-1010	158	136	label	label	NOUN
cuesj-1010	158	137	label	label	NOUN
cuesj-1010	158	138	positive	positive	ADJ
cuesj-1010	158	139	tp	tp	NOUN
cuesj-1010	158	140	(	(	PUNCT
cuesj-1010	158	141	true	true	ADJ
cuesj-1010	158	142	positive	positive	ADJ
cuesj-1010	158	143	)	)	PUNCT
cuesj-1010	158	144	fn	fn	NOUN
cuesj-1010	158	145	(	(	PUNCT
cuesj-1010	158	146	false	false	ADJ
cuesj-1010	158	147	negative	negative	ADJ
cuesj-1010	158	148	)	)	PUNCT
cuesj-1010	158	149	confusion	confusion	NOUN
cuesj-1010	158	150	matrix	matrix	NOUN
cuesj-1010	158	151	negative	negative	ADJ
cuesj-1010	158	152	fp	fp	X
cuesj-1010	158	153	(	(	PUNCT
cuesj-1010	158	154	false	false	ADJ
cuesj-1010	158	155	positive	positive	ADJ
cuesj-1010	158	156	)	)	PUNCT
cuesj-1010	158	157	tn	tn	NOUN
cuesj-1010	158	158	(	(	PUNCT
cuesj-1010	158	159	true	true	ADJ
cuesj-1010	158	160	negative	negative	ADJ
cuesj-1010	158	161	)	)	PUNCT
cuesj-1010	158	162	abdallah	abdallah	PROPN
cuesj-1010	158	163	:	:	PUNCT
cuesj-1010	158	164	vehicle	vehicle	NOUN
cuesj-1010	158	165	detection	detection	NOUN
cuesj-1010	158	166	with	with	ADP
cuesj-1010	158	167	tello	tello	PROPN
cuesj-1010	158	168	drone	drone	PROPN
cuesj-1010	158	169	and	and	CCONJ
cuesj-1010	158	170	yolov5	yolov5	NOUN
cuesj-1010	158	171	6	6	NUM
cuesj-1010	158	172	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	158	173	cuesj	cuesj	ADJ
cuesj-1010	158	174	2024	2024	NUM
cuesj-1010	158	175	,	,	PUNCT
cuesj-1010	158	176	8	8	NUM
cuesj-1010	158	177	(	(	PUNCT
cuesj-1010	158	178	1	1	NUM
cuesj-1010	158	179	):	):	PUNCT
cuesj-1010	158	180	1	1	NUM
cuesj-1010	158	181	-	-	SYM
cuesj-1010	158	182	7	7	NUM
cuesj-1010	158	183	accurate	accurate	ADJ
cuesj-1010	158	184	outcome	outcome	NOUN
cuesj-1010	158	185	.	.	PUNCT
cuesj-1010	159	1	the	the	DET
cuesj-1010	159	2	time	time	NOUN
cuesj-1010	159	3	spent	spend	VERB
cuesj-1010	159	4	on	on	ADP
cuesj-1010	159	5	training	training	NOUN
cuesj-1010	159	6	was	be	AUX
cuesj-1010	159	7	2	2	NUM
cuesj-1010	159	8	h	h	NOUN
cuesj-1010	159	9	and	and	CCONJ
cuesj-1010	159	10	10	10	NUM
cuesj-1010	159	11	min	min	NOUN
cuesj-1010	159	12	.	.	PUNCT
cuesj-1010	160	1	the	the	DET
cuesj-1010	160	2	image	image	NOUN
cuesj-1010	160	3	size	size	NOUN
cuesj-1010	160	4	and	and	CCONJ
cuesj-1010	160	5	epochs	epoch	NOUN
cuesj-1010	160	6	were	be	AUX
cuesj-1010	160	7	the	the	DET
cuesj-1010	160	8	same	same	ADJ
cuesj-1010	160	9	as	as	ADP
cuesj-1010	160	10	in	in	ADP
cuesj-1010	160	11	previous	previous	ADJ
cuesj-1010	160	12	models	model	NOUN
cuesj-1010	160	13	.	.	PUNCT
cuesj-1010	161	1	table	table	NOUN
cuesj-1010	161	2	7	7	NUM
cuesj-1010	161	3	appears	appear	VERB
cuesj-1010	161	4	for	for	ADP
cuesj-1010	161	5	the	the	DET
cuesj-1010	161	6	model	model	NOUN
cuesj-1010	161	7	of	of	ADP
cuesj-1010	161	8	v5x	v5x	NOUN
cuesj-1010	161	9	or	or	CCONJ
cuesj-1010	161	10	the	the	DET
cuesj-1010	161	11	xlarg	xlarg	PROPN
cuesj-1010	161	12	.	.	PUNCT
cuesj-1010	162	1	the	the	DET
cuesj-1010	162	2	train	train	NOUN
cuesj-1010	162	3	achieves	achieve	VERB
cuesj-1010	162	4	what	what	PRON
cuesj-1010	162	5	we	we	PRON
cuesj-1010	162	6	need	need	VERB
cuesj-1010	162	7	here	here	ADV
cuesj-1010	162	8	in	in	ADP
cuesj-1010	162	9	this	this	DET
cuesj-1010	162	10	article	article	NOUN
cuesj-1010	162	11	,	,	PUNCT
cuesj-1010	162	12	even	even	ADV
cuesj-1010	162	13	though	though	SCONJ
cuesj-1010	162	14	it	it	PRON
cuesj-1010	162	15	takes	take	VERB
cuesj-1010	162	16	us	we	PRON
cuesj-1010	162	17	nearly	nearly	ADV
cuesj-1010	162	18	8	8	NUM
cuesj-1010	162	19	h	h	NOUN
cuesj-1010	162	20	to	to	PART
cuesj-1010	162	21	finalize	finalize	VERB
cuesj-1010	162	22	the	the	DET
cuesj-1010	162	23	train	train	NOUN
cuesj-1010	162	24	.	.	PUNCT
cuesj-1010	163	1	we	we	PRON
cuesj-1010	163	2	chose	choose	VERB
cuesj-1010	163	3	v5x	v5x	NOUN
cuesj-1010	163	4	as	as	ADP
cuesj-1010	163	5	the	the	DET
cuesj-1010	163	6	final	final	ADJ
cuesj-1010	163	7	model	model	NOUN
cuesj-1010	163	8	to	to	PART
cuesj-1010	163	9	be	be	AUX
cuesj-1010	163	10	used	use	VERB
cuesj-1010	163	11	with	with	ADP
cuesj-1010	163	12	the	the	DET
cuesj-1010	163	13	drone	drone	NOUN
cuesj-1010	163	14	for	for	ADP
cuesj-1010	163	15	detecting	detect	VERB
cuesj-1010	163	16	vehicles	vehicle	NOUN
cuesj-1010	163	17	.	.	PUNCT
cuesj-1010	164	1	as	as	ADP
cuesj-1010	164	2	a	a	DET
cuesj-1010	164	3	previous	previous	ADJ
cuesj-1010	164	4	type	type	NOUN
cuesj-1010	164	5	,	,	PUNCT
cuesj-1010	164	6	we	we	PRON
cuesj-1010	164	7	used	use	VERB
cuesj-1010	164	8	a	a	DET
cuesj-1010	164	9	416	416	NUM
cuesj-1010	164	10	*	*	SYM
cuesj-1010	164	11	416	416	NUM
cuesj-1010	164	12	image	image	NOUN
cuesj-1010	164	13	size	size	NOUN
cuesj-1010	164	14	and	and	CCONJ
cuesj-1010	164	15	100	100	NUM
cuesj-1010	164	16	epochs	epoch	NOUN
cuesj-1010	164	17	.	.	PUNCT
cuesj-1010	165	1	conclusion	conclusion	NOUN
cuesj-1010	165	2	to	to	PART
cuesj-1010	165	3	summarize	summarize	VERB
cuesj-1010	165	4	what	what	PRON
cuesj-1010	165	5	has	have	AUX
cuesj-1010	165	6	been	be	AUX
cuesj-1010	165	7	discussed	discuss	VERB
cuesj-1010	165	8	in	in	ADP
cuesj-1010	165	9	this	this	DET
cuesj-1010	165	10	article	article	NOUN
cuesj-1010	165	11	,	,	PUNCT
cuesj-1010	165	12	we	we	PRON
cuesj-1010	165	13	present	present	VERB
cuesj-1010	165	14	a	a	DET
cuesj-1010	165	15	system	system	NOUN
cuesj-1010	165	16	that	that	PRON
cuesj-1010	165	17	can	can	AUX
cuesj-1010	165	18	detect	detect	VERB
cuesj-1010	165	19	vehicles	vehicle	NOUN
cuesj-1010	165	20	from	from	ADP
cuesj-1010	165	21	multiple	multiple	ADJ
cuesj-1010	165	22	sides	side	NOUN
cuesj-1010	165	23	.	.	PUNCT
cuesj-1010	166	1	the	the	DET
cuesj-1010	166	2	sides	side	NOUN
cuesj-1010	166	3	are	be	AUX
cuesj-1010	166	4	classified	classify	VERB
cuesj-1010	166	5	in	in	ADP
cuesj-1010	166	6	the	the	DET
cuesj-1010	166	7	dataset	dataset	NOUN
cuesj-1010	166	8	that	that	PRON
cuesj-1010	166	9	was	be	AUX
cuesj-1010	166	10	collected	collect	VERB
cuesj-1010	166	11	from	from	ADP
cuesj-1010	166	12	the	the	DET
cuesj-1010	166	13	roboflow	roboflow	ADJ
cuesj-1010	166	14	universe	universe	NOUN
cuesj-1010	166	15	.	.	PUNCT
cuesj-1010	167	1	first	first	ADV
cuesj-1010	167	2	,	,	PUNCT
cuesj-1010	167	3	we	we	PRON
cuesj-1010	167	4	classified	classify	VERB
cuesj-1010	167	5	images	image	NOUN
cuesj-1010	167	6	into	into	ADP
cuesj-1010	167	7	four	four	NUM
cuesj-1010	167	8	sections	section	NOUN
cuesj-1010	167	9	(	(	PUNCT
cuesj-1010	167	10	vehicle_above	vehicle_above	PROPN
cuesj-1010	167	11	,	,	PUNCT
cuesj-1010	167	12	vehicle_rear	vehicle_rear	ADJ
cuesj-1010	167	13	,	,	PUNCT
cuesj-1010	167	14	vehicle_front	vehicle_front	ADJ
cuesj-1010	167	15	,	,	PUNCT
cuesj-1010	167	16	and	and	CCONJ
cuesj-1010	167	17	vehicle_sides	vehicle_side	NOUN
cuesj-1010	167	18	)	)	PUNCT
cuesj-1010	167	19	,	,	PUNCT
cuesj-1010	167	20	then	then	ADV
cuesj-1010	167	21	annotated	annotate	VERB
cuesj-1010	167	22	all	all	DET
cuesj-1010	167	23	images	image	NOUN
cuesj-1010	167	24	.	.	PUNCT
cuesj-1010	168	1	next	next	ADV
cuesj-1010	168	2	,	,	PUNCT
cuesj-1010	168	3	we	we	PRON
cuesj-1010	168	4	performed	perform	VERB
cuesj-1010	168	5	the	the	DET
cuesj-1010	168	6	pre	pre	ADJ
cuesj-1010	168	7	-	-	ADJ
cuesj-1010	168	8	processing	process	VERB
cuesj-1010	168	9	and	and	CCONJ
cuesj-1010	168	10	saved	save	VERB
cuesj-1010	168	11	our	our	PRON
cuesj-1010	168	12	dataset	dataset	NOUN
cuesj-1010	168	13	in	in	ADP
cuesj-1010	168	14	the	the	DET
cuesj-1010	168	15	format	format	NOUN
cuesj-1010	168	16	of	of	ADP
cuesj-1010	168	17	our	our	PRON
cuesj-1010	168	18	model	model	NOUN
cuesj-1010	168	19	,	,	PUNCT
cuesj-1010	168	20	which	which	PRON
cuesj-1010	168	21	is	be	AUX
cuesj-1010	168	22	yolov5	yolov5	NOUN
cuesj-1010	168	23	through	through	ADP
cuesj-1010	168	24	roboflow	roboflow	NOUN
cuesj-1010	168	25	,	,	PUNCT
cuesj-1010	168	26	which	which	PRON
cuesj-1010	168	27	plays	play	VERB
cuesj-1010	168	28	the	the	DET
cuesj-1010	168	29	main	main	ADJ
cuesj-1010	168	30	role	role	NOUN
cuesj-1010	168	31	in	in	ADP
cuesj-1010	168	32	this	this	DET
cuesj-1010	168	33	study	study	NOUN
cuesj-1010	168	34	.	.	PUNCT
cuesj-1010	169	1	yolov5	yolov5	NOUN
cuesj-1010	169	2	is	be	AUX
cuesj-1010	169	3	one	one	NUM
cuesj-1010	169	4	of	of	ADP
cuesj-1010	169	5	the	the	DET
cuesj-1010	169	6	versions	version	NOUN
cuesj-1010	169	7	of	of	ADP
cuesj-1010	169	8	yolo	yolo	NOUN
cuesj-1010	169	9	that	that	SCONJ
cuesj-1010	169	10	you	you	PRON
cuesj-1010	169	11	only	only	ADV
cuesj-1010	169	12	look	look	VERB
cuesj-1010	169	13	at	at	ADP
cuesj-1010	169	14	.	.	PUNCT
cuesj-1010	170	1	v5	v5	PROPN
cuesj-1010	170	2	has	have	VERB
cuesj-1010	170	3	three	three	NUM
cuesj-1010	170	4	components	component	NOUN
cuesj-1010	170	5	:	:	PUNCT
cuesj-1010	170	6	first	first	ADV
cuesj-1010	170	7	the	the	DET
cuesj-1010	170	8	csp	csp	PROPN
cuesj-1010	170	9	-	-	PUNCT
cuesj-1010	170	10	darknet53	darknet53	NOUN
cuesj-1010	170	11	backbone	backbone	NOUN
cuesj-1010	170	12	,	,	PUNCT
cuesj-1010	170	13	second	second	ADJ
cuesj-1010	170	14	the	the	DET
cuesj-1010	170	15	neck	neck	NOUN
cuesj-1010	170	16	panet	panet	PROPN
cuesj-1010	170	17	model	model	NOUN
cuesj-1010	170	18	,	,	PUNCT
cuesj-1010	170	19	and	and	CCONJ
cuesj-1010	170	20	third	third	DET
cuesj-1010	170	21	the	the	DET
cuesj-1010	170	22	head	head	NOUN
cuesj-1010	170	23	.	.	PUNCT
cuesj-1010	171	1	each	each	DET
cuesj-1010	171	2	component	component	NOUN
cuesj-1010	171	3	has	have	VERB
cuesj-1010	171	4	a	a	DET
cuesj-1010	171	5	specific	specific	ADJ
cuesj-1010	171	6	process	process	NOUN
cuesj-1010	171	7	.	.	PUNCT
cuesj-1010	172	1	v5	v5	PROPN
cuesj-1010	172	2	has	have	VERB
cuesj-1010	172	3	four	four	NUM
cuesj-1010	172	4	main	main	ADJ
cuesj-1010	172	5	types	type	NOUN
cuesj-1010	172	6	(	(	PUNCT
cuesj-1010	172	7	v5s	v5s	NOUN
cuesj-1010	172	8	,	,	PUNCT
cuesj-1010	172	9	v5	v5	PROPN
cuesj-1010	172	10	m	m	PROPN
cuesj-1010	172	11	,	,	PUNCT
cuesj-1010	172	12	v5l	v5l	PROPN
cuesj-1010	172	13	,	,	PUNCT
cuesj-1010	172	14	and	and	CCONJ
cuesj-1010	172	15	v5x	v5x	NOUN
cuesj-1010	172	16	)	)	PUNCT
cuesj-1010	172	17	,	,	PUNCT
cuesj-1010	172	18	each	each	PRON
cuesj-1010	172	19	of	of	ADP
cuesj-1010	172	20	which	which	PRON
cuesj-1010	172	21	has	have	VERB
cuesj-1010	172	22	different	different	ADJ
cuesj-1010	172	23	parameters	parameter	NOUN
cuesj-1010	172	24	.	.	PUNCT
cuesj-1010	173	1	in	in	ADP
cuesj-1010	173	2	this	this	DET
cuesj-1010	173	3	article	article	NOUN
cuesj-1010	173	4	,	,	PUNCT
cuesj-1010	173	5	we	we	PRON
cuesj-1010	173	6	train	train	VERB
cuesj-1010	173	7	all	all	PRON
cuesj-1010	173	8	of	of	ADP
cuesj-1010	173	9	them	they	PRON
cuesj-1010	173	10	and	and	CCONJ
cuesj-1010	173	11	show	show	VERB
cuesj-1010	173	12	the	the	DET
cuesj-1010	173	13	results	result	NOUN
cuesj-1010	173	14	of	of	ADP
cuesj-1010	173	15	the	the	DET
cuesj-1010	173	16	training	training	NOUN
cuesj-1010	173	17	.	.	PUNCT
cuesj-1010	174	1	the	the	DET
cuesj-1010	174	2	eye	eye	NOUN
cuesj-1010	174	3	camera	camera	NOUN
cuesj-1010	174	4	was	be	AUX
cuesj-1010	174	5	a	a	DET
cuesj-1010	174	6	dji	dji	PROPN
cuesj-1010	174	7	tello	tello	NOUN
cuesj-1010	174	8	drone	drone	NOUN
cuesj-1010	174	9	that	that	PRON
cuesj-1010	174	10	can	can	AUX
cuesj-1010	174	11	be	be	AUX
cuesj-1010	174	12	connected	connect	VERB
cuesj-1010	174	13	with	with	ADP
cuesj-1010	174	14	yolo	yolo	ADJ
cuesj-1010	174	15	models	model	NOUN
cuesj-1010	174	16	using	use	VERB
cuesj-1010	174	17	the	the	DET
cuesj-1010	174	18	pytorch	pytorch	NOUN
cuesj-1010	174	19	framework	framework	NOUN
cuesj-1010	174	20	.	.	PUNCT
cuesj-1010	175	1	tello	tello	PROPN
cuesj-1010	175	2	can	can	AUX
cuesj-1010	175	3	detect	detect	VERB
cuesj-1010	175	4	and	and	CCONJ
cuesj-1010	175	5	classify	classify	VERB
cuesj-1010	175	6	the	the	DET
cuesj-1010	175	7	vehicles	vehicle	NOUN
cuesj-1010	175	8	correctly	correctly	ADV
cuesj-1010	175	9	based	base	VERB
cuesj-1010	175	10	on	on	ADP
cuesj-1010	175	11	the	the	DET
cuesj-1010	175	12	yolo	yolo	ADJ
cuesj-1010	175	13	model	model	NOUN
cuesj-1010	175	14	.	.	PUNCT
cuesj-1010	176	1	in	in	ADP
cuesj-1010	176	2	the	the	DET
cuesj-1010	176	3	future	future	NOUN
cuesj-1010	176	4	,	,	PUNCT
cuesj-1010	176	5	the	the	DET
cuesj-1010	176	6	tello	tello	PROPN
cuesj-1010	176	7	drone	drone	NOUN
cuesj-1010	176	8	will	will	AUX
cuesj-1010	176	9	be	be	AUX
cuesj-1010	176	10	integrated	integrate	VERB
cuesj-1010	176	11	with	with	ADP
cuesj-1010	176	12	a	a	DET
cuesj-1010	176	13	tensorflow	tensorflow	NOUN
cuesj-1010	176	14	model	model	NOUN
cuesj-1010	176	15	and	and	CCONJ
cuesj-1010	176	16	a	a	DET
cuesj-1010	176	17	comparative	comparative	ADJ
cuesj-1010	176	18	analysis	analysis	NOUN
cuesj-1010	176	19	will	will	AUX
cuesj-1010	176	20	be	be	AUX
cuesj-1010	176	21	conducted	conduct	VERB
cuesj-1010	176	22	to	to	PART
cuesj-1010	176	23	evaluate	evaluate	VERB
cuesj-1010	176	24	its	its	PRON
cuesj-1010	176	25	performance	performance	NOUN
cuesj-1010	176	26	against	against	ADP
cuesj-1010	176	27	yolov5	yolov5	NOUN
cuesj-1010	176	28	.	.	PUNCT
cuesj-1010	177	1	the	the	DET
cuesj-1010	177	2	objective	objective	NOUN
cuesj-1010	177	3	is	be	AUX
cuesj-1010	177	4	to	to	PART
cuesj-1010	177	5	determine	determine	VERB
cuesj-1010	177	6	which	which	DET
cuesj-1010	177	7	model	model	NOUN
cuesj-1010	177	8	is	be	AUX
cuesj-1010	177	9	better	well	ADV
cuesj-1010	177	10	suited	suited	ADJ
cuesj-1010	177	11	for	for	ADP
cuesj-1010	177	12	drone	drone	NOUN
cuesj-1010	177	13	applications	application	NOUN
cuesj-1010	177	14	.	.	PUNCT
cuesj-1010	178	1	in	in	ADP
cuesj-1010	178	2	addition	addition	NOUN
cuesj-1010	178	3	,	,	PUNCT
cuesj-1010	178	4	there	there	PRON
cuesj-1010	178	5	are	be	VERB
cuesj-1010	178	6	plans	plan	NOUN
cuesj-1010	178	7	to	to	PART
cuesj-1010	178	8	extend	extend	VERB
cuesj-1010	178	9	the	the	DET
cuesj-1010	178	10	testing	testing	NOUN
cuesj-1010	178	11	of	of	ADP
cuesj-1010	178	12	these	these	DET
cuesj-1010	178	13	models	model	NOUN
cuesj-1010	178	14	to	to	ADP
cuesj-1010	178	15	other	other	ADJ
cuesj-1010	178	16	drones	drone	NOUN
cuesj-1010	178	17	for	for	ADP
cuesj-1010	178	18	a	a	DET
cuesj-1010	178	19	more	more	ADV
cuesj-1010	178	20	comprehensive	comprehensive	ADJ
cuesj-1010	178	21	assessment	assessment	NOUN
cuesj-1010	178	22	.	.	PUNCT
cuesj-1010	179	1	references	reference	NOUN
cuesj-1010	179	2	1	1	NUM
cuesj-1010	179	3	.	.	PUNCT
cuesj-1010	179	4	j.	j.	PROPN
cuesj-1010	179	5	lu	lu	PROPN
cuesj-1010	179	6	,	,	PUNCT
cuesj-1010	179	7	c.	c.	PROPN
cuesj-1010	179	8	ma	ma	PROPN
cuesj-1010	179	9	,	,	PUNCT
cuesj-1010	179	10	l.	l.	PROPN
cuesj-1010	179	11	li	li	PROPN
cuesj-1010	179	12	,	,	PUNCT
cuesj-1010	179	13	x.	x.	PROPN
cuesj-1010	179	14	xing	xing	PROPN
cuesj-1010	179	15	,	,	PUNCT
cuesj-1010	179	16	y.	y.	PROPN
cuesj-1010	179	17	zhang	zhang	PROPN
cuesj-1010	179	18	,	,	PUNCT
cuesj-1010	179	19	z.	z.	PROPN
cuesj-1010	179	20	wang	wang	PROPN
cuesj-1010	179	21	and	and	CCONJ
cuesj-1010	179	22	j.	j.	PROPN
cuesj-1010	179	23	xu	xu	PROPN
cuesj-1010	179	24	.	.	PUNCT
cuesj-1010	180	1	a	a	DET
cuesj-1010	180	2	vehicle	vehicle	NOUN
cuesj-1010	180	3	detection	detection	NOUN
cuesj-1010	180	4	method	method	NOUN
cuesj-1010	180	5	for	for	ADP
cuesj-1010	180	6	aerial	aerial	ADJ
cuesj-1010	180	7	image	image	NOUN
cuesj-1010	180	8	based	base	VERB
cuesj-1010	180	9	on	on	ADP
cuesj-1010	180	10	yolo	yolo	PROPN
cuesj-1010	180	11	.	.	PUNCT
cuesj-1010	181	1	journal	journal	PROPN
cuesj-1010	181	2	of	of	ADP
cuesj-1010	181	3	computer	computer	NOUN
cuesj-1010	181	4	and	and	CCONJ
cuesj-1010	181	5	communications	communication	NOUN
cuesj-1010	181	6	,	,	PUNCT
cuesj-1010	181	7	vol	vol	NOUN
cuesj-1010	181	8	.	.	PROPN
cuesj-1010	181	9	6	6	NUM
cuesj-1010	181	10	,	,	PUNCT
cuesj-1010	181	11	pp	pp	ADJ
cuesj-1010	181	12	.	.	PUNCT
cuesj-1010	182	1	98	98	NUM
cuesj-1010	182	2	-	-	SYM
cuesj-1010	182	3	107	107	NUM
cuesj-1010	182	4	,	,	PUNCT
cuesj-1010	182	5	2018	2018	NUM
cuesj-1010	182	6	.	.	PUNCT
cuesj-1010	183	1	2	2	X
cuesj-1010	183	2	.	.	PUNCT
cuesj-1010	183	3	j.	j.	PROPN
cuesj-1010	183	4	nelson	nelson	PROPN
cuesj-1010	183	5	and	and	CCONJ
cuesj-1010	183	6	j.	j.	PROPN
cuesj-1010	183	7	solawetz	solawetz	PROPN
cuesj-1010	183	8	.	.	PUNCT
cuesj-1010	184	1	responding	respond	VERB
cuesj-1010	184	2	to	to	ADP
cuesj-1010	184	3	the	the	DET
cuesj-1010	184	4	controversy	controversy	NOUN
cuesj-1010	184	5	about	about	ADP
cuesj-1010	184	6	yolov5	yolov5	PROPN
cuesj-1010	184	7	.	.	PUNCT
cuesj-1010	185	1	roboflow	roboflow	PROPN
cuesj-1010	185	2	blog	blog	PROPN
cuesj-1010	185	3	,	,	PUNCT
cuesj-1010	185	4	2020	2020	NUM
cuesj-1010	185	5	.	.	PUNCT
cuesj-1010	186	1	available	available	ADJ
cuesj-1010	186	2	from	from	ADP
cuesj-1010	186	3	:	:	PUNCT
cuesj-1010	186	4	https://blog	https://blog	NUM
cuesj-1010	186	5	.	.	PUNCT
cuesj-1010	187	1	roboflow.com/yolov4-versus-yolov5	roboflow.com/yolov4-versus-yolov5	NOUN
cuesj-1010	188	1	[	[	X
cuesj-1010	188	2	last	last	ADV
cuesj-1010	188	3	accessed	access	VERB
cuesj-1010	188	4	on	on	ADP
cuesj-1010	188	5	2023	2023	NUM
cuesj-1010	188	6	jan	jan	NOUN
cuesj-1010	188	7	10	10	NUM
cuesj-1010	188	8	]	]	PUNCT
cuesj-1010	188	9	.	.	PUNCT
cuesj-1010	189	1	3	3	X
cuesj-1010	189	2	.	.	X
cuesj-1010	189	3	w.	w.	PROPN
cuesj-1010	189	4	liu	liu	PROPN
cuesj-1010	189	5	,	,	PUNCT
cuesj-1010	189	6	d.	d.	PROPN
cuesj-1010	189	7	anguelov	anguelov	PROPN
cuesj-1010	189	8	,	,	PUNCT
cuesj-1010	189	9	d.	d.	PROPN
cuesj-1010	189	10	erhan	erhan	PROPN
cuesj-1010	189	11	,	,	PUNCT
cuesj-1010	189	12	c.	c.	PROPN
cuesj-1010	189	13	szegedy	szegedy	PROPN
cuesj-1010	189	14	,	,	PUNCT
cuesj-1010	189	15	s.	s.	PROPN
cuesj-1010	189	16	reed	reed	PROPN
cuesj-1010	189	17	,	,	PUNCT
cuesj-1010	189	18	c.	c.	PROPN
cuesj-1010	189	19	y.	y.	PROPN
cuesj-1010	189	20	fu	fu	PROPN
cuesj-1010	189	21	and	and	CCONJ
cuesj-1010	189	22	a.	a.	PROPN
cuesj-1010	189	23	c.	c.	PROPN
cuesj-1010	189	24	berg	berg	PROPN
cuesj-1010	189	25	.	.	PUNCT
cuesj-1010	190	1	ssd	ssd	PROPN
cuesj-1010	190	2	:	:	PUNCT
cuesj-1010	190	3	single	single	ADJ
cuesj-1010	190	4	shot	shot	NOUN
cuesj-1010	190	5	multibox	multibox	NOUN
cuesj-1010	190	6	detector	detector	NOUN
cuesj-1010	190	7	.	.	PUNCT
cuesj-1010	191	1	in	in	ADP
cuesj-1010	191	2	:	:	PUNCT
cuesj-1010	191	3	computer	computer	NOUN
cuesj-1010	191	4	vision	vision	NOUN
cuesj-1010	191	5	-	-	PUNCT
cuesj-1010	191	6	eccv	eccv	ADV
cuesj-1010	191	7	2016	2016	NUM
cuesj-1010	191	8	:	:	PUNCT
cuesj-1010	191	9	14th	14th	ADJ
cuesj-1010	191	10	european	european	ADJ
cuesj-1010	191	11	conference	conference	PROPN
cuesj-1010	191	12	,	,	PUNCT
cuesj-1010	191	13	amsterdam	amsterdam	PROPN
cuesj-1010	191	14	,	,	PUNCT
cuesj-1010	191	15	the	the	DET
cuesj-1010	191	16	netherlands	netherlands	PROPN
cuesj-1010	191	17	,	,	PUNCT
cuesj-1010	191	18	october	october	PROPN
cuesj-1010	191	19	11	11	NUM
cuesj-1010	191	20	-	-	SYM
cuesj-1010	191	21	14	14	NUM
cuesj-1010	191	22	,	,	PUNCT
cuesj-1010	191	23	2016	2016	NUM
cuesj-1010	191	24	,	,	PUNCT
cuesj-1010	191	25	proceedings	proceeding	NOUN
cuesj-1010	191	26	,	,	PUNCT
cuesj-1010	191	27	part	part	NOUN
cuesj-1010	191	28	i	i	PRON
cuesj-1010	191	29	14	14	NUM
cuesj-1010	191	30	.	.	PUNCT
cuesj-1010	192	1	springer	springer	NOUN
cuesj-1010	192	2	international	international	ADJ
cuesj-1010	192	3	publishing	publishing	NOUN
cuesj-1010	192	4	,	,	PUNCT
cuesj-1010	192	5	berlin	berlin	PROPN
cuesj-1010	192	6	,	,	PUNCT
cuesj-1010	192	7	pp	pp	PROPN
cuesj-1010	192	8	.	.	PUNCT
cuesj-1010	193	1	21	21	NUM
cuesj-1010	193	2	-	-	SYM
cuesj-1010	193	3	37	37	NUM
cuesj-1010	193	4	,	,	PUNCT
cuesj-1010	193	5	2016	2016	NUM
cuesj-1010	193	6	.	.	PUNCT
cuesj-1010	194	1	available	available	ADJ
cuesj-1010	194	2	from	from	ADP
cuesj-1010	194	3	:	:	PUNCT
cuesj-1010	194	4	https://arxiv.org/pdf/1512.02325.pdf%22source%22	https://arxiv.org/pdf/1512.02325.pdf%22source%22	PROPN
cuesj-1010	194	5	[	[	X
cuesj-1010	194	6	last	last	ADV
cuesj-1010	194	7	accessed	access	VERB
cuesj-1010	194	8	on	on	ADP
cuesj-1010	194	9	2023	2023	NUM
cuesj-1010	194	10	jan	jan	NOUN
cuesj-1010	194	11	02	02	NUM
cuesj-1010	194	12	]	]	X
cuesj-1010	194	13	.	.	PUNCT
cuesj-1010	195	1	4	4	X
cuesj-1010	195	2	.	.	PUNCT
cuesj-1010	195	3	j.	j.	PROPN
cuesj-1010	195	4	redmon	redmon	PROPN
cuesj-1010	195	5	,	,	PUNCT
cuesj-1010	195	6	s.	s.	PROPN
cuesj-1010	195	7	divvala	divvala	PROPN
cuesj-1010	195	8	,	,	PUNCT
cuesj-1010	195	9	r.	r.	PROPN
cuesj-1010	195	10	girshick	girshick	PROPN
cuesj-1010	195	11	and	and	CCONJ
cuesj-1010	195	12	a.	a.	NOUN
cuesj-1010	195	13	farhadi	farhadi	NOUN
cuesj-1010	195	14	.	.	PUNCT
cuesj-1010	196	1	you	you	PRON
cuesj-1010	196	2	only	only	ADV
cuesj-1010	196	3	look	look	VERB
cuesj-1010	196	4	once	once	ADV
cuesj-1010	196	5	:	:	PUNCT
cuesj-1010	196	6	unified	unified	ADJ
cuesj-1010	196	7	,	,	PUNCT
cuesj-1010	196	8	real	real	ADJ
cuesj-1010	196	9	-	-	PUNCT
cuesj-1010	196	10	time	time	NOUN
cuesj-1010	196	11	object	object	NOUN
cuesj-1010	196	12	detection	detection	NOUN
cuesj-1010	196	13	.	.	PUNCT
cuesj-1010	197	1	in	in	ADP
cuesj-1010	197	2	:	:	PUNCT
cuesj-1010	197	3	proceedings	proceeding	NOUN
cuesj-1010	197	4	of	of	ADP
cuesj-1010	197	5	the	the	DET
cuesj-1010	197	6	ieee	ieee	NOUN
cuesj-1010	197	7	conference	conference	NOUN
cuesj-1010	197	8	on	on	ADP
cuesj-1010	197	9	computer	computer	NOUN
cuesj-1010	197	10	vision	vision	NOUN
cuesj-1010	197	11	and	and	CCONJ
cuesj-1010	197	12	pattern	pattern	NOUN
cuesj-1010	197	13	recognition	recognition	NOUN
cuesj-1010	197	14	,	,	PUNCT
cuesj-1010	197	15	pp	pp	PROPN
cuesj-1010	197	16	.	.	PUNCT
cuesj-1010	198	1	779	779	NUM
cuesj-1010	198	2	-	-	SYM
cuesj-1010	198	3	788	788	NUM
cuesj-1010	198	4	,	,	PUNCT
cuesj-1010	198	5	2016	2016	NUM
cuesj-1010	198	6	.	.	PUNCT
cuesj-1010	199	1	available	available	ADJ
cuesj-1010	199	2	from	from	ADP
cuesj-1010	199	3	:	:	PUNCT
cuesj-1010	199	4	https://arxiv.org/	https://arxiv.org/	PRON
cuesj-1010	199	5	abs/1506.02640	abs/1506.02640	NOUN
cuesj-1010	199	6	[	[	X
cuesj-1010	199	7	last	last	ADV
cuesj-1010	199	8	accessed	access	VERB
cuesj-1010	199	9	on	on	ADP
cuesj-1010	199	10	2023	2023	NUM
cuesj-1010	199	11	jan	jan	NOUN
cuesj-1010	199	12	07	07	NUM
cuesj-1010	199	13	]	]	PUNCT
cuesj-1010	199	14	.	.	PUNCT
cuesj-1010	200	1	5	5	X
cuesj-1010	200	2	.	.	PUNCT
cuesj-1010	200	3	j.	j.	PROPN
cuesj-1010	200	4	yu	yu	PROPN
cuesj-1010	200	5	,	,	PUNCT
cuesj-1010	200	6	j.	j.	PROPN
cuesj-1010	200	7	xu	xu	PROPN
cuesj-1010	200	8	,	,	PUNCT
cuesj-1010	200	9	y.	y.	PROPN
cuesj-1010	200	10	chen	chen	PROPN
cuesj-1010	200	11	,	,	PUNCT
cuesj-1010	200	12	w.	w.	PROPN
cuesj-1010	200	13	li	li	PROPN
cuesj-1010	200	14	,	,	PUNCT
cuesj-1010	200	15	q.	q.	PROPN
cuesj-1010	200	16	wang	wang	PROPN
cuesj-1010	200	17	,	,	PUNCT
cuesj-1010	200	18	b.	b.	PROPN
cuesj-1010	200	19	i.	i.	PROPN
cuesj-1010	200	20	yoo	yoo	PROPN
cuesj-1010	200	21	and	and	CCONJ
cuesj-1010	200	22	j.	j.	PROPN
cuesj-1010	200	23	j.	j.	PROPN
cuesj-1010	200	24	han	han	PROPN
cuesj-1010	200	25	.	.	PUNCT
cuesj-1010	201	1	learning	learn	VERB
cuesj-1010	201	2	generalized	generalized	ADJ
cuesj-1010	201	3	intersection	intersection	NOUN
cuesj-1010	201	4	over	over	ADP
cuesj-1010	201	5	union	union	NOUN
cuesj-1010	201	6	for	for	ADP
cuesj-1010	201	7	dense	dense	ADJ
cuesj-1010	201	8	pixelwise	pixelwise	ADJ
cuesj-1010	201	9	prediction	prediction	NOUN
cuesj-1010	201	10	.	.	PUNCT
cuesj-1010	202	1	in	in	ADP
cuesj-1010	202	2	:	:	PUNCT
cuesj-1010	202	3	international	international	ADJ
cuesj-1010	202	4	conference	conference	NOUN
cuesj-1010	202	5	on	on	ADP
cuesj-1010	202	6	machine	machine	NOUN
cuesj-1010	202	7	learning	learning	NOUN
cuesj-1010	202	8	,	,	PUNCT
cuesj-1010	202	9	pp	pp	ADP
cuesj-1010	202	10	.	.	PUNCT
cuesj-1010	202	11	12198	12198	NUM
cuesj-1010	202	12	-	-	SYM
cuesj-1010	202	13	12207	12207	NUM
cuesj-1010	202	14	,	,	PUNCT
cuesj-1010	202	15	2021	2021	NUM
cuesj-1010	202	16	.	.	PUNCT
cuesj-1010	202	17	available	available	ADJ
cuesj-1010	202	18	from	from	ADP
cuesj-1010	202	19	:	:	PUNCT
cuesj-1010	202	20	https://	https://	PROPN
cuesj-1010	202	21	shorturl.at/jbgin	shorturl.at/jbgin	VERB
cuesj-1010	203	1	[	[	PUNCT
cuesj-1010	203	2	last	last	ADV
cuesj-1010	203	3	accessed	access	VERB
cuesj-1010	203	4	on	on	ADP
cuesj-1010	203	5	2023	2023	NUM
cuesj-1010	203	6	jan	jan	NOUN
cuesj-1010	203	7	14	14	NUM
cuesj-1010	203	8	]	]	PUNCT
cuesj-1010	203	9	.	.	PUNCT
cuesj-1010	204	1	6	6	X
cuesj-1010	204	2	.	.	PUNCT
cuesj-1010	204	3	h.	h.	PROPN
cuesj-1010	204	4	rezatofighi	rezatofighi	PROPN
cuesj-1010	204	5	,	,	PUNCT
cuesj-1010	204	6	n.	n.	PROPN
cuesj-1010	204	7	tsoi	tsoi	PROPN
cuesj-1010	204	8	,	,	PUNCT
cuesj-1010	204	9	j.	j.	PROPN
cuesj-1010	204	10	gwak	gwak	PROPN
cuesj-1010	204	11	,	,	PUNCT
cuesj-1010	204	12	a.	a.	NOUN
cuesj-1010	204	13	sadeghian	sadeghian	PROPN
cuesj-1010	204	14	,	,	PUNCT
cuesj-1010	204	15	i.	i.	PROPN
cuesj-1010	204	16	reid	reid	PROPN
cuesj-1010	204	17	,	,	PUNCT
cuesj-1010	204	18	and	and	CCONJ
cuesj-1010	204	19	s.	s.	PROPN
cuesj-1010	204	20	savarese	savarese	PROPN
cuesj-1010	204	21	.	.	PUNCT
cuesj-1010	205	1	generalized	generalized	ADJ
cuesj-1010	205	2	intersection	intersection	NOUN
cuesj-1010	205	3	over	over	ADP
cuesj-1010	205	4	union	union	NOUN
cuesj-1010	205	5	:	:	PUNCT
cuesj-1010	205	6	a	a	DET
cuesj-1010	205	7	metric	metric	ADJ
cuesj-1010	205	8	and	and	CCONJ
cuesj-1010	205	9	a	a	DET
cuesj-1010	205	10	loss	loss	NOUN
cuesj-1010	205	11	for	for	ADP
cuesj-1010	205	12	bounding	bound	VERB
cuesj-1010	205	13	box	box	NOUN
cuesj-1010	205	14	regression	regression	NOUN
cuesj-1010	205	15	.	.	PUNCT
cuesj-1010	206	1	in	in	ADP
cuesj-1010	206	2	:	:	PUNCT
cuesj-1010	206	3	proceedings	proceeding	NOUN
cuesj-1010	206	4	of	of	ADP
cuesj-1010	206	5	the	the	DET
cuesj-1010	206	6	ieee	ieee	NOUN
cuesj-1010	206	7	/	/	SYM
cuesj-1010	206	8	cvf	cvf	NOUN
cuesj-1010	206	9	conference	conference	NOUN
cuesj-1010	206	10	on	on	ADP
cuesj-1010	206	11	computer	computer	NOUN
cuesj-1010	206	12	vision	vision	NOUN
cuesj-1010	206	13	and	and	CCONJ
cuesj-1010	206	14	pattern	pattern	NOUN
cuesj-1010	206	15	recognition	recognition	NOUN
cuesj-1010	206	16	,	,	PUNCT
cuesj-1010	206	17	pp	pp	PROPN
cuesj-1010	206	18	.	.	PUNCT
cuesj-1010	207	1	658666	658666	NUM
cuesj-1010	207	2	,	,	PUNCT
cuesj-1010	207	3	2019	2019	NUM
cuesj-1010	207	4	.	.	PUNCT
cuesj-1010	208	1	available	available	ADJ
cuesj-1010	208	2	from	from	ADP
cuesj-1010	208	3	:	:	PUNCT
cuesj-1010	208	4	https://arxiv.org/abs/1902.09630	https://arxiv.org/abs/1902.09630	PROPN
cuesj-1010	209	1	[	[	X
cuesj-1010	209	2	last	last	ADV
cuesj-1010	209	3	accessed	access	VERB
cuesj-1010	209	4	on	on	ADP
cuesj-1010	209	5	2023	2023	NUM
cuesj-1010	209	6	jan	jan	NOUN
cuesj-1010	209	7	14	14	NUM
cuesj-1010	209	8	]	]	PUNCT
cuesj-1010	209	9	.	.	PUNCT
cuesj-1010	210	1	7	7	X
cuesj-1010	210	2	.	.	X
cuesj-1010	210	3	z.	z.	PROPN
cuesj-1010	210	4	huang	huang	PROPN
cuesj-1010	210	5	,	,	PUNCT
cuesj-1010	210	6	h.	h.	PROPN
cuesj-1010	210	7	zhao	zhao	PROPN
cuesj-1010	210	8	,	,	PUNCT
cuesj-1010	210	9	j.	j.	PROPN
cuesj-1010	210	10	zhan	zhan	PROPN
cuesj-1010	210	11	and	and	CCONJ
cuesj-1010	210	12	h.	h.	PROPN
cuesj-1010	210	13	li	li	PROPN
cuesj-1010	210	14	.	.	PUNCT
cuesj-1010	211	1	a	a	DET
cuesj-1010	211	2	multivariate	multivariate	NOUN
cuesj-1010	211	3	intersection	intersection	NOUN
cuesj-1010	211	4	over	over	ADP
cuesj-1010	211	5	union	union	NOUN
cuesj-1010	211	6	of	of	ADP
cuesj-1010	211	7	siamrpn	siamrpn	PROPN
cuesj-1010	211	8	network	network	NOUN
cuesj-1010	211	9	for	for	ADP
cuesj-1010	211	10	visual	visual	ADJ
cuesj-1010	211	11	tracking	tracking	NOUN
cuesj-1010	211	12	.	.	PUNCT
cuesj-1010	212	1	the	the	DET
cuesj-1010	212	2	visual	visual	ADJ
cuesj-1010	212	3	computer	computer	NOUN
cuesj-1010	212	4	,	,	PUNCT
cuesj-1010	212	5	vol	vol	NOUN
cuesj-1010	212	6	.	.	PROPN
cuesj-1010	212	7	38	38	NUM
cuesj-1010	212	8	,	,	PUNCT
cuesj-1010	212	9	pp	pp	ADJ
cuesj-1010	212	10	.	.	PUNCT
cuesj-1010	212	11	2739	2739	NUM
cuesj-1010	212	12	-	-	SYM
cuesj-1010	212	13	2750	2750	NUM
cuesj-1010	212	14	,	,	PUNCT
cuesj-1010	212	15	2022	2022	NUM
cuesj-1010	212	16	.	.	PUNCT
cuesj-1010	212	17	8	8	NUM
cuesj-1010	212	18	.	.	PUNCT
cuesj-1010	213	1	h.	h.	PROPN
cuesj-1010	213	2	k.	k.	PROPN
cuesj-1010	213	3	jung	jung	PROPN
cuesj-1010	213	4	and	and	CCONJ
cuesj-1010	213	5	g.	g.	PROPN
cuesj-1010	213	6	s.	s.	PROPN
cuesj-1010	213	7	choi	choi	PROPN
cuesj-1010	213	8	.	.	PUNCT
cuesj-1010	214	1	improved	improved	ADJ
cuesj-1010	214	2	yolov5	yolov5	NOUN
cuesj-1010	214	3	:	:	PUNCT
cuesj-1010	214	4	efficient	efficient	ADJ
cuesj-1010	214	5	object	object	NOUN
cuesj-1010	214	6	detection	detection	NOUN
cuesj-1010	214	7	using	use	VERB
cuesj-1010	214	8	drone	drone	NOUN
cuesj-1010	214	9	images	image	NOUN
cuesj-1010	214	10	under	under	ADP
cuesj-1010	214	11	various	various	ADJ
cuesj-1010	214	12	conditions	condition	NOUN
cuesj-1010	214	13	.	.	PUNCT
cuesj-1010	215	1	applied	apply	VERB
cuesj-1010	215	2	sciences	science	NOUN
cuesj-1010	215	3	,	,	PUNCT
cuesj-1010	215	4	vol	vol	NOUN
cuesj-1010	215	5	.	.	PROPN
cuesj-1010	215	6	12	12	NUM
cuesj-1010	215	7	,	,	PUNCT
cuesj-1010	215	8	no	no	INTJ
cuesj-1010	215	9	.	.	NOUN
cuesj-1010	215	10	14	14	NUM
cuesj-1010	215	11	,	,	PUNCT
cuesj-1010	215	12	p.	p.	NOUN
cuesj-1010	215	13	7255	7255	NUM
cuesj-1010	215	14	,	,	PUNCT
cuesj-1010	215	15	2022	2022	NUM
cuesj-1010	215	16	.	.	PUNCT
cuesj-1010	216	1	9	9	NUM
cuesj-1010	216	2	.	.	X
cuesj-1010	216	3	i.	i.	PROPN
cuesj-1010	216	4	katsamenis	katsamenis	PROPN
cuesj-1010	216	5	,	,	PUNCT
cuesj-1010	216	6	e.	e.	PROPN
cuesj-1010	216	7	e.	e.	PROPN
cuesj-1010	216	8	karolou	karolou	PROPN
cuesj-1010	216	9	,	,	PUNCT
cuesj-1010	216	10	a.	a.	PROPN
cuesj-1010	216	11	davradou	davradou	PROPN
cuesj-1010	216	12	,	,	PUNCT
cuesj-1010	216	13	e.	e.	PROPN
cuesj-1010	216	14	protopapadakis	protopapadakis	PROPN
cuesj-1010	216	15	,	,	PUNCT
cuesj-1010	216	16	a.	a.	NOUN
cuesj-1010	216	17	doulamis	doulamis	PROPN
cuesj-1010	216	18	,	,	PUNCT
cuesj-1010	216	19	n.	n.	PROPN
cuesj-1010	216	20	doulamis	doulamis	PROPN
cuesj-1010	216	21	and	and	CCONJ
cuesj-1010	216	22	d.	d.	PROPN
cuesj-1010	216	23	kalogeras	kalogeras	PROPN
cuesj-1010	216	24	.	.	PUNCT
cuesj-1010	217	1	tracon	tracon	NOUN
cuesj-1010	217	2	:	:	PUNCT
cuesj-1010	217	3	a	a	DET
cuesj-1010	217	4	novel	novel	NOUN
cuesj-1010	217	5	dataset	dataset	NOUN
cuesj-1010	217	6	for	for	ADP
cuesj-1010	217	7	real	real	ADJ
cuesj-1010	217	8	-	-	PUNCT
cuesj-1010	217	9	time	time	NOUN
cuesj-1010	217	10	traffic	traffic	NOUN
cuesj-1010	217	11	cones	cone	NOUN
cuesj-1010	217	12	detection	detection	NOUN
cuesj-1010	217	13	using	use	VERB
cuesj-1010	217	14	deep	deep	ADJ
cuesj-1010	217	15	learning	learning	NOUN
cuesj-1010	217	16	.	.	PUNCT
cuesj-1010	218	1	in	in	ADP
cuesj-1010	218	2	:	:	PUNCT
cuesj-1010	218	3	novel	novel	ADJ
cuesj-1010	218	4	and	and	CCONJ
cuesj-1010	218	5	intelligent	intelligent	ADJ
cuesj-1010	218	6	digital	digital	ADJ
cuesj-1010	218	7	systems	system	NOUN
cuesj-1010	218	8	conferences	conference	NOUN
cuesj-1010	218	9	.	.	PUNCT
cuesj-1010	219	1	springer	springer	NOUN
cuesj-1010	219	2	international	international	ADJ
cuesj-1010	219	3	publishing	publishing	NOUN
cuesj-1010	219	4	,	,	PUNCT
cuesj-1010	219	5	cham	cham	PROPN
cuesj-1010	219	6	,	,	PUNCT
cuesj-1010	219	7	pp	pp	ADV
cuesj-1010	219	8	.	.	PUNCT
cuesj-1010	220	1	382	382	NUM
cuesj-1010	220	2	-	-	SYM
cuesj-1010	220	3	391	391	NUM
cuesj-1010	220	4	,	,	PUNCT
cuesj-1010	220	5	2022	2022	NUM
cuesj-1010	220	6	.	.	PUNCT
cuesj-1010	221	1	available	available	ADJ
cuesj-1010	221	2	from	from	ADP
cuesj-1010	221	3	:	:	PUNCT
cuesj-1010	221	4	https://	https://	PROPN
cuesj-1010	221	5	arxiv.org/pdf/2205.11830.pdf	arxiv.org/pdf/2205.11830.pdf	PROPN
cuesj-1010	221	6	[	[	AUX
cuesj-1010	221	7	last	last	ADV
cuesj-1010	221	8	accessed	access	VERB
cuesj-1010	221	9	on	on	ADP
cuesj-1010	221	10	2023	2023	NUM
cuesj-1010	221	11	jan	jan	NOUN
cuesj-1010	221	12	12	12	NUM
cuesj-1010	221	13	]	]	PUNCT
cuesj-1010	221	14	.	.	PUNCT
cuesj-1010	222	1	10	10	NUM
cuesj-1010	222	2	.	.	PUNCT
cuesj-1010	222	3	c.	c.	PROPN
cuesj-1010	222	4	y.	y.	PROPN
cuesj-1010	222	5	wang	wang	PROPN
cuesj-1010	222	6	,	,	PUNCT
cuesj-1010	222	7	h.	h.	PROPN
cuesj-1010	222	8	y.	y.	PROPN
cuesj-1010	222	9	mark	mark	PROPN
cuesj-1010	222	10	liao	liao	PROPN
cuesj-1010	222	11	,	,	PUNCT
cuesj-1010	222	12	y.	y.	PROPN
cuesj-1010	222	13	h.	h.	PROPN
cuesj-1010	222	14	wu	wu	PROPN
cuesj-1010	222	15	,	,	PUNCT
cuesj-1010	222	16	p.	p.	PROPN
cuesj-1010	222	17	y.	y.	PROPN
cuesj-1010	222	18	chen	chen	PROPN
cuesj-1010	222	19	,	,	PUNCT
cuesj-1010	222	20	j.	j.	PROPN
cuesj-1010	222	21	w.	w.	PROPN
cuesj-1010	222	22	hsieh	hsieh	PROPN
cuesj-1010	222	23	and	and	CCONJ
cuesj-1010	222	24	i.	i.	PROPN
cuesj-1010	222	25	h	h	PROPN
cuesj-1010	222	26	yeh	yeh	PROPN
cuesj-1010	222	27	.	.	PUNCT
cuesj-1010	223	1	cspnet	cspnet	NOUN
cuesj-1010	223	2	:	:	PUNCT
cuesj-1010	223	3	a	a	DET
cuesj-1010	223	4	new	new	ADJ
cuesj-1010	223	5	backbone	backbone	NOUN
cuesj-1010	223	6	that	that	PRON
cuesj-1010	223	7	can	can	AUX
cuesj-1010	223	8	enhance	enhance	VERB
cuesj-1010	223	9	learning	learn	VERB
cuesj-1010	223	10	capability	capability	NOUN
cuesj-1010	223	11	of	of	ADP
cuesj-1010	223	12	cnn	cnn	PROPN
cuesj-1010	223	13	.	.	PUNCT
cuesj-1010	224	1	in	in	ADP
cuesj-1010	224	2	:	:	PUNCT
cuesj-1010	224	3	proceedings	proceeding	NOUN
cuesj-1010	224	4	of	of	ADP
cuesj-1010	224	5	the	the	DET
cuesj-1010	224	6	ieee	ieee	NOUN
cuesj-1010	224	7	/	/	SYM
cuesj-1010	224	8	cvf	cvf	NOUN
cuesj-1010	224	9	conference	conference	NOUN
cuesj-1010	224	10	on	on	ADP
cuesj-1010	224	11	computer	computer	NOUN
cuesj-1010	224	12	vision	vision	NOUN
cuesj-1010	224	13	and	and	CCONJ
cuesj-1010	224	14	pattern	pattern	NOUN
cuesj-1010	224	15	recognition	recognition	NOUN
cuesj-1010	224	16	workshops	workshop	NOUN
cuesj-1010	224	17	,	,	PUNCT
cuesj-1010	224	18	pp	pp	ADJ
cuesj-1010	224	19	.	.	PUNCT
cuesj-1010	224	20	390	390	NUM
cuesj-1010	224	21	-	-	SYM
cuesj-1010	224	22	391	391	NUM
cuesj-1010	224	23	.	.	PUNCT
cuesj-1010	224	24	2020	2020	NUM
cuesj-1010	224	25	.	.	PUNCT
cuesj-1010	225	1	available	available	ADJ
cuesj-1010	225	2	from	from	ADP
cuesj-1010	225	3	:	:	PUNCT
cuesj-1010	225	4	https://arxiv.org/abs/1911.11929	https://arxiv.org/abs/1911.11929	NOUN
cuesj-1010	225	5	[	[	X
cuesj-1010	225	6	last	last	ADV
cuesj-1010	225	7	accessed	access	VERB
cuesj-1010	225	8	on	on	ADP
cuesj-1010	225	9	2023	2023	NUM
cuesj-1010	225	10	feb	feb	NOUN
cuesj-1010	225	11	06	06	NUM
cuesj-1010	225	12	]	]	PUNCT
cuesj-1010	225	13	.	.	PUNCT
cuesj-1010	226	1	11	11	NUM
cuesj-1010	226	2	.	.	PUNCT
cuesj-1010	227	1	k.	k.	PROPN
cuesj-1010	227	2	wang	wang	PROPN
cuesj-1010	227	3	,	,	PUNCT
cuesj-1010	227	4	j.	j.	PROPN
cuesj-1010	227	5	h.	h.	PROPN
cuesj-1010	227	6	liew	liew	PROPN
cuesj-1010	227	7	,	,	PUNCT
cuesj-1010	227	8	y.	y.	PROPN
cuesj-1010	227	9	zou	zou	PROPN
cuesj-1010	227	10	,	,	PUNCT
cuesj-1010	227	11	d.	d.	PROPN
cuesj-1010	227	12	zhou	zhou	PROPN
cuesj-1010	227	13	and	and	CCONJ
cuesj-1010	227	14	j.	j.	PROPN
cuesj-1010	227	15	feng	feng	PROPN
cuesj-1010	227	16	.	.	PUNCT
cuesj-1010	228	1	panet	panet	NOUN
cuesj-1010	228	2	:	:	PUNCT
cuesj-1010	228	3	few	few	ADJ
cuesj-1010	228	4	-	-	PUNCT
cuesj-1010	228	5	shot	shot	NOUN
cuesj-1010	228	6	image	image	NOUN
cuesj-1010	228	7	semantic	semantic	ADJ
cuesj-1010	228	8	segmentation	segmentation	NOUN
cuesj-1010	228	9	with	with	ADP
cuesj-1010	228	10	prototype	prototype	NOUN
cuesj-1010	228	11	alignment	alignment	NOUN
cuesj-1010	228	12	.	.	PUNCT
cuesj-1010	229	1	in	in	ADP
cuesj-1010	229	2	:	:	PUNCT
cuesj-1010	229	3	proceedings	proceeding	NOUN
cuesj-1010	229	4	of	of	ADP
cuesj-1010	229	5	the	the	DET
cuesj-1010	229	6	ieee	ieee	NOUN
cuesj-1010	229	7	/	/	SYM
cuesj-1010	229	8	cvf	cvf	NOUN
cuesj-1010	229	9	international	international	ADJ
cuesj-1010	229	10	conference	conference	NOUN
cuesj-1010	229	11	on	on	ADP
cuesj-1010	229	12	computer	computer	NOUN
cuesj-1010	229	13	vision	vision	NOUN
cuesj-1010	229	14	,	,	PUNCT
cuesj-1010	229	15	pp	pp	ADP
cuesj-1010	229	16	.	.	PUNCT
cuesj-1010	229	17	9197	9197	NUM
cuesj-1010	229	18	-	-	SYM
cuesj-1010	229	19	9206	9206	NUM
cuesj-1010	229	20	,	,	PUNCT
cuesj-1010	229	21	2019	2019	NUM
cuesj-1010	229	22	.	.	PUNCT
cuesj-1010	230	1	12	12	NUM
cuesj-1010	230	2	.	.	PUNCT
cuesj-1010	230	3	m.	m.	NOUN
cuesj-1010	230	4	horvat	horvat	PROPN
cuesj-1010	230	5	,	,	PUNCT
cuesj-1010	230	6	l.	l.	PROPN
cuesj-1010	230	7	jelečević	jelečević	PROPN
cuesj-1010	230	8	and	and	CCONJ
cuesj-1010	230	9	g.	g.	PROPN
cuesj-1010	230	10	gledec	gledec	PROPN
cuesj-1010	230	11	.	.	PUNCT
cuesj-1010	231	1	a	a	DET
cuesj-1010	231	2	comparative	comparative	ADJ
cuesj-1010	231	3	study	study	NOUN
cuesj-1010	231	4	of	of	ADP
cuesj-1010	231	5	yolov5	yolov5	NOUN
cuesj-1010	231	6	models	model	NOUN
cuesj-1010	231	7	performance	performance	NOUN
cuesj-1010	231	8	for	for	ADP
cuesj-1010	231	9	image	image	NOUN
cuesj-1010	231	10	localization	localization	NOUN
cuesj-1010	231	11	and	and	CCONJ
cuesj-1010	231	12	classification	classification	NOUN
cuesj-1010	231	13	.	.	PUNCT
cuesj-1010	232	1	in	in	ADP
cuesj-1010	232	2	:	:	PUNCT
cuesj-1010	232	3	proceedings	proceeding	NOUN
cuesj-1010	232	4	of	of	ADP
cuesj-1010	232	5	thecentral	thecentral	ADJ
cuesj-1010	232	6	european	european	ADJ
cuesj-1010	232	7	conference	conference	NOUN
cuesj-1010	232	8	on	on	ADP
cuesj-1010	232	9	information	information	NOUN
cuesj-1010	232	10	and	and	CCONJ
cuesj-1010	232	11	intelligent	intelligent	ADJ
cuesj-1010	232	12	systems	system	NOUN
cuesj-1010	232	13	.	.	PUNCT
cuesj-1010	233	1	faculty	faculty	NOUN
cuesj-1010	233	2	of	of	ADP
cuesj-1010	233	3	organization	organization	NOUN
cuesj-1010	233	4	and	and	CCONJ
cuesj-1010	233	5	informatics	informatic	NOUN
cuesj-1010	233	6	varazdin	varazdin	VERB
cuesj-1010	233	7	,	,	PUNCT
cuesj-1010	233	8	pp	pp	ADJ
cuesj-1010	233	9	.	.	PUNCT
cuesj-1010	234	1	349	349	NUM
cuesj-1010	234	2	-	-	SYM
cuesj-1010	234	3	356	356	NUM
cuesj-1010	234	4	,	,	PUNCT
cuesj-1010	234	5	2022	2022	NUM
cuesj-1010	234	6	.	.	PUNCT
cuesj-1010	235	1	available	available	ADJ
cuesj-1010	235	2	from	from	ADP
cuesj-1010	235	3	:	:	PUNCT
cuesj-1010	235	4	https://shorturl.at/hfpt1	https://shorturl.at/hfpt1	PROPN
cuesj-1010	235	5	[	[	PUNCT
cuesj-1010	235	6	last	last	ADV
cuesj-1010	235	7	accessed	access	VERB
cuesj-1010	235	8	on	on	ADP
cuesj-1010	235	9	2023	2023	NUM
cuesj-1010	235	10	feb	feb	NOUN
cuesj-1010	235	11	16	16	NUM
cuesj-1010	235	12	]	]	PUNCT
cuesj-1010	235	13	.	.	PUNCT
cuesj-1010	236	1	13	13	NUM
cuesj-1010	236	2	.	.	X
cuesj-1010	236	3	d.	d.	PROPN
cuesj-1010	236	4	dlužnevskij	dlužnevskij	PROPN
cuesj-1010	236	5	,	,	PUNCT
cuesj-1010	236	6	p.	p.	PROPN
cuesj-1010	236	7	stefanovič	stefanovič	PROPN
cuesj-1010	236	8	and	and	CCONJ
cuesj-1010	236	9	s.	s.	PROPN
cuesj-1010	236	10	ramanauskaite	ramanauskaite	PROPN
cuesj-1010	236	11	.	.	PUNCT
cuesj-1010	237	1	investigation	investigation	NOUN
cuesj-1010	237	2	of	of	ADP
cuesj-1010	237	3	yolov5	yolov5	NOUN
cuesj-1010	237	4	efficiency	efficiency	NOUN
cuesj-1010	237	5	in	in	ADP
cuesj-1010	237	6	iphone	iphone	NOUN
cuesj-1010	237	7	supported	support	VERB
cuesj-1010	237	8	systems	system	NOUN
cuesj-1010	237	9	.	.	PUNCT
cuesj-1010	238	1	baltic	baltic	PROPN
cuesj-1010	238	2	journal	journal	PROPN
cuesj-1010	238	3	of	of	ADP
cuesj-1010	238	4	modern	modern	ADJ
cuesj-1010	238	5	computing	computing	NOUN
cuesj-1010	238	6	,	,	PUNCT
cuesj-1010	238	7	vol	vol	NOUN
cuesj-1010	238	8	.	.	PROPN
cuesj-1010	238	9	9	9	NUM
cuesj-1010	238	10	,	,	PUNCT
cuesj-1010	238	11	pp	pp	ADJ
cuesj-1010	238	12	.	.	PUNCT
cuesj-1010	239	1	333	333	NUM
cuesj-1010	239	2	-	-	SYM
cuesj-1010	239	3	334	334	NUM
cuesj-1010	239	4	,	,	PUNCT
cuesj-1010	239	5	2021	2021	NUM
cuesj-1010	239	6	.	.	PUNCT
cuesj-1010	240	1	14	14	NUM
cuesj-1010	240	2	.	.	PUNCT
cuesj-1010	241	1	s.	s.	PROPN
cuesj-1010	241	2	mshir	mshir	PROPN
cuesj-1010	241	3	.	.	PUNCT
cuesj-1010	242	1	vehicle	vehicle	NOUN
cuesj-1010	242	2	detection	detection	NOUN
cuesj-1010	242	3	dataset	dataset	VERB
cuesj-1010	242	4	.	.	PUNCT
cuesj-1010	243	1	roboflow	roboflow	ADJ
cuesj-1010	243	2	universe	universe	NOUN
cuesj-1010	243	3	,	,	PUNCT
cuesj-1010	243	4	2023	2023	NUM
cuesj-1010	243	5	.	.	PUNCT
cuesj-1010	244	1	available	available	ADJ
cuesj-1010	244	2	from	from	ADP
cuesj-1010	244	3	:	:	PUNCT
cuesj-1010	244	4	https://universe.roboflow.com/shalaw/vehicledetection-rfmsje	https://universe.roboflow.com/shalaw/vehicledetection-rfmsje	NOUN
cuesj-1010	245	1	[	[	X
cuesj-1010	245	2	last	last	ADV
cuesj-1010	245	3	accessed	access	VERB
cuesj-1010	245	4	on	on	ADP
cuesj-1010	245	5	2023	2023	NUM
cuesj-1010	245	6	feb	feb	NOUN
cuesj-1010	245	7	]	]	X
cuesj-1010	245	8	.	.	PUNCT
cuesj-1010	246	1	15	15	NUM
cuesj-1010	246	2	.	.	PUNCT
cuesj-1010	247	1	s.	s.	PROPN
cuesj-1010	247	2	karslioglu	karslioglu	PROPN
cuesj-1010	247	3	.	.	PUNCT
cuesj-1010	248	1	saving	save	VERB
cuesj-1010	248	2	and	and	CCONJ
cuesj-1010	248	3	loading	loading	NOUN
cuesj-1010	248	4	models	model	NOUN
cuesj-1010	248	5	across	across	ADP
cuesj-1010	248	6	devices	device	NOUN
cuesj-1010	248	7	in	in	ADP
cuesj-1010	248	8	pytorch	pytorch	NOUN
cuesj-1010	248	9	.	.	PUNCT
cuesj-1010	249	1	pytorch	pytorch	NOUN
cuesj-1010	249	2	organization	organization	NOUN
cuesj-1010	249	3	,	,	PUNCT
cuesj-1010	249	4	2023	2023	NUM
cuesj-1010	249	5	.	.	PUNCT
cuesj-1010	250	1	available	available	ADJ
cuesj-1010	250	2	from	from	ADP
cuesj-1010	250	3	:	:	PUNCT
cuesj-1010	250	4	https://	https://	PROPN
cuesj-1010	250	5	pytorch.org/tutorials/recipes/recipes/save_load_across_devices	pytorch.org/tutorials/recipes/recipes/save_load_across_devices	PROPN
cuesj-1010	250	6	.	.	PUNCT
cuesj-1010	250	7	html	html	NOUN
cuesj-1010	251	1	[	[	X
cuesj-1010	251	2	last	last	ADV
cuesj-1010	251	3	accessed	access	VERB
cuesj-1010	251	4	on	on	ADP
cuesj-1010	251	5	2023	2023	NUM
cuesj-1010	251	6	feb	feb	NOUN
cuesj-1010	251	7	02	02	NUM
cuesj-1010	251	8	]	]	X
cuesj-1010	251	9	.	.	PUNCT
cuesj-1010	252	1	16	16	NUM
cuesj-1010	252	2	.	.	PUNCT
cuesj-1010	253	1	p.	p.	NOUN
cuesj-1010	253	2	k.	k.	PROPN
cuesj-1010	254	1	yadav	yadav	PROPN
cuesj-1010	254	2	,	,	PUNCT
cuesj-1010	254	3	j.	j.	PROPN
cuesj-1010	254	4	a.	a.	PROPN
cuesj-1010	254	5	thomasson	thomasson	PROPN
cuesj-1010	254	6	,	,	PUNCT
cuesj-1010	254	7	s.	s.	PROPN
cuesj-1010	254	8	w.	w.	PROPN
cuesj-1010	254	9	searcy	searcy	PROPN
cuesj-1010	254	10	,	,	PUNCT
cuesj-1010	254	11	r.	r.	PROPN
cuesj-1010	254	12	g.	g.	PROPN
cuesj-1010	254	13	hardin	hardin	PROPN
cuesj-1010	254	14	,	,	PUNCT
cuesj-1010	254	15	u.	u.	PROPN
cuesj-1010	254	16	braga	braga	PROPN
cuesj-1010	254	17	-	-	PUNCT
cuesj-1010	254	18	neto	neto	PROPN
cuesj-1010	254	19	,	,	PUNCT
cuesj-1010	254	20	s.	s.	PROPN
cuesj-1010	254	21	c.	c.	PROPN
cuesj-1010	254	22	popescu	popescu	PROPN
cuesj-1010	254	23	,	,	PUNCT
cuesj-1010	254	24	d.	d.	PROPN
cuesj-1010	254	25	e.	e.	PROPN
cuesj-1010	254	26	martin	martin	PROPN
cuesj-1010	254	27	,	,	PUNCT
cuesj-1010	254	28	r.	r.	PROPN
cuesj-1010	254	29	rodriguez	rodriguez	PROPN
cuesj-1010	254	30	,	,	PUNCT
cuesj-1010	254	31	k.	k.	PROPN
cuesj-1010	254	32	meza	meza	PROPN
cuesj-1010	254	33	,	,	PUNCT
cuesj-1010	254	34	j.	j.	PROPN
cuesj-1010	254	35	enciso	enciso	PROPN
cuesj-1010	254	36	,	,	PUNCT
cuesj-1010	254	37	j.	j.	PROPN
cuesj-1010	254	38	s.	s.	PROPN
cuesj-1010	254	39	diaz	diaz	PROPN
cuesj-1010	254	40	and	and	CCONJ
cuesj-1010	254	41	t.	t.	PROPN
cuesj-1010	254	42	wang	wang	PROPN
cuesj-1010	254	43	.	.	PUNCT
cuesj-1010	255	1	assessing	assess	VERB
cuesj-1010	255	2	the	the	DET
cuesj-1010	255	3	performance	performance	NOUN
cuesj-1010	255	4	of	of	ADP
cuesj-1010	255	5	yolov5	yolov5	NOUN
cuesj-1010	255	6	algorithm	algorithm	NOUN
cuesj-1010	255	7	for	for	ADP
cuesj-1010	255	8	detecting	detect	VERB
cuesj-1010	255	9	volunteer	volunteer	NOUN
cuesj-1010	255	10	cotton	cotton	NOUN
cuesj-1010	255	11	plants	plant	NOUN
cuesj-1010	255	12	in	in	ADP
cuesj-1010	255	13	corn	corn	NOUN
cuesj-1010	255	14	fields	field	NOUN
cuesj-1010	255	15	at	at	ADP
cuesj-1010	255	16	three	three	NUM
cuesj-1010	255	17	different	different	ADJ
cuesj-1010	255	18	growth	growth	NOUN
cuesj-1010	255	19	stages	stage	NOUN
cuesj-1010	255	20	.	.	PUNCT
cuesj-1010	256	1	artificial	artificial	ADJ
cuesj-1010	256	2	intelligence	intelligence	NOUN
cuesj-1010	256	3	in	in	ADP
cuesj-1010	256	4	agriculture	agriculture	NOUN
cuesj-1010	256	5	,	,	PUNCT
cuesj-1010	256	6	vol	vol	NOUN
cuesj-1010	256	7	.	.	PROPN
cuesj-1010	257	1	6	6	NUM
cuesj-1010	257	2	,	,	PUNCT
cuesj-1010	257	3	pp	pp	ADJ
cuesj-1010	257	4	.	.	PUNCT
cuesj-1010	258	1	292	292	NUM
cuesj-1010	258	2	-	-	SYM
cuesj-1010	258	3	303	303	NUM
cuesj-1010	258	4	,	,	PUNCT
cuesj-1010	258	5	2022	2022	NUM
cuesj-1010	258	6	.	.	PUNCT
cuesj-1010	259	1	table	table	NOUN
cuesj-1010	259	2	7	7	NUM
cuesj-1010	259	3	:	:	PUNCT
cuesj-1010	259	4	indicator	indicator	NOUN
cuesj-1010	259	5	of	of	ADP
cuesj-1010	259	6	yolov5x	yolov5x	PROPN
cuesj-1010	259	7	model	model	NOUN
cuesj-1010	259	8	class	class	NOUN
cuesj-1010	259	9	precision	precision	NOUN
cuesj-1010	259	10	recal	recal	ADJ
cuesj-1010	259	11	f1	f1	NOUN
cuesj-1010	259	12	score	score	NOUN
cuesj-1010	259	13	map	map	NOUN
cuesj-1010	259	14	(	(	PUNCT
cuesj-1010	259	15	0.5	0.5	NUM
cuesj-1010	259	16	)	)	PUNCT
cuesj-1010	259	17	vehicle_above	vehicle_above	NOUN
cuesj-1010	259	18	0.97	0.97	NUM
cuesj-1010	259	19	0.88	0.88	NUM
cuesj-1010	259	20	0.94	0.94	NUM
cuesj-1010	259	21	0.93	0.93	NUM
cuesj-1010	259	22	yolov5x	yolov5x	PROPN
cuesj-1010	259	23	vehicle_rear	vehicle_rear	PROPN
cuesj-1010	259	24	0.94	0.94	NUM
cuesj-1010	259	25	0.94	0.94	NUM
cuesj-1010	259	26	0.92	0.92	NUM
cuesj-1010	259	27	0.93	0.93	NUM
cuesj-1010	259	28	vehicle_front	vehicle_front	NOUN
cuesj-1010	259	29	0.96	0.96	NUM
cuesj-1010	259	30	0.96	0.96	NUM
cuesj-1010	259	31	0.96	0.96	NUM
cuesj-1010	259	32	0.96	0.96	NUM
cuesj-1010	259	33	vehicle_sides	vehicle_side	NOUN
cuesj-1010	259	34	0.94	0.94	NUM
cuesj-1010	259	35	0.95	0.95	NUM
cuesj-1010	259	36	0.94	0.94	NUM
cuesj-1010	259	37	0.95	0.95	NUM
cuesj-1010	259	38	abdallah	abdallah	PROPN
cuesj-1010	259	39	:	:	PUNCT
cuesj-1010	259	40	vehicle	vehicle	NOUN
cuesj-1010	259	41	detection	detection	NOUN
cuesj-1010	259	42	with	with	ADP
cuesj-1010	259	43	tello	tello	PROPN
cuesj-1010	259	44	drone	drone	PROPN
cuesj-1010	259	45	and	and	CCONJ
cuesj-1010	259	46	yolov5	yolov5	NOUN
cuesj-1010	259	47	7	7	NUM
cuesj-1010	259	48	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-1010	259	49	cuesj	cuesj	ADJ
cuesj-1010	259	50	2024	2024	NUM
cuesj-1010	259	51	,	,	PUNCT
cuesj-1010	259	52	8	8	NUM
cuesj-1010	259	53	(	(	PUNCT
cuesj-1010	259	54	1	1	NUM
cuesj-1010	259	55	):	):	PUNCT
cuesj-1010	259	56	1	1	NUM
cuesj-1010	259	57	-	-	SYM
cuesj-1010	259	58	7	7	NUM
cuesj-1010	259	59	17	17	NUM
cuesj-1010	259	60	.	.	PUNCT
cuesj-1010	260	1	j.	j.	PROPN
cuesj-1010	260	2	lowe	lowe	PROPN
cuesj-1010	260	3	.	.	PROPN
cuesj-1010	260	4	precision	precision	PROPN
cuesj-1010	260	5	and	and	CCONJ
cuesj-1010	260	6	recall	recall	NOUN
cuesj-1010	260	7	in	in	ADP
cuesj-1010	260	8	machine	machine	NOUN
cuesj-1010	260	9	learning	learning	NOUN
cuesj-1010	260	10	.	.	PUNCT
cuesj-1010	261	1	roboflow	roboflow	ADJ
cuesj-1010	261	2	blog	blog	PROPN
cuesj-1010	261	3	,	,	PUNCT
cuesj-1010	261	4	2022	2022	NUM
cuesj-1010	261	5	.	.	PUNCT
cuesj-1010	262	1	available	available	ADJ
cuesj-1010	262	2	from	from	ADP
cuesj-1010	262	3	:	:	PUNCT
cuesj-1010	262	4	https://blog.roboflow.com/precision-andrecall	https://blog.roboflow.com/precision-andrecall	PROPN
cuesj-1010	263	1	[	[	X
cuesj-1010	263	2	last	last	ADV
cuesj-1010	263	3	accessed	access	VERB
cuesj-1010	263	4	on	on	ADP
cuesj-1010	263	5	2023	2023	NUM
cuesj-1010	263	6	feb	feb	NOUN
cuesj-1010	263	7	03	03	NUM
cuesj-1010	263	8	]	]	PUNCT
cuesj-1010	263	9	.	.	PUNCT
cuesj-1010	264	1	18	18	NUM
cuesj-1010	264	2	.	.	PUNCT
cuesj-1010	264	3	f.	f.	PROPN
cuesj-1010	264	4	damien	damien	PROPN
cuesj-1010	264	5	,	,	PUNCT
cuesj-1010	264	6	m.	m.	PROPN
cuesj-1010	264	7	u.	u.	PROPN
cuesj-1010	264	8	javaid	javaid	PROPN
cuesj-1010	264	9	,	,	PUNCT
cuesj-1010	264	10	n.	n.	NOUN
cuesj-1010	264	11	posocco	posocco	NOUN
cuesj-1010	264	12	and	and	CCONJ
cuesj-1010	264	13	s.	s.	PROPN
cuesj-1010	264	14	tihon	tihon	PROPN
cuesj-1010	264	15	.	.	PUNCT
cuesj-1010	265	1	anomaly	anomaly	PROPN
cuesj-1010	265	2	detection	detection	NOUN
cuesj-1010	265	3	:	:	PUNCT
cuesj-1010	265	4	how	how	SCONJ
cuesj-1010	265	5	to	to	PART
cuesj-1010	265	6	artificially	artificially	ADV
cuesj-1010	265	7	increase	increase	VERB
cuesj-1010	265	8	your	your	PRON
cuesj-1010	265	9	f1	f1	NOUN
cuesj-1010	265	10	-	-	PUNCT
cuesj-1010	265	11	score	score	NOUN
cuesj-1010	265	12	with	with	ADP
cuesj-1010	265	13	a	a	DET
cuesj-1010	265	14	biased	biased	ADJ
cuesj-1010	265	15	evaluation	evaluation	NOUN
cuesj-1010	265	16	protocol	protocol	NOUN
cuesj-1010	265	17	.	.	PUNCT
cuesj-1010	266	1	in	in	ADP
cuesj-1010	266	2	:	:	PUNCT
cuesj-1010	266	3	joint	joint	ADJ
cuesj-1010	266	4	european	european	ADJ
cuesj-1010	266	5	conference	conference	NOUN
cuesj-1010	266	6	on	on	ADP
cuesj-1010	266	7	machine	machine	NOUN
cuesj-1010	266	8	learning	learning	NOUN
cuesj-1010	266	9	and	and	CCONJ
cuesj-1010	266	10	knowledge	knowledge	NOUN
cuesj-1010	266	11	discovery	discovery	NOUN
cuesj-1010	266	12	in	in	ADP
cuesj-1010	266	13	databases	database	NOUN
cuesj-1010	266	14	.	.	PUNCT
cuesj-1010	267	1	springer	springer	NOUN
cuesj-1010	267	2	international	international	ADJ
cuesj-1010	267	3	publishing	publishing	NOUN
cuesj-1010	267	4	,	,	PUNCT
cuesj-1010	267	5	cham	cham	PROPN
cuesj-1010	267	6	,	,	PUNCT
cuesj-1010	267	7	pp	pp	ADJ
cuesj-1010	267	8	.	.	PUNCT
cuesj-1010	268	1	3	3	NUM
cuesj-1010	268	2	-	-	SYM
cuesj-1010	268	3	18	18	NUM
cuesj-1010	268	4	,	,	PUNCT
cuesj-1010	268	5	2021	2021	NUM
cuesj-1010	268	6	.	.	PUNCT
cuesj-1010	269	1	available	available	ADJ
cuesj-1010	269	2	from	from	ADP
cuesj-1010	269	3	:	:	PUNCT
cuesj-1010	269	4	https://arxiv.org/pdf/2106.16020.pdf	https://arxiv.org/pdf/2106.16020.pdf	PROPN
cuesj-1010	270	1	[	[	X
cuesj-1010	270	2	last	last	ADV
cuesj-1010	270	3	accessed	access	VERB
cuesj-1010	270	4	on	on	ADP
cuesj-1010	270	5	2023	2023	NUM
cuesj-1010	270	6	feb	feb	NOUN
cuesj-1010	270	7	12	12	NUM
cuesj-1010	270	8	]	]	PUNCT
cuesj-1010	270	9	.	.	PUNCT
cuesj-1010	271	1	19	19	NUM
cuesj-1010	271	2	.	.	PUNCT
cuesj-1010	271	3	j.	j.	PROPN
cuesj-1010	271	4	solawetz	solawetz	PROPN
cuesj-1010	271	5	.	.	PUNCT
cuesj-1010	272	1	what	what	PRON
cuesj-1010	272	2	is	be	AUX
cuesj-1010	272	3	mean	mean	ADJ
cuesj-1010	272	4	average	average	ADJ
cuesj-1010	272	5	precision	precision	NOUN
cuesj-1010	272	6	(	(	PUNCT
cuesj-1010	272	7	map	map	NOUN
cuesj-1010	272	8	)	)	PUNCT
cuesj-1010	272	9	in	in	ADP
cuesj-1010	272	10	object	object	NOUN
cuesj-1010	272	11	detection	detection	NOUN
cuesj-1010	272	12	?	?	PUNCT
cuesj-1010	273	1	roboflow	roboflow	ADJ
cuesj-1010	273	2	blog	blog	PROPN
cuesj-1010	273	3	,	,	PUNCT
cuesj-1010	273	4	2020	2020	NUM
cuesj-1010	273	5	.	.	PUNCT
cuesj-1010	274	1	available	available	ADJ
cuesj-1010	274	2	from	from	ADP
cuesj-1010	274	3	:	:	PUNCT
cuesj-1010	274	4	https://blog	https://blog	NUM
cuesj-1010	274	5	.	.	PUNCT
cuesj-1010	275	1	roboflow.com/mean-average-precision	roboflow.com/mean-average-precision	NOUN
cuesj-1010	275	2	[	[	X
cuesj-1010	275	3	last	last	ADV
cuesj-1010	275	4	accessed	access	VERB
cuesj-1010	275	5	on	on	ADP
cuesj-1010	275	6	2023	2023	NUM
cuesj-1010	275	7	feb	feb	NOUN
cuesj-1010	275	8	14	14	NUM
cuesj-1010	275	9	]	]	PUNCT
cuesj-1010	275	10	.	.	PUNCT
cuesj-1010	276	1	20	20	NUM
cuesj-1010	276	2	.	.	PUNCT
cuesj-1010	276	3	q.	q.	PROPN
cuesj-1010	276	4	h.	h.	PROPN
cuesj-1010	276	5	phan	phan	PROPN
cuesj-1010	276	6	,	,	PUNCT
cuesj-1010	276	7	v.	v.	ADP
cuesj-1010	276	8	t.	t.	PROPN
cuesj-1010	276	9	nguyen	nguyen	PROPN
cuesj-1010	276	10	,	,	PUNCT
cuesj-1010	276	11	c.	c.	PROPN
cuesj-1010	276	12	h.	h.	PROPN
cuesj-1010	276	13	lien	lien	PROPN
cuesj-1010	276	14	,	,	PUNCT
cuesj-1010	276	15	t.	t.	PROPN
cuesj-1010	276	16	p.	p.	PROPN
cuesj-1010	276	17	duong	duong	PROPN
cuesj-1010	276	18	,	,	PUNCT
cuesj-1010	276	19	m.t	m.t	PROPN
cuesj-1010	276	20	.	.	PROPN
cuesj-1010	276	21	k.	k.	PROPN
cuesj-1010	276	22	hou	hou	PROPN
cuesj-1010	276	23	and	and	CCONJ
cuesj-1010	277	1	n.	n.	PROPN
cuesj-1010	277	2	b.	b.	PROPN
cuesj-1010	277	3	le	le	PROPN
cuesj-1010	277	4	.	.	PUNCT
cuesj-1010	278	1	classification	classification	NOUN
cuesj-1010	278	2	of	of	ADP
cuesj-1010	278	3	tomato	tomato	NOUN
cuesj-1010	278	4	fruit	fruit	NOUN
cuesj-1010	278	5	using	use	VERB
cuesj-1010	278	6	yolov5	yolov5	NOUN
cuesj-1010	278	7	and	and	CCONJ
cuesj-1010	278	8	convolutional	convolutional	ADJ
cuesj-1010	278	9	neural	neural	ADJ
cuesj-1010	278	10	network	network	NOUN
cuesj-1010	278	11	models	model	NOUN
cuesj-1010	278	12	.	.	PUNCT
cuesj-1010	279	1	plants	plant	NOUN
cuesj-1010	279	2	,	,	PUNCT
cuesj-1010	279	3	vol	vol	NOUN
cuesj-1010	279	4	.	.	PROPN
cuesj-1010	279	5	12	12	NUM
cuesj-1010	279	6	,	,	PUNCT
cuesj-1010	279	7	p.	p.	NOUN
cuesj-1010	279	8	790	790	NUM
cuesj-1010	279	9	,	,	PUNCT
cuesj-1010	279	10	2023	2023	NUM
cuesj-1010	279	11	.	.	PUNCT