id	sid	tid	token	lemma	pos
easat-5690	1	1	edelweiss	edelweiss	PROPN
easat-5690	1	2	applied	apply	VERB
easat-5690	1	3	science	science	NOUN
easat-5690	1	4	and	and	CCONJ
easat-5690	1	5	technology	technology	NOUN
easat-5690	1	6	issn	issn	PROPN
easat-5690	1	7	:	:	PUNCT
easat-5690	1	8	2576	2576	NUM
easat-5690	1	9	-	-	SYM
easat-5690	1	10	8484	8484	NUM
easat-5690	1	11	vol	vol	NOUN
easat-5690	1	12	.	.	PROPN
easat-5690	2	1	9	9	NUM
easat-5690	2	2	,	,	PUNCT
easat-5690	2	3	no	no	INTJ
easat-5690	2	4	.	.	NOUN
easat-5690	2	5	3	3	NUM
easat-5690	2	6	,	,	PUNCT
easat-5690	2	7	1777	1777	NUM
easat-5690	2	8	-	-	SYM
easat-5690	2	9	1794	1794	NUM
easat-5690	2	10	2025	2025	NUM
easat-5690	2	11	publisher	publisher	NOUN
easat-5690	2	12	:	:	PUNCT
easat-5690	2	13	learning	learn	VERB
easat-5690	2	14	gate	gate	NOUN
easat-5690	2	15	doi	doi	PROPN
easat-5690	2	16	:	:	PUNCT
easat-5690	2	17	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	2	18	©	©	PROPN
easat-5690	2	19	2025	2025	NUM
easat-5690	2	20	by	by	ADP
easat-5690	2	21	the	the	DET
easat-5690	2	22	authors	author	NOUN
easat-5690	2	23	;	;	PUNCT
easat-5690	2	24	licensee	licensee	PROPN
easat-5690	2	25	learning	learning	NOUN
easat-5690	2	26	gate	gate	NOUN
easat-5690	3	1	©	©	PROPN
easat-5690	3	2	2025	2025	NUM
easat-5690	3	3	by	by	ADP
easat-5690	3	4	the	the	DET
easat-5690	3	5	authors	author	NOUN
easat-5690	3	6	;	;	PUNCT
easat-5690	3	7	licensee	licensee	PROPN
easat-5690	3	8	learning	learning	NOUN
easat-5690	3	9	gate	gate	NOUN
easat-5690	3	10	history	history	NOUN
easat-5690	3	11	:	:	PUNCT
easat-5690	3	12	received	receive	VERB
easat-5690	3	13	:	:	PUNCT
easat-5690	3	14	10	10	NUM
easat-5690	3	15	january	january	PROPN
easat-5690	3	16	2025	2025	NUM
easat-5690	3	17	;	;	PUNCT
easat-5690	3	18	revised	revise	VERB
easat-5690	3	19	:	:	PUNCT
easat-5690	3	20	28	28	NUM
easat-5690	3	21	february	february	NOUN
easat-5690	3	22	2025	2025	NUM
easat-5690	3	23	;	;	PUNCT
easat-5690	3	24	accepted	accept	VERB
easat-5690	3	25	:	:	PUNCT
easat-5690	3	26	3	3	NUM
easat-5690	3	27	march	march	NOUN
easat-5690	3	28	2025	2025	NUM
easat-5690	3	29	;	;	PUNCT
easat-5690	3	30	published	publish	VERB
easat-5690	3	31	:	:	PUNCT
easat-5690	3	32	22	22	NUM
easat-5690	3	33	march	march	NOUN
easat-5690	3	34	2025	2025	NUM
easat-5690	3	35	*	*	PUNCT
easat-5690	3	36	correspondence	correspondence	NOUN
easat-5690	3	37	:	:	PUNCT
easat-5690	3	38	2240053955@qq.com	2240053955@qq.com	NUM
easat-5690	3	39	a	a	DET
easat-5690	3	40	diseases	disease	NOUN
easat-5690	3	41	and	and	CCONJ
easat-5690	3	42	pests	pest	VERB
easat-5690	3	43	identification	identification	NOUN
easat-5690	3	44	system	system	NOUN
easat-5690	3	45	in	in	ADP
easat-5690	3	46	tea	tea	NOUN
easat-5690	3	47	leaves	leave	NOUN
easat-5690	3	48	using	use	VERB
easat-5690	3	49	improved	improved	ADJ
easat-5690	3	50	yolov5	yolov5	NOUN
easat-5690	3	51	deep	deep	ADJ
easat-5690	3	52	learning	learning	NOUN
easat-5690	3	53	model	model	NOUN
easat-5690	3	54	xianghong	xianghong	PROPN
easat-5690	3	55	deng1	deng1	PROPN
easat-5690	3	56	,	,	PUNCT
easat-5690	3	57	zhiwei	zhiwei	PROPN
easat-5690	3	58	zhou1	zhou1	PROPN
easat-5690	3	59	,	,	PUNCT
easat-5690	3	60	mingze	mingze	ADJ
easat-5690	3	61	lei2	lei2	NOUN
easat-5690	3	62	,	,	PUNCT
easat-5690	4	1	xuwen	xuwen	PROPN
easat-5690	4	2	zheng2	zheng2	PROPN
easat-5690	4	3	*	*	SYM
easat-5690	4	4	1hunan	1hunan	NUM
easat-5690	4	5	mechanical	mechanical	ADJ
easat-5690	4	6	&	&	CCONJ
easat-5690	4	7	electrical	electrical	ADJ
easat-5690	4	8	polytechnic	polytechnic	ADJ
easat-5690	4	9	,	,	PUNCT
easat-5690	4	10	changsha410151	changsha410151	PROPN
easat-5690	4	11	,	,	PUNCT
easat-5690	4	12	china	china	PROPN
easat-5690	4	13	.	.	PUNCT
easat-5690	5	1	2faculty	2faculty	NUM
easat-5690	5	2	of	of	ADP
easat-5690	5	3	engineering	engineering	NOUN
easat-5690	5	4	,	,	PUNCT
easat-5690	5	5	mahasarakham	mahasarakham	PROPN
easat-5690	5	6	university	university	PROPN
easat-5690	5	7	,	,	PUNCT
easat-5690	5	8	mahasarakham44150	mahasarakham44150	PROPN
easat-5690	5	9	,	,	PUNCT
easat-5690	5	10	thailand	thailand	PROPN
easat-5690	5	11	;	;	PUNCT
easat-5690	5	12	2240053955@qq.com	2240053955@qq.com	NUM
easat-5690	5	13	(	(	PUNCT
easat-5690	5	14	x.z	x.z	PROPN
easat-5690	5	15	.	.	PROPN
easat-5690	5	16	)	)	PUNCT
easat-5690	5	17	.	.	PUNCT
easat-5690	6	1	abstract	abstract	ADJ
easat-5690	6	2	:	:	PUNCT
easat-5690	6	3	people	people	NOUN
easat-5690	6	4	all	all	PRON
easat-5690	6	5	throughout	throughout	ADP
easat-5690	6	6	the	the	DET
easat-5690	6	7	world	world	NOUN
easat-5690	6	8	enjoy	enjoy	VERB
easat-5690	6	9	tea	tea	NOUN
easat-5690	6	10	.	.	PUNCT
easat-5690	7	1	the	the	DET
easat-5690	7	2	traditional	traditional	ADJ
easat-5690	7	3	detection	detection	NOUN
easat-5690	7	4	of	of	ADP
easat-5690	7	5	tea	tea	NOUN
easat-5690	7	6	diseases	disease	NOUN
easat-5690	7	7	relies	rely	VERB
easat-5690	7	8	heavily	heavily	ADV
easat-5690	7	9	on	on	ADP
easat-5690	7	10	farm	farm	NOUN
easat-5690	7	11	experts	expert	NOUN
easat-5690	7	12	,	,	PUNCT
easat-5690	7	13	which	which	PRON
easat-5690	7	14	often	often	ADV
easat-5690	7	15	takes	take	VERB
easat-5690	7	16	a	a	DET
easat-5690	7	17	lot	lot	NOUN
easat-5690	7	18	of	of	ADP
easat-5690	7	19	time	time	NOUN
easat-5690	7	20	.	.	PUNCT
easat-5690	8	1	computer	computer	NOUN
easat-5690	8	2	vision	vision	NOUN
easat-5690	8	3	technology	technology	NOUN
easat-5690	8	4	and	and	CCONJ
easat-5690	8	5	artificial	artificial	ADJ
easat-5690	8	6	intelligence	intelligence	NOUN
easat-5690	8	7	can	can	AUX
easat-5690	8	8	automatically	automatically	ADV
easat-5690	8	9	detect	detect	VERB
easat-5690	8	10	diseases	disease	NOUN
easat-5690	8	11	and	and	CCONJ
easat-5690	8	12	control	control	VERB
easat-5690	8	13	their	their	PRON
easat-5690	8	14	spread	spread	NOUN
easat-5690	8	15	in	in	ADP
easat-5690	8	16	tea	tea	NOUN
easat-5690	8	17	leaves	leave	NOUN
easat-5690	8	18	in	in	ADP
easat-5690	8	19	real	real	ADJ
easat-5690	8	20	time	time	NOUN
easat-5690	8	21	.	.	PUNCT
easat-5690	9	1	this	this	DET
easat-5690	9	2	study	study	NOUN
easat-5690	9	3	suggested	suggest	VERB
easat-5690	9	4	a	a	DET
easat-5690	9	5	deep	deep	ADJ
easat-5690	9	6	learning	learning	NOUN
easat-5690	9	7	-	-	PUNCT
easat-5690	9	8	based	base	VERB
easat-5690	9	9	version	version	NOUN
easat-5690	9	10	of	of	ADP
easat-5690	9	11	the	the	DET
easat-5690	9	12	yolov5	yolov5	NOUN
easat-5690	9	13	algorithm	algorithm	NOUN
easat-5690	9	14	.	.	PUNCT
easat-5690	10	1	it	it	PRON
easat-5690	10	2	then	then	ADV
easat-5690	10	3	improved	improve	VERB
easat-5690	10	4	the	the	DET
easat-5690	10	5	model	model	NOUN
easat-5690	10	6	structure	structure	NOUN
easat-5690	10	7	of	of	ADP
easat-5690	10	8	yolov5	yolov5	NOUN
easat-5690	10	9	and	and	CCONJ
easat-5690	10	10	integrated	integrate	VERB
easat-5690	10	11	omni	omni	ADJ
easat-5690	10	12	dimensional	dimensional	ADJ
easat-5690	10	13	dynamic	dynamic	ADJ
easat-5690	10	14	convolution	convolution	NOUN
easat-5690	10	15	(	(	PUNCT
easat-5690	10	16	odconv	odconv	NOUN
easat-5690	10	17	)	)	PUNCT
easat-5690	10	18	and	and	CCONJ
easat-5690	10	19	convolutional	convolutional	ADJ
easat-5690	10	20	block	block	NOUN
easat-5690	10	21	attention	attention	NOUN
easat-5690	10	22	module	module	NOUN
easat-5690	10	23	(	(	PUNCT
easat-5690	10	24	cbam	cbam	NOUN
easat-5690	10	25	)	)	PUNCT
easat-5690	10	26	attention	attention	NOUN
easat-5690	10	27	mechanisms	mechanism	NOUN
easat-5690	10	28	into	into	ADP
easat-5690	10	29	yolov5	yolov5	NOUN
easat-5690	10	30	.	.	PUNCT
easat-5690	11	1	the	the	DET
easat-5690	11	2	findings	finding	NOUN
easat-5690	11	3	demonstrate	demonstrate	VERB
easat-5690	11	4	that	that	SCONJ
easat-5690	11	5	the	the	DET
easat-5690	11	6	enhanced	enhanced	ADJ
easat-5690	11	7	yolov5	yolov5	PROPN
easat-5690	11	8	's	's	PART
easat-5690	11	9	accuracy	accuracy	NOUN
easat-5690	11	10	is	be	AUX
easat-5690	11	11	better	well	ADJ
easat-5690	11	12	than	than	ADP
easat-5690	11	13	that	that	PRON
easat-5690	11	14	of	of	ADP
easat-5690	11	15	previous	previous	ADJ
easat-5690	11	16	approaches	approach	NOUN
easat-5690	11	17	and	and	CCONJ
easat-5690	11	18	5.06	5.06	NUM
easat-5690	11	19	%	%	NOUN
easat-5690	11	20	greater	great	ADJ
easat-5690	11	21	than	than	SCONJ
easat-5690	11	22	it	it	PRON
easat-5690	11	23	was	be	AUX
easat-5690	11	24	prior	prior	ADV
easat-5690	11	25	to	to	ADP
easat-5690	11	26	the	the	DET
easat-5690	11	27	enhancement	enhancement	NOUN
easat-5690	11	28	.	.	PUNCT
easat-5690	12	1	using	use	VERB
easat-5690	12	2	the	the	DET
easat-5690	12	3	improved	improved	ADJ
easat-5690	12	4	yolov5	yolov5	NOUN
easat-5690	12	5	detection	detection	NOUN
easat-5690	12	6	algorithm	algorithm	NOUN
easat-5690	12	7	to	to	PART
easat-5690	12	8	achieve	achieve	VERB
easat-5690	12	9	a	a	DET
easat-5690	12	10	high	high	ADJ
easat-5690	12	11	-	-	PUNCT
easat-5690	12	12	precision	precision	NOUN
easat-5690	12	13	model	model	NOUN
easat-5690	12	14	for	for	ADP
easat-5690	12	15	identifying	identify	VERB
easat-5690	12	16	tea	tea	NOUN
easat-5690	12	17	pests	pest	NOUN
easat-5690	12	18	and	and	CCONJ
easat-5690	12	19	diseases	disease	NOUN
easat-5690	12	20	and	and	CCONJ
easat-5690	12	21	combining	combine	VERB
easat-5690	12	22	it	it	PRON
easat-5690	12	23	with	with	ADP
easat-5690	12	24	the	the	DET
easat-5690	12	25	pyside6	pyside6	PROPN
easat-5690	12	26	library	library	NOUN
easat-5690	12	27	to	to	PART
easat-5690	12	28	design	design	VERB
easat-5690	12	29	an	an	DET
easat-5690	12	30	interface	interface	NOUN
easat-5690	12	31	recognition	recognition	NOUN
easat-5690	12	32	system	system	NOUN
easat-5690	12	33	,	,	PUNCT
easat-5690	12	34	the	the	DET
easat-5690	12	35	development	development	NOUN
easat-5690	12	36	of	of	ADP
easat-5690	12	37	the	the	DET
easat-5690	12	38	target	target	NOUN
easat-5690	12	39	detection	detection	NOUN
easat-5690	12	40	recognition	recognition	NOUN
easat-5690	12	41	page	page	NOUN
easat-5690	12	42	is	be	AUX
easat-5690	12	43	completed	complete	VERB
easat-5690	12	44	.	.	PUNCT
easat-5690	13	1	in	in	ADP
easat-5690	13	2	conclusion	conclusion	NOUN
easat-5690	13	3	,	,	PUNCT
easat-5690	13	4	this	this	DET
easat-5690	13	5	work	work	NOUN
easat-5690	13	6	offers	offer	VERB
easat-5690	13	7	a	a	DET
easat-5690	13	8	useful	useful	ADJ
easat-5690	13	9	deep	deep	ADJ
easat-5690	13	10	learning	learning	NOUN
easat-5690	13	11	-	-	PUNCT
easat-5690	13	12	based	base	VERB
easat-5690	13	13	technique	technique	NOUN
easat-5690	13	14	for	for	ADP
easat-5690	13	15	the	the	DET
easat-5690	13	16	quick	quick	ADJ
easat-5690	13	17	and	and	CCONJ
easat-5690	13	18	precise	precise	ADJ
easat-5690	13	19	diagnosis	diagnosis	NOUN
easat-5690	13	20	of	of	ADP
easat-5690	13	21	tea	tea	NOUN
easat-5690	13	22	disorders	disorder	NOUN
easat-5690	13	23	in	in	ADP
easat-5690	13	24	the	the	DET
easat-5690	13	25	area	area	NOUN
easat-5690	13	26	of	of	ADP
easat-5690	13	27	autonomous	autonomous	ADJ
easat-5690	13	28	tea	tea	NOUN
easat-5690	13	29	disease	disease	NOUN
easat-5690	13	30	detection	detection	NOUN
easat-5690	13	31	.	.	PUNCT
easat-5690	14	1	keywords	keyword	NOUN
easat-5690	14	2	:	:	PUNCT
easat-5690	14	3	aattention	aattention	NOUN
easat-5690	14	4	mechanisms	mechanism	NOUN
easat-5690	14	5	,	,	PUNCT
easat-5690	14	6	deep	deep	ADJ
easat-5690	14	7	learning	learning	NOUN
easat-5690	14	8	,	,	PUNCT
easat-5690	14	9	smart	smart	ADJ
easat-5690	14	10	agriculture	agriculture	NOUN
easat-5690	14	11	,	,	PUNCT
easat-5690	14	12	tea	tea	NOUN
easat-5690	14	13	pests	pest	NOUN
easat-5690	14	14	and	and	CCONJ
easat-5690	14	15	diseases	disease	NOUN
easat-5690	14	16	.	.	PUNCT
easat-5690	15	1	1	1	X
easat-5690	15	2	.	.	X
easat-5690	15	3	introduction	introduction	NOUN
easat-5690	15	4	people	people	NOUN
easat-5690	15	5	everywhere	everywhere	ADV
easat-5690	15	6	all	all	PRON
easat-5690	15	7	like	like	VERB
easat-5690	15	8	to	to	PART
easat-5690	15	9	drink	drink	VERB
easat-5690	15	10	tea	tea	NOUN
easat-5690	15	11	,	,	PUNCT
easat-5690	15	12	which	which	PRON
easat-5690	15	13	is	be	AUX
easat-5690	15	14	extremely	extremely	ADV
easat-5690	15	15	health	health	NOUN
easat-5690	15	16	beneficial	beneficial	ADJ
easat-5690	15	17	.	.	PUNCT
easat-5690	16	1	in	in	ADP
easat-5690	16	2	china	china	PROPN
easat-5690	16	3	,	,	PUNCT
easat-5690	16	4	tea	tea	NOUN
easat-5690	16	5	is	be	AUX
easat-5690	16	6	one	one	NUM
easat-5690	16	7	of	of	ADP
easat-5690	16	8	the	the	DET
easat-5690	16	9	traditional	traditional	ADJ
easat-5690	16	10	beverages	beverage	NOUN
easat-5690	17	1	[	[	X
easat-5690	17	2	1	1	NUM
easat-5690	17	3	]	]	PUNCT
easat-5690	17	4	.	.	PUNCT
easat-5690	18	1	therefore	therefore	ADV
easat-5690	18	2	,	,	PUNCT
easat-5690	18	3	tea	tea	NOUN
easat-5690	18	4	is	be	AUX
easat-5690	18	5	an	an	DET
easat-5690	18	6	important	important	ADJ
easat-5690	18	7	cash	cash	NOUN
easat-5690	18	8	crop	crop	NOUN
easat-5690	18	9	,	,	PUNCT
easat-5690	18	10	and	and	CCONJ
easat-5690	18	11	tea	tea	NOUN
easat-5690	18	12	trees	tree	NOUN
easat-5690	18	13	are	be	AUX
easat-5690	18	14	also	also	ADV
easat-5690	18	15	planted	plant	VERB
easat-5690	18	16	in	in	ADP
easat-5690	18	17	a	a	DET
easat-5690	18	18	large	large	ADJ
easat-5690	18	19	area	area	NOUN
easat-5690	18	20	on	on	ADP
easat-5690	18	21	the	the	DET
easat-5690	18	22	earth.however	earth.however	NOUN
easat-5690	18	23	,	,	PUNCT
easat-5690	18	24	tea	tea	NOUN
easat-5690	18	25	infections	infection	NOUN
easat-5690	18	26	can	can	AUX
easat-5690	18	27	also	also	ADV
easat-5690	18	28	lower	lower	VERB
easat-5690	18	29	tea	tea	NOUN
easat-5690	18	30	quality	quality	NOUN
easat-5690	18	31	and	and	CCONJ
easat-5690	18	32	result	result	NOUN
easat-5690	18	33	in	in	ADP
easat-5690	18	34	significant	significant	ADJ
easat-5690	18	35	financial	financial	ADJ
easat-5690	18	36	losses	loss	NOUN
easat-5690	18	37	for	for	ADP
easat-5690	18	38	tea	tea	NOUN
easat-5690	18	39	gardens	garden	NOUN
easat-5690	18	40	during	during	ADP
easat-5690	18	41	the	the	DET
easat-5690	18	42	planting	planting	NOUN
easat-5690	18	43	of	of	ADP
easat-5690	18	44	tea	tea	NOUN
easat-5690	18	45	trees	tree	NOUN
easat-5690	18	46	.	.	PUNCT
easat-5690	19	1	according	accord	VERB
easat-5690	19	2	to	to	ADP
easat-5690	19	3	statistics	statistic	NOUN
easat-5690	19	4	,	,	PUNCT
easat-5690	19	5	there	there	PRON
easat-5690	19	6	are	be	VERB
easat-5690	19	7	more	more	ADJ
easat-5690	19	8	than	than	ADP
easat-5690	19	9	900	900	NUM
easat-5690	19	10	kinds	kind	NOUN
easat-5690	19	11	of	of	ADP
easat-5690	19	12	tea	tea	NOUN
easat-5690	19	13	tree	tree	NOUN
easat-5690	19	14	pests	pest	NOUN
easat-5690	19	15	and	and	CCONJ
easat-5690	19	16	diseases	disease	NOUN
easat-5690	19	17	recorded	record	VERB
easat-5690	19	18	.	.	PUNCT
easat-5690	20	1	when	when	SCONJ
easat-5690	20	2	tea	tea	NOUN
easat-5690	20	3	leaves	leave	NOUN
easat-5690	20	4	are	be	AUX
easat-5690	20	5	healthy	healthy	ADJ
easat-5690	20	6	,	,	PUNCT
easat-5690	20	7	they	they	PRON
easat-5690	20	8	have	have	VERB
easat-5690	20	9	a	a	DET
easat-5690	20	10	vivid	vivid	ADJ
easat-5690	20	11	color	color	NOUN
easat-5690	20	12	,	,	PUNCT
easat-5690	20	13	but	but	CCONJ
easat-5690	20	14	when	when	SCONJ
easat-5690	20	15	they	they	PRON
easat-5690	20	16	are	be	AUX
easat-5690	20	17	affected	affect	VERB
easat-5690	20	18	by	by	ADP
easat-5690	20	19	diseases	disease	NOUN
easat-5690	20	20	,	,	PUNCT
easat-5690	20	21	the	the	DET
easat-5690	20	22	color	color	NOUN
easat-5690	20	23	of	of	ADP
easat-5690	20	24	the	the	DET
easat-5690	20	25	tea	tea	NOUN
easat-5690	20	26	leaves	leave	NOUN
easat-5690	20	27	will	will	AUX
easat-5690	20	28	change	change	VERB
easat-5690	20	29	significantly	significantly	ADV
easat-5690	20	30	.	.	PUNCT
easat-5690	21	1	plants	plant	NOUN
easat-5690	21	2	are	be	AUX
easat-5690	21	3	susceptible	susceptible	ADJ
easat-5690	21	4	to	to	ADP
easat-5690	21	5	many	many	ADJ
easat-5690	21	6	illnesses	illness	NOUN
easat-5690	21	7	as	as	ADP
easat-5690	21	8	a	a	DET
easat-5690	21	9	result	result	NOUN
easat-5690	21	10	of	of	ADP
easat-5690	21	11	unfavorable	unfavorable	ADJ
easat-5690	21	12	seasonal	seasonal	ADJ
easat-5690	21	13	and	and	CCONJ
easat-5690	21	14	environmental	environmental	ADJ
easat-5690	21	15	factors	factor	NOUN
easat-5690	21	16	[	[	X
easat-5690	21	17	2	2	NUM
easat-5690	21	18	]	]	PUNCT
easat-5690	21	19	.	.	PUNCT
easat-5690	22	1	farmers	farmer	NOUN
easat-5690	22	2	in	in	ADP
easat-5690	22	3	the	the	DET
easat-5690	22	4	fields	field	NOUN
easat-5690	22	5	judge	judge	NOUN
easat-5690	22	6	and	and	CCONJ
easat-5690	22	7	identify	identify	VERB
easat-5690	22	8	tea	tea	NOUN
easat-5690	22	9	diseases	disease	NOUN
easat-5690	22	10	with	with	ADP
easat-5690	22	11	their	their	PRON
easat-5690	22	12	naked	naked	ADJ
easat-5690	22	13	eyes	eye	NOUN
easat-5690	22	14	and	and	CCONJ
easat-5690	22	15	past	past	ADJ
easat-5690	22	16	experience	experience	NOUN
easat-5690	22	17	[	[	X
easat-5690	22	18	3].the	3].the	DET
easat-5690	22	19	precise	precise	ADJ
easat-5690	22	20	identification	identification	NOUN
easat-5690	22	21	and	and	CCONJ
easat-5690	22	22	detection	detection	NOUN
easat-5690	22	23	of	of	ADP
easat-5690	22	24	tea	tea	NOUN
easat-5690	22	25	-	-	PUNCT
easat-5690	22	26	related	relate	VERB
easat-5690	22	27	illnesses	illness	NOUN
easat-5690	22	28	,	,	PUNCT
easat-5690	22	29	coupled	couple	VERB
easat-5690	22	30	with	with	ADP
easat-5690	22	31	the	the	DET
easat-5690	22	32	prompt	prompt	ADJ
easat-5690	22	33	adoption	adoption	NOUN
easat-5690	22	34	of	of	ADP
easat-5690	22	35	prevention	prevention	NOUN
easat-5690	22	36	and	and	CCONJ
easat-5690	22	37	control	control	NOUN
easat-5690	22	38	measures	measure	NOUN
easat-5690	22	39	,	,	PUNCT
easat-5690	22	40	holds	hold	VERB
easat-5690	22	41	immense	immense	ADJ
easat-5690	22	42	significance	significance	NOUN
easat-5690	22	43	in	in	ADP
easat-5690	22	44	minimizing	minimize	VERB
easat-5690	22	45	production	production	NOUN
easat-5690	22	46	losses	loss	NOUN
easat-5690	22	47	,	,	PUNCT
easat-5690	22	48	enhancing	enhance	VERB
easat-5690	22	49	tea	tea	NOUN
easat-5690	22	50	quality	quality	NOUN
easat-5690	22	51	,	,	PUNCT
easat-5690	22	52	and	and	CCONJ
easat-5690	22	53	augmenting	augment	VERB
easat-5690	22	54	the	the	DET
easat-5690	22	55	income	income	NOUN
easat-5690	22	56	of	of	ADP
easat-5690	22	57	tea	tea	NOUN
easat-5690	22	58	farmers	farmer	NOUN
easat-5690	23	1	[	[	X
easat-5690	23	2	4].computer	4].computer	ADJ
easat-5690	23	3	image	image	NOUN
easat-5690	23	4	processing	processing	NOUN
easat-5690	23	5	technology	technology	NOUN
easat-5690	23	6	has	have	AUX
easat-5690	23	7	been	be	AUX
easat-5690	23	8	widely	widely	ADV
easat-5690	23	9	used	use	VERB
easat-5690	23	10	in	in	ADP
easat-5690	23	11	recent	recent	ADJ
easat-5690	23	12	years	year	NOUN
easat-5690	23	13	to	to	PART
easat-5690	23	14	address	address	VERB
easat-5690	23	15	a	a	DET
easat-5690	23	16	variety	variety	NOUN
easat-5690	23	17	of	of	ADP
easat-5690	23	18	agricultural	agricultural	ADJ
easat-5690	23	19	science	science	NOUN
easat-5690	23	20	issues	issue	NOUN
easat-5690	23	21	due	due	ADP
easat-5690	23	22	to	to	ADP
easat-5690	23	23	the	the	DET
easat-5690	23	24	quick	quick	ADJ
easat-5690	23	25	development	development	NOUN
easat-5690	23	26	of	of	ADP
easat-5690	23	27	smart	smart	ADJ
easat-5690	23	28	and	and	CCONJ
easat-5690	23	29	precision	precision	NOUN
easat-5690	23	30	agriculture	agriculture	NOUN
easat-5690	23	31	[	[	X
easat-5690	23	32	5].many	5].many	NUM
easat-5690	23	33	machine	machine	NOUN
easat-5690	23	34	learning	learn	VERB
easat-5690	23	35	algorithms	algorithm	NOUN
easat-5690	23	36	have	have	AUX
easat-5690	23	37	been	be	AUX
easat-5690	23	38	used	use	VERB
easat-5690	23	39	for	for	ADP
easat-5690	23	40	tea	tea	NOUN
easat-5690	23	41	leaf	leaf	NOUN
easat-5690	23	42	disease	disease	NOUN
easat-5690	23	43	classification	classification	NOUN
easat-5690	23	44	[	[	X
easat-5690	23	45	6].deep	6].deep	NUM
easat-5690	23	46	learning	learning	NOUN
easat-5690	23	47	is	be	AUX
easat-5690	23	48	a	a	DET
easat-5690	23	49	branch	branch	NOUN
easat-5690	23	50	of	of	ADP
easat-5690	23	51	machine	machine	NOUN
easat-5690	23	52	learning	learning	NOUN
easat-5690	23	53	that	that	PRON
easat-5690	23	54	involves	involve	VERB
easat-5690	23	55	writing	write	VERB
easat-5690	23	56	algorithms	algorithm	NOUN
easat-5690	23	57	that	that	PRON
easat-5690	23	58	mimic	mimic	VERB
easat-5690	23	59	human	human	ADJ
easat-5690	23	60	brain	brain	NOUN
easat-5690	23	61	functions	function	NOUN
easat-5690	23	62	.	.	PUNCT
easat-5690	24	1	the	the	DET
easat-5690	24	2	convolutional	convolutional	ADJ
easat-5690	24	3	neural	neural	ADJ
easat-5690	24	4	network	network	NOUN
easat-5690	24	5	(	(	PUNCT
easat-5690	24	6	cnn	cnn	PROPN
easat-5690	24	7	)	)	PUNCT
easat-5690	24	8	is	be	AUX
easat-5690	24	9	now	now	ADV
easat-5690	24	10	the	the	DET
easat-5690	24	11	most	most	ADV
easat-5690	24	12	widely	widely	ADV
easat-5690	24	13	used	use	VERB
easat-5690	24	14	deep	deep	ADJ
easat-5690	24	15	learning	learning	NOUN
easat-5690	24	16	technique	technique	NOUN
easat-5690	24	17	[	[	X
easat-5690	24	18	7].deep	7].deep	NUM
easat-5690	24	19	learning	learning	NOUN
easat-5690	24	20	(	(	PUNCT
easat-5690	24	21	dl	dl	NOUN
easat-5690	24	22	)	)	PUNCT
easat-5690	24	23	has	have	AUX
easat-5690	24	24	become	become	VERB
easat-5690	24	25	more	more	ADV
easat-5690	24	26	significant	significant	ADJ
easat-5690	24	27	because	because	SCONJ
easat-5690	24	28	of	of	ADP
easat-5690	24	29	its	its	PRON
easat-5690	24	30	powerful	powerful	ADJ
easat-5690	24	31	computing	computing	NOUN
easat-5690	24	32	capabilities	capability	NOUN
easat-5690	24	33	[	[	X
easat-5690	24	34	8	8	NUM
easat-5690	24	35	]	]	PUNCT
easat-5690	24	36	.	.	PUNCT
easat-5690	25	1	among	among	ADP
easat-5690	25	2	conventional	conventional	ADJ
easat-5690	25	3	techniques	technique	NOUN
easat-5690	25	4	,	,	PUNCT
easat-5690	25	5	deep	deep	ADJ
easat-5690	25	6	learning	learning	NOUN
easat-5690	25	7	is	be	AUX
easat-5690	25	8	distinct	distinct	ADJ
easat-5690	25	9	because	because	SCONJ
easat-5690	25	10	it	it	PRON
easat-5690	25	11	eliminates	eliminate	VERB
easat-5690	25	12	the	the	DET
easat-5690	25	13	need	need	NOUN
easat-5690	25	14	for	for	ADP
easat-5690	25	15	a	a	DET
easat-5690	25	16	laborious	laborious	ADJ
easat-5690	25	17	manual	manual	ADJ
easat-5690	25	18	feature	feature	NOUN
easat-5690	25	19	extraction	extraction	NOUN
easat-5690	25	20	procedure	procedure	NOUN
easat-5690	25	21	and	and	CCONJ
easat-5690	25	22	enables	enable	VERB
easat-5690	25	23	the	the	DET
easat-5690	25	24	network	network	NOUN
easat-5690	25	25	to	to	PART
easat-5690	25	26	learn	learn	VERB
easat-5690	25	27	discriminative	discriminative	NOUN
easat-5690	25	28	features	feature	NOUN
easat-5690	25	29	straight	straight	ADV
easat-5690	25	30	from	from	ADP
easat-5690	25	31	pixel	pixel	PROPN
easat-5690	25	32	intensities	intensity	NOUN
easat-5690	26	1	[	[	X
easat-5690	26	2	9].to	9].to	NUM
easat-5690	26	3	get	get	VERB
easat-5690	26	4	highlevel	highlevel	ADJ
easat-5690	26	5	properties	property	NOUN
easat-5690	26	6	,	,	PUNCT
easat-5690	26	7	deep	deep	ADJ
easat-5690	26	8	learning	learning	NOUN
easat-5690	26	9	models	model	NOUN
easat-5690	26	10	employ	employ	VERB
easat-5690	26	11	many	many	ADJ
easat-5690	26	12	layers	layer	NOUN
easat-5690	26	13	that	that	PRON
easat-5690	26	14	process	process	VERB
easat-5690	26	15	data	datum	NOUN
easat-5690	26	16	through	through	ADP
easat-5690	26	17	nonlinear	nonlinear	ADJ
easat-5690	26	18	1778	1778	NUM
easat-5690	26	19	edelweiss	edelweiss	PROPN
easat-5690	26	20	applied	apply	VERB
easat-5690	26	21	science	science	NOUN
easat-5690	26	22	and	and	CCONJ
easat-5690	26	23	technology	technology	NOUN
easat-5690	26	24	issn	issn	PROPN
easat-5690	26	25	:	:	PUNCT
easat-5690	26	26	2576	2576	NUM
easat-5690	26	27	-	-	SYM
easat-5690	26	28	8484	8484	NUM
easat-5690	26	29	vol	vol	NOUN
easat-5690	26	30	.	.	PROPN
easat-5690	27	1	9	9	NUM
easat-5690	27	2	,	,	PUNCT
easat-5690	27	3	no	no	INTJ
easat-5690	27	4	.	.	NOUN
easat-5690	28	1	3	3	NUM
easat-5690	28	2	:	:	SYM
easat-5690	28	3	1777	1777	NUM
easat-5690	28	4	-	-	SYM
easat-5690	28	5	1794	1794	NUM
easat-5690	28	6	,	,	PUNCT
easat-5690	28	7	2025	2025	NUM
easat-5690	28	8	doi	doi	NOUN
easat-5690	28	9	:	:	PUNCT
easat-5690	28	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	28	11	©	©	PROPN
easat-5690	28	12	2025	2025	NUM
easat-5690	28	13	by	by	ADP
easat-5690	28	14	the	the	DET
easat-5690	28	15	authors	author	NOUN
easat-5690	28	16	;	;	PUNCT
easat-5690	28	17	licensee	licensee	PROPN
easat-5690	28	18	learning	learn	VERB
easat-5690	28	19	gate	gate	NOUN
easat-5690	28	20	transformations	transformation	NOUN
easat-5690	28	21	[	[	X
easat-5690	28	22	10	10	NUM
easat-5690	28	23	]	]	PUNCT
easat-5690	28	24	.	.	PUNCT
easat-5690	29	1	deep	deep	ADJ
easat-5690	29	2	learning	learn	VERB
easat-5690	29	3	algorithms	algorithm	NOUN
easat-5690	29	4	are	be	AUX
easat-5690	29	5	highly	highly	ADV
easat-5690	29	6	accurate	accurate	ADJ
easat-5690	29	7	and	and	CCONJ
easat-5690	29	8	have	have	VERB
easat-5690	29	9	layers	layer	NOUN
easat-5690	29	10	with	with	ADP
easat-5690	29	11	phases	phase	NOUN
easat-5690	29	12	for	for	ADP
easat-5690	29	13	feature	feature	NOUN
easat-5690	29	14	extraction	extraction	NOUN
easat-5690	29	15	[	[	X
easat-5690	29	16	11	11	NUM
easat-5690	29	17	]	]	PUNCT
easat-5690	29	18	.	.	PUNCT
easat-5690	30	1	in	in	ADP
easat-5690	30	2	the	the	DET
easat-5690	30	3	study	study	NOUN
easat-5690	30	4	,	,	PUNCT
easat-5690	30	5	an	an	DET
easat-5690	30	6	automatic	automatic	ADJ
easat-5690	30	7	system	system	NOUN
easat-5690	30	8	was	be	AUX
easat-5690	30	9	developed	develop	VERB
easat-5690	30	10	using	use	VERB
easat-5690	30	11	support	support	NOUN
easat-5690	30	12	vector	vector	NOUN
easat-5690	30	13	machine	machine	NOUN
easat-5690	30	14	(	(	PUNCT
easat-5690	30	15	svm	svm	PROPN
easat-5690	30	16	)	)	PUNCT
easat-5690	30	17	to	to	PART
easat-5690	30	18	detect	detect	VERB
easat-5690	30	19	three	three	NUM
easat-5690	30	20	different	different	ADJ
easat-5690	30	21	types	type	NOUN
easat-5690	30	22	of	of	ADP
easat-5690	30	23	tea	tea	NOUN
easat-5690	30	24	diseases	disease	NOUN
easat-5690	30	25	with	with	ADP
easat-5690	30	26	a	a	DET
easat-5690	30	27	rather	rather	ADV
easat-5690	30	28	limited	limited	ADJ
easat-5690	30	29	set	set	NOUN
easat-5690	30	30	of	of	ADP
easat-5690	30	31	characteristics[3	characteristics[3	PROPN
easat-5690	30	32	]	]	PUNCT
easat-5690	30	33	.	.	PUNCT
easat-5690	31	1	researchers	researcher	NOUN
easat-5690	31	2	proposed	propose	VERB
easat-5690	31	3	the	the	DET
easat-5690	31	4	use	use	NOUN
easat-5690	31	5	of	of	ADP
easat-5690	31	6	convolutional	convolutional	ADJ
easat-5690	31	7	neural	neural	ADJ
easat-5690	31	8	networks	network	NOUN
easat-5690	31	9	(	(	PUNCT
easat-5690	31	10	cnn	cnn	PROPN
easat-5690	31	11	)	)	PUNCT
easat-5690	31	12	for	for	ADP
easat-5690	31	13	the	the	DET
easat-5690	31	14	detection	detection	NOUN
easat-5690	31	15	of	of	ADP
easat-5690	31	16	tea	tea	NOUN
easat-5690	31	17	diseases	disease	NOUN
easat-5690	31	18	.	.	PUNCT
easat-5690	32	1	konstantinos	konstantinos	PROPN
easat-5690	32	2	p.	p.	PROPN
easat-5690	32	3	ferentinos[12]used	ferentinos[12]use	VERB
easat-5690	32	4	deep	deep	ADJ
easat-5690	32	5	learning	learning	NOUN
easat-5690	32	6	models	model	NOUN
easat-5690	32	7	,	,	PUNCT
easat-5690	32	8	vgg	vgg	PROPN
easat-5690	32	9	(	(	PUNCT
easat-5690	32	10	highest	high	ADJ
easat-5690	32	11	success	success	NOUN
easat-5690	32	12	rate	rate	NOUN
easat-5690	32	13	)	)	PUNCT
easat-5690	32	14	and	and	CCONJ
easat-5690	32	15	alexnet	alexnet	PROPN
easat-5690	32	16	owtbn	owtbn	NOUN
easat-5690	32	17	architectures	architecture	NOUN
easat-5690	32	18	(	(	PUNCT
easat-5690	32	19	lowest	low	ADJ
easat-5690	32	20	final	final	ADJ
easat-5690	32	21	average	average	ADJ
easat-5690	32	22	testing	testing	NOUN
easat-5690	32	23	error),for	error),for	ADP
easat-5690	32	24	plant	plant	NOUN
easat-5690	32	25	disease	disease	NOUN
easat-5690	32	26	detection	detection	NOUN
easat-5690	32	27	and	and	CCONJ
easat-5690	32	28	diagnosis	diagnosis	NOUN
easat-5690	32	29	.	.	PUNCT
easat-5690	33	1	the	the	DET
easat-5690	33	2	system	system	NOUN
easat-5690	33	3	uses	use	VERB
easat-5690	33	4	three	three	NUM
easat-5690	33	5	models	model	NOUN
easat-5690	33	6	,	,	PUNCT
easat-5690	33	7	two	two	NUM
easat-5690	33	8	of	of	ADP
easat-5690	33	9	which	which	PRON
easat-5690	33	10	are	be	AUX
easat-5690	33	11	pre	pre	ADJ
easat-5690	33	12	-	-	VERB
easat-5690	33	13	trained	trained	ADJ
easat-5690	33	14	.	.	PUNCT
easat-5690	34	1	they	they	PRON
easat-5690	34	2	are	be	AUX
easat-5690	34	3	faster	fast	ADJ
easat-5690	34	4	rcnn	rcnn	PROPN
easat-5690	34	5	(	(	PUNCT
easat-5690	34	6	inception	inception	NOUN
easat-5690	34	7	-	-	PUNCT
easat-5690	34	8	v2	v2	NOUN
easat-5690	34	9	)	)	PUNCT
easat-5690	34	10	and	and	CCONJ
easat-5690	34	11	vgg16	vgg16	PROPN
easat-5690	34	12	.	.	PUNCT
easat-5690	35	1	the	the	DET
easat-5690	35	2	other	other	ADJ
easat-5690	35	3	one	one	NOUN
easat-5690	35	4	is	be	AUX
easat-5690	35	5	a	a	DET
easat-5690	35	6	manually	manually	ADV
easat-5690	35	7	trained	train	VERB
easat-5690	35	8	model	model	NOUN
easat-5690	35	9	,	,	PUNCT
easat-5690	35	10	specifically	specifically	ADV
easat-5690	35	11	a	a	DET
easat-5690	35	12	sequential	sequential	ADJ
easat-5690	35	13	model	model	NOUN
easat-5690	35	14	or	or	CCONJ
easat-5690	35	15	cnn[13].in	cnn[13].in	NOUN
easat-5690	35	16	this	this	DET
easat-5690	35	17	study	study	NOUN
easat-5690	35	18	,	,	PUNCT
easat-5690	35	19	the	the	DET
easat-5690	35	20	model	model	NOUN
easat-5690	35	21	based	base	VERB
easat-5690	35	22	on	on	ADP
easat-5690	35	23	a	a	DET
easat-5690	35	24	faster	fast	ADJ
easat-5690	35	25	region	region	NOUN
easat-5690	35	26	-	-	PUNCT
easat-5690	35	27	based	base	VERB
easat-5690	35	28	convolutional	convolutional	ADJ
easat-5690	35	29	neural	neural	ADJ
easat-5690	35	30	network	network	NOUN
easat-5690	35	31	(	(	PUNCT
easat-5690	35	32	faster	fast	ADJ
easat-5690	35	33	r	r	X
easat-5690	35	34	-	-	PUNCT
easat-5690	35	35	cnn	cnn	NOUN
easat-5690	35	36	)	)	PUNCT
easat-5690	35	37	is	be	AUX
easat-5690	35	38	trained	train	VERB
easat-5690	35	39	using	use	VERB
easat-5690	35	40	over	over	ADP
easat-5690	35	41	1000	1000	NUM
easat-5690	35	42	photos	photo	NOUN
easat-5690	35	43	of	of	ADP
easat-5690	35	44	tea	tea	NOUN
easat-5690	35	45	leaves	leave	NOUN
easat-5690	35	46	[	[	X
easat-5690	35	47	14].this	14].this	NUM
easat-5690	35	48	study	study	NOUN
easat-5690	35	49	presented	present	VERB
easat-5690	35	50	ax	ax	NOUN
easat-5690	35	51	-	-	PUNCT
easat-5690	35	52	retinanet	retinanet	NOUN
easat-5690	35	53	,	,	PUNCT
easat-5690	35	54	an	an	DET
easat-5690	35	55	improved	improved	ADJ
easat-5690	35	56	variant	variant	NOUN
easat-5690	35	57	of	of	ADP
easat-5690	35	58	retinanet	retinanet	NOUN
easat-5690	35	59	,	,	PUNCT
easat-5690	35	60	as	as	ADP
easat-5690	35	61	an	an	DET
easat-5690	35	62	enhanced	enhanced	ADJ
easat-5690	35	63	target	target	NOUN
easat-5690	35	64	detection	detection	NOUN
easat-5690	35	65	and	and	CCONJ
easat-5690	35	66	recognition	recognition	NOUN
easat-5690	35	67	network	network	NOUN
easat-5690	35	68	for	for	ADP
easat-5690	35	69	automatically	automatically	ADV
easat-5690	35	70	detecting	detect	VERB
easat-5690	35	71	and	and	CCONJ
easat-5690	35	72	recognizing	recognize	VERB
easat-5690	35	73	tea	tea	NOUN
easat-5690	35	74	illnesses	illness	NOUN
easat-5690	35	75	in	in	ADP
easat-5690	35	76	natural	natural	ADJ
easat-5690	35	77	scene	scene	NOUN
easat-5690	35	78	photos	photo	NOUN
easat-5690	35	79	[	[	X
easat-5690	35	80	4	4	NUM
easat-5690	35	81	]	]	PUNCT
easat-5690	35	82	.	.	PUNCT
easat-5690	36	1	in	in	ADP
easat-5690	36	2	order	order	NOUN
easat-5690	36	3	to	to	PART
easat-5690	36	4	identify	identify	VERB
easat-5690	36	5	tomato	tomato	NOUN
easat-5690	36	6	diseases	disease	NOUN
easat-5690	36	7	and	and	CCONJ
easat-5690	36	8	pinpoint	pinpoint	VERB
easat-5690	36	9	the	the	DET
easat-5690	36	10	areas	area	NOUN
easat-5690	36	11	of	of	ADP
easat-5690	36	12	damaged	damage	VERB
easat-5690	36	13	tomato	tomato	NOUN
easat-5690	36	14	leaves	leave	NOUN
easat-5690	36	15	,	,	PUNCT
easat-5690	36	16	this	this	DET
easat-5690	36	17	study	study	NOUN
easat-5690	36	18	enhanced	enhance	VERB
easat-5690	36	19	the	the	DET
easat-5690	36	20	faster	fast	ADJ
easat-5690	36	21	rcnn	rcnn	PROPN
easat-5690	36	22	algorithm	algorithm	NOUN
easat-5690	37	1	[	[	X
easat-5690	37	2	15].with	15].with	PROPN
easat-5690	37	3	the	the	DET
easat-5690	37	4	objective	objective	NOUN
easat-5690	37	5	of	of	ADP
easat-5690	37	6	enhancing	enhance	VERB
easat-5690	37	7	the	the	DET
easat-5690	37	8	detection	detection	NOUN
easat-5690	37	9	performance	performance	NOUN
easat-5690	37	10	for	for	ADP
easat-5690	37	11	blurry	blurry	ADJ
easat-5690	37	12	,	,	PUNCT
easat-5690	37	13	occluded	occluded	ADJ
easat-5690	37	14	,	,	PUNCT
easat-5690	37	15	and	and	CCONJ
easat-5690	37	16	small	small	ADJ
easat-5690	37	17	diseased	diseased	ADJ
easat-5690	37	18	leaves	leave	NOUN
easat-5690	37	19	,	,	PUNCT
easat-5690	37	20	a	a	DET
easat-5690	37	21	deep	deep	ADJ
easat-5690	37	22	learning	learning	NOUN
easat-5690	37	23	framework	framework	NOUN
easat-5690	37	24	known	know	VERB
easat-5690	37	25	as	as	ADP
easat-5690	37	26	"	"	PUNCT
easat-5690	37	27	faster	fast	ADV
easat-5690	37	28	regionbased	regionbase	VERB
easat-5690	37	29	convolutional	convolutional	ADJ
easat-5690	37	30	neural	neural	ADJ
easat-5690	37	31	network	network	NOUN
easat-5690	37	32	"	"	PUNCT
easat-5690	37	33	is	be	AUX
easat-5690	37	34	employed	employ	VERB
easat-5690	37	35	for	for	ADP
easat-5690	37	36	detecting	detect	VERB
easat-5690	37	37	tlb	tlb	NOUN
easat-5690	37	38	(	(	PUNCT
easat-5690	37	39	tea	tea	NOUN
easat-5690	37	40	leaf	leaf	NOUN
easat-5690	37	41	blight	blight	NOUN
easat-5690	37	42	)	)	PUNCT
easat-5690	38	1	leaves[16].to	leaves[16].to	PUNCT
easat-5690	38	2	recognize	recognize	VERB
easat-5690	38	3	tree	tree	NOUN
easat-5690	38	4	leaves	leave	NOUN
easat-5690	38	5	,	,	PUNCT
easat-5690	38	6	two	two	NUM
easat-5690	38	7	deep	deep	ADJ
easat-5690	38	8	learning	learning	NOUN
easat-5690	38	9	models	model	NOUN
easat-5690	38	10	—	—	PUNCT
easat-5690	38	11	faster	fast	ADJ
easat-5690	38	12	r	r	NOUN
easat-5690	38	13	-	-	PUNCT
easat-5690	38	14	cnn	cnn	PROPN
easat-5690	38	15	and	and	CCONJ
easat-5690	38	16	yolov5	yolov5	NOUN
easat-5690	38	17	—	—	PUNCT
easat-5690	38	18	representing	represent	VERB
easat-5690	38	19	twostage	twostage	NOUN
easat-5690	38	20	and	and	CCONJ
easat-5690	38	21	one	one	NUM
easat-5690	38	22	-	-	PUNCT
easat-5690	38	23	stage	stage	NOUN
easat-5690	38	24	techniques	technique	NOUN
easat-5690	38	25	are	be	AUX
easat-5690	38	26	used	use	VERB
easat-5690	38	27	[	[	PUNCT
easat-5690	38	28	17].the	17].the	NUM
easat-5690	38	29	bifpn	bifpn	PROPN
easat-5690	38	30	feature	feature	NOUN
easat-5690	38	31	fusion	fusion	NOUN
easat-5690	38	32	network	network	NOUN
easat-5690	38	33	and	and	CCONJ
easat-5690	38	34	adaptive	adaptive	ADJ
easat-5690	38	35	spatial	spatial	ADJ
easat-5690	38	36	feature	feature	NOUN
easat-5690	38	37	fusion	fusion	NOUN
easat-5690	38	38	(	(	PUNCT
easat-5690	38	39	asff	asff	NOUN
easat-5690	38	40	)	)	PUNCT
easat-5690	38	41	technologies	technology	NOUN
easat-5690	38	42	have	have	AUX
easat-5690	38	43	been	be	AUX
easat-5690	38	44	used	use	VERB
easat-5690	38	45	to	to	PART
easat-5690	38	46	improve	improve	VERB
easat-5690	38	47	the	the	DET
easat-5690	38	48	multi	multi	ADJ
easat-5690	38	49	-	-	ADJ
easat-5690	38	50	scale	scale	ADJ
easat-5690	38	51	feature	feature	NOUN
easat-5690	38	52	fusion	fusion	NOUN
easat-5690	38	53	of	of	ADP
easat-5690	38	54	tea	tea	NOUN
easat-5690	38	55	illnesses	illness	NOUN
easat-5690	38	56	,	,	PUNCT
easat-5690	38	57	increasing	increase	VERB
easat-5690	38	58	the	the	DET
easat-5690	38	59	model	model	NOUN
easat-5690	38	60	's	's	PART
easat-5690	38	61	resistance	resistance	NOUN
easat-5690	38	62	to	to	PART
easat-5690	38	63	interference	interference	VERB
easat-5690	38	64	from	from	ADP
easat-5690	38	65	complicated	complicated	ADJ
easat-5690	38	66	backgrounds	background	NOUN
easat-5690	38	67	lin	lin	PROPN
easat-5690	38	68	,	,	PUNCT
easat-5690	38	69	et	et	PROPN
easat-5690	38	70	al	al	PROPN
easat-5690	38	71	.	.	PUNCT
easat-5690	39	1	[	[	X
easat-5690	39	2	18	18	NUM
easat-5690	39	3	]	]	PUNCT
easat-5690	39	4	.	.	PUNCT
easat-5690	40	1	z.	z.	PROPN
easat-5690	40	2	xue	xue	PROPN
easat-5690	40	3	,	,	PUNCT
easat-5690	40	4	r.	r.	PROPN
easat-5690	40	5	xu	xu	PROPN
easat-5690	40	6	,	,	PUNCT
easat-5690	40	7	d.	d.	PROPN
easat-5690	40	8	bai	bai	PROPN
easat-5690	40	9	and	and	CCONJ
easat-5690	40	10	h.	h.	PROPN
easat-5690	40	11	lin	lin	PROPN
easat-5690	41	1	[	[	X
easat-5690	41	2	19	19	NUM
easat-5690	41	3	]	]	PUNCT
easat-5690	41	4	replaced	replace	VERB
easat-5690	41	5	the	the	DET
easat-5690	41	6	original	original	ADJ
easat-5690	41	7	yolov5	yolov5	NOUN
easat-5690	41	8	's	's	PART
easat-5690	41	9	spatial	spatial	ADJ
easat-5690	41	10	pyramid	pyramid	NOUN
easat-5690	41	11	pooling	pool	VERB
easat-5690	41	12	fast	fast	ADJ
easat-5690	41	13	(	(	PUNCT
easat-5690	41	14	sppf	sppf	ADJ
easat-5690	41	15	)	)	PUNCT
easat-5690	41	16	module	module	NOUN
easat-5690	41	17	with	with	ADP
easat-5690	41	18	a	a	DET
easat-5690	41	19	receptive	receptive	ADJ
easat-5690	41	20	field	field	NOUN
easat-5690	41	21	block	block	NOUN
easat-5690	41	22	(	(	PUNCT
easat-5690	41	23	rfb	rfb	PROPN
easat-5690	41	24	)	)	PUNCT
easat-5690	41	25	module	module	NOUN
easat-5690	41	26	and	and	CCONJ
easat-5690	41	27	incorporated	incorporate	VERB
easat-5690	41	28	self	self	NOUN
easat-5690	41	29	-	-	PUNCT
easat-5690	41	30	attention	attention	NOUN
easat-5690	41	31	and	and	CCONJ
easat-5690	41	32	convolution	convolution	NOUN
easat-5690	41	33	(	(	PUNCT
easat-5690	41	34	acmix	acmix	NOUN
easat-5690	41	35	)	)	PUNCT
easat-5690	41	36	and	and	CCONJ
easat-5690	41	37	convolutional	convolutional	ADJ
easat-5690	41	38	block	block	NOUN
easat-5690	41	39	attention	attention	NOUN
easat-5690	41	40	module	module	NOUN
easat-5690	41	41	(	(	PUNCT
easat-5690	41	42	cbam	cbam	NOUN
easat-5690	41	43	)	)	PUNCT
easat-5690	41	44	into	into	ADP
easat-5690	41	45	yolov5.this	yolov5.this	PROPN
easat-5690	41	46	article	article	NOUN
easat-5690	41	47	uses	use	VERB
easat-5690	41	48	the	the	DET
easat-5690	41	49	enhanced	enhanced	ADJ
easat-5690	41	50	yolov5	yolov5	NOUN
easat-5690	41	51	to	to	PART
easat-5690	41	52	identify	identify	VERB
easat-5690	41	53	peach	peach	NOUN
easat-5690	41	54	tree	tree	NOUN
easat-5690	41	55	leaf	leaf	NOUN
easat-5690	41	56	diseases	disease	NOUN
easat-5690	41	57	[	[	X
easat-5690	41	58	20].the	20].the	DET
easat-5690	41	59	original	original	ADJ
easat-5690	41	60	bottleneck	bottleneck	NOUN
easat-5690	41	61	csp	csp	PROPN
easat-5690	41	62	module	module	NOUN
easat-5690	41	63	in	in	ADP
easat-5690	41	64	the	the	DET
easat-5690	41	65	main	main	ADJ
easat-5690	41	66	trunk	trunk	NOUN
easat-5690	41	67	of	of	ADP
easat-5690	41	68	the	the	DET
easat-5690	41	69	yolov5s	yolov5s	PROPN
easat-5690	41	70	network	network	NOUN
easat-5690	41	71	architecture	architecture	NOUN
easat-5690	41	72	is	be	AUX
easat-5690	41	73	replaced	replace	VERB
easat-5690	41	74	by	by	ADP
easat-5690	41	75	the	the	DET
easat-5690	41	76	bottleneckcsp-2	bottleneckcsp-2	PROPN
easat-5690	41	77	module	module	NOUN
easat-5690	41	78	,	,	PUNCT
easat-5690	41	79	which	which	PRON
easat-5690	41	80	has	have	AUX
easat-5690	41	81	been	be	AUX
easat-5690	41	82	enhanced	enhance	VERB
easat-5690	41	83	and	and	CCONJ
easat-5690	41	84	redesigned	redesign	VERB
easat-5690	41	85	[	[	PUNCT
easat-5690	41	86	21].the	21].the	NUM
easat-5690	41	87	convolutional	convolutional	ADJ
easat-5690	41	88	neural	neural	ADJ
easat-5690	41	89	network	network	NOUN
easat-5690	41	90	's	's	PART
easat-5690	41	91	calculations	calculation	NOUN
easat-5690	41	92	are	be	AUX
easat-5690	41	93	reduced	reduce	VERB
easat-5690	41	94	in	in	ADP
easat-5690	41	95	the	the	DET
easat-5690	41	96	backbone	backbone	NOUN
easat-5690	41	97	network	network	NOUN
easat-5690	41	98	by	by	ADP
easat-5690	41	99	substituting	substitute	VERB
easat-5690	41	100	the	the	DET
easat-5690	41	101	involution	involution	NOUN
easat-5690	41	102	bottleneck	bottleneck	NOUN
easat-5690	41	103	module	module	NOUN
easat-5690	41	104	for	for	ADP
easat-5690	41	105	the	the	DET
easat-5690	41	106	bottleneck	bottleneck	NOUN
easat-5690	41	107	module	module	NOUN
easat-5690	41	108	in	in	ADP
easat-5690	41	109	the	the	DET
easat-5690	41	110	c3	c3	NOUN
easat-5690	41	111	module	module	NOUN
easat-5690	41	112	[	[	X
easat-5690	41	113	22].the	22].the	DET
easat-5690	41	114	article	article	NOUN
easat-5690	41	115	makes	make	VERB
easat-5690	41	116	a	a	DET
easat-5690	41	117	selection	selection	NOUN
easat-5690	41	118	of	of	ADP
easat-5690	41	119	the	the	DET
easat-5690	41	120	yolov5	yolov5	NOUN
easat-5690	41	121	weakly	weakly	ADV
easat-5690	41	122	supervised	supervise	VERB
easat-5690	41	123	model	model	NOUN
easat-5690	41	124	and	and	CCONJ
easat-5690	41	125	subsequently	subsequently	ADV
easat-5690	41	126	discovers	discover	VERB
easat-5690	41	127	,	,	PUNCT
easat-5690	41	128	through	through	ADP
easat-5690	41	129	rigorous	rigorous	ADJ
easat-5690	41	130	experiments	experiment	NOUN
easat-5690	41	131	,	,	PUNCT
easat-5690	41	132	that	that	PRON
easat-5690	41	133	incorporating	incorporate	VERB
easat-5690	41	134	the	the	DET
easat-5690	41	135	gam	gam	NOUN
easat-5690	41	136	attention	attention	NOUN
easat-5690	41	137	mechanism	mechanism	NOUN
easat-5690	41	138	into	into	ADP
easat-5690	41	139	the	the	DET
easat-5690	41	140	yolov5	yolov5	NOUN
easat-5690	41	141	network	network	NOUN
easat-5690	41	142	framework	framework	NOUN
easat-5690	41	143	significantly	significantly	ADV
easat-5690	41	144	enhances	enhance	VERB
easat-5690	41	145	the	the	DET
easat-5690	41	146	ability	ability	NOUN
easat-5690	41	147	to	to	PART
easat-5690	41	148	identify	identify	VERB
easat-5690	41	149	apolygus	apolygus	NOUN
easat-5690	41	150	lucorum	lucorum	PROPN
easat-5690	42	1	[	[	X
easat-5690	42	2	23].this	23].this	NUM
easat-5690	42	3	research	research	NOUN
easat-5690	42	4	proposes	propose	VERB
easat-5690	42	5	an	an	DET
easat-5690	42	6	enhanced	enhanced	ADJ
easat-5690	42	7	se	se	NOUN
easat-5690	42	8	-	-	ADJ
easat-5690	42	9	yolov5	yolov5	NOUN
easat-5690	42	10	network	network	NOUN
easat-5690	42	11	model	model	NOUN
easat-5690	42	12	for	for	ADP
easat-5690	42	13	tomato	tomato	NOUN
easat-5690	42	14	viral	viral	ADJ
easat-5690	42	15	illness	illness	NOUN
easat-5690	42	16	detection	detection	NOUN
easat-5690	42	17	[	[	X
easat-5690	42	18	24	24	NUM
easat-5690	42	19	]	]	PUNCT
easat-5690	42	20	.	.	PUNCT
easat-5690	43	1	a	a	DET
easat-5690	43	2	shallow	shallow	ADJ
easat-5690	43	3	feature	feature	NOUN
easat-5690	43	4	layer	layer	NOUN
easat-5690	43	5	is	be	AUX
easat-5690	43	6	the	the	DET
easat-5690	43	7	basis	basis	NOUN
easat-5690	43	8	for	for	ADP
easat-5690	43	9	the	the	DET
easat-5690	43	10	enhanced	enhanced	ADJ
easat-5690	43	11	yolo	yolo	ADJ
easat-5690	43	12	v5	v5	PROPN
easat-5690	43	13	algorithm[25].by	algorithm[25].by	PROPN
easat-5690	43	14	employing	employ	VERB
easat-5690	43	15	shufflenetv2	shufflenetv2	NOUN
easat-5690	43	16	as	as	ADP
easat-5690	43	17	the	the	DET
easat-5690	43	18	enhanced	enhanced	ADJ
easat-5690	43	19	backbone	backbone	NOUN
easat-5690	43	20	network	network	NOUN
easat-5690	43	21	,	,	PUNCT
easat-5690	43	22	the	the	DET
easat-5690	43	23	model	model	NOUN
easat-5690	43	24	structure	structure	NOUN
easat-5690	43	25	is	be	AUX
easat-5690	43	26	subsequently	subsequently	ADV
easat-5690	43	27	finetuned	finetune	VERB
easat-5690	43	28	for	for	ADP
easat-5690	43	29	simplification	simplification	NOUN
easat-5690	43	30	.	.	PUNCT
easat-5690	44	1	during	during	ADP
easat-5690	44	2	the	the	DET
easat-5690	44	3	feature	feature	NOUN
easat-5690	44	4	fusion	fusion	NOUN
easat-5690	44	5	phase	phase	NOUN
easat-5690	44	6	,	,	PUNCT
easat-5690	44	7	the	the	DET
easat-5690	44	8	cbam	cbam	NOUN
easat-5690	44	9	module	module	NOUN
easat-5690	44	10	is	be	AUX
easat-5690	44	11	introduced	introduce	VERB
easat-5690	44	12	to	to	PART
easat-5690	44	13	further	far	ADV
easat-5690	44	14	augment	augment	VERB
easat-5690	44	15	the	the	DET
easat-5690	44	16	effective	effective	ADJ
easat-5690	44	17	feature	feature	NOUN
easat-5690	44	18	information	information	NOUN
easat-5690	44	19	pertaining	pertain	VERB
easat-5690	44	20	to	to	AUX
easat-5690	44	21	lychees[26].a	lychees[26].a	VERB
easat-5690	44	22	multi	multi	ADJ
easat-5690	44	23	-	-	ADJ
easat-5690	44	24	attention	attention	ADJ
easat-5690	44	25	mechanism	mechanism	NOUN
easat-5690	44	26	module	module	NOUN
easat-5690	44	27	,	,	PUNCT
easat-5690	44	28	consisting	consist	VERB
easat-5690	44	29	of	of	ADP
easat-5690	44	30	the	the	DET
easat-5690	44	31	convolutional	convolutional	ADJ
easat-5690	44	32	block	block	NOUN
easat-5690	44	33	attention	attention	NOUN
easat-5690	44	34	module	module	NOUN
easat-5690	44	35	(	(	PUNCT
easat-5690	44	36	cbam	cbam	NOUN
easat-5690	44	37	)	)	PUNCT
easat-5690	44	38	and	and	CCONJ
easat-5690	44	39	efficient	efficient	ADJ
easat-5690	44	40	channel	channel	NOUN
easat-5690	44	41	attention	attention	NOUN
easat-5690	44	42	(	(	PUNCT
easat-5690	44	43	eca	eca	PROPN
easat-5690	44	44	)	)	PUNCT
easat-5690	44	45	,	,	PUNCT
easat-5690	44	46	was	be	AUX
easat-5690	44	47	placed	place	VERB
easat-5690	44	48	into	into	ADP
easat-5690	44	49	the	the	DET
easat-5690	44	50	network	network	NOUN
easat-5690	44	51	's	's	PART
easat-5690	44	52	neck	neck	NOUN
easat-5690	44	53	[	[	X
easat-5690	44	54	27].the	27].the	DET
easat-5690	44	55	proposed	propose	VERB
easat-5690	44	56	method	method	NOUN
easat-5690	44	57	incorporates	incorporate	VERB
easat-5690	44	58	an	an	DET
easat-5690	44	59	efficient	efficient	ADJ
easat-5690	44	60	channel	channel	NOUN
easat-5690	44	61	attention	attention	NOUN
easat-5690	44	62	module	module	NOUN
easat-5690	44	63	(	(	PUNCT
easat-5690	44	64	eca	eca	NOUN
easat-5690	44	65	)	)	PUNCT
easat-5690	44	66	within	within	ADP
easat-5690	44	67	the	the	DET
easat-5690	44	68	c3	c3	NOUN
easat-5690	44	69	module	module	NOUN
easat-5690	44	70	of	of	ADP
easat-5690	44	71	the	the	DET
easat-5690	44	72	yolov5s	yolov5s	PROPN
easat-5690	44	73	network	network	NOUN
easat-5690	44	74	model	model	NOUN
easat-5690	44	75	's	's	PART
easat-5690	44	76	backbone	backbone	NOUN
easat-5690	44	77	structure	structure	NOUN
easat-5690	44	78	.	.	PUNCT
easat-5690	45	1	additionally	additionally	ADV
easat-5690	45	2	,	,	PUNCT
easat-5690	45	3	a	a	DET
easat-5690	45	4	global	global	ADJ
easat-5690	45	5	attention	attention	NOUN
easat-5690	45	6	mechanism	mechanism	NOUN
easat-5690	45	7	module	module	NOUN
easat-5690	45	8	(	(	PUNCT
easat-5690	45	9	gam	gam	NOUN
easat-5690	45	10	)	)	PUNCT
easat-5690	45	11	is	be	AUX
easat-5690	45	12	interposed	interpose	VERB
easat-5690	45	13	between	between	ADP
easat-5690	45	14	the	the	DET
easat-5690	45	15	neck	neck	NOUN
easat-5690	45	16	structure	structure	NOUN
easat-5690	45	17	and	and	CCONJ
easat-5690	45	18	the	the	DET
easat-5690	45	19	head	head	NOUN
easat-5690	45	20	structure[28].in	structure[28].in	PRON
easat-5690	45	21	order	order	NOUN
easat-5690	45	22	to	to	PART
easat-5690	45	23	address	address	VERB
easat-5690	45	24	the	the	DET
easat-5690	45	25	high	high	ADJ
easat-5690	45	26	miss	miss	NOUN
easat-5690	45	27	rate	rate	NOUN
easat-5690	45	28	of	of	ADP
easat-5690	45	29	small	small	ADJ
easat-5690	45	30	targets	target	NOUN
easat-5690	45	31	in	in	ADP
easat-5690	45	32	wheat	wheat	NOUN
easat-5690	45	33	population	population	NOUN
easat-5690	45	34	photos	photo	NOUN
easat-5690	45	35	,	,	PUNCT
easat-5690	45	36	an	an	DET
easat-5690	45	37	integrated	integrate	VERB
easat-5690	45	38	transformer	transformer	ADJ
easat-5690	45	39	small	small	ADJ
easat-5690	45	40	-	-	PUNCT
easat-5690	45	41	target	target	NOUN
easat-5690	45	42	detection	detection	NOUN
easat-5690	45	43	head	head	NOUN
easat-5690	45	44	(	(	PUNCT
easat-5690	45	45	tsdh	tsdh	NOUN
easat-5690	45	46	)	)	PUNCT
easat-5690	45	47	was	be	AUX
easat-5690	45	48	added	add	VERB
easat-5690	45	49	concurrently[29].with	concurrently[29].with	ADJ
easat-5690	45	50	compact	compact	ADJ
easat-5690	45	51	,	,	PUNCT
easat-5690	45	52	dense	dense	ADJ
easat-5690	45	53	,	,	PUNCT
easat-5690	45	54	and	and	CCONJ
easat-5690	45	55	overlapping	overlap	VERB
easat-5690	45	56	crop	crop	NOUN
easat-5690	45	57	disease	disease	NOUN
easat-5690	45	58	targets	target	NOUN
easat-5690	45	59	,	,	PUNCT
easat-5690	45	60	an	an	DET
easat-5690	45	61	enhanced	enhance	VERB
easat-5690	45	62	yolov5s	yolov5s	PROPN
easat-5690	45	63	model	model	NOUN
easat-5690	45	64	was	be	AUX
easat-5690	45	65	put	put	VERB
easat-5690	45	66	forth	forth	ADP
easat-5690	45	67	to	to	PART
easat-5690	45	68	detect	detect	VERB
easat-5690	45	69	a	a	DET
easat-5690	45	70	number	number	NOUN
easat-5690	45	71	of	of	ADP
easat-5690	45	72	prevalent	prevalent	ADJ
easat-5690	45	73	crop	crop	NOUN
easat-5690	45	74	illnesses[30].a	illnesses[30].a	PROPN
easat-5690	45	75	.	.	PUNCT
easat-5690	46	1	wang	wang	PROPN
easat-5690	46	2	,	,	PUNCT
easat-5690	46	3	t.	t.	PROPN
easat-5690	46	4	peng	peng	PROPN
easat-5690	46	5	,	,	PUNCT
easat-5690	46	6	h.	h.	PROPN
easat-5690	46	7	cao	cao	PROPN
easat-5690	46	8	,	,	PUNCT
easat-5690	46	9	y.	y.	PROPN
easat-5690	46	10	xu	xu	PROPN
easat-5690	46	11	,	,	PUNCT
easat-5690	46	12	x.	x.	PROPN
easat-5690	46	13	wei	wei	PROPN
easat-5690	46	14	and	and	CCONJ
easat-5690	46	15	b.	b.	PROPN
easat-5690	46	16	cui[31]proposed	cui[31]propose	VERB
easat-5690	46	17	tia	tia	PROPN
easat-5690	46	18	yolov5	yolov5	PROPN
easat-5690	46	19	,	,	PUNCT
easat-5690	46	20	which	which	PRON
easat-5690	46	21	adds	add	VERB
easat-5690	46	22	a	a	DET
easat-5690	46	23	transformer	transformer	NOUN
easat-5690	46	24	encoder	encoder	NOUN
easat-5690	46	25	block	block	NOUN
easat-5690	46	26	to	to	ADP
easat-5690	46	27	the	the	DET
easat-5690	46	28	backbone	backbone	NOUN
easat-5690	46	29	network	network	NOUN
easat-5690	46	30	to	to	PART
easat-5690	46	31	improve	improve	VERB
easat-5690	46	32	the	the	DET
easat-5690	46	33	model	model	NOUN
easat-5690	46	34	's	's	PART
easat-5690	46	35	sensitivity	sensitivity	NOUN
easat-5690	46	36	to	to	ADP
easat-5690	46	37	weeds	weed	NOUN
easat-5690	46	38	and	and	CCONJ
easat-5690	46	39	proposes	propose	VERB
easat-5690	46	40	a	a	DET
easat-5690	46	41	channel	channel	NOUN
easat-5690	46	42	feature	feature	NOUN
easat-5690	46	43	fusion	fusion	NOUN
easat-5690	46	44	with	with	ADP
easat-5690	46	45	inner	inner	ADJ
easat-5690	46	46	convolution	convolution	NOUN
easat-5690	46	47	(	(	PUNCT
easat-5690	46	48	cffi	cffi	NOUN
easat-5690	46	49	)	)	PUNCT
easat-5690	46	50	strategy	strategy	NOUN
easat-5690	46	51	for	for	ADP
easat-5690	46	52	channel	channel	NOUN
easat-5690	46	53	feature	feature	NOUN
easat-5690	46	54	fusion	fusion	NOUN
easat-5690	46	55	.	.	PUNCT
easat-5690	47	1	the	the	DET
easat-5690	47	2	yolov5	yolov5	NOUN
easat-5690	47	3	model	model	NOUN
easat-5690	47	4	has	have	AUX
easat-5690	47	5	been	be	AUX
easat-5690	47	6	improved	improve	VERB
easat-5690	47	7	by	by	ADP
easat-5690	47	8	utilizing	utilize	VERB
easat-5690	47	9	a	a	DET
easat-5690	47	10	coordinate	coordinate	ADJ
easat-5690	47	11	attention	attention	NOUN
easat-5690	47	12	(	(	PUNCT
easat-5690	47	13	ca	ca	NOUN
easat-5690	47	14	)	)	PUNCT
easat-5690	47	15	module	module	NOUN
easat-5690	47	16	and	and	CCONJ
easat-5690	47	17	a	a	DET
easat-5690	47	18	regression	regression	NOUN
easat-5690	47	19	loss	loss	NOUN
easat-5690	47	20	function	function	NOUN
easat-5690	47	21	for	for	ADP
easat-5690	47	22	bounding	bound	VERB
easat-5690	47	23	boxes	box	NOUN
easat-5690	47	24	to	to	PART
easat-5690	47	25	detect	detect	VERB
easat-5690	47	26	and	and	CCONJ
easat-5690	47	27	accurately	accurately	ADV
easat-5690	47	28	count	count	VERB
easat-5690	47	29	pod	pod	ADJ
easat-5690	47	30	targets	target	NOUN
easat-5690	47	31	on	on	ADP
easat-5690	47	32	live	live	ADJ
easat-5690	47	33	plants	plant	NOUN
easat-5690	47	34	[	[	X
easat-5690	47	35	32].to	32].to	NUM
easat-5690	47	36	facilitate	facilitate	NOUN
easat-5690	47	37	information	information	NOUN
easat-5690	47	38	interchange	interchange	NOUN
easat-5690	47	39	between	between	ADP
easat-5690	47	40	channels	channel	NOUN
easat-5690	47	41	and	and	CCONJ
easat-5690	47	42	refine	refine	VERB
easat-5690	47	43	features	feature	NOUN
easat-5690	47	44	,	,	PUNCT
easat-5690	47	45	s.	s.	PROPN
easat-5690	47	46	zhu	zhu	PROPN
easat-5690	47	47	,	,	PUNCT
easat-5690	47	48	w.	w.	PROPN
easat-5690	47	49	ma	ma	PROPN
easat-5690	47	50	,	,	PUNCT
easat-5690	47	51	j.	j.	PROPN
easat-5690	47	52	wang	wang	PROPN
easat-5690	47	53	,	,	PUNCT
easat-5690	47	54	m.	m.	PROPN
easat-5690	47	55	yang	yang	PROPN
easat-5690	47	56	,	,	PUNCT
easat-5690	47	57	y.	y.	PROPN
easat-5690	47	58	wang	wang	PROPN
easat-5690	47	59	,	,	PUNCT
easat-5690	47	60	and	and	CCONJ
easat-5690	47	61	c.	c.	PROPN
easat-5690	47	62	1779	1779	NUM
easat-5690	47	63	edelweiss	edelweiss	PROPN
easat-5690	47	64	applied	apply	VERB
easat-5690	47	65	science	science	NOUN
easat-5690	47	66	and	and	CCONJ
easat-5690	47	67	technology	technology	NOUN
easat-5690	47	68	issn	issn	PROPN
easat-5690	47	69	:	:	PUNCT
easat-5690	47	70	2576	2576	NUM
easat-5690	47	71	-	-	SYM
easat-5690	47	72	8484	8484	NUM
easat-5690	47	73	vol	vol	NOUN
easat-5690	47	74	.	.	PROPN
easat-5690	48	1	9	9	NUM
easat-5690	48	2	,	,	PUNCT
easat-5690	48	3	no	no	INTJ
easat-5690	48	4	.	.	NOUN
easat-5690	49	1	3	3	NUM
easat-5690	49	2	:	:	SYM
easat-5690	49	3	1777	1777	NUM
easat-5690	49	4	-	-	SYM
easat-5690	49	5	1794	1794	NUM
easat-5690	49	6	,	,	PUNCT
easat-5690	49	7	2025	2025	NUM
easat-5690	49	8	doi	doi	NOUN
easat-5690	49	9	:	:	PUNCT
easat-5690	49	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	49	11	©	©	PROPN
easat-5690	49	12	2025	2025	NUM
easat-5690	49	13	by	by	ADP
easat-5690	49	14	the	the	DET
easat-5690	49	15	authors	author	NOUN
easat-5690	49	16	;	;	PUNCT
easat-5690	49	17	licensee	licensee	PROPN
easat-5690	49	18	learning	learning	NOUN
easat-5690	49	19	gate	gate	NOUN
easat-5690	49	20	wang[33]introduced	wang[33]introduce	VERB
easat-5690	49	21	hierarchical	hierarchical	ADJ
easat-5690	49	22	mixed	mixed	ADJ
easat-5690	49	23	-	-	PUNCT
easat-5690	49	24	scale	scale	NOUN
easat-5690	49	25	units	unit	NOUN
easat-5690	49	26	(	(	PUNCT
easat-5690	49	27	hmus	hmus	PROPN
easat-5690	49	28	)	)	PUNCT
easat-5690	49	29	within	within	ADP
easat-5690	49	30	the	the	DET
easat-5690	49	31	neck	neck	NOUN
easat-5690	49	32	network	network	NOUN
easat-5690	49	33	.	.	PUNCT
easat-5690	50	1	by	by	ADP
easat-5690	50	2	incorporating	incorporate	VERB
easat-5690	50	3	cam	cam	NOUN
easat-5690	50	4	into	into	ADP
easat-5690	50	5	the	the	DET
easat-5690	50	6	original	original	ADJ
easat-5690	50	7	yolov5s	yolov5s	PROPN
easat-5690	50	8	network	network	NOUN
easat-5690	50	9	,	,	PUNCT
easat-5690	50	10	the	the	DET
easat-5690	50	11	model	model	NOUN
easat-5690	50	12	achieves	achieve	VERB
easat-5690	50	13	greater	great	ADJ
easat-5690	50	14	accuracy	accuracy	NOUN
easat-5690	50	15	in	in	ADP
easat-5690	50	16	locating	locate	VERB
easat-5690	50	17	and	and	CCONJ
easat-5690	50	18	identifying	identify	VERB
easat-5690	50	19	fruits	fruit	NOUN
easat-5690	50	20	within	within	ADP
easat-5690	50	21	dense	dense	ADJ
easat-5690	50	22	images	image	NOUN
easat-5690	50	23	[	[	X
easat-5690	50	24	34].the	34].the	NUM
easat-5690	50	25	yolov5	yolov5	NOUN
easat-5690	50	26	model	model	NOUN
easat-5690	50	27	is	be	AUX
easat-5690	50	28	enhanced	enhance	VERB
easat-5690	50	29	and	and	CCONJ
easat-5690	50	30	used	use	VERB
easat-5690	50	31	for	for	ADP
easat-5690	50	32	the	the	DET
easat-5690	50	33	detection	detection	NOUN
easat-5690	50	34	of	of	ADP
easat-5690	50	35	thick	thick	ADJ
easat-5690	50	36	and	and	CCONJ
easat-5690	50	37	small	small	ADJ
easat-5690	50	38	tea	tea	NOUN
easat-5690	50	39	shoots	shoot	NOUN
easat-5690	50	40	[	[	X
easat-5690	50	41	35	35	NUM
easat-5690	50	42	]	]	PUNCT
easat-5690	50	43	.	.	PUNCT
easat-5690	51	1	the	the	DET
easat-5690	51	2	categorization	categorization	NOUN
easat-5690	51	3	and	and	CCONJ
easat-5690	51	4	identification	identification	NOUN
easat-5690	51	5	of	of	ADP
easat-5690	51	6	illnesses	illness	NOUN
easat-5690	51	7	and	and	CCONJ
easat-5690	51	8	pests	pest	NOUN
easat-5690	51	9	in	in	ADP
easat-5690	51	10	tea	tea	NOUN
easat-5690	51	11	leaves	leave	NOUN
easat-5690	51	12	has	have	AUX
easat-5690	51	13	improved	improve	VERB
easat-5690	51	14	somewhat	somewhat	ADV
easat-5690	51	15	thanks	thank	NOUN
easat-5690	51	16	to	to	ADP
easat-5690	51	17	deep	deep	ADJ
easat-5690	51	18	learning	learning	NOUN
easat-5690	51	19	models	model	NOUN
easat-5690	51	20	.	.	PUNCT
easat-5690	52	1	on	on	ADP
easat-5690	52	2	the	the	DET
easat-5690	52	3	other	other	ADJ
easat-5690	52	4	hand	hand	NOUN
easat-5690	52	5	,	,	PUNCT
easat-5690	52	6	the	the	DET
easat-5690	52	7	model	model	NOUN
easat-5690	52	8	's	's	PART
easat-5690	52	9	performance	performance	NOUN
easat-5690	52	10	and	and	CCONJ
easat-5690	52	11	parameters	parameter	NOUN
easat-5690	52	12	can	can	AUX
easat-5690	52	13	still	still	ADV
easat-5690	52	14	be	be	AUX
easat-5690	52	15	improved	improve	VERB
easat-5690	52	16	.	.	PUNCT
easat-5690	53	1	based	base	VERB
easat-5690	53	2	on	on	ADP
easat-5690	53	3	the	the	DET
easat-5690	53	4	above	above	ADJ
easat-5690	53	5	research	research	NOUN
easat-5690	53	6	,	,	PUNCT
easat-5690	53	7	this	this	DET
easat-5690	53	8	paper	paper	NOUN
easat-5690	53	9	proposes	propose	VERB
easat-5690	53	10	according	accord	VERB
easat-5690	53	11	to	to	ADP
easat-5690	53	12	the	the	DET
easat-5690	53	13	characteristics	characteristic	NOUN
easat-5690	53	14	of	of	ADP
easat-5690	53	15	several	several	ADJ
easat-5690	53	16	tea	tea	NOUN
easat-5690	53	17	diseases	disease	NOUN
easat-5690	53	18	,	,	PUNCT
easat-5690	53	19	the	the	DET
easat-5690	53	20	identification	identification	NOUN
easat-5690	53	21	model	model	NOUN
easat-5690	53	22	of	of	ADP
easat-5690	53	23	diseases	disease	NOUN
easat-5690	53	24	and	and	CCONJ
easat-5690	53	25	insect	insect	NOUN
easat-5690	53	26	pests	pest	NOUN
easat-5690	53	27	is	be	AUX
easat-5690	53	28	constructed	construct	VERB
easat-5690	53	29	to	to	PART
easat-5690	53	30	accurately	accurately	ADV
easat-5690	53	31	identify	identify	VERB
easat-5690	53	32	the	the	DET
easat-5690	53	33	types	type	NOUN
easat-5690	53	34	of	of	ADP
easat-5690	53	35	tea	tea	NOUN
easat-5690	53	36	diseases	disease	NOUN
easat-5690	53	37	and	and	CCONJ
easat-5690	53	38	insect	insect	NOUN
easat-5690	53	39	pests	pest	NOUN
easat-5690	53	40	,	,	PUNCT
easat-5690	53	41	and	and	CCONJ
easat-5690	53	42	then	then	ADV
easat-5690	53	43	provide	provide	VERB
easat-5690	53	44	reasonable	reasonable	ADJ
easat-5690	53	45	suggestions	suggestion	NOUN
easat-5690	53	46	for	for	SCONJ
easat-5690	53	47	farmers	farmer	NOUN
easat-5690	53	48	to	to	PART
easat-5690	53	49	actively	actively	ADV
easat-5690	53	50	prevent	prevent	VERB
easat-5690	53	51	diseases	disease	NOUN
easat-5690	53	52	and	and	CCONJ
easat-5690	53	53	insect	insect	NOUN
easat-5690	53	54	pests	pest	NOUN
easat-5690	53	55	,	,	PUNCT
easat-5690	53	56	and	and	CCONJ
easat-5690	53	57	solve	solve	VERB
easat-5690	53	58	the	the	DET
easat-5690	53	59	problem	problem	NOUN
easat-5690	53	60	of	of	ADP
easat-5690	53	61	intelligent	intelligent	ADJ
easat-5690	53	62	tea	tea	NOUN
easat-5690	53	63	garden	garden	NOUN
easat-5690	53	64	disease	disease	NOUN
easat-5690	53	65	and	and	CCONJ
easat-5690	53	66	insect	insect	NOUN
easat-5690	53	67	pests	pest	NOUN
easat-5690	53	68	’	'	PUNCT
easat-5690	53	69	identification	identification	NOUN
easat-5690	53	70	.	.	PUNCT
easat-5690	54	1	therefore	therefore	ADV
easat-5690	54	2	,	,	PUNCT
easat-5690	54	3	in	in	ADP
easat-5690	54	4	order	order	NOUN
easat-5690	54	5	to	to	PART
easat-5690	54	6	balance	balance	VERB
easat-5690	54	7	detection	detection	NOUN
easat-5690	54	8	speed	speed	NOUN
easat-5690	54	9	and	and	CCONJ
easat-5690	54	10	accuracy	accuracy	NOUN
easat-5690	54	11	,	,	PUNCT
easat-5690	54	12	we	we	PRON
easat-5690	54	13	have	have	AUX
easat-5690	54	14	built	build	VERB
easat-5690	54	15	a	a	DET
easat-5690	54	16	lightweight	lightweight	ADJ
easat-5690	54	17	network	network	NOUN
easat-5690	54	18	model	model	NOUN
easat-5690	54	19	based	base	VERB
easat-5690	54	20	on	on	ADP
easat-5690	54	21	the	the	DET
easat-5690	54	22	design	design	NOUN
easat-5690	54	23	of	of	ADP
easat-5690	54	24	yolov5.the	yolov5.the	DET
easat-5690	54	25	following	follow	VERB
easat-5690	54	26	are	be	AUX
easat-5690	54	27	this	this	DET
easat-5690	54	28	paper	paper	NOUN
easat-5690	54	29	's	's	PART
easat-5690	54	30	primary	primary	ADJ
easat-5690	54	31	contributions	contribution	NOUN
easat-5690	54	32	:	:	PUNCT
easat-5690	54	33	(	(	PUNCT
easat-5690	54	34	1	1	X
easat-5690	54	35	)	)	PUNCT
easat-5690	54	36	the	the	DET
easat-5690	54	37	backbone	backbone	NOUN
easat-5690	54	38	structure	structure	NOUN
easat-5690	54	39	module	module	NOUN
easat-5690	54	40	has	have	AUX
easat-5690	54	41	been	be	AUX
easat-5690	54	42	improved	improve	VERB
easat-5690	54	43	,	,	PUNCT
easat-5690	54	44	and	and	CCONJ
easat-5690	54	45	csp	csp	PROPN
easat-5690	54	46	has	have	AUX
easat-5690	54	47	been	be	AUX
easat-5690	54	48	upgraded	upgrade	VERB
easat-5690	54	49	to	to	ADP
easat-5690	54	50	cspk	cspk	PROPN
easat-5690	54	51	,	,	PUNCT
easat-5690	54	52	which	which	PRON
easat-5690	54	53	enhances	enhance	VERB
easat-5690	54	54	feature	feature	NOUN
easat-5690	54	55	propagation	propagation	NOUN
easat-5690	54	56	,	,	PUNCT
easat-5690	54	57	aids	aid	NOUN
easat-5690	54	58	in	in	ADP
easat-5690	54	59	better	well	ADJ
easat-5690	54	60	information	information	NOUN
easat-5690	54	61	transmission	transmission	NOUN
easat-5690	54	62	across	across	ADP
easat-5690	54	63	different	different	ADJ
easat-5690	54	64	layers	layer	NOUN
easat-5690	54	65	of	of	ADP
easat-5690	54	66	the	the	DET
easat-5690	54	67	network	network	NOUN
easat-5690	54	68	,	,	PUNCT
easat-5690	54	69	thereby	thereby	ADV
easat-5690	54	70	improving	improve	VERB
easat-5690	54	71	performance	performance	NOUN
easat-5690	54	72	.	.	PUNCT
easat-5690	55	1	(	(	PUNCT
easat-5690	55	2	2	2	X
easat-5690	55	3	)	)	PUNCT
easat-5690	55	4	in	in	ADP
easat-5690	55	5	the	the	DET
easat-5690	55	6	yolov5	yolov5	NOUN
easat-5690	55	7	model	model	NOUN
easat-5690	55	8	,	,	PUNCT
easat-5690	55	9	we	we	PRON
easat-5690	55	10	have	have	AUX
easat-5690	55	11	incorporated	incorporate	VERB
easat-5690	55	12	the	the	DET
easat-5690	55	13	convolutional	convolutional	ADJ
easat-5690	55	14	block	block	NOUN
easat-5690	55	15	attention	attention	NOUN
easat-5690	55	16	module	module	NOUN
easat-5690	55	17	(	(	PUNCT
easat-5690	55	18	cbam	cbam	NOUN
easat-5690	55	19	)	)	PUNCT
easat-5690	55	20	to	to	PART
easat-5690	55	21	enhance	enhance	VERB
easat-5690	55	22	the	the	DET
easat-5690	55	23	recognizability	recognizability	NOUN
easat-5690	55	24	of	of	ADP
easat-5690	55	25	feature	feature	NOUN
easat-5690	55	26	images	image	NOUN
easat-5690	55	27	.	.	PUNCT
easat-5690	56	1	secondly	secondly	ADV
easat-5690	56	2	,	,	PUNCT
easat-5690	56	3	owing	owe	VERB
easat-5690	56	4	to	to	ADP
easat-5690	56	5	the	the	DET
easat-5690	56	6	extensive	extensive	ADJ
easat-5690	56	7	area	area	NOUN
easat-5690	56	8	of	of	ADP
easat-5690	56	9	tea	tea	NOUN
easat-5690	56	10	lesion	lesion	NOUN
easat-5690	56	11	and	and	CCONJ
easat-5690	56	12	strong	strong	ADJ
easat-5690	56	13	background	background	NOUN
easat-5690	56	14	contrast	contrast	NOUN
easat-5690	56	15	,	,	PUNCT
easat-5690	56	16	the	the	DET
easat-5690	56	17	yolov5	yolov5	NOUN
easat-5690	56	18	backbone	backbone	NOUN
easat-5690	56	19	network	network	NOUN
easat-5690	56	20	has	have	AUX
easat-5690	56	21	introduced	introduce	VERB
easat-5690	56	22	the	the	DET
easat-5690	56	23	cbam	cbam	NOUN
easat-5690	56	24	attention	attention	NOUN
easat-5690	56	25	mechanism	mechanism	NOUN
easat-5690	56	26	,	,	PUNCT
easat-5690	56	27	which	which	PRON
easat-5690	56	28	improves	improve	VERB
easat-5690	56	29	the	the	DET
easat-5690	56	30	directionality	directionality	NOUN
easat-5690	56	31	of	of	ADP
easat-5690	56	32	tea	tea	NOUN
easat-5690	56	33	lesion	lesion	NOUN
easat-5690	56	34	recognition	recognition	NOUN
easat-5690	56	35	,	,	PUNCT
easat-5690	56	36	resulting	result	VERB
easat-5690	56	37	in	in	ADP
easat-5690	56	38	faster	fast	ADJ
easat-5690	56	39	convergence	convergence	NOUN
easat-5690	56	40	speed	speed	NOUN
easat-5690	56	41	and	and	CCONJ
easat-5690	56	42	enhanced	enhanced	ADJ
easat-5690	56	43	reasoning	reasoning	NOUN
easat-5690	56	44	and	and	CCONJ
easat-5690	56	45	training	training	NOUN
easat-5690	56	46	capabilities	capability	NOUN
easat-5690	56	47	of	of	ADP
easat-5690	56	48	the	the	DET
easat-5690	56	49	detection	detection	NOUN
easat-5690	56	50	algorithm	algorithm	NOUN
easat-5690	56	51	.	.	PUNCT
easat-5690	57	1	(	(	PUNCT
easat-5690	57	2	3	3	X
easat-5690	57	3	)	)	PUNCT
easat-5690	57	4	odc	odc	VERB
easat-5690	57	5	,	,	PUNCT
easat-5690	57	6	dynamic	dynamic	ADJ
easat-5690	57	7	convolution	convolution	NOUN
easat-5690	57	8	enhances	enhance	VERB
easat-5690	57	9	the	the	DET
easat-5690	57	10	adaptability	adaptability	NOUN
easat-5690	57	11	of	of	ADP
easat-5690	57	12	features	feature	NOUN
easat-5690	57	13	,	,	PUNCT
easat-5690	57	14	while	while	SCONJ
easat-5690	57	15	attention	attention	NOUN
easat-5690	57	16	mechanisms	mechanism	NOUN
easat-5690	57	17	enhance	enhance	VERB
easat-5690	57	18	the	the	DET
easat-5690	57	19	network	network	NOUN
easat-5690	57	20	's	's	PART
easat-5690	57	21	ability	ability	NOUN
easat-5690	57	22	to	to	PART
easat-5690	57	23	focus	focus	VERB
easat-5690	57	24	on	on	ADP
easat-5690	57	25	the	the	DET
easat-5690	57	26	specific	specific	ADJ
easat-5690	57	27	information	information	NOUN
easat-5690	57	28	required	require	VERB
easat-5690	57	29	for	for	ADP
easat-5690	57	30	the	the	DET
easat-5690	57	31	current	current	ADJ
easat-5690	57	32	task	task	NOUN
easat-5690	57	33	.	.	PUNCT
easat-5690	58	1	(	(	PUNCT
easat-5690	58	2	4)a	4)a	NUM
easat-5690	58	3	visual	visual	ADJ
easat-5690	58	4	interface	interface	NOUN
easat-5690	58	5	for	for	ADP
easat-5690	58	6	detecting	detect	VERB
easat-5690	58	7	and	and	CCONJ
easat-5690	58	8	recognizing	recognize	VERB
easat-5690	58	9	tea	tea	NOUN
easat-5690	58	10	disease	disease	NOUN
easat-5690	58	11	images	image	NOUN
easat-5690	58	12	has	have	AUX
easat-5690	58	13	been	be	AUX
easat-5690	58	14	designed	design	VERB
easat-5690	58	15	,	,	PUNCT
easat-5690	58	16	and	and	CCONJ
easat-5690	58	17	the	the	DET
easat-5690	58	18	functions	function	NOUN
easat-5690	58	19	of	of	ADP
easat-5690	58	20	tea	tea	NOUN
easat-5690	58	21	disease	disease	NOUN
easat-5690	58	22	and	and	CCONJ
easat-5690	58	23	pest	pest	NOUN
easat-5690	58	24	diseases	disease	NOUN
easat-5690	58	25	include	include	VERB
easat-5690	58	26	importing	import	VERB
easat-5690	58	27	and	and	CCONJ
easat-5690	58	28	initializing	initialize	VERB
easat-5690	58	29	training	training	NOUN
easat-5690	58	30	models	model	NOUN
easat-5690	58	31	;	;	PUNCT
easat-5690	58	32	adjusting	adjust	VERB
easat-5690	58	33	confidence	confidence	NOUN
easat-5690	58	34	and	and	CCONJ
easat-5690	58	35	iou	iou	NOUN
easat-5690	58	36	(	(	PUNCT
easat-5690	58	37	intersection	intersection	NOUN
easat-5690	58	38	over	over	ADP
easat-5690	58	39	union	union	NOUN
easat-5690	58	40	)	)	PUNCT
easat-5690	58	41	thresholds	threshold	NOUN
easat-5690	58	42	,	,	PUNCT
easat-5690	58	43	image	image	NOUN
easat-5690	58	44	upload	upload	NOUN
easat-5690	58	45	,	,	PUNCT
easat-5690	58	46	detection	detection	NOUN
easat-5690	58	47	,	,	PUNCT
easat-5690	58	48	visualization	visualization	NOUN
easat-5690	58	49	of	of	ADP
easat-5690	58	50	results	result	NOUN
easat-5690	58	51	,	,	PUNCT
easat-5690	58	52	export	export	NOUN
easat-5690	58	53	of	of	ADP
easat-5690	58	54	results	result	NOUN
easat-5690	58	55	,	,	PUNCT
easat-5690	58	56	and	and	CCONJ
easat-5690	58	57	ending	end	VERB
easat-5690	58	58	the	the	DET
easat-5690	58	59	detection	detection	NOUN
easat-5690	58	60	;	;	PUNCT
easat-5690	58	61	video	video	NOUN
easat-5690	58	62	upload	upload	NOUN
easat-5690	58	63	,	,	PUNCT
easat-5690	58	64	detection	detection	NOUN
easat-5690	58	65	,	,	PUNCT
easat-5690	58	66	visualization	visualization	NOUN
easat-5690	58	67	of	of	ADP
easat-5690	58	68	results	result	NOUN
easat-5690	58	69	,	,	PUNCT
easat-5690	58	70	export	export	NOUN
easat-5690	58	71	of	of	ADP
easat-5690	58	72	results	result	NOUN
easat-5690	58	73	,	,	PUNCT
easat-5690	58	74	and	and	CCONJ
easat-5690	58	75	ending	end	VERB
easat-5690	58	76	the	the	DET
easat-5690	58	77	detection	detection	NOUN
easat-5690	58	78	.	.	PUNCT
easat-5690	59	1	the	the	DET
easat-5690	59	2	following	follow	VERB
easat-5690	59	3	is	be	AUX
easat-5690	59	4	the	the	DET
easat-5690	59	5	structure	structure	NOUN
easat-5690	59	6	of	of	ADP
easat-5690	59	7	the	the	DET
easat-5690	59	8	paper	paper	NOUN
easat-5690	59	9	:	:	PUNCT
easat-5690	59	10	the	the	DET
easat-5690	59	11	tea	tea	NOUN
easat-5690	59	12	plant	plant	NOUN
easat-5690	59	13	dataset	dataset	NOUN
easat-5690	59	14	,	,	PUNCT
easat-5690	59	15	the	the	DET
easat-5690	59	16	enhanced	enhanced	ADJ
easat-5690	59	17	yolov5s	yolov5s	PROPN
easat-5690	59	18	model	model	NOUN
easat-5690	59	19	,	,	PUNCT
easat-5690	59	20	the	the	DET
easat-5690	59	21	experimental	experimental	ADJ
easat-5690	59	22	setup	setup	NOUN
easat-5690	59	23	,	,	PUNCT
easat-5690	59	24	and	and	CCONJ
easat-5690	59	25	the	the	DET
easat-5690	59	26	evaluation	evaluation	NOUN
easat-5690	59	27	measures	measure	NOUN
easat-5690	59	28	are	be	AUX
easat-5690	59	29	all	all	PRON
easat-5690	59	30	covered	cover	VERB
easat-5690	59	31	in	in	ADP
easat-5690	59	32	detail	detail	NOUN
easat-5690	59	33	in	in	ADP
easat-5690	59	34	section	section	NOUN
easat-5690	59	35	2	2	NUM
easat-5690	59	36	.	.	PUNCT
easat-5690	60	1	the	the	DET
easat-5690	60	2	findings	finding	NOUN
easat-5690	60	3	are	be	AUX
easat-5690	60	4	examined	examine	VERB
easat-5690	60	5	and	and	CCONJ
easat-5690	60	6	discussed	discuss	VERB
easat-5690	60	7	in	in	ADP
easat-5690	60	8	section	section	NOUN
easat-5690	60	9	3	3	NUM
easat-5690	60	10	.	.	PUNCT
easat-5690	61	1	our	our	PRON
easat-5690	61	2	tea	tea	NOUN
easat-5690	61	3	plant	plant	NOUN
easat-5690	61	4	disease	disease	NOUN
easat-5690	61	5	and	and	CCONJ
easat-5690	61	6	pest	pest	VERB
easat-5690	61	7	detection	detection	NOUN
easat-5690	61	8	method	method	NOUN
easat-5690	61	9	is	be	AUX
easat-5690	61	10	shown	show	VERB
easat-5690	61	11	in	in	ADP
easat-5690	61	12	section	section	NOUN
easat-5690	61	13	4	4	NUM
easat-5690	61	14	.	.	PUNCT
easat-5690	62	1	the	the	DET
easat-5690	62	2	conclusions	conclusion	NOUN
easat-5690	62	3	are	be	AUX
easat-5690	62	4	summed	sum	VERB
easat-5690	62	5	up	up	ADP
easat-5690	62	6	in	in	ADP
easat-5690	62	7	the	the	DET
easat-5690	62	8	last	last	ADJ
easat-5690	62	9	section	section	NOUN
easat-5690	62	10	.	.	PUNCT
easat-5690	63	1	2	2	X
easat-5690	63	2	.	.	X
easat-5690	63	3	materials	material	NOUN
easat-5690	63	4	and	and	CCONJ
easat-5690	63	5	methods	method	NOUN
easat-5690	63	6	2.1	2.1	NUM
easat-5690	63	7	.	.	PUNCT
easat-5690	64	1	dataset	dataset	VERB
easat-5690	64	2	all	all	DET
easat-5690	64	3	the	the	DET
easat-5690	64	4	photos	photo	NOUN
easat-5690	64	5	of	of	ADP
easat-5690	64	6	tea	tea	NOUN
easat-5690	64	7	plant	plant	NOUN
easat-5690	64	8	diseases	disease	NOUN
easat-5690	64	9	and	and	CCONJ
easat-5690	64	10	pests	pest	NOUN
easat-5690	64	11	utilized	utilize	VERB
easat-5690	64	12	in	in	ADP
easat-5690	64	13	this	this	DET
easat-5690	64	14	study	study	NOUN
easat-5690	64	15	were	be	AUX
easat-5690	64	16	captured	capture	VERB
easat-5690	64	17	outdoors	outdoors	ADV
easat-5690	64	18	in	in	ADP
easat-5690	64	19	natural	natural	ADJ
easat-5690	64	20	environments	environment	NOUN
easat-5690	64	21	.	.	PUNCT
easat-5690	65	1	the	the	DET
easat-5690	65	2	images	image	NOUN
easat-5690	65	3	were	be	AUX
easat-5690	65	4	captured	capture	VERB
easat-5690	65	5	at	at	ADP
easat-5690	65	6	the	the	DET
easat-5690	65	7	xiangcha	xiangcha	PROPN
easat-5690	65	8	tea	tea	NOUN
easat-5690	65	9	farm	farm	NOUN
easat-5690	65	10	in	in	ADP
easat-5690	65	11	changsha	changsha	PROPN
easat-5690	65	12	,	,	PUNCT
easat-5690	65	13	hunan	hunan	PROPN
easat-5690	65	14	province	province	PROPN
easat-5690	65	15	,	,	PUNCT
easat-5690	65	16	china	china	PROPN
easat-5690	65	17	.	.	PUNCT
easat-5690	66	1	image	image	NOUN
easat-5690	66	2	collection	collection	NOUN
easat-5690	66	3	was	be	AUX
easat-5690	66	4	performed	perform	VERB
easat-5690	66	5	under	under	ADP
easat-5690	66	6	clear	clear	ADJ
easat-5690	66	7	weather	weather	NOUN
easat-5690	66	8	conditions	condition	NOUN
easat-5690	66	9	,	,	PUNCT
easat-5690	66	10	with	with	ADP
easat-5690	66	11	adequate	adequate	ADJ
easat-5690	66	12	light	light	NOUN
easat-5690	66	13	and	and	CCONJ
easat-5690	66	14	a	a	DET
easat-5690	66	15	favourable	favourable	ADJ
easat-5690	66	16	collection	collection	NOUN
easat-5690	66	17	environment	environment	NOUN
easat-5690	66	18	.	.	PUNCT
easat-5690	67	1	since	since	SCONJ
easat-5690	67	2	our	our	PRON
easat-5690	67	3	photos	photo	NOUN
easat-5690	67	4	were	be	AUX
easat-5690	67	5	taken	take	VERB
easat-5690	67	6	in	in	ADP
easat-5690	67	7	natural	natural	ADJ
easat-5690	67	8	settings	setting	NOUN
easat-5690	67	9	,	,	PUNCT
easat-5690	67	10	we	we	PRON
easat-5690	67	11	captured	capture	VERB
easat-5690	67	12	both	both	DET
easat-5690	67	13	images	image	NOUN
easat-5690	67	14	with	with	ADP
easat-5690	67	15	dense	dense	ADJ
easat-5690	67	16	targets	target	NOUN
easat-5690	67	17	and	and	CCONJ
easat-5690	67	18	images	image	NOUN
easat-5690	67	19	with	with	ADP
easat-5690	67	20	fewer	few	ADJ
easat-5690	67	21	targets	target	NOUN
easat-5690	67	22	.	.	PUNCT
easat-5690	68	1	phenomena	phenomenon	NOUN
easat-5690	68	2	such	such	ADJ
easat-5690	68	3	as	as	ADP
easat-5690	68	4	occlusion	occlusion	NOUN
easat-5690	68	5	by	by	ADP
easat-5690	68	6	leaves	leave	NOUN
easat-5690	68	7	,	,	PUNCT
easat-5690	68	8	overlapping	overlap	VERB
easat-5690	68	9	leaves	leave	NOUN
easat-5690	68	10	,	,	PUNCT
easat-5690	68	11	and	and	CCONJ
easat-5690	68	12	large	large	ADJ
easat-5690	68	13	areas	area	NOUN
easat-5690	68	14	of	of	ADP
easat-5690	68	15	vegetation	vegetation	NOUN
easat-5690	68	16	are	be	AUX
easat-5690	68	17	present	present	ADJ
easat-5690	68	18	in	in	ADP
easat-5690	68	19	the	the	DET
easat-5690	68	20	photos	photo	NOUN
easat-5690	68	21	.	.	PUNCT
easat-5690	69	1	there	there	PRON
easat-5690	69	2	are	be	VERB
easat-5690	69	3	also	also	ADV
easat-5690	69	4	shadows	shadow	NOUN
easat-5690	69	5	and	and	CCONJ
easat-5690	69	6	reflections	reflection	NOUN
easat-5690	69	7	on	on	ADP
easat-5690	69	8	the	the	DET
easat-5690	69	9	leaves	leave	NOUN
easat-5690	69	10	in	in	ADP
easat-5690	69	11	the	the	DET
easat-5690	69	12	images	image	NOUN
easat-5690	69	13	.	.	PUNCT
easat-5690	70	1	a	a	DET
easat-5690	70	2	few	few	ADJ
easat-5690	70	3	typical	typical	ADJ
easat-5690	70	4	samples	sample	NOUN
easat-5690	70	5	from	from	ADP
easat-5690	70	6	the	the	DET
easat-5690	70	7	dataset	dataset	NOUN
easat-5690	70	8	are	be	AUX
easat-5690	70	9	displayed	display	VERB
easat-5690	70	10	in	in	ADP
easat-5690	70	11	figure	figure	NOUN
easat-5690	70	12	1	1	NUM
easat-5690	70	13	.	.	PUNCT
easat-5690	70	14	1780	1780	NUM
easat-5690	70	15	edelweiss	edelweiss	PROPN
easat-5690	70	16	applied	apply	VERB
easat-5690	70	17	science	science	NOUN
easat-5690	70	18	and	and	CCONJ
easat-5690	70	19	technology	technology	NOUN
easat-5690	70	20	issn	issn	PROPN
easat-5690	70	21	:	:	PUNCT
easat-5690	70	22	2576	2576	NUM
easat-5690	70	23	-	-	SYM
easat-5690	70	24	8484	8484	NUM
easat-5690	70	25	vol	vol	NOUN
easat-5690	70	26	.	.	PROPN
easat-5690	71	1	9	9	NUM
easat-5690	71	2	,	,	PUNCT
easat-5690	71	3	no	no	INTJ
easat-5690	71	4	.	.	NOUN
easat-5690	72	1	3	3	NUM
easat-5690	72	2	:	:	SYM
easat-5690	72	3	1777	1777	NUM
easat-5690	72	4	-	-	SYM
easat-5690	72	5	1794	1794	NUM
easat-5690	72	6	,	,	PUNCT
easat-5690	72	7	2025	2025	NUM
easat-5690	72	8	doi	doi	NOUN
easat-5690	72	9	:	:	PUNCT
easat-5690	72	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	72	11	©	©	PROPN
easat-5690	72	12	2025	2025	NUM
easat-5690	72	13	by	by	ADP
easat-5690	72	14	the	the	DET
easat-5690	72	15	authors	author	NOUN
easat-5690	72	16	;	;	PUNCT
easat-5690	72	17	licensee	licensee	PROPN
easat-5690	72	18	learning	learn	VERB
easat-5690	72	19	gate	gate	NOUN
easat-5690	72	20	figure	figure	NOUN
easat-5690	72	21	1	1	NUM
easat-5690	72	22	.	.	PUNCT
easat-5690	73	1	some	some	DET
easat-5690	73	2	illustrative	illustrative	ADJ
easat-5690	73	3	examples	example	NOUN
easat-5690	73	4	from	from	ADP
easat-5690	73	5	our	our	PRON
easat-5690	73	6	dataset	dataset	NOUN
easat-5690	73	7	.	.	PUNCT
easat-5690	74	1	the	the	DET
easat-5690	74	2	photos	photo	NOUN
easat-5690	74	3	of	of	ADP
easat-5690	74	4	tea	tea	NOUN
easat-5690	74	5	plant	plant	NOUN
easat-5690	74	6	diseases	disease	NOUN
easat-5690	74	7	and	and	CCONJ
easat-5690	74	8	pests	pest	NOUN
easat-5690	74	9	are	be	AUX
easat-5690	74	10	stored	store	VERB
easat-5690	74	11	in	in	ADP
easat-5690	74	12	jpg	jpg	NOUN
easat-5690	74	13	format	format	NOUN
easat-5690	74	14	.	.	PUNCT
easat-5690	75	1	to	to	PART
easat-5690	75	2	label	label	VERB
easat-5690	75	3	the	the	DET
easat-5690	75	4	data	datum	NOUN
easat-5690	75	5	,	,	PUNCT
easat-5690	75	6	we	we	PRON
easat-5690	75	7	employed	employ	VERB
easat-5690	75	8	labelimg	labelimg	NOUN
easat-5690	75	9	[	[	X
easat-5690	75	10	36	36	NUM
easat-5690	75	11	]	]	PUNCT
easat-5690	75	12	.	.	PUNCT
easat-5690	76	1	the	the	DET
easat-5690	76	2	label	label	NOUN
easat-5690	76	3	files	file	NOUN
easat-5690	76	4	are	be	AUX
easat-5690	76	5	ultimately	ultimately	ADV
easat-5690	76	6	saved	save	VERB
easat-5690	76	7	in	in	ADP
easat-5690	76	8	voc	voc	NOUN
easat-5690	76	9	format	format	NOUN
easat-5690	76	10	.	.	PUNCT
easat-5690	77	1	labeling	labeling	NOUN
easat-5690	77	2	boxes	box	NOUN
easat-5690	77	3	are	be	AUX
easat-5690	77	4	added	add	VERB
easat-5690	77	5	and	and	CCONJ
easat-5690	77	6	corresponding	correspond	VERB
easat-5690	77	7	labeling	labeling	NOUN
easat-5690	77	8	files	file	NOUN
easat-5690	77	9	are	be	AUX
easat-5690	77	10	generated	generate	VERB
easat-5690	77	11	for	for	ADP
easat-5690	77	12	the	the	DET
easat-5690	77	13	regions	region	NOUN
easat-5690	77	14	of	of	ADP
easat-5690	77	15	the	the	DET
easat-5690	77	16	image	image	NOUN
easat-5690	77	17	suffering	suffer	VERB
easat-5690	77	18	from	from	ADP
easat-5690	77	19	tea	tea	NOUN
easat-5690	77	20	diseases	disease	NOUN
easat-5690	77	21	.	.	PUNCT
easat-5690	78	1	the	the	DET
easat-5690	78	2	image	image	NOUN
easat-5690	78	3	labeling	labeling	NOUN
easat-5690	78	4	is	be	AUX
easat-5690	78	5	shown	show	VERB
easat-5690	78	6	in	in	ADP
easat-5690	78	7	1781	1781	NUM
easat-5690	78	8	edelweiss	edelweiss	PROPN
easat-5690	78	9	applied	apply	VERB
easat-5690	78	10	science	science	NOUN
easat-5690	78	11	and	and	CCONJ
easat-5690	78	12	technology	technology	NOUN
easat-5690	78	13	issn	issn	PROPN
easat-5690	78	14	:	:	PUNCT
easat-5690	78	15	2576	2576	NUM
easat-5690	78	16	-	-	SYM
easat-5690	78	17	8484	8484	NUM
easat-5690	78	18	vol	vol	NOUN
easat-5690	78	19	.	.	PROPN
easat-5690	79	1	9	9	NUM
easat-5690	79	2	,	,	PUNCT
easat-5690	79	3	no	no	INTJ
easat-5690	79	4	.	.	NOUN
easat-5690	80	1	3	3	NUM
easat-5690	80	2	:	:	SYM
easat-5690	80	3	1777	1777	NUM
easat-5690	80	4	-	-	SYM
easat-5690	80	5	1794	1794	NUM
easat-5690	80	6	,	,	PUNCT
easat-5690	80	7	2025	2025	NUM
easat-5690	80	8	doi	doi	NOUN
easat-5690	80	9	:	:	PUNCT
easat-5690	80	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	80	11	©	©	PROPN
easat-5690	80	12	2025	2025	NUM
easat-5690	80	13	by	by	ADP
easat-5690	80	14	the	the	DET
easat-5690	80	15	authors	author	NOUN
easat-5690	80	16	;	;	PUNCT
easat-5690	80	17	licensee	licensee	PROPN
easat-5690	80	18	learning	learn	VERB
easat-5690	80	19	gate	gate	NOUN
easat-5690	80	20	figure	figure	NOUN
easat-5690	80	21	2	2	NUM
easat-5690	80	22	.	.	PUNCT
easat-5690	80	23	to	to	PART
easat-5690	80	24	increase	increase	VERB
easat-5690	80	25	the	the	DET
easat-5690	80	26	size	size	NOUN
easat-5690	80	27	of	of	ADP
easat-5690	80	28	the	the	DET
easat-5690	80	29	final	final	ADJ
easat-5690	80	30	dataset	dataset	NOUN
easat-5690	80	31	,	,	PUNCT
easat-5690	80	32	we	we	PRON
easat-5690	80	33	employed	employ	VERB
easat-5690	80	34	data	datum	NOUN
easat-5690	80	35	augmentation	augmentation	NOUN
easat-5690	80	36	.	.	PUNCT
easat-5690	81	1	lastly	lastly	ADV
easat-5690	81	2	,	,	PUNCT
easat-5690	81	3	an	an	DET
easat-5690	81	4	8:1:1	8:1:1	PROPN
easat-5690	81	5	randomization	randomization	NOUN
easat-5690	81	6	process	process	NOUN
easat-5690	81	7	was	be	AUX
easat-5690	81	8	used	use	VERB
easat-5690	81	9	to	to	PART
easat-5690	81	10	divide	divide	VERB
easat-5690	81	11	the	the	DET
easat-5690	81	12	annotated	annotate	VERB
easat-5690	81	13	photos	photo	NOUN
easat-5690	81	14	into	into	ADP
easat-5690	81	15	training	training	NOUN
easat-5690	81	16	,	,	PUNCT
easat-5690	81	17	validation	validation	NOUN
easat-5690	81	18	,	,	PUNCT
easat-5690	81	19	and	and	CCONJ
easat-5690	81	20	testing	testing	NOUN
easat-5690	81	21	sets	set	NOUN
easat-5690	81	22	.	.	PUNCT
easat-5690	82	1	by	by	ADP
easat-5690	82	2	employing	employ	VERB
easat-5690	82	3	random	random	ADJ
easat-5690	82	4	scaling	scaling	NOUN
easat-5690	82	5	,	,	PUNCT
easat-5690	82	6	cropping	cropping	NOUN
easat-5690	82	7	,	,	PUNCT
easat-5690	82	8	and	and	CCONJ
easat-5690	82	9	layout	layout	VERB
easat-5690	82	10	techniques	technique	NOUN
easat-5690	82	11	for	for	ADP
easat-5690	82	12	stitching	stitching	NOUN
easat-5690	82	13	,	,	PUNCT
easat-5690	82	14	mosaic	mosaic	ADJ
easat-5690	82	15	data	datum	NOUN
easat-5690	82	16	augmentation	augmentation	NOUN
easat-5690	82	17	creates	create	VERB
easat-5690	82	18	a	a	DET
easat-5690	82	19	single	single	ADJ
easat-5690	82	20	image	image	NOUN
easat-5690	82	21	out	out	ADP
easat-5690	82	22	of	of	ADP
easat-5690	82	23	four	four	NUM
easat-5690	82	24	.	.	PUNCT
easat-5690	83	1	1782	1782	NUM
easat-5690	84	1	edelweiss	edelweiss	PROPN
easat-5690	84	2	applied	apply	VERB
easat-5690	84	3	science	science	NOUN
easat-5690	84	4	and	and	CCONJ
easat-5690	84	5	technology	technology	NOUN
easat-5690	84	6	issn	issn	PROPN
easat-5690	84	7	:	:	PUNCT
easat-5690	84	8	2576	2576	NUM
easat-5690	84	9	-	-	SYM
easat-5690	84	10	8484	8484	NUM
easat-5690	84	11	vol	vol	NOUN
easat-5690	84	12	.	.	PROPN
easat-5690	85	1	9	9	NUM
easat-5690	85	2	,	,	PUNCT
easat-5690	85	3	no	no	INTJ
easat-5690	85	4	.	.	NOUN
easat-5690	86	1	3	3	NUM
easat-5690	86	2	:	:	SYM
easat-5690	86	3	1777	1777	NUM
easat-5690	86	4	-	-	SYM
easat-5690	86	5	1794	1794	NUM
easat-5690	86	6	,	,	PUNCT
easat-5690	86	7	2025	2025	NUM
easat-5690	86	8	doi	doi	NOUN
easat-5690	86	9	:	:	PUNCT
easat-5690	86	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	86	11	©	©	PROPN
easat-5690	86	12	2025	2025	NUM
easat-5690	86	13	by	by	ADP
easat-5690	86	14	the	the	DET
easat-5690	86	15	authors	author	NOUN
easat-5690	86	16	;	;	PUNCT
easat-5690	86	17	licensee	licensee	PROPN
easat-5690	86	18	learning	learn	VERB
easat-5690	86	19	gate	gate	NOUN
easat-5690	86	20	figure	figure	NOUN
easat-5690	86	21	2	2	NUM
easat-5690	86	22	.	.	PUNCT
easat-5690	86	23	image	image	NOUN
easat-5690	86	24	labeling	labeling	NOUN
easat-5690	86	25	fabrication	fabrication	NOUN
easat-5690	86	26	.	.	PUNCT
easat-5690	87	1	2.2	2.2	NUM
easat-5690	87	2	.	.	PUNCT
easat-5690	88	1	original	original	ADJ
easat-5690	88	2	yolov5s	yolov5s	PROPN
easat-5690	88	3	the	the	DET
easat-5690	88	4	yolo	yolo	ADJ
easat-5690	88	5	algorithm	algorithm	NOUN
easat-5690	88	6	stands	stand	VERB
easat-5690	88	7	for	for	SCONJ
easat-5690	88	8	"	"	PUNCT
easat-5690	88	9	you	you	PRON
easat-5690	88	10	only	only	ADV
easat-5690	88	11	look	look	VERB
easat-5690	88	12	once	once	ADV
easat-5690	88	13	,	,	PUNCT
easat-5690	88	14	"	"	PUNCT
easat-5690	88	15	and	and	CCONJ
easat-5690	88	16	it	it	PRON
easat-5690	88	17	utilizes	utilize	VERB
easat-5690	88	18	a	a	DET
easat-5690	88	19	single	single	ADJ
easat-5690	88	20	convolutional	convolutional	ADJ
easat-5690	88	21	neural	neural	ADJ
easat-5690	88	22	network	network	NOUN
easat-5690	88	23	(	(	PUNCT
easat-5690	88	24	cnn	cnn	PROPN
easat-5690	88	25	)	)	PUNCT
easat-5690	88	26	model	model	NOUN
easat-5690	88	27	to	to	PART
easat-5690	88	28	achieve	achieve	VERB
easat-5690	88	29	end	end	NOUN
easat-5690	88	30	-	-	PUNCT
easat-5690	88	31	to	to	ADP
easat-5690	88	32	-	-	PUNCT
easat-5690	88	33	end	end	NOUN
easat-5690	88	34	object	object	NOUN
easat-5690	88	35	detection	detection	NOUN
easat-5690	88	36	.	.	PUNCT
easat-5690	89	1	yolov5	yolov5	NOUN
easat-5690	89	2	can	can	AUX
easat-5690	89	3	be	be	AUX
easat-5690	89	4	divided	divide	VERB
easat-5690	89	5	into	into	ADP
easat-5690	89	6	yolov5n	yolov5n	PROPN
easat-5690	89	7	,	,	PUNCT
easat-5690	89	8	yolov5s	yolov5s	PROPN
easat-5690	89	9	,	,	PUNCT
easat-5690	89	10	yolov5	yolov5	PROPN
easat-5690	89	11	m	m	PROPN
easat-5690	89	12	,	,	PUNCT
easat-5690	89	13	yolov5l	yolov5l	PROPN
easat-5690	89	14	,	,	PUNCT
easat-5690	89	15	and	and	CCONJ
easat-5690	89	16	yolov5x	yolov5x	PROPN
easat-5690	89	17	based	base	VERB
easat-5690	89	18	on	on	ADP
easat-5690	89	19	the	the	DET
easat-5690	89	20	various	various	ADJ
easat-5690	89	21	network	network	NOUN
easat-5690	89	22	widths	width	NOUN
easat-5690	89	23	and	and	CCONJ
easat-5690	89	24	depths.based	depths.base	VERB
easat-5690	89	25	on	on	ADP
easat-5690	89	26	yolov5s-5.3	yolov5s-5.3	NUM
easat-5690	89	27	,	,	PUNCT
easat-5690	89	28	we	we	PRON
easat-5690	89	29	present	present	VERB
easat-5690	89	30	an	an	DET
easat-5690	89	31	upgraded	upgrade	VERB
easat-5690	89	32	version	version	NOUN
easat-5690	89	33	of	of	ADP
easat-5690	89	34	the	the	DET
easat-5690	89	35	yolov5s	yolov5s	PROPN
easat-5690	89	36	disease	disease	NOUN
easat-5690	89	37	and	and	CCONJ
easat-5690	89	38	pest	pest	VERB
easat-5690	89	39	detection	detection	NOUN
easat-5690	89	40	model	model	NOUN
easat-5690	89	41	in	in	ADP
easat-5690	89	42	this	this	DET
easat-5690	89	43	study	study	NOUN
easat-5690	89	44	.	.	PUNCT
easat-5690	90	1	on	on	ADP
easat-5690	90	2	some	some	DET
easat-5690	90	3	researchers	researcher	NOUN
easat-5690	90	4	'	'	PART
easat-5690	90	5	devices	device	NOUN
easat-5690	90	6	,	,	PUNCT
easat-5690	90	7	the	the	DET
easat-5690	90	8	image	image	NOUN
easat-5690	90	9	inference	inference	NOUN
easat-5690	90	10	speed	speed	NOUN
easat-5690	90	11	of	of	ADP
easat-5690	90	12	the	the	DET
easat-5690	90	13	yolov5s	yolov5s	PROPN
easat-5690	90	14	model	model	NOUN
easat-5690	90	15	has	have	AUX
easat-5690	90	16	reached	reach	VERB
easat-5690	90	17	455fps	455fps	NUM
easat-5690	90	18	,	,	PUNCT
easat-5690	90	19	which	which	PRON
easat-5690	90	20	is	be	AUX
easat-5690	90	21	a	a	DET
easat-5690	90	22	significant	significant	ADJ
easat-5690	90	23	advantage	advantage	NOUN
easat-5690	90	24	that	that	PRON
easat-5690	90	25	has	have	AUX
easat-5690	90	26	been	be	AUX
easat-5690	90	27	widely	widely	ADV
easat-5690	90	28	used	use	VERB
easat-5690	90	29	by	by	ADP
easat-5690	90	30	many	many	ADJ
easat-5690	90	31	scholars[37].the	scholars[37].the	ADJ
easat-5690	90	32	yolov5	yolov5	NOUN
easat-5690	90	33	algorithm	algorithm	PROPN
easat-5690	90	34	utilizes	utilize	VERB
easat-5690	90	35	deep	deep	ADJ
easat-5690	90	36	learning	learn	VERB
easat-5690	90	37	techniques	technique	NOUN
easat-5690	90	38	to	to	PART
easat-5690	90	39	swiftly	swiftly	ADV
easat-5690	90	40	and	and	CCONJ
easat-5690	90	41	accurately	accurately	ADV
easat-5690	90	42	identify	identify	VERB
easat-5690	90	43	objects	object	NOUN
easat-5690	90	44	in	in	ADP
easat-5690	90	45	images	image	NOUN
easat-5690	90	46	,	,	PUNCT
easat-5690	90	47	while	while	SCONJ
easat-5690	90	48	annotating	annotate	VERB
easat-5690	90	49	their	their	PRON
easat-5690	90	50	positions	position	NOUN
easat-5690	90	51	and	and	CCONJ
easat-5690	90	52	class	class	NOUN
easat-5690	90	53	information	information	NOUN
easat-5690	90	54	,	,	PUNCT
easat-5690	90	55	as	as	SCONJ
easat-5690	90	56	shown	show	VERB
easat-5690	90	57	in	in	ADP
easat-5690	90	58	1783	1783	NUM
easat-5690	90	59	edelweiss	edelweiss	PROPN
easat-5690	90	60	applied	apply	VERB
easat-5690	90	61	science	science	NOUN
easat-5690	90	62	and	and	CCONJ
easat-5690	90	63	technology	technology	NOUN
easat-5690	90	64	issn	issn	PROPN
easat-5690	90	65	:	:	PUNCT
easat-5690	90	66	2576	2576	NUM
easat-5690	90	67	-	-	SYM
easat-5690	90	68	8484	8484	NUM
easat-5690	90	69	vol	vol	NOUN
easat-5690	90	70	.	.	PROPN
easat-5690	91	1	9	9	NUM
easat-5690	91	2	,	,	PUNCT
easat-5690	91	3	no	no	INTJ
easat-5690	91	4	.	.	NOUN
easat-5690	92	1	3	3	NUM
easat-5690	92	2	:	:	SYM
easat-5690	92	3	1777	1777	NUM
easat-5690	92	4	-	-	SYM
easat-5690	92	5	1794	1794	NUM
easat-5690	92	6	,	,	PUNCT
easat-5690	92	7	2025	2025	NUM
easat-5690	92	8	doi	doi	NOUN
easat-5690	92	9	:	:	PUNCT
easat-5690	92	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	92	11	©	©	PROPN
easat-5690	92	12	2025	2025	NUM
easat-5690	92	13	by	by	ADP
easat-5690	92	14	the	the	DET
easat-5690	92	15	authors	author	NOUN
easat-5690	92	16	;	;	PUNCT
easat-5690	92	17	licensee	licensee	PROPN
easat-5690	92	18	learning	learn	VERB
easat-5690	92	19	gate	gate	PROPN
easat-5690	92	20	figure	figure	NOUN
easat-5690	92	21	3	3	NUM
easat-5690	92	22	.	.	PUNCT
easat-5690	92	23	figure	figure	NOUN
easat-5690	92	24	3	3	NUM
easat-5690	92	25	.	.	PUNCT
easat-5690	93	1	the	the	DET
easat-5690	93	2	network	network	NOUN
easat-5690	93	3	structure	structure	NOUN
easat-5690	93	4	of	of	ADP
easat-5690	93	5	yolov5s-5.3	yolov5s-5.3	PROPN
easat-5690	93	6	model	model	NOUN
easat-5690	93	7	.	.	PUNCT
easat-5690	94	1	as	as	SCONJ
easat-5690	94	2	shown	show	VERB
easat-5690	94	3	in	in	ADP
easat-5690	94	4	fig	fig	NOUN
easat-5690	94	5	.	.	PUNCT
easat-5690	95	1	3	3	NUM
easat-5690	95	2	,	,	PUNCT
easat-5690	95	3	the	the	DET
easat-5690	95	4	yolov5	yolov5	NOUN
easat-5690	95	5	version	version	NOUN
easat-5690	95	6	5.3	5.3	NUM
easat-5690	95	7	network	network	NOUN
easat-5690	95	8	structure	structure	NOUN
easat-5690	95	9	model	model	NOUN
easat-5690	95	10	includes	include	VERB
easat-5690	95	11	three	three	NUM
easat-5690	95	12	parts	part	NOUN
easat-5690	95	13	:	:	PUNCT
easat-5690	95	14	backbone	backbone	NOUN
easat-5690	95	15	function	function	NOUN
easat-5690	95	16	:	:	PUNCT
easat-5690	95	17	feature	feature	NOUN
easat-5690	95	18	extraction	extraction	NOUN
easat-5690	95	19	.	.	PUNCT
easat-5690	96	1	neck	neck	NOUN
easat-5690	96	2	function	function	NOUN
easat-5690	96	3	:	:	PUNCT
easat-5690	96	4	mix	mix	VERB
easat-5690	96	5	and	and	CCONJ
easat-5690	96	6	combine	combine	VERB
easat-5690	96	7	features	feature	NOUN
easat-5690	96	8	in	in	ADP
easat-5690	96	9	one	one	NUM
easat-5690	96	10	wave	wave	NOUN
easat-5690	96	11	,	,	PUNCT
easat-5690	96	12	and	and	CCONJ
easat-5690	96	13	transfer	transfer	VERB
easat-5690	96	14	these	these	DET
easat-5690	96	15	features	feature	NOUN
easat-5690	96	16	to	to	ADP
easat-5690	96	17	the	the	DET
easat-5690	96	18	prediction	prediction	NOUN
easat-5690	96	19	layer	layer	NOUN
easat-5690	96	20	.	.	PUNCT
easat-5690	97	1	head	head	NOUN
easat-5690	97	2	function	function	NOUN
easat-5690	97	3	:	:	PUNCT
easat-5690	97	4	make	make	VERB
easat-5690	97	5	final	final	ADJ
easat-5690	97	6	prediction	prediction	NOUN
easat-5690	97	7	output	output	NOUN
easat-5690	97	8	.	.	PUNCT
easat-5690	98	1	the	the	DET
easat-5690	98	2	input	input	NOUN
easat-5690	98	3	,	,	PUNCT
easat-5690	98	4	backbone	backbone	NOUN
easat-5690	98	5	,	,	PUNCT
easat-5690	98	6	neck	neck	NOUN
easat-5690	98	7	,	,	PUNCT
easat-5690	98	8	and	and	CCONJ
easat-5690	98	9	prediction	prediction	NOUN
easat-5690	98	10	are	be	AUX
easat-5690	98	11	the	the	DET
easat-5690	98	12	four	four	NUM
easat-5690	98	13	primary	primary	ADJ
easat-5690	98	14	parts	part	NOUN
easat-5690	98	15	of	of	ADP
easat-5690	98	16	the	the	DET
easat-5690	98	17	yolov5	yolov5	NOUN
easat-5690	98	18	network	network	NOUN
easat-5690	98	19	topology	topology	NOUN
easat-5690	98	20	.	.	PUNCT
easat-5690	99	1	yolov5	yolov5	PROPN
easat-5690	99	2	's	's	PART
easat-5690	99	3	input	input	NOUN
easat-5690	99	4	part	part	NOUN
easat-5690	99	5	serves	serve	VERB
easat-5690	99	6	as	as	ADP
easat-5690	99	7	the	the	DET
easat-5690	99	8	network	network	NOUN
easat-5690	99	9	's	's	PART
easat-5690	99	10	entrance	entrance	NOUN
easat-5690	99	11	point	point	NOUN
easat-5690	99	12	.	.	PUNCT
easat-5690	100	1	here	here	ADV
easat-5690	100	2	,	,	PUNCT
easat-5690	100	3	the	the	DET
easat-5690	100	4	input	input	NOUN
easat-5690	100	5	data	datum	NOUN
easat-5690	100	6	is	be	AUX
easat-5690	100	7	randomly	randomly	ADV
easat-5690	100	8	cropped	crop	VERB
easat-5690	100	9	and	and	CCONJ
easat-5690	100	10	1784	1784	NUM
easat-5690	100	11	edelweiss	edelweiss	PROPN
easat-5690	100	12	applied	apply	VERB
easat-5690	100	13	science	science	NOUN
easat-5690	100	14	and	and	CCONJ
easat-5690	100	15	technology	technology	NOUN
easat-5690	100	16	issn	issn	PROPN
easat-5690	100	17	:	:	PUNCT
easat-5690	100	18	2576	2576	NUM
easat-5690	100	19	-	-	SYM
easat-5690	100	20	8484	8484	NUM
easat-5690	100	21	vol	vol	NOUN
easat-5690	100	22	.	.	PROPN
easat-5690	101	1	9	9	NUM
easat-5690	101	2	,	,	PUNCT
easat-5690	101	3	no	no	INTJ
easat-5690	101	4	.	.	NOUN
easat-5690	102	1	3	3	NUM
easat-5690	102	2	:	:	SYM
easat-5690	102	3	1777	1777	NUM
easat-5690	102	4	-	-	SYM
easat-5690	102	5	1794	1794	NUM
easat-5690	102	6	,	,	PUNCT
easat-5690	102	7	2025	2025	NUM
easat-5690	102	8	doi	doi	NOUN
easat-5690	102	9	:	:	PUNCT
easat-5690	102	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	102	11	©	©	PROPN
easat-5690	102	12	2025	2025	NUM
easat-5690	102	13	by	by	ADP
easat-5690	102	14	the	the	DET
easat-5690	102	15	authors	author	NOUN
easat-5690	102	16	;	;	PUNCT
easat-5690	102	17	licensee	licensee	PROPN
easat-5690	102	18	learning	learning	NOUN
easat-5690	102	19	gate	gate	NOUN
easat-5690	102	20	concatenated	concatenate	VERB
easat-5690	102	21	using	use	VERB
easat-5690	102	22	the	the	DET
easat-5690	102	23	mosaic	mosaic	ADJ
easat-5690	102	24	data	datum	NOUN
easat-5690	102	25	augmentation	augmentation	NOUN
easat-5690	102	26	approach	approach	NOUN
easat-5690	102	27	.	.	PUNCT
easat-5690	103	1	the	the	DET
easat-5690	103	2	backbone	backbone	NOUN
easat-5690	103	3	network	network	NOUN
easat-5690	103	4	is	be	AUX
easat-5690	103	5	the	the	DET
easat-5690	103	6	feature	feature	NOUN
easat-5690	103	7	extraction	extraction	NOUN
easat-5690	103	8	part	part	NOUN
easat-5690	103	9	of	of	ADP
easat-5690	103	10	yolov5	yolov5	NOUN
easat-5690	103	11	,	,	PUNCT
easat-5690	103	12	and	and	CCONJ
easat-5690	103	13	its	its	PRON
easat-5690	103	14	feature	feature	NOUN
easat-5690	103	15	extraction	extraction	NOUN
easat-5690	103	16	capability	capability	NOUN
easat-5690	103	17	directly	directly	ADV
easat-5690	103	18	impacts	impact	VERB
easat-5690	103	19	the	the	DET
easat-5690	103	20	overall	overall	ADJ
easat-5690	103	21	performance	performance	NOUN
easat-5690	103	22	of	of	ADP
easat-5690	103	23	the	the	DET
easat-5690	103	24	network	network	NOUN
easat-5690	103	25	.	.	PUNCT
easat-5690	104	1	the	the	DET
easat-5690	104	2	input	input	NOUN
easat-5690	104	3	feature	feature	NOUN
easat-5690	104	4	map	map	NOUN
easat-5690	104	5	is	be	AUX
easat-5690	104	6	split	split	VERB
easat-5690	104	7	into	into	ADP
easat-5690	104	8	two	two	NUM
easat-5690	104	9	halves	half	NOUN
easat-5690	104	10	by	by	ADP
easat-5690	104	11	yolov5	yolov5	NOUN
easat-5690	104	12	's	's	PART
easat-5690	104	13	use	use	NOUN
easat-5690	104	14	of	of	ADP
easat-5690	104	15	the	the	DET
easat-5690	104	16	cspnet	cspnet	NOUN
easat-5690	104	17	(	(	PUNCT
easat-5690	104	18	cross	cross	NOUN
easat-5690	104	19	stage	stage	NOUN
easat-5690	104	20	partial	partial	ADJ
easat-5690	104	21	network	network	NOUN
easat-5690	104	22	)	)	PUNCT
easat-5690	104	23	structure	structure	NOUN
easat-5690	104	24	during	during	ADP
easat-5690	104	25	the	the	DET
easat-5690	104	26	feature	feature	NOUN
easat-5690	104	27	extraction	extraction	NOUN
easat-5690	104	28	stage	stage	NOUN
easat-5690	104	29	.	.	PUNCT
easat-5690	105	1	one	one	NUM
easat-5690	105	2	portion	portion	NOUN
easat-5690	105	3	is	be	AUX
easat-5690	105	4	directly	directly	ADV
easat-5690	105	5	down	down	ADV
easat-5690	105	6	sampled	sample	VERB
easat-5690	105	7	,	,	PUNCT
easat-5690	105	8	while	while	SCONJ
easat-5690	105	9	the	the	DET
easat-5690	105	10	other	other	ADJ
easat-5690	105	11	is	be	AUX
easat-5690	105	12	processed	process	VERB
easat-5690	105	13	through	through	ADP
easat-5690	105	14	a	a	DET
easat-5690	105	15	sequence	sequence	NOUN
easat-5690	105	16	of	of	ADP
easat-5690	105	17	convolutional	convolutional	ADJ
easat-5690	105	18	layers	layer	NOUN
easat-5690	105	19	.	.	PUNCT
easat-5690	106	1	ultimately	ultimately	ADV
easat-5690	106	2	,	,	PUNCT
easat-5690	106	3	feature	feature	NOUN
easat-5690	106	4	maps	map	NOUN
easat-5690	106	5	from	from	ADP
easat-5690	106	6	these	these	DET
easat-5690	106	7	two	two	NUM
easat-5690	106	8	sections	section	NOUN
easat-5690	106	9	are	be	AUX
easat-5690	106	10	combined	combine	VERB
easat-5690	106	11	.	.	PUNCT
easat-5690	107	1	this	this	DET
easat-5690	107	2	design	design	NOUN
easat-5690	107	3	enables	enable	VERB
easat-5690	107	4	the	the	DET
easat-5690	107	5	network	network	NOUN
easat-5690	107	6	to	to	PART
easat-5690	107	7	have	have	VERB
easat-5690	107	8	stronger	strong	ADJ
easat-5690	107	9	non	non	ADJ
easat-5690	107	10	-	-	ADJ
easat-5690	107	11	linear	linear	ADJ
easat-5690	107	12	expression	expression	NOUN
easat-5690	107	13	ability	ability	NOUN
easat-5690	107	14	,	,	PUNCT
easat-5690	107	15	which	which	PRON
easat-5690	107	16	can	can	AUX
easat-5690	107	17	better	well	ADV
easat-5690	107	18	handle	handle	VERB
easat-5690	107	19	complex	complex	ADJ
easat-5690	107	20	backgrounds	background	NOUN
easat-5690	107	21	and	and	CCONJ
easat-5690	107	22	diverse	diverse	ADJ
easat-5690	107	23	objects	object	NOUN
easat-5690	107	24	in	in	ADP
easat-5690	107	25	object	object	NOUN
easat-5690	107	26	detection	detection	NOUN
easat-5690	107	27	tasks	task	NOUN
easat-5690	107	28	,	,	PUNCT
easat-5690	107	29	as	as	SCONJ
easat-5690	107	30	shown	show	VERB
easat-5690	107	31	in	in	ADP
easat-5690	107	32	figure	figure	NOUN
easat-5690	107	33	4	4	NUM
easat-5690	107	34	.	.	SYM
easat-5690	107	35	1785	1785	NUM
easat-5690	107	36	edelweiss	edelweiss	PROPN
easat-5690	107	37	applied	apply	VERB
easat-5690	107	38	science	science	NOUN
easat-5690	107	39	and	and	CCONJ
easat-5690	107	40	technology	technology	NOUN
easat-5690	107	41	issn	issn	PROPN
easat-5690	107	42	:	:	PUNCT
easat-5690	107	43	2576	2576	NUM
easat-5690	107	44	-	-	SYM
easat-5690	107	45	8484	8484	NUM
easat-5690	107	46	vol	vol	NOUN
easat-5690	107	47	.	.	PROPN
easat-5690	108	1	9	9	NUM
easat-5690	108	2	,	,	PUNCT
easat-5690	108	3	no	no	INTJ
easat-5690	108	4	.	.	NOUN
easat-5690	109	1	3	3	NUM
easat-5690	109	2	:	:	SYM
easat-5690	109	3	1777	1777	NUM
easat-5690	109	4	-	-	SYM
easat-5690	109	5	1794	1794	NUM
easat-5690	109	6	,	,	PUNCT
easat-5690	109	7	2025	2025	NUM
easat-5690	109	8	doi	doi	NOUN
easat-5690	109	9	:	:	PUNCT
easat-5690	109	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	109	11	©	©	PROPN
easat-5690	109	12	2025	2025	NUM
easat-5690	109	13	by	by	ADP
easat-5690	109	14	the	the	DET
easat-5690	109	15	authors	author	NOUN
easat-5690	109	16	;	;	PUNCT
easat-5690	109	17	licensee	licensee	PROPN
easat-5690	109	18	learning	learn	VERB
easat-5690	109	19	gate	gate	NOUN
easat-5690	109	20	figure	figure	NOUN
easat-5690	109	21	4	4	NUM
easat-5690	109	22	.	.	PUNCT
easat-5690	110	1	the	the	DET
easat-5690	110	2	component	component	NOUN
easat-5690	110	3	networks	network	NOUN
easat-5690	110	4	in	in	ADP
easat-5690	110	5	the	the	DET
easat-5690	110	6	original	original	ADJ
easat-5690	110	7	yolov5s	yolov5s	PROPN
easat-5690	110	8	.	.	PUNCT
easat-5690	111	1	the	the	DET
easat-5690	111	2	convolutional	convolutional	ADJ
easat-5690	111	3	module	module	NOUN
easat-5690	111	4	is	be	AUX
easat-5690	111	5	a	a	DET
easat-5690	111	6	commonly	commonly	ADV
easat-5690	111	7	used	use	VERB
easat-5690	111	8	fundamental	fundamental	ADJ
easat-5690	111	9	block	block	NOUN
easat-5690	111	10	in	in	ADP
easat-5690	111	11	convolutional	convolutional	ADJ
easat-5690	111	12	neural	neural	ADJ
easat-5690	111	13	networks	network	NOUN
easat-5690	111	14	,	,	PUNCT
easat-5690	111	15	consisting	consist	VERB
easat-5690	111	16	primarily	primarily	ADV
easat-5690	111	17	of	of	ADP
easat-5690	111	18	convolutional	convolutional	ADJ
easat-5690	111	19	layers	layer	NOUN
easat-5690	111	20	,	,	PUNCT
easat-5690	111	21	batch	batch	VERB
easat-5690	111	22	normalization	normalization	NOUN
easat-5690	111	23	(	(	PUNCT
easat-5690	111	24	bn	bn	NOUN
easat-5690	111	25	)	)	PUNCT
easat-5690	111	26	layers	layer	NOUN
easat-5690	111	27	,	,	PUNCT
easat-5690	111	28	and	and	CCONJ
easat-5690	111	29	activation	activation	NOUN
easat-5690	111	30	functions	function	NOUN
easat-5690	111	31	.	.	PUNCT
easat-5690	112	1	one	one	NUM
easat-5690	112	2	of	of	ADP
easat-5690	112	3	the	the	DET
easat-5690	112	4	fundamental	fundamental	ADJ
easat-5690	112	5	layers	layer	NOUN
easat-5690	112	6	in	in	ADP
easat-5690	112	7	a	a	DET
easat-5690	112	8	convolutional	convolutional	ADJ
easat-5690	112	9	neural	neural	ADJ
easat-5690	112	10	network	network	NOUN
easat-5690	112	11	is	be	AUX
easat-5690	112	12	the	the	DET
easat-5690	112	13	convolutional	convolutional	ADJ
easat-5690	112	14	layer	layer	NOUN
easat-5690	112	15	,	,	PUNCT
easat-5690	112	16	which	which	PRON
easat-5690	112	17	serves	serve	VERB
easat-5690	112	18	to	to	PART
easat-5690	112	19	extract	extract	VERB
easat-5690	112	20	local	local	ADJ
easat-5690	112	21	spatial	spatial	ADJ
easat-5690	112	22	information	information	NOUN
easat-5690	112	23	from	from	ADP
easat-5690	112	24	input	input	NOUN
easat-5690	112	25	features	feature	NOUN
easat-5690	112	26	.	.	PUNCT
easat-5690	113	1	following	follow	VERB
easat-5690	113	2	the	the	DET
easat-5690	113	3	convolutional	convolutional	ADJ
easat-5690	113	4	layer	layer	NOUN
easat-5690	113	5	,	,	PUNCT
easat-5690	113	6	a	a	DET
easat-5690	113	7	bn	bn	NOUN
easat-5690	113	8	layer	layer	NOUN
easat-5690	113	9	is	be	AUX
easat-5690	113	10	added	add	VERB
easat-5690	113	11	,	,	PUNCT
easat-5690	113	12	functioning	function	VERB
easat-5690	113	13	as	as	ADP
easat-5690	113	14	a	a	DET
easat-5690	113	15	normalization	normalization	NOUN
easat-5690	113	16	layer	layer	NOUN
easat-5690	113	17	that	that	PRON
easat-5690	113	18	standardizes	standardize	VERB
easat-5690	113	19	the	the	DET
easat-5690	113	20	distribution	distribution	NOUN
easat-5690	113	21	of	of	ADP
easat-5690	113	22	feature	feature	NOUN
easat-5690	113	23	values	value	NOUN
easat-5690	113	24	within	within	ADP
easat-5690	113	25	the	the	DET
easat-5690	113	26	neural	neural	ADJ
easat-5690	113	27	network	network	NOUN
easat-5690	113	28	.	.	PUNCT
easat-5690	114	1	the	the	DET
easat-5690	114	2	activation	activation	NOUN
easat-5690	114	3	function	function	NOUN
easat-5690	114	4	is	be	AUX
easat-5690	114	5	a	a	DET
easat-5690	114	6	nonlinear	nonlinear	ADJ
easat-5690	114	7	function	function	NOUN
easat-5690	114	8	that	that	PRON
easat-5690	114	9	introduces	introduce	VERB
easat-5690	114	10	nonlinear	nonlinear	ADJ
easat-5690	114	11	transformation	transformation	NOUN
easat-5690	114	12	capabilities	capability	NOUN
easat-5690	114	13	into	into	ADP
easat-5690	114	14	the	the	DET
easat-5690	114	15	neural	neural	ADJ
easat-5690	114	16	network	network	NOUN
easat-5690	114	17	.	.	PUNCT
easat-5690	115	1	common	common	ADJ
easat-5690	115	2	activation	activation	NOUN
easat-5690	115	3	functions	function	NOUN
easat-5690	115	4	include	include	VERB
easat-5690	115	5	sigmoid	sigmoid	NOUN
easat-5690	115	6	,	,	PUNCT
easat-5690	115	7	relu	relu	NOUN
easat-5690	115	8	,	,	PUNCT
easat-5690	115	9	leakyrelu	leakyrelu	NOUN
easat-5690	115	10	,	,	PUNCT
easat-5690	115	11	elu	elu	PROPN
easat-5690	115	12	,	,	PUNCT
easat-5690	115	13	and	and	CCONJ
easat-5690	115	14	others	other	NOUN
easat-5690	115	15	.	.	PUNCT
easat-5690	116	1	the	the	DET
easat-5690	116	2	three	three	NUM
easat-5690	116	3	detect	detect	NOUN
easat-5690	116	4	detectors	detector	NOUN
easat-5690	116	5	that	that	PRON
easat-5690	116	6	make	make	VERB
easat-5690	116	7	up	up	ADP
easat-5690	116	8	the	the	DET
easat-5690	116	9	head	head	NOUN
easat-5690	116	10	's	's	PART
easat-5690	116	11	main	main	ADJ
easat-5690	116	12	body	body	NOUN
easat-5690	116	13	use	use	VERB
easat-5690	116	14	grid	grid	NOUN
easat-5690	116	15	-	-	PUNCT
easat-5690	116	16	based	base	VERB
easat-5690	116	17	anchoring	anchoring	NOUN
easat-5690	116	18	to	to	PART
easat-5690	116	19	identify	identify	VERB
easat-5690	116	20	objects	object	NOUN
easat-5690	116	21	on	on	ADP
easat-5690	116	22	feature	feature	NOUN
easat-5690	116	23	maps	map	NOUN
easat-5690	116	24	with	with	ADP
easat-5690	116	25	varying	vary	VERB
easat-5690	116	26	scales	scale	NOUN
easat-5690	116	27	.	.	PUNCT
easat-5690	117	1	2.3	2.3	NUM
easat-5690	117	2	.	.	PUNCT
easat-5690	117	3	improved	improve	VERB
easat-5690	117	4	yolov5s	yolov5s	NOUN
easat-5690	117	5	the	the	DET
easat-5690	117	6	algorithm	algorithm	NOUN
easat-5690	117	7	must	must	AUX
easat-5690	117	8	be	be	AUX
easat-5690	117	9	enhanced	enhance	VERB
easat-5690	117	10	in	in	ADP
easat-5690	117	11	order	order	NOUN
easat-5690	117	12	to	to	PART
easat-5690	117	13	increase	increase	VERB
easat-5690	117	14	the	the	DET
easat-5690	117	15	accuracy	accuracy	NOUN
easat-5690	117	16	of	of	ADP
easat-5690	117	17	the	the	DET
easat-5690	117	18	yolov5	yolov5	NOUN
easat-5690	117	19	series	series	NOUN
easat-5690	117	20	models	model	NOUN
easat-5690	117	21	.	.	PUNCT
easat-5690	118	1	•	•	NUM
easat-5690	118	2	the	the	DET
easat-5690	118	3	network	network	NOUN
easat-5690	118	4	's	's	PART
easat-5690	118	5	feature	feature	NOUN
easat-5690	118	6	extraction	extraction	NOUN
easat-5690	118	7	capability	capability	NOUN
easat-5690	118	8	is	be	AUX
easat-5690	118	9	enhanced	enhance	VERB
easat-5690	118	10	by	by	ADP
easat-5690	118	11	improving	improve	VERB
easat-5690	118	12	the	the	DET
easat-5690	118	13	feature	feature	NOUN
easat-5690	118	14	extraction	extraction	NOUN
easat-5690	118	15	structure	structure	NOUN
easat-5690	118	16	;	;	PUNCT
easat-5690	118	17	•	•	NUM
easat-5690	118	18	a	a	DET
easat-5690	118	19	tiny	tiny	ADJ
easat-5690	118	20	target	target	NOUN
easat-5690	118	21	detection	detection	NOUN
easat-5690	118	22	layer	layer	NOUN
easat-5690	118	23	is	be	AUX
easat-5690	118	24	added	add	VERB
easat-5690	118	25	to	to	PART
easat-5690	118	26	enhance	enhance	VERB
easat-5690	118	27	the	the	DET
easat-5690	118	28	detection	detection	NOUN
easat-5690	118	29	performance	performance	NOUN
easat-5690	118	30	of	of	ADP
easat-5690	118	31	small	small	ADJ
easat-5690	118	32	targets	target	NOUN
easat-5690	118	33	.	.	PUNCT
easat-5690	119	1	•	•	ADV
easat-5690	119	2	enhance	enhance	VERB
easat-5690	119	3	the	the	DET
easat-5690	119	4	model	model	NOUN
easat-5690	119	5	's	's	PART
easat-5690	119	6	accuracy	accuracy	NOUN
easat-5690	119	7	and	and	CCONJ
easat-5690	119	8	the	the	DET
easat-5690	119	9	backbone	backbone	NOUN
easat-5690	119	10	network	network	NOUN
easat-5690	119	11	.	.	PUNCT
easat-5690	120	1	as	as	SCONJ
easat-5690	120	2	shown	show	VERB
easat-5690	120	3	in	in	ADP
easat-5690	120	4	fig	fig	NOUN
easat-5690	120	5	.	.	PUNCT
easat-5690	121	1	5	5	NUM
easat-5690	121	2	,	,	PUNCT
easat-5690	121	3	the	the	DET
easat-5690	121	4	improvements	improvement	NOUN
easat-5690	121	5	to	to	ADP
easat-5690	121	6	this	this	DET
easat-5690	121	7	model	model	NOUN
easat-5690	121	8	include	include	VERB
easat-5690	121	9	changing	change	VERB
easat-5690	121	10	csp	csp	PROPN
easat-5690	121	11	to	to	PART
easat-5690	121	12	cspk	cspk	PROPN
easat-5690	121	13	,	,	PUNCT
easat-5690	121	14	enhancing	enhance	VERB
easat-5690	121	15	the	the	DET
easat-5690	121	16	convergence	convergence	NOUN
easat-5690	121	17	of	of	ADP
easat-5690	121	18	feature	feature	NOUN
easat-5690	121	19	maps	map	NOUN
easat-5690	121	20	,	,	PUNCT
easat-5690	121	21	preventing	prevent	VERB
easat-5690	121	22	feature	feature	NOUN
easat-5690	121	23	loss	loss	NOUN
easat-5690	121	24	,	,	PUNCT
easat-5690	121	25	and	and	CCONJ
easat-5690	121	26	adding	add	VERB
easat-5690	121	27	bn	bn	ADP
easat-5690	121	28	layers	layer	NOUN
easat-5690	121	29	and	and	CCONJ
easat-5690	121	30	the	the	DET
easat-5690	121	31	leakyrelu	leakyrelu	ADJ
easat-5690	121	32	activation	activation	NOUN
easat-5690	121	33	function	function	VERB
easat-5690	121	34	to	to	PART
easat-5690	121	35	improve	improve	VERB
easat-5690	121	36	the	the	DET
easat-5690	121	37	stability	stability	NOUN
easat-5690	121	38	and	and	CCONJ
easat-5690	121	39	generalization	generalization	NOUN
easat-5690	121	40	performance	performance	NOUN
easat-5690	121	41	of	of	ADP
easat-5690	121	42	the	the	DET
easat-5690	121	43	model	model	NOUN
easat-5690	121	44	.	.	PUNCT
easat-5690	122	1	this	this	PRON
easat-5690	122	2	will	will	AUX
easat-5690	122	3	allow	allow	VERB
easat-5690	122	4	the	the	DET
easat-5690	122	5	model	model	NOUN
easat-5690	122	6	to	to	PART
easat-5690	122	7	pick	pick	VERB
easat-5690	122	8	up	up	ADP
easat-5690	122	9	additional	additional	ADJ
easat-5690	122	10	features	feature	NOUN
easat-5690	122	11	.	.	PUNCT
easat-5690	123	1	and	and	CCONJ
easat-5690	123	2	by	by	ADP
easat-5690	123	3	adding	add	VERB
easat-5690	123	4	odconv2d	odconv2d	NOUN
easat-5690	123	5	and	and	CCONJ
easat-5690	123	6	cdam	cdam	NOUN
easat-5690	123	7	to	to	ADP
easat-5690	123	8	the	the	DET
easat-5690	123	9	model	model	NOUN
easat-5690	123	10	's	's	PART
easat-5690	123	11	head	head	NOUN
easat-5690	123	12	to	to	PART
easat-5690	123	13	further	far	ADV
easat-5690	123	14	enhance	enhance	VERB
easat-5690	123	15	feature	feature	NOUN
easat-5690	123	16	adaptability	adaptability	NOUN
easat-5690	123	17	and	and	CCONJ
easat-5690	123	18	improve	improve	VERB
easat-5690	123	19	the	the	DET
easat-5690	123	20	performance	performance	NOUN
easat-5690	123	21	of	of	ADP
easat-5690	123	22	focusing	focus	VERB
easat-5690	123	23	on	on	ADP
easat-5690	123	24	essential	essential	ADJ
easat-5690	123	25	features	feature	NOUN
easat-5690	123	26	,	,	PUNCT
easat-5690	123	27	as	as	SCONJ
easat-5690	123	28	shown	show	VERB
easat-5690	123	29	in	in	ADP
easat-5690	123	30	1786	1786	NUM
easat-5690	123	31	edelweiss	edelweiss	PROPN
easat-5690	123	32	applied	apply	VERB
easat-5690	123	33	science	science	NOUN
easat-5690	123	34	and	and	CCONJ
easat-5690	123	35	technology	technology	NOUN
easat-5690	123	36	issn	issn	PROPN
easat-5690	123	37	:	:	PUNCT
easat-5690	123	38	2576	2576	NUM
easat-5690	123	39	-	-	SYM
easat-5690	123	40	8484	8484	NUM
easat-5690	123	41	vol	vol	NOUN
easat-5690	123	42	.	.	PROPN
easat-5690	124	1	9	9	NUM
easat-5690	124	2	,	,	PUNCT
easat-5690	124	3	no	no	INTJ
easat-5690	124	4	.	.	NOUN
easat-5690	125	1	3	3	NUM
easat-5690	125	2	:	:	SYM
easat-5690	125	3	1777	1777	NUM
easat-5690	125	4	-	-	SYM
easat-5690	125	5	1794	1794	NUM
easat-5690	125	6	,	,	PUNCT
easat-5690	125	7	2025	2025	NUM
easat-5690	125	8	doi	doi	NOUN
easat-5690	125	9	:	:	PUNCT
easat-5690	125	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	125	11	©	©	PROPN
easat-5690	125	12	2025	2025	NUM
easat-5690	125	13	by	by	ADP
easat-5690	125	14	the	the	DET
easat-5690	125	15	authors	author	NOUN
easat-5690	125	16	;	;	PUNCT
easat-5690	125	17	licensee	licensee	PROPN
easat-5690	125	18	learning	learn	VERB
easat-5690	125	19	gate	gate	PROPN
easat-5690	125	20	figure	figure	NOUN
easat-5690	125	21	5	5	NUM
easat-5690	125	22	.	.	PUNCT
easat-5690	125	23	figure	figure	NOUN
easat-5690	125	24	5	5	NUM
easat-5690	125	25	.	.	PUNCT
easat-5690	126	1	the	the	DET
easat-5690	126	2	network	network	NOUN
easat-5690	126	3	structure	structure	NOUN
easat-5690	126	4	of	of	ADP
easat-5690	126	5	the	the	DET
easat-5690	126	6	improved	improved	ADJ
easat-5690	126	7	yolov5s	yolov5s	PROPN
easat-5690	126	8	.	.	PUNCT
easat-5690	127	1	in	in	ADP
easat-5690	127	2	neural	neural	ADJ
easat-5690	127	3	network	network	NOUN
easat-5690	127	4	architectures	architecture	NOUN
easat-5690	127	5	,	,	PUNCT
easat-5690	127	6	csp	csp	PROPN
easat-5690	127	7	(	(	PUNCT
easat-5690	127	8	cross	cross	NOUN
easat-5690	127	9	stage	stage	NOUN
easat-5690	127	10	partial	partial	NOUN
easat-5690	127	11	)	)	PUNCT
easat-5690	127	12	is	be	AUX
easat-5690	127	13	a	a	DET
easat-5690	127	14	connection	connection	NOUN
easat-5690	127	15	approach	approach	NOUN
easat-5690	127	16	that	that	PRON
easat-5690	127	17	is	be	AUX
easat-5690	127	18	mostly	mostly	ADV
easat-5690	127	19	used	use	VERB
easat-5690	127	20	to	to	PART
easat-5690	127	21	improve	improve	VERB
easat-5690	127	22	the	the	DET
easat-5690	127	23	information	information	NOUN
easat-5690	127	24	transmission	transmission	NOUN
easat-5690	127	25	and	and	CCONJ
easat-5690	127	26	interaction	interaction	NOUN
easat-5690	127	27	of	of	ADP
easat-5690	127	28	features	feature	NOUN
easat-5690	127	29	across	across	ADP
easat-5690	127	30	several	several	ADJ
easat-5690	127	31	stages	stage	NOUN
easat-5690	127	32	or	or	CCONJ
easat-5690	127	33	scales	scale	NOUN
easat-5690	127	34	.	.	PUNCT
easat-5690	128	1	1787	1787	NUM
easat-5690	128	2	edelweiss	edelweiss	PROPN
easat-5690	128	3	applied	apply	VERB
easat-5690	128	4	science	science	NOUN
easat-5690	128	5	and	and	CCONJ
easat-5690	128	6	technology	technology	NOUN
easat-5690	128	7	issn	issn	PROPN
easat-5690	128	8	:	:	PUNCT
easat-5690	128	9	2576	2576	NUM
easat-5690	128	10	-	-	SYM
easat-5690	128	11	8484	8484	NUM
easat-5690	128	12	vol	vol	NOUN
easat-5690	128	13	.	.	PROPN
easat-5690	129	1	9	9	NUM
easat-5690	129	2	,	,	PUNCT
easat-5690	129	3	no	no	INTJ
easat-5690	129	4	.	.	NOUN
easat-5690	130	1	3	3	NUM
easat-5690	130	2	:	:	SYM
easat-5690	130	3	1777	1777	NUM
easat-5690	130	4	-	-	SYM
easat-5690	130	5	1794	1794	NUM
easat-5690	130	6	,	,	PUNCT
easat-5690	130	7	2025	2025	NUM
easat-5690	130	8	doi	doi	NOUN
easat-5690	130	9	:	:	PUNCT
easat-5690	130	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	130	11	©	©	PROPN
easat-5690	130	12	2025	2025	NUM
easat-5690	130	13	by	by	ADP
easat-5690	130	14	the	the	DET
easat-5690	130	15	authors	author	NOUN
easat-5690	130	16	;	;	PUNCT
easat-5690	130	17	licensee	licensee	PROPN
easat-5690	130	18	learning	learning	NOUN
easat-5690	130	19	gate	gate	VERB
easat-5690	130	20	the	the	DET
easat-5690	130	21	term	term	NOUN
easat-5690	130	22	"	"	PUNCT
easat-5690	130	23	conv2d	conv2d	NOUN
easat-5690	130	24	"	"	PUNCT
easat-5690	130	25	describes	describe	VERB
easat-5690	130	26	a	a	DET
easat-5690	130	27	convolution	convolution	NOUN
easat-5690	130	28	procedure	procedure	NOUN
easat-5690	130	29	that	that	PRON
easat-5690	130	30	is	be	AUX
easat-5690	130	31	frequently	frequently	ADV
easat-5690	130	32	used	use	VERB
easat-5690	130	33	to	to	PART
easat-5690	130	34	process	process	VERB
easat-5690	130	35	picture	picture	NOUN
easat-5690	130	36	data	datum	NOUN
easat-5690	130	37	when	when	SCONJ
easat-5690	130	38	it	it	PRON
easat-5690	130	39	is	be	AUX
easat-5690	130	40	applied	apply	VERB
easat-5690	130	41	to	to	ADP
easat-5690	130	42	two	two	NUM
easat-5690	130	43	-	-	PUNCT
easat-5690	130	44	dimensional	dimensional	ADJ
easat-5690	130	45	data	datum	NOUN
easat-5690	130	46	.	.	PUNCT
easat-5690	131	1	it	it	PRON
easat-5690	131	2	takes	take	VERB
easat-5690	131	3	in	in	ADP
easat-5690	131	4	as	as	ADP
easat-5690	131	5	input	input	NOUN
easat-5690	131	6	a	a	DET
easat-5690	131	7	three	three	NUM
easat-5690	131	8	-	-	PUNCT
easat-5690	131	9	dimensional	dimensional	ADJ
easat-5690	131	10	tensor	tensor	NOUN
easat-5690	131	11	,	,	PUNCT
easat-5690	131	12	of	of	ADP
easat-5690	131	13	which	which	PRON
easat-5690	131	14	the	the	DET
easat-5690	131	15	first	first	ADJ
easat-5690	131	16	and	and	CCONJ
easat-5690	131	17	second	second	ADJ
easat-5690	131	18	dimensions	dimension	NOUN
easat-5690	131	19	stand	stand	VERB
easat-5690	131	20	for	for	ADP
easat-5690	131	21	the	the	DET
easat-5690	131	22	image	image	NOUN
easat-5690	131	23	's	's	PART
easat-5690	131	24	height	height	NOUN
easat-5690	131	25	and	and	CCONJ
easat-5690	131	26	breadth	breadth	NOUN
easat-5690	131	27	,	,	PUNCT
easat-5690	131	28	and	and	CCONJ
easat-5690	131	29	the	the	DET
easat-5690	131	30	third	third	ADJ
easat-5690	131	31	dimension	dimension	NOUN
easat-5690	131	32	for	for	ADP
easat-5690	131	33	the	the	DET
easat-5690	131	34	image	image	NOUN
easat-5690	131	35	's	's	PART
easat-5690	131	36	channel	channel	NOUN
easat-5690	131	37	count	count	NOUN
easat-5690	131	38	.	.	PUNCT
easat-5690	132	1	csp	csp	PROPN
easat-5690	132	2	plays	play	VERB
easat-5690	132	3	a	a	DET
easat-5690	132	4	crucial	crucial	ADJ
easat-5690	132	5	role	role	NOUN
easat-5690	132	6	in	in	ADP
easat-5690	132	7	boosting	boost	VERB
easat-5690	132	8	the	the	DET
easat-5690	132	9	network	network	NOUN
easat-5690	132	10	's	's	PART
easat-5690	132	11	representational	representational	ADJ
easat-5690	132	12	capacity	capacity	NOUN
easat-5690	132	13	through	through	ADP
easat-5690	132	14	improved	improve	VERB
easat-5690	132	15	feature	feature	NOUN
easat-5690	132	16	extraction	extraction	NOUN
easat-5690	132	17	and	and	CCONJ
easat-5690	132	18	representation	representation	NOUN
easat-5690	132	19	,	,	PUNCT
easat-5690	132	20	which	which	PRON
easat-5690	132	21	leads	lead	VERB
easat-5690	132	22	to	to	ADP
easat-5690	132	23	improved	improved	ADJ
easat-5690	132	24	performance	performance	NOUN
easat-5690	132	25	in	in	ADP
easat-5690	132	26	a	a	DET
easat-5690	132	27	variety	variety	NOUN
easat-5690	132	28	of	of	ADP
easat-5690	132	29	computer	computer	NOUN
easat-5690	132	30	vision	vision	NOUN
easat-5690	132	31	applications	application	NOUN
easat-5690	132	32	.	.	PUNCT
easat-5690	133	1	the	the	DET
easat-5690	133	2	cspk	cspk	PROPN
easat-5690	133	3	structure	structure	NOUN
easat-5690	133	4	can	can	AUX
easat-5690	133	5	enhance	enhance	VERB
easat-5690	133	6	feature	feature	NOUN
easat-5690	133	7	propagation	propagation	NOUN
easat-5690	133	8	more	more	ADV
easat-5690	133	9	,	,	PUNCT
easat-5690	133	10	aiding	aid	VERB
easat-5690	133	11	in	in	ADP
easat-5690	133	12	better	well	ADJ
easat-5690	133	13	information	information	NOUN
easat-5690	133	14	transmission	transmission	NOUN
easat-5690	133	15	across	across	ADP
easat-5690	133	16	different	different	ADJ
easat-5690	133	17	layers	layer	NOUN
easat-5690	133	18	of	of	ADP
easat-5690	133	19	the	the	DET
easat-5690	133	20	network	network	NOUN
easat-5690	133	21	,	,	PUNCT
easat-5690	133	22	thereby	thereby	ADV
easat-5690	133	23	improving	improve	VERB
easat-5690	133	24	performance	performance	NOUN
easat-5690	133	25	,	,	PUNCT
easat-5690	133	26	as	as	SCONJ
easat-5690	133	27	shown	show	VERB
easat-5690	133	28	in	in	ADP
easat-5690	133	29	figure	figure	NOUN
easat-5690	133	30	6	6	NUM
easat-5690	133	31	.	.	SYM
easat-5690	133	32	1788	1788	NUM
easat-5690	133	33	edelweiss	edelweiss	PROPN
easat-5690	133	34	applied	apply	VERB
easat-5690	133	35	science	science	NOUN
easat-5690	133	36	and	and	CCONJ
easat-5690	133	37	technology	technology	NOUN
easat-5690	133	38	issn	issn	PROPN
easat-5690	133	39	:	:	PUNCT
easat-5690	133	40	2576	2576	NUM
easat-5690	133	41	-	-	SYM
easat-5690	133	42	8484	8484	NUM
easat-5690	133	43	vol	vol	NOUN
easat-5690	133	44	.	.	PROPN
easat-5690	134	1	9	9	NUM
easat-5690	134	2	,	,	PUNCT
easat-5690	134	3	no	no	INTJ
easat-5690	134	4	.	.	NOUN
easat-5690	135	1	3	3	NUM
easat-5690	135	2	:	:	SYM
easat-5690	135	3	1777	1777	NUM
easat-5690	135	4	-	-	SYM
easat-5690	135	5	1794	1794	NUM
easat-5690	135	6	,	,	PUNCT
easat-5690	135	7	2025	2025	NUM
easat-5690	135	8	doi	doi	NOUN
easat-5690	135	9	:	:	PUNCT
easat-5690	135	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	135	11	©	©	PROPN
easat-5690	135	12	2025	2025	NUM
easat-5690	135	13	by	by	ADP
easat-5690	135	14	the	the	DET
easat-5690	135	15	authors	author	NOUN
easat-5690	135	16	;	;	PUNCT
easat-5690	135	17	licensee	licensee	PROPN
easat-5690	135	18	learning	learn	VERB
easat-5690	135	19	gate	gate	PROPN
easat-5690	135	20	figure	figure	NOUN
easat-5690	135	21	6	6	NUM
easat-5690	135	22	.	.	PUNCT
easat-5690	135	23	comparison	comparison	NOUN
easat-5690	135	24	of	of	ADP
easat-5690	135	25	csp	csp	PROPN
easat-5690	135	26	and	and	CCONJ
easat-5690	135	27	cspk	cspk	PROPN
easat-5690	135	28	structure	structure	NOUN
easat-5690	135	29	.	.	PUNCT
easat-5690	136	1	adding	add	VERB
easat-5690	136	2	odconv2d	odconv2d	NOUN
easat-5690	136	3	and	and	CCONJ
easat-5690	136	4	cbam	cbam	NOUN
easat-5690	136	5	(	(	PUNCT
easat-5690	136	6	see	see	VERB
easat-5690	136	7	figure	figure	NOUN
easat-5690	136	8	7	7	NUM
easat-5690	136	9	)	)	PUNCT
easat-5690	136	10	to	to	ADP
easat-5690	136	11	the	the	DET
easat-5690	136	12	model	model	NOUN
easat-5690	136	13	's	's	PART
easat-5690	136	14	head	head	NOUN
easat-5690	136	15	to	to	PART
easat-5690	136	16	further	far	ADV
easat-5690	136	17	enhance	enhance	VERB
easat-5690	136	18	feature	feature	NOUN
easat-5690	136	19	adaptability	adaptability	NOUN
easat-5690	136	20	and	and	CCONJ
easat-5690	136	21	improve	improve	VERB
easat-5690	136	22	the	the	DET
easat-5690	136	23	performance	performance	NOUN
easat-5690	136	24	of	of	ADP
easat-5690	136	25	focusing	focus	VERB
easat-5690	136	26	on	on	ADP
easat-5690	136	27	essential	essential	ADJ
easat-5690	136	28	features	feature	NOUN
easat-5690	136	29	.	.	PUNCT
easat-5690	137	1	dynamic	dynamic	ADJ
easat-5690	137	2	convolution	convolution	NOUN
easat-5690	137	3	refers	refer	VERB
easat-5690	137	4	to	to	ADP
easat-5690	137	5	the	the	DET
easat-5690	137	6	concept	concept	NOUN
easat-5690	137	7	where	where	SCONJ
easat-5690	137	8	the	the	DET
easat-5690	137	9	weights	weight	NOUN
easat-5690	137	10	of	of	ADP
easat-5690	137	11	the	the	DET
easat-5690	137	12	convolutional	convolutional	ADJ
easat-5690	137	13	kernel	kernel	NOUN
easat-5690	137	14	are	be	AUX
easat-5690	137	15	not	not	PART
easat-5690	137	16	fixed	fix	VERB
easat-5690	137	17	during	during	ADP
easat-5690	137	18	the	the	DET
easat-5690	137	19	convolution	convolution	NOUN
easat-5690	137	20	process	process	NOUN
easat-5690	137	21	.	.	PUNCT
easat-5690	138	1	instead	instead	ADV
easat-5690	138	2	,	,	PUNCT
easat-5690	138	3	they	they	PRON
easat-5690	138	4	can	can	AUX
easat-5690	138	5	be	be	AUX
easat-5690	138	6	dynamically	dynamically	ADV
easat-5690	138	7	adjusted	adjust	VERB
easat-5690	138	8	based	base	VERB
easat-5690	138	9	on	on	ADP
easat-5690	138	10	the	the	DET
easat-5690	138	11	input	input	NOUN
easat-5690	138	12	data	datum	NOUN
easat-5690	138	13	.	.	PUNCT
easat-5690	139	1	this	this	DET
easat-5690	139	2	adaptability	adaptability	NOUN
easat-5690	139	3	of	of	ADP
easat-5690	139	4	the	the	DET
easat-5690	139	5	convolutional	convolutional	ADJ
easat-5690	139	6	kernel	kernel	NOUN
easat-5690	139	7	allows	allow	VERB
easat-5690	139	8	it	it	PRON
easat-5690	139	9	to	to	PART
easat-5690	139	10	better	well	ADV
easat-5690	139	11	capture	capture	VERB
easat-5690	139	12	the	the	DET
easat-5690	139	13	features	feature	NOUN
easat-5690	139	14	present	present	ADJ
easat-5690	139	15	in	in	ADP
easat-5690	139	16	the	the	DET
easat-5690	139	17	input	input	NOUN
easat-5690	139	18	data	datum	NOUN
easat-5690	139	19	,	,	PUNCT
easat-5690	139	20	thereby	thereby	ADV
easat-5690	139	21	enhancing	enhance	VERB
easat-5690	139	22	the	the	DET
easat-5690	139	23	performance	performance	NOUN
easat-5690	139	24	of	of	ADP
easat-5690	139	25	convolutional	convolutional	ADJ
easat-5690	139	26	networks	network	NOUN
easat-5690	139	27	.	.	PUNCT
easat-5690	140	1	attention	attention	NOUN
easat-5690	140	2	mechanisms	mechanism	NOUN
easat-5690	140	3	are	be	AUX
easat-5690	140	4	crucial	crucial	ADJ
easat-5690	140	5	neural	neural	ADJ
easat-5690	140	6	network	network	NOUN
easat-5690	140	7	modules	module	NOUN
easat-5690	140	8	that	that	PRON
easat-5690	140	9	enable	enable	VERB
easat-5690	140	10	networks	network	NOUN
easat-5690	140	11	to	to	PART
easat-5690	140	12	focus	focus	VERB
easat-5690	140	13	more	more	ADV
easat-5690	140	14	effectively	effectively	ADV
easat-5690	140	15	on	on	ADP
easat-5690	140	16	relevant	relevant	ADJ
easat-5690	140	17	information	information	NOUN
easat-5690	140	18	when	when	SCONJ
easat-5690	140	19	processing	process	VERB
easat-5690	140	20	sequential	sequential	ADJ
easat-5690	140	21	data	datum	NOUN
easat-5690	140	22	.	.	PUNCT
easat-5690	141	1	attention	attention	NOUN
easat-5690	141	2	mechanisms	mechanism	NOUN
easat-5690	141	3	guarantee	guarantee	VERB
easat-5690	141	4	that	that	SCONJ
easat-5690	141	5	the	the	DET
easat-5690	141	6	network	network	NOUN
easat-5690	141	7	gives	give	VERB
easat-5690	141	8	more	more	ADJ
easat-5690	141	9	attention	attention	NOUN
easat-5690	141	10	to	to	ADP
easat-5690	141	11	crucial	crucial	ADJ
easat-5690	141	12	information	information	NOUN
easat-5690	141	13	during	during	ADP
easat-5690	141	14	data	data	NOUN
easat-5690	141	15	processing	processing	NOUN
easat-5690	141	16	and	and	CCONJ
easat-5690	141	17	transmission	transmission	NOUN
easat-5690	141	18	by	by	ADP
easat-5690	141	19	allocating	allocate	VERB
easat-5690	141	20	varying	vary	VERB
easat-5690	141	21	weights	weight	NOUN
easat-5690	141	22	to	to	ADP
easat-5690	141	23	various	various	ADJ
easat-5690	141	24	input	input	NOUN
easat-5690	141	25	data	datum	NOUN
easat-5690	141	26	components	component	NOUN
easat-5690	141	27	according	accord	VERB
easat-5690	141	28	to	to	ADP
easat-5690	141	29	their	their	PRON
easat-5690	141	30	importance	importance	NOUN
easat-5690	141	31	.	.	PUNCT
easat-5690	142	1	1789	1789	NUM
easat-5690	142	2	edelweiss	edelweiss	PROPN
easat-5690	142	3	applied	apply	VERB
easat-5690	142	4	science	science	NOUN
easat-5690	142	5	and	and	CCONJ
easat-5690	142	6	technology	technology	NOUN
easat-5690	142	7	issn	issn	PROPN
easat-5690	142	8	:	:	PUNCT
easat-5690	142	9	2576	2576	NUM
easat-5690	142	10	-	-	SYM
easat-5690	142	11	8484	8484	NUM
easat-5690	142	12	vol	vol	NOUN
easat-5690	142	13	.	.	PROPN
easat-5690	143	1	9	9	NUM
easat-5690	143	2	,	,	PUNCT
easat-5690	143	3	no	no	INTJ
easat-5690	143	4	.	.	NOUN
easat-5690	144	1	3	3	NUM
easat-5690	144	2	:	:	SYM
easat-5690	144	3	1777	1777	NUM
easat-5690	144	4	-	-	SYM
easat-5690	144	5	1794	1794	NUM
easat-5690	144	6	,	,	PUNCT
easat-5690	144	7	2025	2025	NUM
easat-5690	144	8	doi	doi	NOUN
easat-5690	144	9	:	:	PUNCT
easat-5690	144	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	144	11	©	©	PROPN
easat-5690	144	12	2025	2025	NUM
easat-5690	144	13	by	by	ADP
easat-5690	144	14	the	the	DET
easat-5690	144	15	authors	author	NOUN
easat-5690	144	16	;	;	PUNCT
easat-5690	144	17	licensee	licensee	PROPN
easat-5690	144	18	learning	learn	VERB
easat-5690	144	19	gate	gate	NOUN
easat-5690	144	20	figure	figure	NOUN
easat-5690	144	21	7	7	NUM
easat-5690	144	22	.	.	PUNCT
easat-5690	144	23	diagram	diagram	NOUN
easat-5690	144	24	of	of	ADP
easat-5690	144	25	cbam	cbam	NOUN
easat-5690	144	26	.	.	PUNCT
easat-5690	145	1	the	the	DET
easat-5690	145	2	channel	channel	PROPN
easat-5690	145	3	submodule	submodule	NOUN
easat-5690	145	4	uses	use	VERB
easat-5690	145	5	the	the	DET
easat-5690	145	6	shared	share	VERB
easat-5690	145	7	network	network	NOUN
easat-5690	145	8	's	's	PART
easat-5690	145	9	max	max	PROPN
easat-5690	145	10	pooling	pooling	NOUN
easat-5690	145	11	and	and	CCONJ
easat-5690	145	12	average	average	ADJ
easat-5690	145	13	pooling	pooling	NOUN
easat-5690	145	14	outputs	output	NOUN
easat-5690	145	15	,	,	PUNCT
easat-5690	145	16	as	as	SCONJ
easat-5690	145	17	seen	see	VERB
easat-5690	145	18	in	in	ADP
easat-5690	145	19	figure	figure	NOUN
easat-5690	145	20	7	7	NUM
easat-5690	145	21	,	,	PUNCT
easat-5690	145	22	while	while	SCONJ
easat-5690	145	23	the	the	DET
easat-5690	145	24	spatial	spatial	ADJ
easat-5690	145	25	submodule	submodule	NOUN
easat-5690	145	26	uses	use	VERB
easat-5690	145	27	two	two	NUM
easat-5690	145	28	similar	similar	ADJ
easat-5690	145	29	outputs	output	NOUN
easat-5690	145	30	that	that	PRON
easat-5690	145	31	are	be	AUX
easat-5690	145	32	pooled	pool	VERB
easat-5690	145	33	along	along	ADP
easat-5690	145	34	the	the	DET
easat-5690	145	35	channel	channel	NOUN
easat-5690	145	36	axis	axis	NOUN
easat-5690	145	37	and	and	CCONJ
easat-5690	145	38	then	then	ADV
easat-5690	145	39	sends	send	VERB
easat-5690	145	40	them	they	PRON
easat-5690	145	41	to	to	ADP
easat-5690	145	42	the	the	DET
easat-5690	145	43	convolutional	convolutional	ADJ
easat-5690	145	44	layer.to	layer.to	PRON
easat-5690	145	45	learn	learn	VERB
easat-5690	145	46	the	the	DET
easat-5690	145	47	attention	attention	NOUN
easat-5690	145	48	of	of	ADP
easat-5690	145	49	convolutional	convolutional	ADJ
easat-5690	145	50	kernels	kernel	NOUN
easat-5690	145	51	along	along	ADP
easat-5690	145	52	the	the	DET
easat-5690	145	53	four	four	NUM
easat-5690	145	54	dimensions	dimension	NOUN
easat-5690	145	55	of	of	ADP
easat-5690	145	56	kernel	kernel	PROPN
easat-5690	145	57	space	space	NOUN
easat-5690	145	58	in	in	ADP
easat-5690	145	59	any	any	DET
easat-5690	145	60	convolutional	convolutional	ADJ
easat-5690	145	61	layer	layer	NOUN
easat-5690	145	62	,	,	PUNCT
easat-5690	145	63	odconv	odconv	NOUN
easat-5690	145	64	employs	employ	VERB
easat-5690	145	65	a	a	DET
easat-5690	145	66	revolutionary	revolutionary	ADJ
easat-5690	145	67	multidimensional	multidimensional	ADJ
easat-5690	145	68	attention	attention	NOUN
easat-5690	145	69	mechanism	mechanism	NOUN
easat-5690	145	70	and	and	CCONJ
easat-5690	145	71	parallel	parallel	ADJ
easat-5690	145	72	technique	technique	NOUN
easat-5690	145	73	.	.	PUNCT
easat-5690	146	1	the	the	DET
easat-5690	146	2	number	number	NOUN
easat-5690	146	3	of	of	ADP
easat-5690	146	4	convolution	convolution	NOUN
easat-5690	146	5	kernel	kernel	NOUN
easat-5690	146	6	input	input	NOUN
easat-5690	146	7	channels	channel	NOUN
easat-5690	146	8	,	,	PUNCT
easat-5690	146	9	the	the	DET
easat-5690	146	10	convolution	convolution	NOUN
easat-5690	146	11	kernel	kernel	NOUN
easat-5690	146	12	's	's	PART
easat-5690	146	13	receptive	receptive	ADJ
easat-5690	146	14	field	field	NOUN
easat-5690	146	15	,	,	PUNCT
easat-5690	146	16	the	the	DET
easat-5690	146	17	number	number	NOUN
easat-5690	146	18	of	of	ADP
easat-5690	146	19	convolution	convolution	NOUN
easat-5690	146	20	kernel	kernel	NOUN
easat-5690	146	21	output	output	NOUN
easat-5690	146	22	channels	channel	NOUN
easat-5690	146	23	,	,	PUNCT
easat-5690	146	24	and	and	CCONJ
easat-5690	146	25	the	the	DET
easat-5690	146	26	number	number	NOUN
easat-5690	146	27	of	of	ADP
easat-5690	146	28	convolution	convolution	NOUN
easat-5690	146	29	kernels	kernel	NOUN
easat-5690	146	30	are	be	AUX
easat-5690	146	31	the	the	DET
easat-5690	146	32	four	four	NUM
easat-5690	146	33	distinct	distinct	ADJ
easat-5690	146	34	forms	form	NOUN
easat-5690	146	35	of	of	ADP
easat-5690	146	36	attention.as	attention.as	X
easat-5690	146	37	an	an	DET
easat-5690	146	38	alternative	alternative	NOUN
easat-5690	146	39	to	to	ADP
easat-5690	146	40	conventional	conventional	ADJ
easat-5690	146	41	convolution	convolution	NOUN
easat-5690	146	42	,	,	PUNCT
easat-5690	146	43	odconv	odconv	NOUN
easat-5690	146	44	can	can	AUX
easat-5690	146	45	be	be	AUX
easat-5690	146	46	inserted	insert	VERB
easat-5690	146	47	into	into	ADP
easat-5690	146	48	many	many	ADJ
easat-5690	146	49	cnn	cnn	PROPN
easat-5690	146	50	architectures	architecture	NOUN
easat-5690	146	51	.	.	PUNCT
easat-5690	147	1	that	that	PRON
easat-5690	147	2	is	be	AUX
easat-5690	147	3	to	to	PART
easat-5690	147	4	say	say	VERB
easat-5690	147	5	,	,	PUNCT
easat-5690	147	6	this	this	DET
easat-5690	147	7	module	module	NOUN
easat-5690	147	8	is	be	AUX
easat-5690	147	9	plug-and-play.as	plug-and-play.as	NUM
easat-5690	147	10	a	a	DET
easat-5690	147	11	result	result	NOUN
easat-5690	147	12	,	,	PUNCT
easat-5690	147	13	while	while	SCONJ
easat-5690	147	14	both	both	DET
easat-5690	147	15	dynamic	dynamic	ADJ
easat-5690	147	16	convolution	convolution	NOUN
easat-5690	147	17	and	and	CCONJ
easat-5690	147	18	attention	attention	NOUN
easat-5690	147	19	mechanisms	mechanism	NOUN
easat-5690	147	20	contribute	contribute	VERB
easat-5690	147	21	to	to	ADP
easat-5690	147	22	improving	improve	VERB
easat-5690	147	23	the	the	DET
easat-5690	147	24	performance	performance	NOUN
easat-5690	147	25	of	of	ADP
easat-5690	147	26	neural	neural	ADJ
easat-5690	147	27	networks	network	NOUN
easat-5690	147	28	,	,	PUNCT
easat-5690	147	29	they	they	PRON
easat-5690	147	30	serve	serve	VERB
easat-5690	147	31	different	different	ADJ
easat-5690	147	32	purposes	purpose	NOUN
easat-5690	147	33	.	.	PUNCT
easat-5690	148	1	dynamic	dynamic	ADJ
easat-5690	148	2	convolution	convolution	NOUN
easat-5690	148	3	enhances	enhance	VERB
easat-5690	148	4	the	the	DET
easat-5690	148	5	adaptability	adaptability	NOUN
easat-5690	148	6	of	of	ADP
easat-5690	148	7	features	feature	NOUN
easat-5690	148	8	,	,	PUNCT
easat-5690	148	9	while	while	SCONJ
easat-5690	148	10	attention	attention	NOUN
easat-5690	148	11	mechanisms	mechanism	NOUN
easat-5690	148	12	enhance	enhance	VERB
easat-5690	148	13	the	the	DET
easat-5690	148	14	network	network	NOUN
easat-5690	148	15	's	's	PART
easat-5690	148	16	ability	ability	NOUN
easat-5690	148	17	to	to	PART
easat-5690	148	18	focus	focus	VERB
easat-5690	148	19	on	on	ADP
easat-5690	148	20	the	the	DET
easat-5690	148	21	specific	specific	ADJ
easat-5690	148	22	information	information	NOUN
easat-5690	148	23	required	require	VERB
easat-5690	148	24	for	for	ADP
easat-5690	148	25	the	the	DET
easat-5690	148	26	current	current	ADJ
easat-5690	148	27	task	task	NOUN
easat-5690	148	28	.	.	PUNCT
easat-5690	149	1	2.4	2.4	NUM
easat-5690	149	2	.	.	PUNCT
easat-5690	150	1	experiment	experiment	NOUN
easat-5690	150	2	setup	setup	NOUN
easat-5690	150	3	this	this	DET
easat-5690	150	4	experiment	experiment	NOUN
easat-5690	150	5	is	be	AUX
easat-5690	150	6	based	base	VERB
easat-5690	150	7	on	on	ADP
easat-5690	150	8	a	a	DET
easat-5690	150	9	windows	window	NOUN
easat-5690	150	10	operating	operating	NOUN
easat-5690	150	11	system	system	NOUN
easat-5690	150	12	,	,	PUNCT
easat-5690	150	13	using	use	VERB
easat-5690	150	14	pytorch	pytorch	NOUN
easat-5690	150	15	learning	learn	VERB
easat-5690	150	16	framework	framework	NOUN
easat-5690	150	17	to	to	PART
easat-5690	150	18	load	load	VERB
easat-5690	150	19	the	the	DET
easat-5690	150	20	model	model	NOUN
easat-5690	150	21	,	,	PUNCT
easat-5690	150	22	apply	apply	VERB
easat-5690	150	23	python	python	NOUN
easat-5690	150	24	language	language	NOUN
easat-5690	150	25	to	to	PART
easat-5690	150	26	write	write	VERB
easat-5690	150	27	the	the	DET
easat-5690	150	28	program	program	NOUN
easat-5690	150	29	,	,	PUNCT
easat-5690	150	30	and	and	CCONJ
easat-5690	150	31	save	save	VERB
easat-5690	150	32	the	the	DET
easat-5690	150	33	model	model	NOUN
easat-5690	150	34	parameters	parameter	NOUN
easat-5690	150	35	with	with	ADP
easat-5690	150	36	the	the	DET
easat-5690	150	37	best	good	ADJ
easat-5690	150	38	effect	effect	NOUN
easat-5690	150	39	during	during	ADP
easat-5690	150	40	the	the	DET
easat-5690	150	41	training	training	NOUN
easat-5690	150	42	process	process	NOUN
easat-5690	150	43	.	.	PUNCT
easat-5690	151	1	table	table	NOUN
easat-5690	151	2	1	1	NUM
easat-5690	151	3	displays	display	VERB
easat-5690	151	4	the	the	DET
easat-5690	151	5	experimental	experimental	ADJ
easat-5690	151	6	platform	platform	NOUN
easat-5690	151	7	's	's	PART
easat-5690	151	8	configuration	configuration	NOUN
easat-5690	151	9	table.for	table.for	ADP
easat-5690	151	10	the	the	DET
easat-5690	151	11	improved	improved	ADJ
easat-5690	151	12	yolov5s	yolov5s	NOUN
easat-5690	151	13	(	(	PUNCT
easat-5690	151	14	including	include	VERB
easat-5690	151	15	other	other	ADJ
easat-5690	151	16	comparative	comparative	ADJ
easat-5690	151	17	models	model	NOUN
easat-5690	151	18	)	)	PUNCT
easat-5690	151	19	used	use	VERB
easat-5690	151	20	in	in	ADP
easat-5690	151	21	this	this	DET
easat-5690	151	22	paper	paper	NOUN
easat-5690	151	23	,	,	PUNCT
easat-5690	151	24	the	the	DET
easat-5690	151	25	input	input	NOUN
easat-5690	151	26	images	image	NOUN
easat-5690	151	27	were	be	AUX
easat-5690	151	28	of	of	ADP
easat-5690	151	29	640	640	NUM
easat-5690	151	30	×	×	NOUN
easat-5690	151	31	640	640	NUM
easat-5690	151	32	pixels	pixel	NOUN
easat-5690	151	33	,	,	PUNCT
easat-5690	151	34	with	with	ADP
easat-5690	151	35	a	a	DET
easat-5690	151	36	batch	batch	NOUN
easat-5690	151	37	size	size	NOUN
easat-5690	151	38	of	of	ADP
easat-5690	151	39	16	16	NUM
easat-5690	151	40	,	,	PUNCT
easat-5690	151	41	momentum	momentum	NOUN
easat-5690	151	42	of	of	ADP
easat-5690	151	43	0.937	0.937	NUM
easat-5690	151	44	,	,	PUNCT
easat-5690	151	45	weight	weight	NOUN
easat-5690	151	46	decay	decay	NOUN
easat-5690	151	47	of	of	ADP
easat-5690	151	48	0.0005	0.0005	NUM
easat-5690	151	49	,	,	PUNCT
easat-5690	151	50	iou	iou	NOUN
easat-5690	151	51	of	of	ADP
easat-5690	151	52	0.2	0.2	NUM
easat-5690	151	53	,	,	PUNCT
easat-5690	151	54	hsv_h	hsv_h	PROPN
easat-5690	151	55	of	of	ADP
easat-5690	151	56	0.015	0.015	NUM
easat-5690	151	57	,	,	PUNCT
easat-5690	151	58	hsv_s	hsv_s	NOUN
easat-5690	151	59	of	of	ADP
easat-5690	151	60	0.7	0.7	NUM
easat-5690	151	61	,	,	PUNCT
easat-5690	151	62	hsv_v	hsv_v	NOUN
easat-5690	151	63	of	of	ADP
easat-5690	151	64	0.4	0.4	NUM
easat-5690	151	65	,	,	PUNCT
easat-5690	151	66	mosaic	mosaic	PROPN
easat-5690	151	67	of	of	ADP
easat-5690	151	68	1.0	1.0	NUM
easat-5690	151	69	,	,	PUNCT
easat-5690	151	70	scale	scale	NOUN
easat-5690	151	71	of	of	ADP
easat-5690	151	72	0.5	0.5	NUM
easat-5690	151	73	,	,	PUNCT
easat-5690	151	74	translate	translate	NOUN
easat-5690	151	75	of	of	ADP
easat-5690	151	76	0.1	0.1	NUM
easat-5690	151	77	,	,	PUNCT
easat-5690	151	78	mixup	mixup	NOUN
easat-5690	151	79	of	of	ADP
easat-5690	151	80	0.0	0.0	NUM
easat-5690	151	81	,	,	PUNCT
easat-5690	151	82	and	and	CCONJ
easat-5690	151	83	a	a	DET
easat-5690	151	84	training	training	NOUN
easat-5690	151	85	time	time	NOUN
easat-5690	151	86	of	of	ADP
easat-5690	151	87	150	150	NUM
easat-5690	151	88	epochs	epoch	NOUN
easat-5690	151	89	.	.	PUNCT
easat-5690	152	1	random	random	ADJ
easat-5690	152	2	initialization	initialization	NOUN
easat-5690	152	3	techniques	technique	NOUN
easat-5690	152	4	were	be	AUX
easat-5690	152	5	used	use	VERB
easat-5690	152	6	to	to	PART
easat-5690	152	7	initialize	initialize	VERB
easat-5690	152	8	the	the	DET
easat-5690	152	9	weights	weight	NOUN
easat-5690	152	10	,	,	PUNCT
easat-5690	152	11	allowing	allow	VERB
easat-5690	152	12	all	all	DET
easat-5690	152	13	models	model	NOUN
easat-5690	152	14	to	to	PART
easat-5690	152	15	be	be	AUX
easat-5690	152	16	trained	train	VERB
easat-5690	152	17	from	from	ADP
easat-5690	152	18	scratch	scratch	NOUN
easat-5690	152	19	.	.	PUNCT
easat-5690	153	1	table	table	NOUN
easat-5690	153	2	1	1	NUM
easat-5690	153	3	.	.	PUNCT
easat-5690	154	1	the	the	DET
easat-5690	154	2	experiment	experiment	NOUN
easat-5690	154	3	platform	platform	NOUN
easat-5690	154	4	's	's	PART
easat-5690	154	5	configuration	configuration	NOUN
easat-5690	154	6	table	table	NOUN
easat-5690	154	7	.	.	PUNCT
easat-5690	155	1	setting	set	VERB
easat-5690	155	2	for	for	ADP
easat-5690	155	3	experiments	experiment	NOUN
easat-5690	155	4	specifics	specific	NOUN
easat-5690	155	5	mirroring	mirror	VERB
easat-5690	155	6	pytorch1.7.0	pytorch1.7.0	PROPN
easat-5690	155	7	python3.8(ubuntu18.04	python3.8(ubuntu18.04	PROPN
easat-5690	155	8	)	)	PUNCT
easat-5690	155	9	cuda	cuda	NOUN
easat-5690	155	10	11.0	11.0	NUM
easat-5690	155	11	gpu	gpu	PROPN
easat-5690	155	12	tesla	tesla	PROPN
easat-5690	155	13	t4(16	t4(16	PROPN
easat-5690	155	14	gb	gb	NOUN
easat-5690	155	15	)	)	PUNCT
easat-5690	155	16	*	*	PUNCT
easat-5690	155	17	1	1	NUM
easat-5690	155	18	cpu	cpu	NOUN
easat-5690	155	19	8	8	NUM
easat-5690	155	20	vcpu	vcpu	NOUN
easat-5690	155	21	intel	intel	PROPN
easat-5690	155	22	xeon	xeon	PROPN
easat-5690	155	23	processor	processor	PROPN
easat-5690	155	24	(	(	PUNCT
easat-5690	155	25	skylake	skylake	NOUN
easat-5690	155	26	,	,	PUNCT
easat-5690	155	27	ibrs	ibrs	ADJ
easat-5690	155	28	)	)	PUNCT
easat-5690	155	29	memory	memory	NOUN
easat-5690	155	30	56	56	NUM
easat-5690	155	31	gb	gb	ADP
easat-5690	155	32	2.5	2.5	NUM
easat-5690	155	33	.	.	PUNCT
easat-5690	156	1	evaluation	evaluation	NOUN
easat-5690	156	2	metrics	metric	NOUN
easat-5690	156	3	we	we	PRON
easat-5690	156	4	will	will	AUX
easat-5690	156	5	employ	employ	VERB
easat-5690	156	6	a	a	DET
easat-5690	156	7	number	number	NOUN
easat-5690	156	8	of	of	ADP
easat-5690	156	9	standard	standard	ADJ
easat-5690	156	10	evaluation	evaluation	NOUN
easat-5690	156	11	measures	measure	NOUN
easat-5690	156	12	to	to	PART
easat-5690	156	13	evaluate	evaluate	VERB
easat-5690	156	14	the	the	DET
easat-5690	156	15	models	model	NOUN
easat-5690	156	16	'	'	PART
easat-5690	156	17	performance	performance	NOUN
easat-5690	156	18	in	in	ADP
easat-5690	156	19	our	our	PRON
easat-5690	156	20	investigation[38	investigation[38	NOUN
easat-5690	156	21	]	]	X
easat-5690	156	22	.	.	PUNCT
easat-5690	157	1	2.5.1	2.5.1	NUM
easat-5690	157	2	.	.	PUNCT
easat-5690	157	3	confusion	confusion	NOUN
easat-5690	157	4	matrix	matrix	NOUN
easat-5690	157	5	the	the	DET
easat-5690	157	6	four	four	NUM
easat-5690	157	7	categories	category	NOUN
easat-5690	157	8	of	of	ADP
easat-5690	157	9	categorization	categorization	NOUN
easat-5690	157	10	task	task	NOUN
easat-5690	157	11	prediction	prediction	NOUN
easat-5690	157	12	outputs	output	NOUN
easat-5690	157	13	in	in	ADP
easat-5690	157	14	machine	machine	NOUN
easat-5690	157	15	learning	learning	NOUN
easat-5690	157	16	and	and	CCONJ
easat-5690	157	17	deep	deep	ADJ
easat-5690	157	18	learning	learning	NOUN
easat-5690	157	19	are	be	AUX
easat-5690	157	20	known	know	VERB
easat-5690	157	21	as	as	ADP
easat-5690	157	22	confusion	confusion	NOUN
easat-5690	157	23	matrices	matrix	NOUN
easat-5690	157	24	and	and	CCONJ
easat-5690	157	25	include	include	VERB
easat-5690	157	26	the	the	DET
easat-5690	157	27	following	follow	VERB
easat-5690	157	28	:	:	PUNCT
easat-5690	157	29	one	one	NUM
easat-5690	157	30	tool	tool	NOUN
easat-5690	157	31	for	for	ADP
easat-5690	157	32	visualizing	visualize	VERB
easat-5690	157	33	and	and	CCONJ
easat-5690	157	34	summarizing	summarize	VERB
easat-5690	157	35	classification	classification	NOUN
easat-5690	157	36	problem	problem	NOUN
easat-5690	157	37	performance	performance	NOUN
easat-5690	157	38	is	be	AUX
easat-5690	157	39	the	the	DET
easat-5690	157	40	confusion	confusion	NOUN
easat-5690	157	41	matrix[39	matrix[39	NOUN
easat-5690	157	42	]	]	PUNCT
easat-5690	157	43	.	.	PUNCT
easat-5690	158	1	1790	1790	NUM
easat-5690	158	2	edelweiss	edelweiss	PROPN
easat-5690	158	3	applied	apply	VERB
easat-5690	158	4	science	science	NOUN
easat-5690	158	5	and	and	CCONJ
easat-5690	158	6	technology	technology	NOUN
easat-5690	158	7	issn	issn	PROPN
easat-5690	158	8	:	:	PUNCT
easat-5690	158	9	2576	2576	NUM
easat-5690	158	10	-	-	SYM
easat-5690	158	11	8484	8484	NUM
easat-5690	158	12	vol	vol	NOUN
easat-5690	158	13	.	.	PROPN
easat-5690	159	1	9	9	NUM
easat-5690	159	2	,	,	PUNCT
easat-5690	159	3	no	no	INTJ
easat-5690	159	4	.	.	NOUN
easat-5690	160	1	3	3	NUM
easat-5690	160	2	:	:	SYM
easat-5690	160	3	1777	1777	NUM
easat-5690	160	4	-	-	SYM
easat-5690	160	5	1794	1794	NUM
easat-5690	160	6	,	,	PUNCT
easat-5690	160	7	2025	2025	NUM
easat-5690	160	8	doi	doi	NOUN
easat-5690	160	9	:	:	PUNCT
easat-5690	160	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	160	11	©	©	PROPN
easat-5690	160	12	2025	2025	NUM
easat-5690	160	13	by	by	ADP
easat-5690	160	14	the	the	DET
easat-5690	160	15	authors	author	NOUN
easat-5690	160	16	;	;	PUNCT
easat-5690	160	17	licensee	licensee	PROPN
easat-5690	160	18	learning	learning	NOUN
easat-5690	160	19	gate	gate	VERB
easat-5690	160	20	true	true	ADJ
easat-5690	160	21	positive	positive	ADJ
easat-5690	160	22	(	(	PUNCT
easat-5690	160	23	tp	tp	NOUN
easat-5690	160	24	)	)	PUNCT
easat-5690	161	1	[	[	X
easat-5690	161	2	2	2	NUM
easat-5690	161	3	]	]	PUNCT
easat-5690	161	4	:	:	PUNCT
easat-5690	161	5	both	both	CCONJ
easat-5690	161	6	the	the	DET
easat-5690	161	7	label	label	NOUN
easat-5690	161	8	value	value	NOUN
easat-5690	161	9	and	and	CCONJ
easat-5690	161	10	the	the	DET
easat-5690	161	11	anticipated	anticipate	VERB
easat-5690	161	12	value	value	NOUN
easat-5690	161	13	are	be	AUX
easat-5690	161	14	positive	positive	ADJ
easat-5690	161	15	cases	case	NOUN
easat-5690	161	16	,	,	PUNCT
easat-5690	161	17	suggesting	suggest	VERB
easat-5690	161	18	that	that	SCONJ
easat-5690	161	19	the	the	DET
easat-5690	161	20	prediction	prediction	NOUN
easat-5690	161	21	is	be	AUX
easat-5690	161	22	accurate	accurate	ADJ
easat-5690	161	23	.	.	PUNCT
easat-5690	162	1	false	false	ADJ
easat-5690	162	2	negative	negative	ADJ
easat-5690	162	3	(	(	PUNCT
easat-5690	162	4	fn	fn	NOUN
easat-5690	162	5	):	):	PUNCT
easat-5690	162	6	an	an	DET
easat-5690	162	7	inaccurate	inaccurate	ADJ
easat-5690	162	8	prediction	prediction	NOUN
easat-5690	162	9	is	be	AUX
easat-5690	162	10	made	make	VERB
easat-5690	162	11	when	when	SCONJ
easat-5690	162	12	the	the	DET
easat-5690	162	13	label	label	NOUN
easat-5690	162	14	value	value	NOUN
easat-5690	162	15	is	be	AUX
easat-5690	162	16	positive	positive	ADJ
easat-5690	162	17	and	and	CCONJ
easat-5690	162	18	the	the	DET
easat-5690	162	19	projected	project	VERB
easat-5690	162	20	value	value	NOUN
easat-5690	162	21	is	be	AUX
easat-5690	162	22	negative	negative	ADJ
easat-5690	162	23	.	.	PUNCT
easat-5690	163	1	false	false	ADJ
easat-5690	163	2	positive	positive	ADJ
easat-5690	163	3	(	(	PUNCT
easat-5690	163	4	fp	fp	NOUN
easat-5690	163	5	):	):	PUNCT
easat-5690	163	6	an	an	DET
easat-5690	163	7	inaccurate	inaccurate	ADJ
easat-5690	163	8	prediction	prediction	NOUN
easat-5690	163	9	is	be	AUX
easat-5690	163	10	indicated	indicate	VERB
easat-5690	163	11	when	when	SCONJ
easat-5690	163	12	the	the	DET
easat-5690	163	13	label	label	NOUN
easat-5690	163	14	value	value	NOUN
easat-5690	163	15	is	be	AUX
easat-5690	163	16	a	a	DET
easat-5690	163	17	negative	negative	ADJ
easat-5690	163	18	example	example	NOUN
easat-5690	163	19	despite	despite	SCONJ
easat-5690	163	20	the	the	DET
easat-5690	163	21	prediction	prediction	NOUN
easat-5690	163	22	being	be	AUX
easat-5690	163	23	a	a	DET
easat-5690	163	24	positive	positive	ADJ
easat-5690	163	25	example	example	NOUN
easat-5690	163	26	.	.	PUNCT
easat-5690	164	1	true	true	ADJ
easat-5690	164	2	negative	negative	ADJ
easat-5690	164	3	(	(	PUNCT
easat-5690	164	4	tn	tn	NOUN
easat-5690	164	5	):	):	PUNCT
easat-5690	164	6	this	this	PRON
easat-5690	164	7	indicates	indicate	VERB
easat-5690	164	8	that	that	SCONJ
easat-5690	164	9	the	the	DET
easat-5690	164	10	prediction	prediction	NOUN
easat-5690	164	11	is	be	AUX
easat-5690	164	12	accurate	accurate	ADJ
easat-5690	164	13	as	as	ADP
easat-5690	164	14	both	both	CCONJ
easat-5690	164	15	the	the	DET
easat-5690	164	16	label	label	NOUN
easat-5690	164	17	value	value	NOUN
easat-5690	164	18	and	and	CCONJ
easat-5690	164	19	the	the	DET
easat-5690	164	20	predicted	predict	VERB
easat-5690	164	21	value	value	NOUN
easat-5690	164	22	are	be	AUX
easat-5690	164	23	negative	negative	ADJ
easat-5690	164	24	cases	case	NOUN
easat-5690	164	25	.	.	PUNCT
easat-5690	165	1	2.5.2	2.5.2	NUM
easat-5690	165	2	.	.	PUNCT
easat-5690	165	3	precision	precision	PROPN
easat-5690	165	4	precision(p	precision(p	PROPN
easat-5690	165	5	)	)	PUNCT
easat-5690	165	6	refers	refer	VERB
easat-5690	165	7	to	to	ADP
easat-5690	165	8	precision	precision	NOUN
easat-5690	165	9	,	,	PUNCT
easat-5690	165	10	and	and	CCONJ
easat-5690	165	11	its	its	PRON
easat-5690	165	12	definition	definition	NOUN
easat-5690	165	13	is	be	AUX
easat-5690	165	14	as	as	SCONJ
easat-5690	165	15	follows	follow	VERB
easat-5690	165	16	:	:	PUNCT
easat-5690	165	17	100	100	NUM
easat-5690	165	18	%	%	NOUN
easat-5690	165	19	tpfp	tpfp	NOUN
easat-5690	165	20	tpprecision	tpprecision	NOUN
easat-5690	165	21			NOUN
easat-5690	166	1	+	+	X
easat-5690	166	2	=	=	SYM
easat-5690	166	3	(	(	PUNCT
easat-5690	166	4	1	1	NUM
easat-5690	166	5	)	)	PUNCT
easat-5690	166	6	accuracy	accuracy	NOUN
easat-5690	166	7	's	's	PART
easat-5690	166	8	denominator	denominator	NOUN
easat-5690	166	9	comprises	comprise	VERB
easat-5690	166	10	tp	tp	ADP
easat-5690	166	11	(	(	PUNCT
easat-5690	166	12	true	true	ADJ
easat-5690	166	13	positives	positive	NOUN
easat-5690	166	14	)	)	PUNCT
easat-5690	166	15	and	and	CCONJ
easat-5690	166	16	fp	fp	INTJ
easat-5690	166	17	(	(	PUNCT
easat-5690	166	18	false	false	ADJ
easat-5690	166	19	positives	positive	NOUN
easat-5690	166	20	)	)	PUNCT
easat-5690	166	21	.	.	PUNCT
easat-5690	167	1	tp	tp	NOUN
easat-5690	167	2	counts	count	VERB
easat-5690	167	3	cases	case	NOUN
easat-5690	167	4	predicted	predict	VERB
easat-5690	167	5	positive	positive	ADJ
easat-5690	167	6	that	that	PRON
easat-5690	167	7	are	be	AUX
easat-5690	167	8	actually	actually	ADV
easat-5690	167	9	positive	positive	ADJ
easat-5690	167	10	,	,	PUNCT
easat-5690	167	11	while	while	SCONJ
easat-5690	167	12	fp	fp	PROPN
easat-5690	167	13	counts	count	NOUN
easat-5690	167	14	wrongly	wrongly	ADV
easat-5690	167	15	predicted	predict	VERB
easat-5690	167	16	positives	positive	NOUN
easat-5690	167	17	(	(	PUNCT
easat-5690	167	18	actual	actual	ADJ
easat-5690	167	19	negatives	negative	NOUN
easat-5690	167	20	)	)	PUNCT
easat-5690	167	21	.	.	PUNCT
easat-5690	168	1	accuracy	accuracy	NOUN
easat-5690	168	2	reflects	reflect	VERB
easat-5690	168	3	the	the	DET
easat-5690	168	4	correlation	correlation	NOUN
easat-5690	168	5	between	between	ADP
easat-5690	168	6	predicted	predict	VERB
easat-5690	168	7	positives	positive	NOUN
easat-5690	168	8	and	and	CCONJ
easat-5690	168	9	actual	actual	ADJ
easat-5690	168	10	positives	positive	NOUN
easat-5690	168	11	/	/	SYM
easat-5690	168	12	negatives	negative	NOUN
easat-5690	168	13	.	.	PUNCT
easat-5690	169	1	higher	high	ADJ
easat-5690	169	2	accuracy	accuracy	NOUN
easat-5690	169	3	means	mean	VERB
easat-5690	169	4	lower	low	ADJ
easat-5690	169	5	fp	fp	NOUN
easat-5690	169	6	,	,	PUNCT
easat-5690	169	7	fewer	few	ADJ
easat-5690	169	8	misclassified	misclassified	ADJ
easat-5690	169	9	cases	case	NOUN
easat-5690	169	10	,	,	PUNCT
easat-5690	169	11	and	and	CCONJ
easat-5690	169	12	higher	high	ADJ
easat-5690	169	13	purity	purity	NOUN
easat-5690	169	14	in	in	ADP
easat-5690	169	15	predicted	predict	VERB
easat-5690	169	16	positives	positive	NOUN
easat-5690	169	17	,	,	PUNCT
easat-5690	169	18	translating	translate	VERB
easat-5690	169	19	to	to	ADP
easat-5690	169	20	fewer	few	ADJ
easat-5690	169	21	false	false	ADJ
easat-5690	169	22	positives	positive	NOUN
easat-5690	169	23	.	.	PUNCT
easat-5690	170	1	2.5.3	2.5.3	NUM
easat-5690	170	2	.	.	PUNCT
easat-5690	170	3	recall	recall	PROPN
easat-5690	170	4	recall(r	recall(r	PROPN
easat-5690	170	5	)	)	PUNCT
easat-5690	170	6	refers	refer	VERB
easat-5690	170	7	to	to	ADP
easat-5690	170	8	the	the	DET
easat-5690	170	9	recall	recall	NOUN
easat-5690	170	10	rate	rate	NOUN
easat-5690	170	11	,	,	PUNCT
easat-5690	170	12	and	and	CCONJ
easat-5690	170	13	the	the	DET
easat-5690	170	14	definition	definition	NOUN
easat-5690	170	15	of	of	ADP
easat-5690	170	16	recall	recall	NOUN
easat-5690	170	17	is	be	AUX
easat-5690	170	18	as	as	SCONJ
easat-5690	170	19	follows	follow	VERB
easat-5690	170	20	:	:	PUNCT
easat-5690	170	21	100	100	NUM
easat-5690	170	22	%	%	NOUN
easat-5690	170	23	tpfn	tpfn	NOUN
easat-5690	170	24	tprecall	tprecall	NOUN
easat-5690	170	25			NOUN
easat-5690	171	1	+	+	CCONJ
easat-5690	171	2	=	=	SYM
easat-5690	171	3	(	(	PUNCT
easat-5690	171	4	2	2	NUM
easat-5690	171	5	)	)	PUNCT
easat-5690	171	6	according	accord	VERB
easat-5690	171	7	to	to	ADP
easat-5690	171	8	this	this	DET
easat-5690	171	9	definition	definition	NOUN
easat-5690	171	10	,	,	PUNCT
easat-5690	171	11	the	the	DET
easat-5690	171	12	denominator	denominator	NOUN
easat-5690	171	13	of	of	ADP
easat-5690	171	14	recall	recall	NOUN
easat-5690	171	15	is	be	AUX
easat-5690	171	16	the	the	DET
easat-5690	171	17	sum	sum	NOUN
easat-5690	171	18	of	of	ADP
easat-5690	171	19	tp	tp	NOUN
easat-5690	171	20	and	and	CCONJ
easat-5690	171	21	fn	fn	NOUN
easat-5690	171	22	,	,	PUNCT
easat-5690	171	23	where	where	SCONJ
easat-5690	171	24	tp	tp	NOUN
easat-5690	171	25	is	be	AUX
easat-5690	171	26	the	the	DET
easat-5690	171	27	number	number	NOUN
easat-5690	171	28	of	of	ADP
easat-5690	171	29	positive	positive	ADJ
easat-5690	171	30	cases	case	NOUN
easat-5690	171	31	predicted	predict	VERB
easat-5690	171	32	,	,	PUNCT
easat-5690	171	33	and	and	CCONJ
easat-5690	171	34	the	the	DET
easat-5690	171	35	actual	actual	ADJ
easat-5690	171	36	value	value	NOUN
easat-5690	171	37	is	be	AUX
easat-5690	171	38	also	also	ADV
easat-5690	171	39	positive	positive	ADJ
easat-5690	171	40	;	;	PUNCT
easat-5690	171	41	fn	fn	NOUN
easat-5690	171	42	is	be	AUX
easat-5690	171	43	the	the	DET
easat-5690	171	44	number	number	NOUN
easat-5690	171	45	of	of	ADP
easat-5690	171	46	negative	negative	ADJ
easat-5690	171	47	cases	case	NOUN
easat-5690	171	48	predicted	predict	VERB
easat-5690	171	49	,	,	PUNCT
easat-5690	171	50	but	but	CCONJ
easat-5690	171	51	the	the	DET
easat-5690	171	52	actual	actual	ADJ
easat-5690	171	53	value	value	NOUN
easat-5690	171	54	is	be	AUX
easat-5690	171	55	positive	positive	ADJ
easat-5690	171	56	.	.	PUNCT
easat-5690	172	1	analyzing	analyze	VERB
easat-5690	172	2	the	the	DET
easat-5690	172	3	formula	formula	NOUN
easat-5690	172	4	reveals	reveal	VERB
easat-5690	172	5	that	that	SCONJ
easat-5690	172	6	recall	recall	NOUN
easat-5690	172	7	encompasses	encompass	VERB
easat-5690	172	8	both	both	PRON
easat-5690	172	9	predicted	predict	VERB
easat-5690	172	10	positive	positive	ADJ
easat-5690	172	11	and	and	CCONJ
easat-5690	172	12	negative	negative	ADJ
easat-5690	172	13	cases	case	NOUN
easat-5690	172	14	alongside	alongside	ADP
easat-5690	172	15	actual	actual	ADJ
easat-5690	172	16	positive	positive	ADJ
easat-5690	172	17	cases	case	NOUN
easat-5690	172	18	.	.	PUNCT
easat-5690	173	1	as	as	ADP
easat-5690	173	2	recall	recall	NOUN
easat-5690	173	3	increases	increase	NOUN
easat-5690	173	4	,	,	PUNCT
easat-5690	173	5	false	false	ADJ
easat-5690	173	6	negatives	negative	NOUN
easat-5690	173	7	(	(	PUNCT
easat-5690	173	8	fn	fn	NOUN
easat-5690	173	9	)	)	PUNCT
easat-5690	173	10	decrease	decrease	NOUN
easat-5690	173	11	,	,	PUNCT
easat-5690	173	12	indicating	indicate	VERB
easat-5690	173	13	a	a	DET
easat-5690	173	14	reduction	reduction	NOUN
easat-5690	173	15	in	in	ADP
easat-5690	173	16	positive	positive	ADJ
easat-5690	173	17	cases	case	NOUN
easat-5690	173	18	misclassified	misclassifie	VERB
easat-5690	173	19	as	as	ADP
easat-5690	173	20	negative	negative	ADJ
easat-5690	173	21	.	.	PUNCT
easat-5690	174	1	this	this	PRON
easat-5690	174	2	translates	translate	VERB
easat-5690	174	3	to	to	ADP
easat-5690	174	4	identifying	identify	VERB
easat-5690	174	5	a	a	DET
easat-5690	174	6	greater	great	ADJ
easat-5690	174	7	number	number	NOUN
easat-5690	174	8	of	of	ADP
easat-5690	174	9	positive	positive	ADJ
easat-5690	174	10	cases	case	NOUN
easat-5690	174	11	,	,	PUNCT
easat-5690	174	12	signifying	signify	VERB
easat-5690	174	13	fewer	few	ADJ
easat-5690	174	14	missed	miss	VERB
easat-5690	174	15	detections	detection	NOUN
easat-5690	174	16	.	.	PUNCT
easat-5690	175	1	2.5.4	2.5.4	NUM
easat-5690	175	2	.	.	PUNCT
easat-5690	176	1	p	p	X
easat-5690	176	2	-	-	PUNCT
easat-5690	176	3	r	r	NOUN
easat-5690	176	4	curve	curve	NOUN
easat-5690	176	5	precise	precise	ADJ
easat-5690	176	6	and	and	CCONJ
easat-5690	176	7	recall	recall	NOUN
easat-5690	176	8	coordinates	coordinate	NOUN
easat-5690	176	9	,	,	PUNCT
easat-5690	176	10	respectively	respectively	ADV
easat-5690	176	11	,	,	PUNCT
easat-5690	176	12	surround	surround	VERB
easat-5690	176	13	the	the	DET
easat-5690	176	14	p	p	NOUN
easat-5690	176	15	-	-	PUNCT
easat-5690	176	16	r	r	NOUN
easat-5690	176	17	curve	curve	NOUN
easat-5690	176	18	.	.	PUNCT
easat-5690	177	1	the	the	DET
easat-5690	177	2	lines	line	NOUN
easat-5690	177	3	of	of	ADP
easat-5690	177	4	different	different	ADJ
easat-5690	177	5	colors	color	NOUN
easat-5690	177	6	represent	represent	VERB
easat-5690	177	7	the	the	DET
easat-5690	177	8	pr	pr	NOUN
easat-5690	177	9	curves	curve	NOUN
easat-5690	177	10	of	of	ADP
easat-5690	177	11	different	different	ADJ
easat-5690	177	12	categories	category	NOUN
easat-5690	177	13	,	,	PUNCT
easat-5690	177	14	while	while	SCONJ
easat-5690	177	15	the	the	DET
easat-5690	177	16	thick	thick	ADJ
easat-5690	177	17	blue	blue	ADJ
easat-5690	177	18	lines	line	NOUN
easat-5690	177	19	represent	represent	VERB
easat-5690	177	20	the	the	DET
easat-5690	177	21	average	average	ADJ
easat-5690	177	22	pr	pr	NOUN
easat-5690	177	23	curves	curve	NOUN
easat-5690	177	24	of	of	ADP
easat-5690	177	25	all	all	DET
easat-5690	177	26	categories	category	NOUN
easat-5690	177	27	.	.	PUNCT
easat-5690	178	1	when	when	SCONJ
easat-5690	178	2	assessing	assess	VERB
easat-5690	178	3	a	a	DET
easat-5690	178	4	model	model	NOUN
easat-5690	178	5	's	's	PART
easat-5690	178	6	prediction	prediction	NOUN
easat-5690	178	7	outcomes	outcome	NOUN
easat-5690	178	8	,	,	PUNCT
easat-5690	178	9	the	the	DET
easat-5690	178	10	region	region	NOUN
easat-5690	178	11	bounded	bound	VERB
easat-5690	178	12	by	by	ADP
easat-5690	178	13	the	the	DET
easat-5690	178	14	axes	axis	NOUN
easat-5690	178	15	and	and	CCONJ
easat-5690	178	16	the	the	DET
easat-5690	178	17	p	p	NOUN
easat-5690	178	18	-	-	PUNCT
easat-5690	178	19	r	r	NOUN
easat-5690	178	20	curve	curve	NOUN
easat-5690	178	21	can	can	AUX
easat-5690	178	22	be	be	AUX
easat-5690	178	23	used	use	VERB
easat-5690	178	24	as	as	ADP
easat-5690	178	25	a	a	DET
easat-5690	178	26	reference	reference	NOUN
easat-5690	178	27	.	.	PUNCT
easat-5690	179	1	one	one	PRON
easat-5690	179	2	can	can	AUX
easat-5690	179	3	conclude	conclude	VERB
easat-5690	179	4	that	that	SCONJ
easat-5690	179	5	the	the	DET
easat-5690	179	6	prediction	prediction	NOUN
easat-5690	179	7	results	result	NOUN
easat-5690	179	8	of	of	ADP
easat-5690	179	9	one	one	NUM
easat-5690	179	10	model	model	NOUN
easat-5690	179	11	are	be	AUX
easat-5690	179	12	unquestionably	unquestionably	ADV
easat-5690	179	13	superior	superior	ADJ
easat-5690	179	14	to	to	ADP
easat-5690	179	15	those	those	PRON
easat-5690	179	16	of	of	ADP
easat-5690	179	17	the	the	DET
easat-5690	179	18	other	other	ADJ
easat-5690	179	19	if	if	SCONJ
easat-5690	179	20	the	the	DET
easat-5690	179	21	p	p	NOUN
easat-5690	179	22	-	-	PUNCT
easat-5690	179	23	r	r	NOUN
easat-5690	179	24	curve	curve	NOUN
easat-5690	179	25	of	of	ADP
easat-5690	179	26	one	one	NUM
easat-5690	179	27	model	model	NOUN
easat-5690	179	28	fully	fully	ADV
easat-5690	179	29	encloses	enclose	VERB
easat-5690	179	30	the	the	DET
easat-5690	179	31	p	p	NOUN
easat-5690	179	32	-	-	PUNCT
easat-5690	179	33	r	r	NOUN
easat-5690	179	34	curve	curve	NOUN
easat-5690	179	35	of	of	ADP
easat-5690	179	36	the	the	DET
easat-5690	179	37	other	other	ADJ
easat-5690	179	38	model	model	NOUN
easat-5690	179	39	.	.	PUNCT
easat-5690	180	1	2.5.5	2.5.5	X
easat-5690	180	2	.	.	X
easat-5690	180	3	ap	ap	PROPN
easat-5690	180	4	ap	ap	PROPN
easat-5690	181	1	(	(	PUNCT
easat-5690	181	2	average	average	ADJ
easat-5690	181	3	precision	precision	NOUN
easat-5690	181	4	):	):	PUNCT
easat-5690	181	5	in	in	ADP
easat-5690	181	6	spite	spite	NOUN
easat-5690	181	7	of	of	ADP
easat-5690	181	8	its	its	PRON
easat-5690	181	9	name	name	NOUN
easat-5690	181	10	,	,	PUNCT
easat-5690	181	11	ap	ap	PROPN
easat-5690	181	12	is	be	AUX
easat-5690	181	13	not	not	PART
easat-5690	181	14	determined	determine	VERB
easat-5690	181	15	by	by	ADP
easat-5690	181	16	averaging	average	VERB
easat-5690	181	17	precision	precision	NOUN
easat-5690	181	18	numbers	number	NOUN
easat-5690	181	19	.	.	PUNCT
easat-5690	182	1	rather	rather	ADV
easat-5690	182	2	,	,	PUNCT
easat-5690	182	3	it	it	PRON
easat-5690	182	4	uses	use	VERB
easat-5690	182	5	integration	integration	NOUN
easat-5690	182	6	to	to	PART
easat-5690	182	7	find	find	VERB
easat-5690	182	8	the	the	DET
easat-5690	182	9	area	area	NOUN
easat-5690	182	10	under	under	ADP
easat-5690	182	11	the	the	DET
easat-5690	182	12	pr	pr	NOUN
easat-5690	182	13	curve	curve	NOUN
easat-5690	182	14	(	(	PUNCT
easat-5690	182	15	precision	precision	NOUN
easat-5690	182	16	-	-	PUNCT
easat-5690	182	17	recall	recall	NOUN
easat-5690	182	18	curve	curve	NOUN
easat-5690	182	19	)	)	PUNCT
easat-5690	182	20	for	for	ADP
easat-5690	182	21	every	every	DET
easat-5690	182	22	class	class	NOUN
easat-5690	182	23	.	.	PUNCT
easat-5690	183	1	a	a	DET
easat-5690	183	2	model	model	NOUN
easat-5690	183	3	's	's	PART
easat-5690	183	4	overall	overall	ADJ
easat-5690	183	5	precision	precision	NOUN
easat-5690	183	6	and	and	CCONJ
easat-5690	183	7	recall	recall	NOUN
easat-5690	183	8	are	be	AUX
easat-5690	183	9	comparatively	comparatively	ADV
easat-5690	183	10	greater	great	ADJ
easat-5690	183	11	if	if	SCONJ
easat-5690	183	12	its	its	PRON
easat-5690	183	13	ap	ap	PROPN
easat-5690	183	14	is	be	AUX
easat-5690	183	15	higher	high	ADJ
easat-5690	183	16	,	,	PUNCT
easat-5690	183	17	which	which	PRON
easat-5690	183	18	means	mean	VERB
easat-5690	183	19	the	the	DET
easat-5690	183	20	area	area	NOUN
easat-5690	183	21	bounded	bound	VERB
easat-5690	183	22	by	by	ADP
easat-5690	183	23	the	the	DET
easat-5690	183	24	pr	pr	NOUN
easat-5690	183	25	curve	curve	NOUN
easat-5690	183	26	and	and	CCONJ
easat-5690	183	27	the	the	DET
easat-5690	183	28	axes	axis	NOUN
easat-5690	183	29	is	be	AUX
easat-5690	183	30	larger	large	ADJ
easat-5690	183	31	.	.	PUNCT
easat-5690	184	1	=	=	NUM
easat-5690	184	2	1	1	NUM
easat-5690	184	3	0	0	NUM
easat-5690	184	4	p(r)drap	p(r)drap	NOUN
easat-5690	184	5	(	(	PUNCT
easat-5690	185	1	3	3	NUM
easat-5690	185	2	)	)	PUNCT
easat-5690	185	3	2.5.6	2.5.6	NUM
easat-5690	185	4	.	.	PUNCT
easat-5690	185	5	map	map	NOUN
easat-5690	185	6	determine	determine	VERB
easat-5690	185	7	the	the	DET
easat-5690	185	8	average	average	ADJ
easat-5690	185	9	ap	ap	PROPN
easat-5690	185	10	value	value	NOUN
easat-5690	185	11	for	for	ADP
easat-5690	185	12	each	each	DET
easat-5690	185	13	category	category	NOUN
easat-5690	185	14	using	use	VERB
easat-5690	185	15	the	the	DET
easat-5690	185	16	map	map	NOUN
easat-5690	185	17	(	(	PUNCT
easat-5690	185	18	mean	mean	NOUN
easat-5690	185	19	of	of	ADP
easat-5690	185	20	average	average	ADJ
easat-5690	185	21	precision	precision	NOUN
easat-5690	185	22	)	)	PUNCT
easat-5690	185	23	formula	formula	NOUN
easat-5690	185	24	.	.	PUNCT
easat-5690	186	1	the	the	DET
easat-5690	186	2	accuracy	accuracy	NOUN
easat-5690	186	3	of	of	ADP
easat-5690	186	4	predictions	prediction	NOUN
easat-5690	186	5	for	for	ADP
easat-5690	186	6	each	each	DET
easat-5690	186	7	category	category	NOUN
easat-5690	186	8	can	can	AUX
easat-5690	186	9	be	be	AUX
easat-5690	186	10	represented	represent	VERB
easat-5690	186	11	by	by	ADP
easat-5690	186	12	ap	ap	PROPN
easat-5690	186	13	,	,	PUNCT
easat-5690	186	14	and	and	CCONJ
easat-5690	186	15	the	the	DET
easat-5690	186	16	accuracy	accuracy	NOUN
easat-5690	186	17	of	of	ADP
easat-5690	186	18	the	the	DET
easat-5690	186	19	overall	overall	ADJ
easat-5690	186	20	model	model	NOUN
easat-5690	186	21	is	be	AUX
easat-5690	186	22	represented	represent	VERB
easat-5690	186	23	by	by	ADP
easat-5690	186	24	map	map	NOUN
easat-5690	186	25	,	,	PUNCT
easat-5690	186	26	which	which	PRON
easat-5690	186	27	is	be	AUX
easat-5690	186	28	the	the	DET
easat-5690	186	29	average	average	NOUN
easat-5690	186	30	of	of	ADP
easat-5690	186	31	the	the	DET
easat-5690	186	32	aps	ap	NOUN
easat-5690	186	33	for	for	ADP
easat-5690	186	34	all	all	DET
easat-5690	186	35	categories	category	NOUN
easat-5690	186	36	.	.	PUNCT
easat-5690	187	1	the	the	DET
easat-5690	187	2	region	region	NOUN
easat-5690	187	3	that	that	PRON
easat-5690	187	4	the	the	DET
easat-5690	187	5	pr	pr	NOUN
easat-5690	187	6	curve	curve	NOUN
easat-5690	187	7	and	and	CCONJ
easat-5690	187	8	coordinate	coordinate	VERB
easat-5690	187	9	axis	axis	NOUN
easat-5690	187	10	enclose	enclose	ADJ
easat-5690	187	11	increases	increase	NOUN
easat-5690	187	12	with	with	ADP
easat-5690	187	13	map	map	NOUN
easat-5690	187	14	.	.	PUNCT
easat-5690	188	1	we	we	PRON
easat-5690	188	2	frequently	frequently	ADV
easat-5690	188	3	discuss	discuss	VERB
easat-5690	188	4	the	the	DET
easat-5690	188	5	accuracy	accuracy	NOUN
easat-5690	188	6	1791	1791	NUM
easat-5690	188	7	edelweiss	edelweiss	PROPN
easat-5690	188	8	applied	apply	VERB
easat-5690	188	9	science	science	NOUN
easat-5690	188	10	and	and	CCONJ
easat-5690	188	11	technology	technology	NOUN
easat-5690	188	12	issn	issn	PROPN
easat-5690	188	13	:	:	PUNCT
easat-5690	188	14	2576	2576	NUM
easat-5690	188	15	-	-	SYM
easat-5690	188	16	8484	8484	NUM
easat-5690	188	17	vol	vol	NOUN
easat-5690	188	18	.	.	PROPN
easat-5690	189	1	9	9	NUM
easat-5690	189	2	,	,	PUNCT
easat-5690	189	3	no	no	INTJ
easat-5690	189	4	.	.	NOUN
easat-5690	190	1	3	3	NUM
easat-5690	190	2	:	:	SYM
easat-5690	190	3	1777	1777	NUM
easat-5690	190	4	-	-	SYM
easat-5690	190	5	1794	1794	NUM
easat-5690	190	6	,	,	PUNCT
easat-5690	190	7	2025	2025	NUM
easat-5690	190	8	doi	doi	NOUN
easat-5690	190	9	:	:	PUNCT
easat-5690	190	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	190	11	©	©	PROPN
easat-5690	190	12	2025	2025	NUM
easat-5690	190	13	by	by	ADP
easat-5690	190	14	the	the	DET
easat-5690	190	15	authors	author	NOUN
easat-5690	190	16	;	;	PUNCT
easat-5690	190	17	licensee	licensee	NOUN
easat-5690	190	18	learning	learn	VERB
easat-5690	190	19	gate	gate	NOUN
easat-5690	190	20	of	of	ADP
easat-5690	190	21	a	a	DET
easat-5690	190	22	particular	particular	ADJ
easat-5690	190	23	object	object	NOUN
easat-5690	190	24	detection	detection	NOUN
easat-5690	190	25	technique	technique	NOUN
easat-5690	190	26	,	,	PUNCT
easat-5690	190	27	commonly	commonly	ADV
easat-5690	190	28	known	know	VERB
easat-5690	190	29	as	as	ADP
easat-5690	190	30	map	map	NOUN
easat-5690	190	31	.	.	PUNCT
easat-5690	191	1	n	n	CCONJ
easat-5690	191	2	n	n	PROPN
easat-5690	191	3	1i	1i	NUM
easat-5690	191	4	api	api	NOUN
easat-5690	191	5	map	map	NOUN
easat-5690	191	6			X
easat-5690	192	1	=	=	PUNCT
easat-5690	192	2	=	=	SYM
easat-5690	192	3	(	(	PUNCT
easat-5690	192	4	4	4	NUM
easat-5690	192	5	)	)	PUNCT
easat-5690	192	6	2.5.7	2.5.7	NUM
easat-5690	192	7	.	.	PUNCT
easat-5690	193	1	map@0.5	map@0.5	VERB
easat-5690	193	2	the	the	DET
easat-5690	193	3	yolo	yolo	ADJ
easat-5690	193	4	model	model	NOUN
easat-5690	193	5	displays	display	VERB
easat-5690	193	6	map@0.5	map@0.5	PROPN
easat-5690	193	7	.	.	PROPN
easat-5690	193	8	when	when	SCONJ
easat-5690	193	9	the	the	DET
easat-5690	193	10	iou	iou	NOUN
easat-5690	193	11	(	(	PUNCT
easat-5690	193	12	intersection	intersection	NOUN
easat-5690	193	13	over	over	ADP
easat-5690	193	14	union	union	NOUN
easat-5690	193	15	)	)	PUNCT
easat-5690	193	16	criterion	criterion	NOUN
easat-5690	193	17	is	be	AUX
easat-5690	193	18	0.5	0.5	NUM
easat-5690	193	19	,	,	PUNCT
easat-5690	193	20	this	this	DET
easat-5690	193	21	expression	expression	NOUN
easat-5690	193	22	shows	show	VERB
easat-5690	193	23	the	the	DET
easat-5690	193	24	value	value	NOUN
easat-5690	193	25	of	of	ADP
easat-5690	193	26	map	map	NOUN
easat-5690	193	27	.	.	PUNCT
easat-5690	194	1	the	the	DET
easat-5690	194	2	object	object	NOUN
easat-5690	194	3	prediction	prediction	NOUN
easat-5690	194	4	is	be	AUX
easat-5690	194	5	deemed	deem	VERB
easat-5690	194	6	accurate	accurate	ADJ
easat-5690	194	7	when	when	SCONJ
easat-5690	194	8	the	the	DET
easat-5690	194	9	iou	iou	NOUN
easat-5690	194	10	between	between	ADP
easat-5690	194	11	the	the	DET
easat-5690	194	12	predicted	predict	VERB
easat-5690	194	13	box	box	NOUN
easat-5690	194	14	and	and	CCONJ
easat-5690	194	15	the	the	DET
easat-5690	194	16	annotated	annotate	VERB
easat-5690	194	17	box	box	NOUN
easat-5690	194	18	is	be	AUX
easat-5690	194	19	larger	large	ADJ
easat-5690	194	20	than	than	ADP
easat-5690	194	21	0.5	0.5	NUM
easat-5690	194	22	;	;	PUNCT
easat-5690	194	23	map	map	NOUN
easat-5690	194	24	is	be	AUX
easat-5690	194	25	then	then	ADV
easat-5690	194	26	computed	compute	VERB
easat-5690	194	27	using	use	VERB
easat-5690	194	28	this	this	DET
easat-5690	194	29	assumption	assumption	NOUN
easat-5690	194	30	.	.	PUNCT
easat-5690	195	1	map@0.5	map@0.5	NOUN
easat-5690	195	2	,	,	PUNCT
easat-5690	195	3	generally	generally	ADV
easat-5690	195	4	speaking	speak	VERB
easat-5690	195	5	.	.	PUNCT
easat-5690	196	1	it	it	PRON
easat-5690	196	2	is	be	AUX
easat-5690	196	3	among	among	ADP
easat-5690	196	4	the	the	DET
easat-5690	196	5	metrics	metric	NOUN
easat-5690	196	6	used	use	VERB
easat-5690	196	7	to	to	PART
easat-5690	196	8	assess	assess	VERB
easat-5690	196	9	the	the	DET
easat-5690	196	10	yolo	yolo	ADJ
easat-5690	196	11	model	model	NOUN
easat-5690	196	12	.	.	PUNCT
easat-5690	197	1	2.5.8	2.5.8	X
easat-5690	197	2	.	.	PUNCT
easat-5690	198	1	map@	map@	VERB
easat-5690	199	1	[	[	X
easat-5690	199	2	0.5:0.95	0.5:0.95	X
easat-5690	199	3	]	]	PUNCT
easat-5690	199	4	another	another	DET
easat-5690	199	5	way	way	NOUN
easat-5690	199	6	to	to	PART
easat-5690	199	7	express	express	VERB
easat-5690	199	8	the	the	DET
easat-5690	199	9	mean	mean	ADJ
easat-5690	199	10	average	average	ADJ
easat-5690	199	11	precision	precision	NOUN
easat-5690	199	12	at	at	ADP
easat-5690	199	13	different	different	ADJ
easat-5690	199	14	intersection	intersection	NOUN
easat-5690	199	15	over	over	ADP
easat-5690	199	16	union	union	NOUN
easat-5690	199	17	(	(	PUNCT
easat-5690	199	18	iou	iou	NOUN
easat-5690	199	19	)	)	PUNCT
easat-5690	199	20	thresholds	threshold	NOUN
easat-5690	199	21	in	in	ADP
easat-5690	199	22	the	the	DET
easat-5690	199	23	yolo	yolo	ADJ
easat-5690	199	24	model	model	NOUN
easat-5690	199	25	is	be	AUX
easat-5690	199	26	as	as	ADV
easat-5690	199	27	map	map	NOUN
easat-5690	199	28	@	@	ADP
easat-5690	200	1	[	[	X
easat-5690	200	2	0.5:0.95	0.5:0.95	X
easat-5690	200	3	]	]	PUNCT
easat-5690	200	4	.	.	PUNCT
easat-5690	201	1	ten	ten	NUM
easat-5690	201	2	iou	iou	NOUN
easat-5690	201	3	thresholds	threshold	NOUN
easat-5690	201	4	are	be	AUX
easat-5690	201	5	set	set	VERB
easat-5690	201	6	with	with	ADP
easat-5690	201	7	a	a	DET
easat-5690	201	8	step	step	NOUN
easat-5690	201	9	size	size	NOUN
easat-5690	201	10	of	of	ADP
easat-5690	201	11	0.05	0.05	NUM
easat-5690	201	12	within	within	ADP
easat-5690	201	13	the	the	DET
easat-5690	201	14	range	range	NOUN
easat-5690	201	15	of	of	ADP
easat-5690	201	16	[	[	X
easat-5690	201	17	0.5	0.5	NUM
easat-5690	201	18	,	,	PUNCT
easat-5690	201	19	0.95	0.95	NUM
easat-5690	201	20	]	]	PUNCT
easat-5690	201	21	,	,	PUNCT
easat-5690	201	22	and	and	CCONJ
easat-5690	201	23	the	the	DET
easat-5690	201	24	average	average	ADJ
easat-5690	201	25	precision	precision	NOUN
easat-5690	201	26	at	at	ADP
easat-5690	201	27	each	each	PRON
easat-5690	201	28	of	of	ADP
easat-5690	201	29	these	these	DET
easat-5690	201	30	thresholds	threshold	NOUN
easat-5690	201	31	is	be	AUX
easat-5690	201	32	computed	compute	VERB
easat-5690	201	33	.	.	PUNCT
easat-5690	202	1	the	the	DET
easat-5690	202	2	map	map	NOUN
easat-5690	202	3	@	@	ADP
easat-5690	203	1	[	[	X
easat-5690	203	2	0.5:0.95	0.5:0.95	X
easat-5690	203	3	]	]	X
easat-5690	203	4	is	be	AUX
easat-5690	203	5	then	then	ADV
easat-5690	203	6	calculated	calculate	VERB
easat-5690	203	7	by	by	ADP
easat-5690	203	8	taking	take	VERB
easat-5690	203	9	the	the	DET
easat-5690	203	10	mean	mean	NOUN
easat-5690	203	11	of	of	ADP
easat-5690	203	12	these	these	DET
easat-5690	203	13	values	value	NOUN
easat-5690	203	14	.	.	PUNCT
easat-5690	204	1	3	3	X
easat-5690	204	2	.	.	NOUN
easat-5690	204	3	results	result	NOUN
easat-5690	204	4	and	and	CCONJ
easat-5690	204	5	discussion	discussion	NOUN
easat-5690	204	6	3.1	3.1	NUM
easat-5690	204	7	.	.	PUNCT
easat-5690	205	1	the	the	DET
easat-5690	205	2	diseases	disease	NOUN
easat-5690	205	3	and	and	CCONJ
easat-5690	205	4	pests	pest	NOUN
easat-5690	205	5	in	in	ADP
easat-5690	205	6	tea	tea	NOUN
easat-5690	205	7	leaves	leave	VERB
easat-5690	205	8	detection	detection	NOUN
easat-5690	205	9	from	from	ADP
easat-5690	205	10	the	the	DET
easat-5690	205	11	figure	figure	NOUN
easat-5690	205	12	8	8	NUM
easat-5690	205	13	below	below	ADV
easat-5690	205	14	,	,	PUNCT
easat-5690	205	15	it	it	PRON
easat-5690	205	16	can	can	AUX
easat-5690	205	17	be	be	AUX
easat-5690	205	18	seen	see	VERB
easat-5690	205	19	that	that	SCONJ
easat-5690	205	20	as	as	ADP
easat-5690	205	21	the	the	DET
easat-5690	205	22	number	number	NOUN
easat-5690	205	23	of	of	ADP
easat-5690	205	24	training	training	NOUN
easat-5690	205	25	iterations	iteration	NOUN
easat-5690	205	26	increases	increase	NOUN
easat-5690	205	27	,	,	PUNCT
easat-5690	205	28	the	the	DET
easat-5690	205	29	training	training	NOUN
easat-5690	205	30	validation	validation	NOUN
easat-5690	205	31	loss	loss	NOUN
easat-5690	205	32	of	of	ADP
easat-5690	205	33	the	the	DET
easat-5690	205	34	model	model	NOUN
easat-5690	205	35	gradually	gradually	ADV
easat-5690	205	36	decreases	decrease	VERB
easat-5690	205	37	,	,	PUNCT
easat-5690	205	38	indicating	indicate	VERB
easat-5690	205	39	that	that	SCONJ
easat-5690	205	40	the	the	DET
easat-5690	205	41	model	model	NOUN
easat-5690	205	42	continuously	continuously	ADV
easat-5690	205	43	learns	learn	VERB
easat-5690	205	44	more	more	ADV
easat-5690	205	45	accurate	accurate	ADJ
easat-5690	205	46	features	feature	NOUN
easat-5690	205	47	.	.	PUNCT
easat-5690	206	1	following	follow	VERB
easat-5690	206	2	training	training	NOUN
easat-5690	206	3	,	,	PUNCT
easat-5690	206	4	we	we	PRON
easat-5690	206	5	assessed	assess	VERB
easat-5690	206	6	the	the	DET
easat-5690	206	7	model	model	NOUN
easat-5690	206	8	using	use	VERB
easat-5690	206	9	the	the	DET
easat-5690	206	10	dataset	dataset	NOUN
easat-5690	206	11	's	's	PART
easat-5690	206	12	validation	validation	NOUN
easat-5690	206	13	set	set	NOUN
easat-5690	206	14	and	and	CCONJ
easat-5690	206	15	got	get	VERB
easat-5690	206	16	the	the	DET
easat-5690	206	17	following	follow	VERB
easat-5690	206	18	outcomes	outcome	NOUN
easat-5690	206	19	.	.	PUNCT
easat-5690	207	1	1792	1792	NUM
easat-5690	207	2	edelweiss	edelweiss	PROPN
easat-5690	207	3	applied	apply	VERB
easat-5690	207	4	science	science	NOUN
easat-5690	207	5	and	and	CCONJ
easat-5690	207	6	technology	technology	NOUN
easat-5690	207	7	issn	issn	PROPN
easat-5690	207	8	:	:	PUNCT
easat-5690	207	9	2576	2576	NUM
easat-5690	207	10	-	-	SYM
easat-5690	207	11	8484	8484	NUM
easat-5690	207	12	vol	vol	NOUN
easat-5690	207	13	.	.	PROPN
easat-5690	208	1	9	9	NUM
easat-5690	208	2	,	,	PUNCT
easat-5690	208	3	no	no	INTJ
easat-5690	208	4	.	.	NOUN
easat-5690	209	1	3	3	NUM
easat-5690	209	2	:	:	SYM
easat-5690	209	3	1777	1777	NUM
easat-5690	209	4	-	-	SYM
easat-5690	209	5	1794	1794	NUM
easat-5690	209	6	,	,	PUNCT
easat-5690	209	7	2025	2025	NUM
easat-5690	209	8	doi	doi	NOUN
easat-5690	209	9	:	:	PUNCT
easat-5690	209	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	209	11	©	©	PROPN
easat-5690	209	12	2025	2025	NUM
easat-5690	209	13	by	by	ADP
easat-5690	209	14	the	the	DET
easat-5690	209	15	authors	author	NOUN
easat-5690	209	16	;	;	PUNCT
easat-5690	209	17	licensee	licensee	PROPN
easat-5690	209	18	learning	learn	VERB
easat-5690	209	19	gate	gate	PROPN
easat-5690	209	20	figure	figure	NOUN
easat-5690	209	21	8	8	NUM
easat-5690	209	22	.	.	PUNCT
easat-5690	210	1	the	the	DET
easat-5690	210	2	training	training	NOUN
easat-5690	210	3	results	result	NOUN
easat-5690	210	4	.	.	PUNCT
easat-5690	211	1	consequently	consequently	ADV
easat-5690	211	2	,	,	PUNCT
easat-5690	211	3	we	we	PRON
easat-5690	211	4	decided	decide	VERB
easat-5690	211	5	to	to	PART
easat-5690	211	6	choose	choose	VERB
easat-5690	211	7	the	the	DET
easat-5690	211	8	weight	weight	NOUN
easat-5690	211	9	that	that	PRON
easat-5690	211	10	yielded	yield	VERB
easat-5690	211	11	the	the	DET
easat-5690	211	12	best	good	ADJ
easat-5690	211	13	accuracy	accuracy	NOUN
easat-5690	211	14	as	as	ADP
easat-5690	211	15	our	our	PRON
easat-5690	211	16	ideal	ideal	ADJ
easat-5690	211	17	weight	weight	NOUN
easat-5690	211	18	.	.	PUNCT
easat-5690	212	1	the	the	DET
easat-5690	212	2	pr	pr	NOUN
easat-5690	212	3	curve	curve	NOUN
easat-5690	212	4	of	of	ADP
easat-5690	212	5	our	our	PRON
easat-5690	212	6	enhanced	enhanced	ADJ
easat-5690	212	7	yolov5	yolov5	NOUN
easat-5690	212	8	model	model	NOUN
easat-5690	212	9	on	on	ADP
easat-5690	212	10	the	the	DET
easat-5690	212	11	validation	validation	NOUN
easat-5690	212	12	data	datum	NOUN
easat-5690	212	13	is	be	AUX
easat-5690	212	14	displayed	display	VERB
easat-5690	212	15	in	in	ADP
easat-5690	212	16	figure	figure	NOUN
easat-5690	212	17	9	9	NUM
easat-5690	212	18	.	.	PUNCT
easat-5690	212	19	figure	figure	NOUN
easat-5690	212	20	9	9	NUM
easat-5690	212	21	shows	show	VERB
easat-5690	212	22	that	that	SCONJ
easat-5690	212	23	the	the	DET
easat-5690	212	24	model	model	NOUN
easat-5690	212	25	's	's	PART
easat-5690	212	26	accuracy	accuracy	NOUN
easat-5690	212	27	and	and	CCONJ
easat-5690	212	28	recall	recall	NOUN
easat-5690	212	29	were	be	AUX
easat-5690	212	30	both	both	ADV
easat-5690	212	31	high	high	ADJ
easat-5690	212	32	.	.	PUNCT
easat-5690	213	1	figure	figure	NOUN
easat-5690	213	2	9	9	NUM
easat-5690	213	3	.	.	PUNCT
easat-5690	214	1	the	the	DET
easat-5690	214	2	precision	precision	NOUN
easat-5690	214	3	-	-	PUNCT
easat-5690	214	4	recall	recall	NOUN
easat-5690	214	5	curve	curve	NOUN
easat-5690	214	6	.	.	PUNCT
easat-5690	215	1	1793	1793	NUM
easat-5690	215	2	edelweiss	edelweiss	PROPN
easat-5690	215	3	applied	apply	VERB
easat-5690	215	4	science	science	NOUN
easat-5690	215	5	and	and	CCONJ
easat-5690	215	6	technology	technology	NOUN
easat-5690	215	7	issn	issn	PROPN
easat-5690	215	8	:	:	PUNCT
easat-5690	215	9	2576	2576	NUM
easat-5690	215	10	-	-	SYM
easat-5690	215	11	8484	8484	NUM
easat-5690	215	12	vol	vol	NOUN
easat-5690	215	13	.	.	PROPN
easat-5690	216	1	9	9	NUM
easat-5690	216	2	,	,	PUNCT
easat-5690	216	3	no	no	INTJ
easat-5690	216	4	.	.	NOUN
easat-5690	217	1	3	3	NUM
easat-5690	217	2	:	:	SYM
easat-5690	217	3	1777	1777	NUM
easat-5690	217	4	-	-	SYM
easat-5690	217	5	1794	1794	NUM
easat-5690	217	6	,	,	PUNCT
easat-5690	217	7	2025	2025	NUM
easat-5690	217	8	doi	doi	NOUN
easat-5690	217	9	:	:	PUNCT
easat-5690	217	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	217	11	©	©	PROPN
easat-5690	217	12	2025	2025	NUM
easat-5690	217	13	by	by	ADP
easat-5690	217	14	the	the	DET
easat-5690	217	15	authors	author	NOUN
easat-5690	217	16	;	;	PUNCT
easat-5690	217	17	licensee	licensee	PROPN
easat-5690	217	18	learning	learning	NOUN
easat-5690	217	19	gate	gate	NOUN
easat-5690	217	20	as	as	SCONJ
easat-5690	217	21	shown	show	VERB
easat-5690	217	22	in	in	ADP
easat-5690	217	23	figure	figure	NOUN
easat-5690	217	24	10(a	10(a	NUM
easat-5690	217	25	)	)	PUNCT
easat-5690	217	26	and	and	CCONJ
easat-5690	217	27	figure	figure	VERB
easat-5690	217	28	10(b	10(b	NUM
easat-5690	217	29	)	)	PUNCT
easat-5690	217	30	,	,	PUNCT
easat-5690	217	31	the	the	DET
easat-5690	217	32	enhanced	enhance	VERB
easat-5690	217	33	yolov5s	yolov5s	PROPN
easat-5690	217	34	model	model	NOUN
easat-5690	217	35	outperforms	outperform	VERB
easat-5690	217	36	its	its	PRON
easat-5690	217	37	original	original	ADJ
easat-5690	217	38	counterpart	counterpart	NOUN
easat-5690	217	39	.	.	PUNCT
easat-5690	218	1	1794	1794	NUM
easat-5690	218	2	edelweiss	edelweiss	PROPN
easat-5690	218	3	applied	apply	VERB
easat-5690	218	4	science	science	NOUN
easat-5690	218	5	and	and	CCONJ
easat-5690	218	6	technology	technology	NOUN
easat-5690	218	7	issn	issn	PROPN
easat-5690	218	8	:	:	PUNCT
easat-5690	218	9	2576	2576	NUM
easat-5690	218	10	-	-	SYM
easat-5690	218	11	8484	8484	NUM
easat-5690	218	12	vol	vol	NOUN
easat-5690	218	13	.	.	PROPN
easat-5690	219	1	9	9	NUM
easat-5690	219	2	,	,	PUNCT
easat-5690	219	3	no	no	INTJ
easat-5690	219	4	.	.	NOUN
easat-5690	220	1	3	3	NUM
easat-5690	220	2	:	:	SYM
easat-5690	220	3	1777	1777	NUM
easat-5690	220	4	-	-	SYM
easat-5690	220	5	1794	1794	NUM
easat-5690	220	6	,	,	PUNCT
easat-5690	220	7	2025	2025	NUM
easat-5690	220	8	doi	doi	NOUN
easat-5690	220	9	:	:	PUNCT
easat-5690	220	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	220	11	©	©	PROPN
easat-5690	220	12	2025	2025	NUM
easat-5690	220	13	by	by	ADP
easat-5690	220	14	the	the	DET
easat-5690	220	15	authors	author	NOUN
easat-5690	220	16	;	;	PUNCT
easat-5690	220	17	licensee	licensee	PROPN
easat-5690	220	18	learning	learn	VERB
easat-5690	220	19	gate	gate	PROPN
easat-5690	220	20	figure	figure	NOUN
easat-5690	220	21	10(a	10(a	NUM
easat-5690	220	22	)	)	PUNCT
easat-5690	220	23	depicts	depict	VERB
easat-5690	220	24	the	the	DET
easat-5690	220	25	box	box	PROPN
easat-5690	220	26	validation	validation	NOUN
easat-5690	220	27	loss	loss	NOUN
easat-5690	220	28	,	,	PUNCT
easat-5690	220	29	which	which	PRON
easat-5690	220	30	quantifies	quantify	VERB
easat-5690	220	31	the	the	DET
easat-5690	220	32	accuracy	accuracy	NOUN
easat-5690	220	33	of	of	ADP
easat-5690	220	34	locating	locate	VERB
easat-5690	220	35	pests	pest	NOUN
easat-5690	220	36	and	and	CCONJ
easat-5690	220	37	diseases	disease	NOUN
easat-5690	220	38	in	in	ADP
easat-5690	220	39	tea	tea	NOUN
easat-5690	220	40	images	image	NOUN
easat-5690	220	41	.	.	PUNCT
easat-5690	221	1	this	this	DET
easat-5690	221	2	metric	metric	NOUN
easat-5690	221	3	exhibits	exhibit	VERB
easat-5690	221	4	a	a	DET
easat-5690	221	5	steady	steady	ADJ
easat-5690	221	6	downward	downward	ADJ
easat-5690	221	7	trend	trend	NOUN
easat-5690	221	8	,	,	PUNCT
easat-5690	221	9	indicative	indicative	ADJ
easat-5690	221	10	of	of	ADP
easat-5690	221	11	the	the	DET
easat-5690	221	12	model	model	NOUN
easat-5690	221	13	's	's	PART
easat-5690	221	14	training	training	NOUN
easat-5690	221	15	performance	performance	NOUN
easat-5690	221	16	.	.	PUNCT
easat-5690	222	1	notably	notably	ADV
easat-5690	222	2	,	,	PUNCT
easat-5690	222	3	the	the	DET
easat-5690	222	4	improved	improved	ADJ
easat-5690	222	5	yolov5s	yolov5s	PROPN
easat-5690	222	6	achieves	achieve	VERB
easat-5690	222	7	a	a	DET
easat-5690	222	8	lower	low	ADJ
easat-5690	222	9	box	box	NOUN
easat-5690	222	10	validation	validation	NOUN
easat-5690	222	11	loss	loss	NOUN
easat-5690	222	12	compared	compare	VERB
easat-5690	222	13	to	to	ADP
easat-5690	222	14	the	the	DET
easat-5690	222	15	original	original	NOUN
easat-5690	222	16	,	,	PUNCT
easat-5690	222	17	highlighting	highlight	VERB
easat-5690	222	18	the	the	DET
easat-5690	222	19	effectiveness	effectiveness	NOUN
easat-5690	222	20	of	of	ADP
easat-5690	222	21	the	the	DET
easat-5690	222	22	deeper	deep	ADJ
easat-5690	222	23	neural	neural	ADJ
easat-5690	222	24	network	network	NOUN
easat-5690	222	25	architecture	architecture	NOUN
easat-5690	222	26	.	.	PUNCT
easat-5690	223	1	as	as	SCONJ
easat-5690	223	2	the	the	DET
easat-5690	223	3	model	model	NOUN
easat-5690	223	4	undergoes	undergo	VERB
easat-5690	223	5	training	training	NOUN
easat-5690	223	6	,	,	PUNCT
easat-5690	223	7	its	its	PRON
easat-5690	223	8	performance	performance	NOUN
easat-5690	223	9	gains	gain	NOUN
easat-5690	223	10	momentum	momentum	NOUN
easat-5690	223	11	.	.	PUNCT
easat-5690	224	1	figure	figure	NOUN
easat-5690	224	2	10(b	10(b	NUM
easat-5690	224	3	)	)	PUNCT
easat-5690	224	4	further	far	ADV
easat-5690	224	5	demonstrates	demonstrate	VERB
easat-5690	224	6	that	that	SCONJ
easat-5690	224	7	the	the	DET
easat-5690	224	8	reduced	reduce	VERB
easat-5690	224	9	box	box	NOUN
easat-5690	224	10	validation	validation	NOUN
easat-5690	224	11	loss	loss	NOUN
easat-5690	224	12	correlates	correlate	VERB
easat-5690	224	13	with	with	ADP
easat-5690	224	14	an	an	DET
easat-5690	224	15	increase	increase	NOUN
easat-5690	224	16	in	in	ADP
easat-5690	224	17	mean	mean	ADJ
easat-5690	224	18	average	average	ADJ
easat-5690	224	19	precision	precision	NOUN
easat-5690	224	20	(	(	PUNCT
easat-5690	224	21	map	map	NOUN
easat-5690	224	22	)	)	PUNCT
easat-5690	224	23	.	.	PUNCT
easat-5690	225	1	specifically	specifically	ADV
easat-5690	225	2	,	,	PUNCT
easat-5690	225	3	the	the	DET
easat-5690	225	4	enhanced	enhance	VERB
easat-5690	225	5	yolov5s	yolov5s	PROPN
easat-5690	225	6	achieves	achieve	VERB
easat-5690	225	7	a	a	DET
easat-5690	225	8	map	map	NOUN
easat-5690	225	9	of	of	ADP
easat-5690	225	10	85.32	85.32	NUM
easat-5690	225	11	%	%	NOUN
easat-5690	225	12	,	,	PUNCT
easat-5690	225	13	surpassing	surpass	VERB
easat-5690	225	14	the	the	DET
easat-5690	225	15	81.3	81.3	NUM
easat-5690	225	16	%	%	NOUN
easat-5690	225	17	of	of	ADP
easat-5690	225	18	the	the	DET
easat-5690	225	19	original	original	ADJ
easat-5690	225	20	model	model	NOUN
easat-5690	225	21	.	.	PUNCT
easat-5690	226	1	figure	figure	NOUN
easat-5690	226	2	10	10	NUM
easat-5690	226	3	.	.	PUNCT
easat-5690	227	1	the	the	DET
easat-5690	227	2	training	training	NOUN
easat-5690	227	3	outcome	outcome	NOUN
easat-5690	227	4	of	of	ADP
easat-5690	227	5	models	model	NOUN
easat-5690	227	6	.	.	PUNCT
easat-5690	228	1	3.2	3.2	NUM
easat-5690	228	2	.	.	PUNCT
easat-5690	228	3	comparison	comparison	NOUN
easat-5690	228	4	of	of	ADP
easat-5690	228	5	the	the	DET
easat-5690	228	6	model	model	NOUN
easat-5690	228	7	's	's	PART
easat-5690	228	8	output	output	NOUN
easat-5690	228	9	we	we	PRON
easat-5690	228	10	used	use	VERB
easat-5690	228	11	the	the	DET
easat-5690	228	12	test	test	NOUN
easat-5690	228	13	set	set	VERB
easat-5690	228	14	to	to	PART
easat-5690	228	15	compare	compare	VERB
easat-5690	228	16	the	the	DET
easat-5690	228	17	performance	performance	NOUN
easat-5690	228	18	of	of	ADP
easat-5690	228	19	the	the	DET
easat-5690	228	20	enhanced	enhanced	ADJ
easat-5690	228	21	yolo	yolo	ADJ
easat-5690	228	22	v5	v5	PROPN
easat-5690	228	23	model	model	NOUN
easat-5690	228	24	suggested	suggest	VERB
easat-5690	228	25	in	in	ADP
easat-5690	228	26	this	this	DET
easat-5690	228	27	study	study	NOUN
easat-5690	228	28	with	with	ADP
easat-5690	228	29	other	other	ADJ
easat-5690	228	30	common	common	ADJ
easat-5690	228	31	convolutional	convolutional	ADJ
easat-5690	228	32	neural	neural	ADJ
easat-5690	228	33	networks	network	NOUN
easat-5690	228	34	,	,	PUNCT
easat-5690	228	35	including	include	VERB
easat-5690	228	36	faster	fast	ADJ
easat-5690	228	37	-	-	PUNCT
easat-5690	228	38	rcnn	rcnn	NOUN
easat-5690	228	39	,	,	PUNCT
easat-5690	228	40	ssd	ssd	PROPN
easat-5690	228	41	,	,	PUNCT
easat-5690	228	42	yolov3	yolov3	PROPN
easat-5690	228	43	,	,	PUNCT
easat-5690	228	44	yolov4	yolov4	PROPN
easat-5690	228	45	,	,	PUNCT
easat-5690	228	46	and	and	CCONJ
easat-5690	228	47	yolo	yolo	ADJ
easat-5690	228	48	v5	v5	PROPN
easat-5690	228	49	,	,	PUNCT
easat-5690	228	50	for	for	ADP
easat-5690	228	51	the	the	DET
easat-5690	228	52	detection	detection	NOUN
easat-5690	228	53	of	of	ADP
easat-5690	228	54	tea	tea	NOUN
easat-5690	228	55	illness	illness	NOUN
easat-5690	228	56	.	.	PUNCT
easat-5690	229	1	the	the	DET
easat-5690	229	2	test	test	NOUN
easat-5690	229	3	set	set	NOUN
easat-5690	229	4	results	result	NOUN
easat-5690	229	5	are	be	AUX
easat-5690	229	6	displayed	display	VERB
easat-5690	229	7	in	in	ADP
easat-5690	229	8	table	table	NOUN
easat-5690	229	9	2	2	NUM
easat-5690	229	10	,	,	PUNCT
easat-5690	229	11	which	which	PRON
easat-5690	229	12	shows	show	VERB
easat-5690	229	13	that	that	SCONJ
easat-5690	229	14	our	our	PRON
easat-5690	229	15	model	model	NOUN
easat-5690	229	16	has	have	AUX
easat-5690	229	17	significantly	significantly	ADV
easat-5690	229	18	outperformed	outperform	VERB
easat-5690	229	19	other	other	ADJ
easat-5690	229	20	models	model	NOUN
easat-5690	229	21	.	.	PUNCT
easat-5690	230	1	the	the	DET
easat-5690	230	2	map	map	NOUN
easat-5690	230	3	has	have	AUX
easat-5690	230	4	increased	increase	VERB
easat-5690	230	5	by	by	ADP
easat-5690	230	6	over	over	ADP
easat-5690	230	7	4	4	NUM
easat-5690	230	8	percentage	percentage	NOUN
easat-5690	230	9	points	point	NOUN
easat-5690	230	10	when	when	SCONJ
easat-5690	230	11	compared	compare	VERB
easat-5690	230	12	to	to	ADP
easat-5690	230	13	the	the	DET
easat-5690	230	14	same	same	ADJ
easat-5690	230	15	series	series	NOUN
easat-5690	230	16	of	of	ADP
easat-5690	230	17	yolo	yolo	ADJ
easat-5690	230	18	v5	v5	PROPN
easat-5690	230	19	,	,	PUNCT
easat-5690	230	20	suggesting	suggest	VERB
easat-5690	230	21	that	that	SCONJ
easat-5690	230	22	this	this	DET
easat-5690	230	23	model	model	NOUN
easat-5690	230	24	is	be	AUX
easat-5690	230	25	useful	useful	ADJ
easat-5690	230	26	for	for	ADP
easat-5690	230	27	detecting	detect	VERB
easat-5690	230	28	tea	tea	NOUN
easat-5690	230	29	sickness	sickness	NOUN
easat-5690	230	30	.	.	PUNCT
easat-5690	231	1	1795	1795	NUM
easat-5690	231	2	edelweiss	edelweiss	PROPN
easat-5690	231	3	applied	apply	VERB
easat-5690	231	4	science	science	NOUN
easat-5690	231	5	and	and	CCONJ
easat-5690	231	6	technology	technology	NOUN
easat-5690	231	7	issn	issn	PROPN
easat-5690	231	8	:	:	PUNCT
easat-5690	231	9	2576	2576	NUM
easat-5690	231	10	-	-	SYM
easat-5690	231	11	8484	8484	NUM
easat-5690	231	12	vol	vol	NOUN
easat-5690	231	13	.	.	PROPN
easat-5690	232	1	9	9	NUM
easat-5690	232	2	,	,	PUNCT
easat-5690	232	3	no	no	INTJ
easat-5690	232	4	.	.	NOUN
easat-5690	233	1	3	3	NUM
easat-5690	233	2	:	:	SYM
easat-5690	233	3	1777	1777	NUM
easat-5690	233	4	-	-	SYM
easat-5690	233	5	1794	1794	NUM
easat-5690	233	6	,	,	PUNCT
easat-5690	233	7	2025	2025	NUM
easat-5690	233	8	doi	doi	NOUN
easat-5690	233	9	:	:	PUNCT
easat-5690	233	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	233	11	©	©	PROPN
easat-5690	233	12	2025	2025	NUM
easat-5690	233	13	by	by	ADP
easat-5690	233	14	the	the	DET
easat-5690	233	15	authors	author	NOUN
easat-5690	233	16	;	;	PUNCT
easat-5690	233	17	licensee	licensee	PROPN
easat-5690	233	18	learning	learn	VERB
easat-5690	233	19	gate	gate	NOUN
easat-5690	233	20	table	table	NOUN
easat-5690	233	21	2	2	NUM
easat-5690	233	22	.	.	PUNCT
easat-5690	234	1	the	the	DET
easat-5690	234	2	results	result	NOUN
easat-5690	234	3	of	of	ADP
easat-5690	234	4	tea	tea	NOUN
easat-5690	234	5	disease	disease	NOUN
easat-5690	234	6	detection	detection	NOUN
easat-5690	234	7	based	base	VERB
easat-5690	234	8	on	on	ADP
easat-5690	234	9	different	different	ADJ
easat-5690	234	10	methods	method	NOUN
easat-5690	234	11	.	.	PUNCT
easat-5690	234	12	’	'	PUNCT
easat-5690	235	1	model	model	NOUN
easat-5690	235	2	precision	precision	PROPN
easat-5690	235	3	recall	recall	VERB
easat-5690	235	4	map（%	map（%	PROPN
easat-5690	235	5	）	）	ADJ
easat-5690	235	6	faster	fast	ADJ
easat-5690	235	7	-	-	PUNCT
easat-5690	235	8	rcnn	rcnn	NOUN
easat-5690	235	9	78.46	78.46	NUM
easat-5690	235	10	%	%	NOUN
easat-5690	235	11	50.45	50.45	NUM
easat-5690	235	12	%	%	NOUN
easat-5690	235	13	73.95	73.95	NUM
easat-5690	235	14	%	%	NOUN
easat-5690	235	15	ssd	ssd	NOUN
easat-5690	235	16	80.70	80.70	NUM
easat-5690	235	17	%	%	NOUN
easat-5690	235	18	51.46	51.46	NUM
easat-5690	235	19	%	%	NOUN
easat-5690	235	20	62.36	62.36	NUM
easat-5690	235	21	%	%	NOUN
easat-5690	235	22	yolo	yolo	PROPN
easat-5690	235	23	v3	v3	PROPN
easat-5690	235	24	86.02	86.02	NUM
easat-5690	235	25	%	%	NOUN
easat-5690	235	26	64.98	64.98	NUM
easat-5690	235	27	%	%	NOUN
easat-5690	235	28	72.74	72.74	NUM
easat-5690	235	29	%	%	NOUN
easat-5690	235	30	yolo	yolo	ADJ
easat-5690	235	31	v4	v4	PROPN
easat-5690	235	32	87.56	87.56	NUM
easat-5690	235	33	%	%	NOUN
easat-5690	235	34	56.23	56.23	NUM
easat-5690	235	35	%	%	NOUN
easat-5690	235	36	66.86	66.86	NUM
easat-5690	235	37	%	%	NOUN
easat-5690	235	38	yolo	yolo	ADJ
easat-5690	235	39	v5	v5	PROPN
easat-5690	235	40	83.52	83.52	NUM
easat-5690	235	41	%	%	NOUN
easat-5690	235	42	71.71	71.71	NUM
easat-5690	235	43	%	%	NOUN
easat-5690	235	44	81.23	81.23	NUM
easat-5690	235	45	%	%	NOUN
easat-5690	235	46	our	our	PRON
easat-5690	235	47	method	method	NOUN
easat-5690	235	48	88.79	88.79	NUM
easat-5690	235	49	%	%	NOUN
easat-5690	235	50	74.97	74.97	NUM
easat-5690	235	51	%	%	NOUN
easat-5690	235	52	85.36	85.36	NUM
easat-5690	235	53	%	%	NOUN
easat-5690	235	54	the	the	DET
easat-5690	235	55	effect	effect	NOUN
easat-5690	235	56	of	of	ADP
easat-5690	235	57	our	our	PRON
easat-5690	235	58	improved	improved	ADJ
easat-5690	235	59	model	model	NOUN
easat-5690	235	60	performance	performance	NOUN
easat-5690	235	61	can	can	AUX
easat-5690	235	62	be	be	AUX
easat-5690	235	63	seen	see	VERB
easat-5690	235	64	more	more	ADV
easat-5690	235	65	intuitively	intuitively	ADV
easat-5690	235	66	in	in	ADP
easat-5690	235	67	figure	figure	NOUN
easat-5690	235	68	11	11	NUM
easat-5690	235	69	.	.	PUNCT
easat-5690	236	1	1796	1796	NUM
easat-5690	236	2	edelweiss	edelweiss	PROPN
easat-5690	236	3	applied	apply	VERB
easat-5690	236	4	science	science	NOUN
easat-5690	236	5	and	and	CCONJ
easat-5690	236	6	technology	technology	NOUN
easat-5690	236	7	issn	issn	PROPN
easat-5690	236	8	:	:	PUNCT
easat-5690	236	9	2576	2576	NUM
easat-5690	236	10	-	-	SYM
easat-5690	236	11	8484	8484	NUM
easat-5690	236	12	vol	vol	NOUN
easat-5690	236	13	.	.	PROPN
easat-5690	237	1	9	9	NUM
easat-5690	237	2	,	,	PUNCT
easat-5690	237	3	no	no	INTJ
easat-5690	237	4	.	.	NOUN
easat-5690	238	1	3	3	NUM
easat-5690	238	2	:	:	SYM
easat-5690	238	3	1777	1777	NUM
easat-5690	238	4	-	-	SYM
easat-5690	238	5	1794	1794	NUM
easat-5690	238	6	,	,	PUNCT
easat-5690	238	7	2025	2025	NUM
easat-5690	238	8	doi	doi	NOUN
easat-5690	238	9	:	:	PUNCT
easat-5690	238	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	238	11	©	©	PROPN
easat-5690	238	12	2025	2025	NUM
easat-5690	238	13	by	by	ADP
easat-5690	238	14	the	the	DET
easat-5690	238	15	authors	author	NOUN
easat-5690	238	16	;	;	PUNCT
easat-5690	238	17	licensee	licensee	PROPN
easat-5690	238	18	learning	learn	VERB
easat-5690	238	19	gate	gate	PROPN
easat-5690	238	20	figure	figure	NOUN
easat-5690	238	21	11	11	NUM
easat-5690	238	22	.	.	PUNCT
easat-5690	239	1	the	the	DET
easat-5690	239	2	performance	performance	NOUN
easat-5690	239	3	of	of	ADP
easat-5690	239	4	models	model	NOUN
easat-5690	239	5	.	.	PUNCT
easat-5690	240	1	4	4	X
easat-5690	240	2	.	.	X
easat-5690	240	3	tea	tea	NOUN
easat-5690	240	4	disease	disease	NOUN
easat-5690	240	5	and	and	CCONJ
easat-5690	240	6	pest	pest	VERB
easat-5690	240	7	detection	detection	NOUN
easat-5690	240	8	system	system	NOUN
easat-5690	240	9	this	this	DET
easat-5690	240	10	system	system	NOUN
easat-5690	240	11	uses	use	VERB
easat-5690	240	12	an	an	DET
easat-5690	240	13	improved	improved	ADJ
easat-5690	240	14	yolov5	yolov5	NOUN
easat-5690	240	15	object	object	NOUN
easat-5690	240	16	detection	detection	NOUN
easat-5690	240	17	model	model	NOUN
easat-5690	240	18	to	to	PART
easat-5690	240	19	train	train	VERB
easat-5690	240	20	a	a	DET
easat-5690	240	21	tea	tea	NOUN
easat-5690	240	22	pest	pest	NOUN
easat-5690	240	23	and	and	CCONJ
easat-5690	240	24	disease	disease	NOUN
easat-5690	240	25	dataset	dataset	VERB
easat-5690	240	26	and	and	CCONJ
easat-5690	240	27	uses	use	VERB
easat-5690	240	28	the	the	DET
easat-5690	240	29	pyside6	pyside6	PROPN
easat-5690	240	30	library	library	NOUN
easat-5690	240	31	to	to	PART
easat-5690	240	32	build	build	VERB
easat-5690	240	33	a	a	DET
easat-5690	240	34	page	page	NOUN
easat-5690	240	35	display	display	NOUN
easat-5690	240	36	system	system	NOUN
easat-5690	240	37	.	.	PUNCT
easat-5690	241	1	the	the	DET
easat-5690	241	2	main	main	ADJ
easat-5690	241	3	libraries	library	NOUN
easat-5690	241	4	used	use	VERB
easat-5690	241	5	in	in	ADP
easat-5690	241	6	the	the	DET
easat-5690	241	7	code	code	NOUN
easat-5690	241	8	include	include	VERB
easat-5690	241	9	pytorch	pytorch	NOUN
easat-5690	241	10	,	,	PUNCT
easat-5690	241	11	numpy	numpy	NOUN
easat-5690	241	12	,	,	PUNCT
easat-5690	241	13	opencv	opencv	NOUN
easat-5690	241	14	,	,	PUNCT
easat-5690	241	15	pyqt	pyqt	PROPN
easat-5690	241	16	,	,	PUNCT
easat-5690	241	17	etc	etc	X
easat-5690	241	18	.	.	X
easat-5690	241	19	,	,	PUNCT
easat-5690	241	20	to	to	PART
easat-5690	241	21	complete	complete	VERB
easat-5690	241	22	the	the	DET
easat-5690	241	23	development	development	NOUN
easat-5690	241	24	of	of	ADP
easat-5690	241	25	a	a	DET
easat-5690	241	26	tea	tea	NOUN
easat-5690	241	27	pest	pest	NOUN
easat-5690	241	28	and	and	CCONJ
easat-5690	241	29	disease	disease	NOUN
easat-5690	241	30	target	target	NOUN
easat-5690	241	31	detection	detection	NOUN
easat-5690	241	32	and	and	CCONJ
easat-5690	241	33	recognition	recognition	NOUN
easat-5690	241	34	page	page	NOUN
easat-5690	241	35	.	.	PUNCT
easat-5690	242	1	pyside6	pyside6	PROPN
easat-5690	242	2	is	be	AUX
easat-5690	242	3	one	one	NUM
easat-5690	242	4	of	of	ADP
easat-5690	242	5	the	the	DET
easat-5690	242	6	gui	gui	NOUN
easat-5690	242	7	(	(	PUNCT
easat-5690	242	8	graphical	graphical	ADJ
easat-5690	242	9	user	user	NOUN
easat-5690	242	10	interface	interface	NOUN
easat-5690	242	11	)	)	PUNCT
easat-5690	242	12	programming	programming	NOUN
easat-5690	242	13	solutions	solution	NOUN
easat-5690	242	14	for	for	ADP
easat-5690	242	15	python	python	NOUN
easat-5690	242	16	language	language	NOUN
easat-5690	242	17	,	,	PUNCT
easat-5690	242	18	which	which	PRON
easat-5690	242	19	can	can	AUX
easat-5690	242	20	quickly	quickly	ADV
easat-5690	242	21	create	create	VERB
easat-5690	242	22	gui	gui	NOUN
easat-5690	242	23	applications	application	NOUN
easat-5690	242	24	for	for	ADP
easat-5690	242	25	python	python	NOUN
easat-5690	242	26	programs	program	NOUN
easat-5690	242	27	.	.	PUNCT
easat-5690	243	1	design	design	VERB
easat-5690	243	2	a	a	DET
easat-5690	243	3	graphical	graphical	ADJ
easat-5690	243	4	interface	interface	NOUN
easat-5690	243	5	using	use	VERB
easat-5690	243	6	qt	qt	NOUN
easat-5690	243	7	designer	designer	NOUN
easat-5690	243	8	,	,	PUNCT
easat-5690	243	9	and	and	CCONJ
easat-5690	243	10	then	then	ADV
easat-5690	243	11	convert	convert	VERB
easat-5690	243	12	the	the	DET
easat-5690	243	13	designed	design	VERB
easat-5690	243	14	ui	ui	NOUN
easat-5690	243	15	file	file	NOUN
easat-5690	243	16	into	into	ADP
easat-5690	243	17	python	python	NOUN
easat-5690	243	18	code	code	NOUN
easat-5690	243	19	using	use	VERB
easat-5690	243	20	pyside6.the	pyside6.the	DET
easat-5690	243	21	graphical	graphical	ADJ
easat-5690	243	22	interface	interface	NOUN
easat-5690	243	23	contains	contain	VERB
easat-5690	243	24	multiple	multiple	ADJ
easat-5690	243	25	ui	ui	PROPN
easat-5690	243	26	controls	control	NOUN
easat-5690	243	27	,	,	PUNCT
easat-5690	243	28	such	such	ADJ
easat-5690	243	29	as	as	ADP
easat-5690	243	30	labels	label	NOUN
easat-5690	243	31	,	,	PUNCT
easat-5690	243	32	buttons	button	NOUN
easat-5690	243	33	,	,	PUNCT
easat-5690	243	34	text	text	NOUN
easat-5690	243	35	boxes	box	NOUN
easat-5690	243	36	,	,	PUNCT
easat-5690	243	37	multiple	multiple	ADJ
easat-5690	243	38	selection	selection	NOUN
easat-5690	243	39	boxes	box	NOUN
easat-5690	243	40	,	,	PUNCT
easat-5690	243	41	etc	etc	X
easat-5690	243	42	.	.	X
easat-5690	243	43	through	through	ADP
easat-5690	243	44	the	the	DET
easat-5690	243	45	signal	signal	ADJ
easat-5690	243	46	slot	slot	NOUN
easat-5690	243	47	mechanism	mechanism	NOUN
easat-5690	243	48	in	in	ADP
easat-5690	243	49	pyside6	pyside6	PROPN
easat-5690	243	50	,	,	PUNCT
easat-5690	243	51	ui	ui	PROPN
easat-5690	243	52	controls	control	NOUN
easat-5690	243	53	can	can	AUX
easat-5690	243	54	be	be	AUX
easat-5690	243	55	connected	connect	VERB
easat-5690	243	56	to	to	PART
easat-5690	243	57	program	program	VERB
easat-5690	243	58	logic	logic	NOUN
easat-5690	243	59	code	code	NOUN
easat-5690	243	60	.	.	PUNCT
easat-5690	244	1	the	the	DET
easat-5690	244	2	supported	support	VERB
easat-5690	244	3	functions	function	NOUN
easat-5690	244	4	of	of	ADP
easat-5690	244	5	the	the	DET
easat-5690	244	6	system	system	NOUN
easat-5690	244	7	include	include	VERB
easat-5690	244	8	model	model	NOUN
easat-5690	244	9	import	import	NOUN
easat-5690	244	10	and	and	CCONJ
easat-5690	244	11	initialization	initialization	NOUN
easat-5690	244	12	;	;	PUNCT
easat-5690	244	13	confidence	confidence	NOUN
easat-5690	244	14	score	score	NOUN
easat-5690	244	15	and	and	CCONJ
easat-5690	244	16	iou	iou	NOUN
easat-5690	244	17	threshold	threshold	NOUN
easat-5690	244	18	adjustment	adjustment	NOUN
easat-5690	244	19	,	,	PUNCT
easat-5690	244	20	image	image	NOUN
easat-5690	244	21	upload	upload	NOUN
easat-5690	244	22	,	,	PUNCT
easat-5690	244	23	detection	detection	NOUN
easat-5690	244	24	,	,	PUNCT
easat-5690	244	25	visualization	visualization	NOUN
easat-5690	244	26	of	of	ADP
easat-5690	244	27	results	result	NOUN
easat-5690	244	28	,	,	PUNCT
easat-5690	244	29	result	result	NOUN
easat-5690	244	30	export	export	NOUN
easat-5690	244	31	,	,	PUNCT
easat-5690	244	32	and	and	CCONJ
easat-5690	244	33	ending	end	VERB
easat-5690	244	34	detection	detection	NOUN
easat-5690	244	35	;	;	PUNCT
easat-5690	244	36	video	video	NOUN
easat-5690	244	37	upload	upload	NOUN
easat-5690	244	38	,	,	PUNCT
easat-5690	244	39	detection	detection	NOUN
easat-5690	244	40	,	,	PUNCT
easat-5690	244	41	visualization	visualization	NOUN
easat-5690	244	42	of	of	ADP
easat-5690	244	43	results	result	NOUN
easat-5690	244	44	,	,	PUNCT
easat-5690	244	45	result	result	NOUN
easat-5690	244	46	export	export	NOUN
easat-5690	244	47	,	,	PUNCT
easat-5690	244	48	and	and	CCONJ
easat-5690	244	49	ending	end	VERB
easat-5690	244	50	detection	detection	NOUN
easat-5690	244	51	;	;	PUNCT
easat-5690	244	52	camera	camera	NOUN
easat-5690	244	53	feed	feed	NOUN
easat-5690	244	54	upload	upload	NOUN
easat-5690	244	55	,	,	PUNCT
easat-5690	244	56	detection	detection	NOUN
easat-5690	244	57	,	,	PUNCT
easat-5690	244	58	visualization	visualization	NOUN
easat-5690	244	59	of	of	ADP
easat-5690	244	60	results	result	NOUN
easat-5690	244	61	,	,	PUNCT
easat-5690	244	62	and	and	CCONJ
easat-5690	244	63	ending	end	VERB
easat-5690	244	64	detection	detection	NOUN
easat-5690	244	65	;	;	PUNCT
easat-5690	244	66	as	as	ADV
easat-5690	244	67	well	well	ADV
easat-5690	244	68	as	as	ADP
easat-5690	244	69	the	the	DET
easat-5690	244	70	list	list	NOUN
easat-5690	244	71	of	of	ADP
easat-5690	244	72	detected	detect	VERB
easat-5690	244	73	objects	object	NOUN
easat-5690	244	74	,	,	PUNCT
easat-5690	244	75	location	location	NOUN
easat-5690	244	76	information	information	NOUN
easat-5690	244	77	,	,	PUNCT
easat-5690	244	78	and	and	CCONJ
easat-5690	244	79	more	more	ADJ
easat-5690	244	80	.	.	PUNCT
easat-5690	245	1	the	the	DET
easat-5690	245	2	software	software	NOUN
easat-5690	245	3	interface	interface	NOUN
easat-5690	245	4	designed	design	VERB
easat-5690	245	5	below	below	ADV
easat-5690	245	6	is	be	AUX
easat-5690	245	7	simple	simple	ADJ
easat-5690	245	8	and	and	CCONJ
easat-5690	245	9	elegant	elegant	ADJ
easat-5690	245	10	overall	overall	ADJ
easat-5690	245	11	.	.	PUNCT
easat-5690	246	1	users	user	NOUN
easat-5690	246	2	can	can	AUX
easat-5690	246	3	click	click	VERB
easat-5690	246	4	the	the	DET
easat-5690	246	5	select	select	ADJ
easat-5690	246	6	image	image	NOUN
easat-5690	246	7	button	button	NOUN
easat-5690	246	8	to	to	PART
easat-5690	246	9	upload	upload	VERB
easat-5690	246	10	a	a	DET
easat-5690	246	11	single	single	ADJ
easat-5690	246	12	image	image	NOUN
easat-5690	246	13	for	for	ADP
easat-5690	246	14	detection	detection	NOUN
easat-5690	246	15	and	and	CCONJ
easat-5690	246	16	recognition	recognition	NOUN
easat-5690	246	17	.	.	PUNCT
easat-5690	247	1	click	click	VERB
easat-5690	247	2	the	the	DET
easat-5690	247	3	detection	detection	NOUN
easat-5690	247	4	result	result	NOUN
easat-5690	247	5	display	display	NOUN
easat-5690	247	6	button	button	NOUN
easat-5690	247	7	again	again	ADV
easat-5690	247	8	to	to	PART
easat-5690	247	9	display	display	VERB
easat-5690	247	10	the	the	DET
easat-5690	247	11	input	input	NOUN
easat-5690	247	12	image	image	NOUN
easat-5690	247	13	detection	detection	NOUN
easat-5690	247	14	results	result	VERB
easat-5690	247	15	in	in	ADP
easat-5690	247	16	the	the	DET
easat-5690	247	17	lower	low	ADJ
easat-5690	247	18	left	left	ADJ
easat-5690	247	19	corner	corner	NOUN
easat-5690	247	20	of	of	ADP
easat-5690	247	21	the	the	DET
easat-5690	247	22	system	system	NOUN
easat-5690	247	23	,	,	PUNCT
easat-5690	247	24	as	as	SCONJ
easat-5690	247	25	shown	show	VERB
easat-5690	247	26	in	in	ADP
easat-5690	247	27	figure	figure	NOUN
easat-5690	247	28	12	12	NUM
easat-5690	247	29	.	.	PUNCT
easat-5690	247	30	1797	1797	PROPN
easat-5690	247	31	edelweiss	edelweiss	PROPN
easat-5690	247	32	applied	apply	VERB
easat-5690	247	33	science	science	NOUN
easat-5690	247	34	and	and	CCONJ
easat-5690	247	35	technology	technology	NOUN
easat-5690	247	36	issn	issn	PROPN
easat-5690	247	37	:	:	PUNCT
easat-5690	247	38	2576	2576	NUM
easat-5690	247	39	-	-	SYM
easat-5690	247	40	8484	8484	NUM
easat-5690	247	41	vol	vol	NOUN
easat-5690	247	42	.	.	PROPN
easat-5690	248	1	9	9	NUM
easat-5690	248	2	,	,	PUNCT
easat-5690	248	3	no	no	INTJ
easat-5690	248	4	.	.	NOUN
easat-5690	249	1	3	3	NUM
easat-5690	249	2	:	:	SYM
easat-5690	249	3	1777	1777	NUM
easat-5690	249	4	-	-	SYM
easat-5690	249	5	1794	1794	NUM
easat-5690	249	6	,	,	PUNCT
easat-5690	249	7	2025	2025	NUM
easat-5690	249	8	doi	doi	NOUN
easat-5690	249	9	:	:	PUNCT
easat-5690	249	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	249	11	©	©	PROPN
easat-5690	249	12	2025	2025	NUM
easat-5690	249	13	by	by	ADP
easat-5690	249	14	the	the	DET
easat-5690	249	15	authors	author	NOUN
easat-5690	249	16	;	;	PUNCT
easat-5690	249	17	licensee	licensee	PROPN
easat-5690	249	18	learning	learn	VERB
easat-5690	249	19	gate	gate	PROPN
easat-5690	249	20	figure	figure	NOUN
easat-5690	249	21	12	12	NUM
easat-5690	249	22	.	.	PUNCT
easat-5690	250	1	tea	tea	NOUN
easat-5690	250	2	disease	disease	NOUN
easat-5690	250	3	and	and	CCONJ
easat-5690	250	4	pest	pest	VERB
easat-5690	250	5	detection	detection	NOUN
easat-5690	250	6	display	display	NOUN
easat-5690	250	7	system	system	NOUN
easat-5690	250	8	.	.	PUNCT
easat-5690	251	1	5	5	X
easat-5690	251	2	.	.	X
easat-5690	251	3	conclusions	conclusion	NOUN
easat-5690	251	4	there	there	PRON
easat-5690	251	5	are	be	VERB
easat-5690	251	6	many	many	ADJ
easat-5690	251	7	different	different	ADJ
easat-5690	251	8	types	type	NOUN
easat-5690	251	9	of	of	ADP
easat-5690	251	10	tea	tea	NOUN
easat-5690	251	11	diseases	disease	NOUN
easat-5690	251	12	and	and	CCONJ
easat-5690	251	13	pests	pest	NOUN
easat-5690	251	14	,	,	PUNCT
easat-5690	251	15	and	and	CCONJ
easat-5690	251	16	at	at	ADP
easat-5690	251	17	the	the	DET
easat-5690	251	18	moment	moment	NOUN
easat-5690	251	19	,	,	PUNCT
easat-5690	251	20	identifying	identify	VERB
easat-5690	251	21	these	these	DET
easat-5690	251	22	issues	issue	NOUN
easat-5690	251	23	primarily	primarily	ADV
easat-5690	251	24	depends	depend	VERB
easat-5690	251	25	on	on	ADP
easat-5690	251	26	the	the	DET
easat-5690	251	27	knowledge	knowledge	NOUN
easat-5690	251	28	of	of	ADP
easat-5690	251	29	experts	expert	NOUN
easat-5690	251	30	.	.	PUNCT
easat-5690	252	1	as	as	ADP
easat-5690	252	2	a	a	DET
easat-5690	252	3	result	result	NOUN
easat-5690	252	4	,	,	PUNCT
easat-5690	252	5	this	this	DET
easat-5690	252	6	paper	paper	NOUN
easat-5690	252	7	suggests	suggest	VERB
easat-5690	252	8	an	an	DET
easat-5690	252	9	ensemble	ensemble	ADJ
easat-5690	252	10	learning	learning	NOUN
easat-5690	252	11	-	-	PUNCT
easat-5690	252	12	based	base	VERB
easat-5690	252	13	approach	approach	NOUN
easat-5690	252	14	for	for	ADP
easat-5690	252	15	identifying	identify	VERB
easat-5690	252	16	tea	tea	NOUN
easat-5690	252	17	plant	plant	NOUN
easat-5690	252	18	diseases	disease	NOUN
easat-5690	252	19	and	and	CCONJ
easat-5690	252	20	pests	pest	NOUN
easat-5690	252	21	.	.	PUNCT
easat-5690	253	1	to	to	PART
easat-5690	253	2	achieve	achieve	VERB
easat-5690	253	3	effective	effective	ADJ
easat-5690	253	4	disease	disease	NOUN
easat-5690	253	5	and	and	CCONJ
easat-5690	253	6	pest	pest	NOUN
easat-5690	253	7	recognition	recognition	NOUN
easat-5690	253	8	,	,	PUNCT
easat-5690	253	9	we	we	PRON
easat-5690	253	10	conducted	conduct	VERB
easat-5690	253	11	the	the	DET
easat-5690	253	12	following	follow	VERB
easat-5690	253	13	work	work	NOUN
easat-5690	253	14	.	.	PUNCT
easat-5690	254	1	first	first	ADV
easat-5690	254	2	,	,	PUNCT
easat-5690	254	3	in	in	ADP
easat-5690	254	4	the	the	DET
easat-5690	254	5	object	object	NOUN
easat-5690	254	6	detection	detection	NOUN
easat-5690	254	7	domain	domain	NOUN
easat-5690	254	8	,	,	PUNCT
easat-5690	254	9	the	the	DET
easat-5690	254	10	popular	popular	ADJ
easat-5690	254	11	yolov5	yolov5	NOUN
easat-5690	254	12	model	model	NOUN
easat-5690	254	13	is	be	AUX
easat-5690	254	14	selected	select	VERB
easat-5690	254	15	.	.	PUNCT
easat-5690	255	1	second	second	ADJ
easat-5690	255	2	,	,	PUNCT
easat-5690	255	3	we	we	PRON
easat-5690	255	4	enhanced	enhance	VERB
easat-5690	255	5	the	the	DET
easat-5690	255	6	yolov5	yolov5	NOUN
easat-5690	255	7	model	model	NOUN
easat-5690	255	8	for	for	ADP
easat-5690	255	9	disease	disease	NOUN
easat-5690	255	10	and	and	CCONJ
easat-5690	255	11	pest	pest	VERB
easat-5690	255	12	identification	identification	NOUN
easat-5690	255	13	in	in	ADP
easat-5690	255	14	three	three	NUM
easat-5690	255	15	ways.we	ways.we	NOUN
easat-5690	255	16	upgraded	upgrade	VERB
easat-5690	255	17	csp	csp	PROPN
easat-5690	255	18	to	to	PART
easat-5690	255	19	cspk	cspk	PROPN
easat-5690	255	20	,	,	PUNCT
easat-5690	255	21	which	which	PRON
easat-5690	255	22	better	well	ADV
easat-5690	255	23	enhances	enhance	VERB
easat-5690	255	24	feature	feature	NOUN
easat-5690	255	25	propagation	propagation	NOUN
easat-5690	255	26	,	,	PUNCT
easat-5690	255	27	aiding	aid	VERB
easat-5690	255	28	in	in	ADP
easat-5690	255	29	better	well	ADJ
easat-5690	255	30	information	information	NOUN
easat-5690	255	31	transmission	transmission	NOUN
easat-5690	255	32	between	between	ADP
easat-5690	255	33	different	different	ADJ
easat-5690	255	34	layers	layer	NOUN
easat-5690	255	35	of	of	ADP
easat-5690	255	36	the	the	DET
easat-5690	255	37	network	network	NOUN
easat-5690	255	38	,	,	PUNCT
easat-5690	255	39	thus	thus	ADV
easat-5690	255	40	improving	improve	VERB
easat-5690	255	41	performance	performance	NOUN
easat-5690	255	42	.	.	PUNCT
easat-5690	256	1	to	to	PART
easat-5690	256	2	help	help	VERB
easat-5690	256	3	the	the	DET
easat-5690	256	4	model	model	NOUN
easat-5690	256	5	better	well	ADJ
easat-5690	256	6	focus	focus	NOUN
easat-5690	256	7	on	on	ADP
easat-5690	256	8	objectives	objective	NOUN
easat-5690	256	9	like	like	ADP
easat-5690	256	10	leaf	leaf	NOUN
easat-5690	256	11	blight	blight	NOUN
easat-5690	256	12	,	,	PUNCT
easat-5690	256	13	we	we	PRON
easat-5690	256	14	also	also	ADV
easat-5690	256	15	included	include	VERB
easat-5690	256	16	the	the	DET
easat-5690	256	17	odconv	odconv	NOUN
easat-5690	256	18	and	and	CCONJ
easat-5690	256	19	cbam	cbam	NOUN
easat-5690	256	20	attention	attention	NOUN
easat-5690	256	21	processes	process	NOUN
easat-5690	256	22	.	.	PUNCT
easat-5690	257	1	the	the	DET
easat-5690	257	2	effectiveness	effectiveness	NOUN
easat-5690	257	3	of	of	ADP
easat-5690	257	4	our	our	PRON
easat-5690	257	5	model	model	NOUN
easat-5690	257	6	was	be	AUX
easat-5690	257	7	finally	finally	ADV
easat-5690	257	8	empirically	empirically	ADV
easat-5690	257	9	verified	verify	VERB
easat-5690	257	10	in	in	ADP
easat-5690	257	11	comparison	comparison	NOUN
easat-5690	257	12	to	to	ADP
easat-5690	257	13	the	the	DET
easat-5690	257	14	original	original	ADJ
easat-5690	257	15	yolov5	yolov5	NOUN
easat-5690	257	16	model	model	NOUN
easat-5690	257	17	,	,	PUNCT
easat-5690	257	18	faster	fast	ADJ
easat-5690	257	19	r	r	NOUN
easat-5690	257	20	-	-	PUNCT
easat-5690	257	21	cnn	cnn	PROPN
easat-5690	257	22	,	,	PUNCT
easat-5690	257	23	ssd	ssd	PROPN
easat-5690	257	24	,	,	PUNCT
easat-5690	257	25	yolov4	yolov4	PROPN
easat-5690	257	26	,	,	PUNCT
easat-5690	257	27	and	and	CCONJ
easat-5690	257	28	yolov3	yolov3	PROPN
easat-5690	257	29	models	model	NOUN
easat-5690	257	30	.	.	PUNCT
easat-5690	258	1	furthermore	furthermore	ADV
easat-5690	258	2	,	,	PUNCT
easat-5690	258	3	we	we	PRON
easat-5690	258	4	thoroughly	thoroughly	ADV
easat-5690	258	5	tested	test	VERB
easat-5690	258	6	the	the	DET
easat-5690	258	7	entire	entire	ADJ
easat-5690	258	8	system	system	NOUN
easat-5690	258	9	,	,	PUNCT
easat-5690	258	10	culminating	culminate	VERB
easat-5690	258	11	in	in	ADP
easat-5690	258	12	the	the	DET
easat-5690	258	13	development	development	NOUN
easat-5690	258	14	of	of	ADP
easat-5690	258	15	a	a	DET
easat-5690	258	16	seamless	seamless	ADJ
easat-5690	258	17	,	,	PUNCT
easat-5690	258	18	high	high	ADJ
easat-5690	258	19	-	-	PUNCT
easat-5690	258	20	precision	precision	NOUN
easat-5690	258	21	object	object	NOUN
easat-5690	258	22	identification	identification	NOUN
easat-5690	258	23	system	system	NOUN
easat-5690	258	24	interface	interface	NOUN
easat-5690	258	25	—	—	PUNCT
easat-5690	258	26	which	which	PRON
easat-5690	258	27	serves	serve	VERB
easat-5690	258	28	as	as	ADP
easat-5690	258	29	the	the	DET
easat-5690	258	30	presentation	presentation	NOUN
easat-5690	258	31	interface	interface	NOUN
easat-5690	258	32	for	for	ADP
easat-5690	258	33	this	this	DET
easat-5690	258	34	paper	paper	NOUN
easat-5690	258	35	—	—	PUNCT
easat-5690	258	36	complete	complete	ADJ
easat-5690	258	37	with	with	ADP
easat-5690	258	38	a	a	DET
easat-5690	258	39	user	user	NOUN
easat-5690	258	40	interface	interface	NOUN
easat-5690	258	41	,	,	PUNCT
easat-5690	258	42	test	test	NOUN
easat-5690	258	43	photos	photo	NOUN
easat-5690	258	44	,	,	PUNCT
easat-5690	258	45	and	and	CCONJ
easat-5690	258	46	videos	video	NOUN
easat-5690	258	47	.	.	PUNCT
easat-5690	259	1	in	in	ADP
easat-5690	259	2	future	future	ADJ
easat-5690	259	3	endeavors	endeavor	NOUN
easat-5690	259	4	,	,	PUNCT
easat-5690	259	5	our	our	PRON
easat-5690	259	6	focus	focus	NOUN
easat-5690	259	7	will	will	AUX
easat-5690	259	8	remain	remain	VERB
easat-5690	259	9	on	on	ADP
easat-5690	259	10	refining	refine	VERB
easat-5690	259	11	the	the	DET
easat-5690	259	12	model	model	NOUN
easat-5690	259	13	and	and	CCONJ
easat-5690	259	14	exploring	explore	VERB
easat-5690	259	15	more	more	ADV
easat-5690	259	16	streamlined	streamlined	ADJ
easat-5690	259	17	1798	1798	NUM
easat-5690	259	18	edelweiss	edelweiss	PROPN
easat-5690	259	19	applied	apply	VERB
easat-5690	259	20	science	science	NOUN
easat-5690	259	21	and	and	CCONJ
easat-5690	259	22	technology	technology	NOUN
easat-5690	259	23	issn	issn	PROPN
easat-5690	259	24	:	:	PUNCT
easat-5690	259	25	2576	2576	NUM
easat-5690	259	26	-	-	SYM
easat-5690	259	27	8484	8484	NUM
easat-5690	259	28	vol	vol	NOUN
easat-5690	259	29	.	.	PROPN
easat-5690	260	1	9	9	NUM
easat-5690	260	2	,	,	PUNCT
easat-5690	260	3	no	no	INTJ
easat-5690	260	4	.	.	NOUN
easat-5690	261	1	3	3	NUM
easat-5690	261	2	:	:	SYM
easat-5690	261	3	1777	1777	NUM
easat-5690	261	4	-	-	SYM
easat-5690	261	5	1794	1794	NUM
easat-5690	261	6	,	,	PUNCT
easat-5690	261	7	2025	2025	NUM
easat-5690	261	8	doi	doi	NOUN
easat-5690	261	9	:	:	PUNCT
easat-5690	261	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	261	11	©	©	PROPN
easat-5690	261	12	2025	2025	NUM
easat-5690	261	13	by	by	ADP
easat-5690	261	14	the	the	DET
easat-5690	261	15	authors	author	NOUN
easat-5690	261	16	;	;	PUNCT
easat-5690	261	17	licensee	licensee	PROPN
easat-5690	261	18	learning	learn	VERB
easat-5690	261	19	gate	gate	PROPN
easat-5690	261	20	approaches	approach	NOUN
easat-5690	261	21	with	with	ADP
easat-5690	261	22	diminished	diminished	ADJ
easat-5690	261	23	parameter	parameter	NOUN
easat-5690	261	24	density	density	NOUN
easat-5690	261	25	.	.	PUNCT
easat-5690	262	1	we	we	PRON
easat-5690	262	2	also	also	ADV
easat-5690	262	3	intend	intend	VERB
easat-5690	262	4	to	to	PART
easat-5690	262	5	delve	delve	VERB
easat-5690	262	6	into	into	ADP
easat-5690	262	7	methods	method	NOUN
easat-5690	262	8	for	for	ADP
easat-5690	262	9	deploying	deploying	NOUN
easat-5690	262	10	models	model	NOUN
easat-5690	262	11	designed	design	VERB
easat-5690	262	12	for	for	ADP
easat-5690	262	13	tea	tea	NOUN
easat-5690	262	14	plant	plant	NOUN
easat-5690	262	15	pest	pest	NOUN
easat-5690	262	16	and	and	CCONJ
easat-5690	262	17	disease	disease	NOUN
easat-5690	262	18	detection	detection	NOUN
easat-5690	262	19	.	.	PUNCT
easat-5690	263	1	to	to	PART
easat-5690	263	2	enhance	enhance	VERB
easat-5690	263	3	the	the	DET
easat-5690	263	4	comprehensiveness	comprehensiveness	NOUN
easat-5690	263	5	of	of	ADP
easat-5690	263	6	our	our	PRON
easat-5690	263	7	research	research	NOUN
easat-5690	263	8	,	,	PUNCT
easat-5690	263	9	it	it	PRON
easat-5690	263	10	is	be	AUX
easat-5690	263	11	imperative	imperative	ADJ
easat-5690	263	12	to	to	PART
easat-5690	263	13	increase	increase	VERB
easat-5690	263	14	the	the	DET
easat-5690	263	15	image	image	NOUN
easat-5690	263	16	capture	capture	NOUN
easat-5690	263	17	at	at	ADP
easat-5690	263	18	varying	vary	VERB
easat-5690	263	19	heights	height	NOUN
easat-5690	263	20	,	,	PUNCT
easat-5690	263	21	encompassing	encompass	VERB
easat-5690	263	22	the	the	DET
easat-5690	263	23	entire	entire	ADJ
easat-5690	263	24	tea	tea	NOUN
easat-5690	263	25	disease	disease	NOUN
easat-5690	263	26	statistics	statistic	NOUN
easat-5690	263	27	process	process	NOUN
easat-5690	263	28	,	,	PUNCT
easat-5690	263	29	and	and	CCONJ
easat-5690	263	30	augment	augment	VERB
easat-5690	263	31	the	the	DET
easat-5690	263	32	training	training	NOUN
easat-5690	263	33	samples	sample	NOUN
easat-5690	263	34	to	to	PART
easat-5690	263	35	boost	boost	VERB
easat-5690	263	36	recognition	recognition	NOUN
easat-5690	263	37	accuracy	accuracy	NOUN
easat-5690	263	38	.	.	PUNCT
easat-5690	264	1	furthermore	furthermore	ADV
easat-5690	264	2	,	,	PUNCT
easat-5690	264	3	we	we	PRON
easat-5690	264	4	envision	envision	VERB
easat-5690	264	5	integrating	integrate	VERB
easat-5690	264	6	detection	detection	NOUN
easat-5690	264	7	methods	method	NOUN
easat-5690	264	8	with	with	ADP
easat-5690	264	9	drones	drone	NOUN
easat-5690	264	10	,	,	PUNCT
easat-5690	264	11	minimizing	minimize	VERB
easat-5690	264	12	model	model	NOUN
easat-5690	264	13	parameter	parameter	NOUN
easat-5690	264	14	counts	count	NOUN
easat-5690	264	15	,	,	PUNCT
easat-5690	264	16	and	and	CCONJ
easat-5690	264	17	achieving	achieve	VERB
easat-5690	264	18	real	real	ADJ
easat-5690	264	19	-	-	PUNCT
easat-5690	264	20	time	time	NOUN
easat-5690	264	21	detection	detection	NOUN
easat-5690	264	22	.	.	PUNCT
easat-5690	265	1	however	however	ADV
easat-5690	265	2	,	,	PUNCT
easat-5690	265	3	it	it	PRON
easat-5690	265	4	's	be	AUX
easat-5690	265	5	worth	worth	ADJ
easat-5690	265	6	noting	note	VERB
easat-5690	265	7	that	that	SCONJ
easat-5690	265	8	swift	swift	ADJ
easat-5690	265	9	detection	detection	NOUN
easat-5690	265	10	necessitates	necessitate	VERB
easat-5690	265	11	specific	specific	ADJ
easat-5690	265	12	hardware	hardware	NOUN
easat-5690	265	13	configurations	configuration	NOUN
easat-5690	265	14	.	.	PUNCT
easat-5690	266	1	we	we	PRON
easat-5690	266	2	will	will	AUX
easat-5690	266	3	strive	strive	VERB
easat-5690	266	4	to	to	PART
easat-5690	266	5	continually	continually	ADV
easat-5690	266	6	enhance	enhance	VERB
easat-5690	266	7	the	the	DET
easat-5690	266	8	model	model	NOUN
easat-5690	266	9	while	while	SCONJ
easat-5690	266	10	extending	extend	VERB
easat-5690	266	11	our	our	PRON
easat-5690	266	12	research	research	NOUN
easat-5690	266	13	on	on	ADP
easat-5690	266	14	diverse	diverse	ADJ
easat-5690	266	15	tea	tea	NOUN
easat-5690	266	16	diseases	disease	NOUN
easat-5690	266	17	and	and	CCONJ
easat-5690	266	18	broadening	broaden	VERB
easat-5690	266	19	its	its	PRON
easat-5690	266	20	applicability	applicability	NOUN
easat-5690	266	21	.	.	PUNCT
easat-5690	267	1	we	we	PRON
easat-5690	267	2	firmly	firmly	ADV
easat-5690	267	3	believe	believe	VERB
easat-5690	267	4	that	that	SCONJ
easat-5690	267	5	this	this	DET
easat-5690	267	6	approach	approach	NOUN
easat-5690	267	7	can	can	AUX
easat-5690	267	8	make	make	VERB
easat-5690	267	9	meaningful	meaningful	ADJ
easat-5690	267	10	contributions	contribution	NOUN
easat-5690	267	11	to	to	ADP
easat-5690	267	12	sustainable	sustainable	ADJ
easat-5690	267	13	,	,	PUNCT
easat-5690	267	14	green	green	ADJ
easat-5690	267	15	,	,	PUNCT
easat-5690	267	16	and	and	CCONJ
easat-5690	267	17	smart	smart	ADJ
easat-5690	267	18	agriculture	agriculture	NOUN
easat-5690	267	19	.	.	PUNCT
easat-5690	268	1	transparency	transparency	NOUN
easat-5690	268	2	:	:	PUNCT
easat-5690	268	3	the	the	DET
easat-5690	268	4	authors	author	NOUN
easat-5690	268	5	confirm	confirm	VERB
easat-5690	268	6	that	that	SCONJ
easat-5690	268	7	the	the	DET
easat-5690	268	8	manuscript	manuscript	NOUN
easat-5690	268	9	is	be	AUX
easat-5690	268	10	an	an	DET
easat-5690	268	11	honest	honest	ADJ
easat-5690	268	12	,	,	PUNCT
easat-5690	268	13	accurate	accurate	ADJ
easat-5690	268	14	,	,	PUNCT
easat-5690	268	15	and	and	CCONJ
easat-5690	268	16	transparent	transparent	ADJ
easat-5690	268	17	account	account	NOUN
easat-5690	268	18	of	of	ADP
easat-5690	268	19	the	the	DET
easat-5690	268	20	study	study	NOUN
easat-5690	268	21	;	;	PUNCT
easat-5690	268	22	that	that	SCONJ
easat-5690	268	23	no	no	DET
easat-5690	268	24	vital	vital	ADJ
easat-5690	268	25	features	feature	NOUN
easat-5690	268	26	of	of	ADP
easat-5690	268	27	the	the	DET
easat-5690	268	28	study	study	NOUN
easat-5690	268	29	have	have	AUX
easat-5690	268	30	been	be	AUX
easat-5690	268	31	omitted	omit	VERB
easat-5690	268	32	;	;	PUNCT
easat-5690	268	33	and	and	CCONJ
easat-5690	268	34	that	that	SCONJ
easat-5690	268	35	any	any	DET
easat-5690	268	36	discrepancies	discrepancy	NOUN
easat-5690	268	37	from	from	ADP
easat-5690	268	38	the	the	DET
easat-5690	268	39	study	study	NOUN
easat-5690	268	40	as	as	SCONJ
easat-5690	268	41	planned	plan	VERB
easat-5690	268	42	have	have	AUX
easat-5690	268	43	been	be	AUX
easat-5690	268	44	explained	explain	VERB
easat-5690	268	45	.	.	PUNCT
easat-5690	269	1	this	this	DET
easat-5690	269	2	study	study	NOUN
easat-5690	269	3	followed	follow	VERB
easat-5690	269	4	all	all	DET
easat-5690	269	5	ethical	ethical	ADJ
easat-5690	269	6	practices	practice	NOUN
easat-5690	269	7	during	during	ADP
easat-5690	269	8	writing	writing	NOUN
easat-5690	269	9	.	.	PUNCT
easat-5690	270	1	acknowledgments	acknowledgment	NOUN
easat-5690	270	2	:	:	PUNCT
easat-5690	270	3	for	for	ADP
easat-5690	270	4	the	the	DET
easat-5690	270	5	facility	facility	NOUN
easat-5690	270	6	support	support	NOUN
easat-5690	270	7	,	,	PUNCT
easat-5690	270	8	the	the	DET
easat-5690	270	9	authors	author	NOUN
easat-5690	270	10	would	would	AUX
easat-5690	270	11	like	like	VERB
easat-5690	270	12	to	to	PART
easat-5690	270	13	thank	thank	VERB
easat-5690	270	14	the	the	DET
easat-5690	270	15	college	college	NOUN
easat-5690	270	16	of	of	ADP
easat-5690	270	17	electrical	electrical	ADJ
easat-5690	270	18	engineering	engineering	NOUN
easat-5690	270	19	,	,	PUNCT
easat-5690	270	20	hunan	hunan	PROPN
easat-5690	270	21	mechanical	mechanical	PROPN
easat-5690	270	22	&	&	CCONJ
easat-5690	270	23	electrical	electrical	ADJ
easat-5690	270	24	polytechnic	polytechnic	ADJ
easat-5690	270	25	,	,	PUNCT
easat-5690	270	26	and	and	CCONJ
easat-5690	270	27	faculty	faculty	NOUN
easat-5690	270	28	of	of	ADP
easat-5690	270	29	engineering	engineering	PROPN
easat-5690	270	30	,	,	PUNCT
easat-5690	270	31	mahasarakham	mahasarakham	PROPN
easat-5690	270	32	university	university	PROPN
easat-5690	270	33	.	.	PUNCT
easat-5690	271	1	copyright	copyright	NOUN
easat-5690	271	2	:	:	PUNCT
easat-5690	271	3	©	©	PROPN
easat-5690	271	4	2025	2025	NUM
easat-5690	271	5	by	by	ADP
easat-5690	271	6	the	the	DET
easat-5690	271	7	authors	author	NOUN
easat-5690	271	8	.	.	PUNCT
easat-5690	272	1	this	this	DET
easat-5690	272	2	open	open	ADJ
easat-5690	272	3	-	-	PUNCT
easat-5690	272	4	access	access	NOUN
easat-5690	272	5	article	article	NOUN
easat-5690	272	6	is	be	AUX
easat-5690	272	7	distributed	distribute	VERB
easat-5690	272	8	under	under	ADP
easat-5690	272	9	the	the	DET
easat-5690	272	10	terms	term	NOUN
easat-5690	272	11	and	and	CCONJ
easat-5690	272	12	conditions	condition	NOUN
easat-5690	272	13	of	of	ADP
easat-5690	272	14	the	the	DET
easat-5690	272	15	creative	creative	ADJ
easat-5690	272	16	commons	common	NOUN
easat-5690	272	17	attribution	attribution	NOUN
easat-5690	272	18	(	(	PUNCT
easat-5690	272	19	cc	cc	NOUN
easat-5690	272	20	by	by	ADP
easat-5690	272	21	)	)	PUNCT
easat-5690	272	22	license	license	NOUN
easat-5690	272	23	(	(	PUNCT
easat-5690	272	24	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-5690	272	25	)	)	PUNCT
easat-5690	272	26	.	.	PUNCT
easat-5690	273	1	references	reference	NOUN
easat-5690	273	2	[	[	X
easat-5690	273	3	1	1	X
easat-5690	273	4	]	]	PUNCT
easat-5690	273	5	s.	s.	PROPN
easat-5690	273	6	xiaoxiao	xiaoxiao	PROPN
easat-5690	273	7	,	,	PUNCT
easat-5690	273	8	m.	m.	NOUN
easat-5690	273	9	shaomin	shaomin	PROPN
easat-5690	273	10	,	,	PUNCT
easat-5690	273	11	x.	x.	NOUN
easat-5690	273	12	yongyu	yongyu	PROPN
easat-5690	273	13	,	,	PUNCT
easat-5690	273	14	c.	c.	PROPN
easat-5690	273	15	zhihao	zhihao	PROPN
easat-5690	273	16	,	,	PUNCT
easat-5690	273	17	and	and	CCONJ
easat-5690	273	18	s.	s.	PROPN
easat-5690	273	19	tingting	tingting	PROPN
easat-5690	273	20	,	,	PUNCT
easat-5690	273	21	"	"	PUNCT
easat-5690	273	22	image	image	NOUN
easat-5690	273	23	recognition	recognition	NOUN
easat-5690	273	24	of	of	ADP
easat-5690	273	25	tea	tea	NOUN
easat-5690	273	26	leaf	leaf	NOUN
easat-5690	273	27	diseases	disease	NOUN
easat-5690	273	28	based	base	VERB
easat-5690	273	29	on	on	ADP
easat-5690	273	30	convolutional	convolutional	ADJ
easat-5690	273	31	neural	neural	ADJ
easat-5690	273	32	network	network	NOUN
easat-5690	273	33	,	,	PUNCT
easat-5690	273	34	"	"	PUNCT
easat-5690	273	35	in	in	ADP
easat-5690	273	36	2018	2018	NUM
easat-5690	273	37	international	international	ADJ
easat-5690	273	38	conference	conference	NOUN
easat-5690	273	39	on	on	ADP
easat-5690	273	40	security	security	NOUN
easat-5690	273	41	,	,	PUNCT
easat-5690	273	42	pattern	pattern	NOUN
easat-5690	273	43	analysis	analysis	NOUN
easat-5690	273	44	,	,	PUNCT
easat-5690	273	45	and	and	CCONJ
easat-5690	273	46	cybernetics	cybernetic	NOUN
easat-5690	273	47	(	(	PUNCT
easat-5690	273	48	spac	spac	PROPN
easat-5690	273	49	)	)	PUNCT
easat-5690	273	50	,	,	PUNCT
easat-5690	273	51	2018	2018	NUM
easat-5690	273	52	:	:	PUNCT
easat-5690	273	53	ieee	ieee	NOUN
easat-5690	273	54	,	,	PUNCT
easat-5690	273	55	pp	pp	ADJ
easat-5690	273	56	.	.	PUNCT
easat-5690	274	1	304	304	NUM
easat-5690	274	2	-	-	SYM
easat-5690	274	3	309	309	NUM
easat-5690	274	4	,	,	PUNCT
easat-5690	274	5	doi	doi	NOUN
easat-5690	274	6	:	:	PUNCT
easat-5690	274	7	https://doi.org/10.1109/spac46244.2018.8965555	https://doi.org/10.1109/spac46244.2018.8965555	NOUN
easat-5690	274	8	.	.	PUNCT
easat-5690	275	1	[	[	X
easat-5690	275	2	2	2	X
easat-5690	275	3	]	]	PUNCT
easat-5690	275	4	h.	h.	NOUN
easat-5690	275	5	ulutas	ulutas	PROPN
easat-5690	275	6	and	and	CCONJ
easat-5690	275	7	v.	v.	ADP
easat-5690	275	8	aslantas	aslanta	NOUN
easat-5690	275	9	,	,	PUNCT
easat-5690	275	10	"	"	PUNCT
easat-5690	275	11	a	a	DET
easat-5690	275	12	fast	fast	ADJ
easat-5690	275	13	and	and	CCONJ
easat-5690	275	14	accurate	accurate	ADJ
easat-5690	275	15	method	method	NOUN
easat-5690	275	16	for	for	ADP
easat-5690	275	17	classifying	classify	VERB
easat-5690	275	18	tomato	tomato	NOUN
easat-5690	275	19	plant	plant	NOUN
easat-5690	275	20	health	health	NOUN
easat-5690	275	21	status	status	NOUN
easat-5690	275	22	using	use	VERB
easat-5690	275	23	machine	machine	NOUN
easat-5690	275	24	learning	learning	NOUN
easat-5690	275	25	and	and	CCONJ
easat-5690	275	26	image	image	NOUN
easat-5690	275	27	processing	processing	NOUN
easat-5690	275	28	,	,	PUNCT
easat-5690	275	29	"	"	PUNCT
easat-5690	275	30	elektronika	elektronika	PROPN
easat-5690	275	31	ir	ir	PROPN
easat-5690	275	32	elektrotechnika	elektrotechnika	NOUN
easat-5690	275	33	,	,	PUNCT
easat-5690	275	34	vol	vol	NOUN
easat-5690	275	35	.	.	PROPN
easat-5690	275	36	29	29	NUM
easat-5690	275	37	,	,	PUNCT
easat-5690	275	38	no	no	INTJ
easat-5690	275	39	.	.	NOUN
easat-5690	275	40	2	2	NUM
easat-5690	275	41	,	,	PUNCT
easat-5690	275	42	pp	pp	ADJ
easat-5690	275	43	.	.	PUNCT
easat-5690	276	1	54	54	NUM
easat-5690	276	2	-	-	SYM
easat-5690	276	3	68	68	NUM
easat-5690	276	4	,	,	PUNCT
easat-5690	276	5	2023	2023	NUM
easat-5690	276	6	.	.	PUNCT
easat-5690	277	1	https://doi.org/10.5755/j02.eie.33866	https://doi.org/10.5755/j02.eie.33866	PROPN
easat-5690	278	1	[	[	X
easat-5690	278	2	3	3	NUM
easat-5690	278	3	]	]	PUNCT
easat-5690	278	4	m.	m.	NOUN
easat-5690	278	5	s.	s.	PROPN
easat-5690	278	6	hossain	hossain	PROPN
easat-5690	278	7	,	,	PUNCT
easat-5690	278	8	r.	r.	PROPN
easat-5690	278	9	m.	m.	PROPN
easat-5690	278	10	mou	mou	PROPN
easat-5690	278	11	,	,	PUNCT
easat-5690	278	12	m.	m.	NOUN
easat-5690	278	13	m.	m.	PROPN
easat-5690	278	14	hasan	hasan	PROPN
easat-5690	278	15	,	,	PUNCT
easat-5690	278	16	s.	s.	PROPN
easat-5690	278	17	chakraborty	chakraborty	PROPN
easat-5690	278	18	,	,	PUNCT
easat-5690	278	19	and	and	CCONJ
easat-5690	278	20	m.	m.	NOUN
easat-5690	278	21	a.	a.	PROPN
easat-5690	278	22	razzak	razzak	PROPN
easat-5690	278	23	,	,	PUNCT
easat-5690	278	24	"	"	PUNCT
easat-5690	278	25	recognition	recognition	NOUN
easat-5690	278	26	and	and	CCONJ
easat-5690	278	27	detection	detection	NOUN
easat-5690	278	28	of	of	ADP
easat-5690	278	29	tea	tea	NOUN
easat-5690	278	30	leaf	leaf	NOUN
easat-5690	278	31	's	's	PART
easat-5690	278	32	diseases	disease	NOUN
easat-5690	278	33	using	use	VERB
easat-5690	278	34	support	support	NOUN
easat-5690	278	35	vector	vector	NOUN
easat-5690	278	36	machine	machine	NOUN
easat-5690	278	37	,	,	PUNCT
easat-5690	278	38	"	"	PUNCT
easat-5690	278	39	in	in	ADP
easat-5690	278	40	ieee	ieee	PROPN
easat-5690	278	41	international	international	ADJ
easat-5690	278	42	colloquium	colloquium	NOUN
easat-5690	278	43	on	on	ADP
easat-5690	278	44	signal	signal	NOUN
easat-5690	278	45	processing	processing	NOUN
easat-5690	278	46	&	&	CCONJ
easat-5690	278	47	its	its	PRON
easat-5690	278	48	applications	application	NOUN
easat-5690	278	49	,	,	PUNCT
easat-5690	278	50	2018	2018	NUM
easat-5690	278	51	,	,	PUNCT
easat-5690	278	52	doi	doi	NOUN
easat-5690	278	53	:	:	PUNCT
easat-5690	278	54	https://doi.org/10.1109/cspa.2018.8368703	https://doi.org/10.1109/cspa.2018.8368703	PROPN
easat-5690	278	55	.	.	PUNCT
easat-5690	279	1	[	[	X
easat-5690	279	2	4	4	X
easat-5690	279	3	]	]	PUNCT
easat-5690	279	4	w.	w.	PROPN
easat-5690	279	5	bao	bao	PROPN
easat-5690	279	6	,	,	PUNCT
easat-5690	279	7	t.	t.	PROPN
easat-5690	279	8	fan	fan	PROPN
easat-5690	279	9	,	,	PUNCT
easat-5690	279	10	g.	g.	PROPN
easat-5690	279	11	hu	hu	PROPN
easat-5690	279	12	,	,	PUNCT
easat-5690	279	13	d.	d.	PROPN
easat-5690	279	14	liang	liang	PROPN
easat-5690	279	15	,	,	PUNCT
easat-5690	279	16	and	and	CCONJ
easat-5690	279	17	h.	h.	PROPN
easat-5690	279	18	li	li	PROPN
easat-5690	279	19	,	,	PUNCT
easat-5690	279	20	"	"	PUNCT
easat-5690	279	21	detection	detection	NOUN
easat-5690	279	22	and	and	CCONJ
easat-5690	279	23	identification	identification	NOUN
easat-5690	279	24	of	of	ADP
easat-5690	279	25	tea	tea	NOUN
easat-5690	279	26	leaf	leaf	NOUN
easat-5690	279	27	diseases	disease	NOUN
easat-5690	279	28	based	base	VERB
easat-5690	279	29	on	on	ADP
easat-5690	279	30	axretinanet	axretinanet	NOUN
easat-5690	279	31	,	,	PUNCT
easat-5690	279	32	"	"	PUNCT
easat-5690	279	33	scientific	scientific	ADJ
easat-5690	279	34	reports	report	NOUN
easat-5690	279	35	,	,	PUNCT
easat-5690	279	36	vol	vol	NOUN
easat-5690	279	37	.	.	PROPN
easat-5690	279	38	12	12	NUM
easat-5690	279	39	,	,	PUNCT
easat-5690	279	40	no	no	INTJ
easat-5690	279	41	.	.	NOUN
easat-5690	279	42	1	1	NUM
easat-5690	279	43	,	,	PUNCT
easat-5690	279	44	p.	p.	NOUN
easat-5690	279	45	2183	2183	NUM
easat-5690	279	46	,	,	PUNCT
easat-5690	279	47	2022	2022	NUM
easat-5690	279	48	.	.	PUNCT
easat-5690	280	1	https://doi.org/10.1038/s41598-022-06181-z	https://doi.org/10.1038/s41598-022-06181-z	PROPN
easat-5690	281	1	[	[	X
easat-5690	281	2	5	5	NUM
easat-5690	281	3	]	]	PUNCT
easat-5690	281	4	j.	j.	PROPN
easat-5690	281	5	chen	chen	PROPN
easat-5690	281	6	,	,	PUNCT
easat-5690	281	7	q.	q.	PROPN
easat-5690	281	8	liu	liu	PROPN
easat-5690	281	9	,	,	PUNCT
easat-5690	281	10	and	and	CCONJ
easat-5690	281	11	l.	l.	PROPN
easat-5690	281	12	gao	gao	PROPN
easat-5690	281	13	,	,	PUNCT
easat-5690	281	14	"	"	PUNCT
easat-5690	281	15	visual	visual	ADJ
easat-5690	281	16	tea	tea	NOUN
easat-5690	281	17	leaf	leaf	NOUN
easat-5690	281	18	disease	disease	NOUN
easat-5690	281	19	recognition	recognition	NOUN
easat-5690	281	20	using	use	VERB
easat-5690	281	21	a	a	DET
easat-5690	281	22	convolutional	convolutional	ADJ
easat-5690	281	23	neural	neural	ADJ
easat-5690	281	24	network	network	NOUN
easat-5690	281	25	model	model	NOUN
easat-5690	281	26	,	,	PUNCT
easat-5690	281	27	"	"	PUNCT
easat-5690	281	28	symmetry	symmetry	NOUN
easat-5690	281	29	,	,	PUNCT
easat-5690	281	30	vol	vol	NOUN
easat-5690	281	31	.	.	PROPN
easat-5690	281	32	11	11	NUM
easat-5690	281	33	,	,	PUNCT
easat-5690	281	34	no	no	INTJ
easat-5690	281	35	.	.	NOUN
easat-5690	281	36	3	3	NUM
easat-5690	281	37	,	,	PUNCT
easat-5690	281	38	p.	p.	NOUN
easat-5690	281	39	343	343	NUM
easat-5690	281	40	,	,	PUNCT
easat-5690	281	41	2019	2019	NUM
easat-5690	281	42	.	.	PUNCT
easat-5690	282	1	https://doi.org/10.3390/sym11030343	https://doi.org/10.3390/sym11030343	NOUN
easat-5690	283	1	[	[	X
easat-5690	283	2	6	6	NUM
easat-5690	283	3	]	]	PUNCT
easat-5690	283	4	r.	r.	PROPN
easat-5690	283	5	latha	latha	PROPN
easat-5690	283	6	et	et	PROPN
easat-5690	283	7	al	al	PROPN
easat-5690	283	8	.	.	PROPN
easat-5690	283	9	,	,	PUNCT
easat-5690	283	10	"	"	PUNCT
easat-5690	283	11	automatic	automatic	ADJ
easat-5690	283	12	detection	detection	NOUN
easat-5690	283	13	of	of	ADP
easat-5690	283	14	tea	tea	NOUN
easat-5690	283	15	leaf	leaf	NOUN
easat-5690	283	16	diseases	disease	NOUN
easat-5690	283	17	using	use	VERB
easat-5690	283	18	deep	deep	ADJ
easat-5690	283	19	convolution	convolution	NOUN
easat-5690	283	20	neural	neural	ADJ
easat-5690	283	21	network	network	NOUN
easat-5690	283	22	,	,	PUNCT
easat-5690	283	23	"	"	PUNCT
easat-5690	283	24	in	in	ADP
easat-5690	283	25	2021	2021	NUM
easat-5690	283	26	international	international	ADJ
easat-5690	283	27	conference	conference	NOUN
easat-5690	283	28	on	on	ADP
easat-5690	283	29	computer	computer	NOUN
easat-5690	283	30	communication	communication	NOUN
easat-5690	283	31	and	and	CCONJ
easat-5690	283	32	informatics	informatic	NOUN
easat-5690	283	33	(	(	PUNCT
easat-5690	283	34	iccci	iccci	NOUN
easat-5690	283	35	)	)	PUNCT
easat-5690	283	36	,	,	PUNCT
easat-5690	283	37	2021	2021	NUM
easat-5690	283	38	:	:	PUNCT
easat-5690	283	39	ieee	ieee	NOUN
easat-5690	283	40	,	,	PUNCT
easat-5690	283	41	pp	pp	ADJ
easat-5690	283	42	.	.	PUNCT
easat-5690	283	43	1	1	NUM
easat-5690	283	44	-	-	SYM
easat-5690	283	45	6	6	NUM
easat-5690	283	46	,	,	PUNCT
easat-5690	283	47	doi	doi	NOUN
easat-5690	283	48	:	:	PUNCT
easat-5690	283	49	https://doi.org/10.1109/iccci50826.2021.9402225	https://doi.org/10.1109/iccci50826.2021.9402225	X
easat-5690	283	50	.	.	PUNCT
easat-5690	284	1	[	[	X
easat-5690	284	2	7	7	X
easat-5690	284	3	]	]	X
easat-5690	284	4	l.	l.	PROPN
easat-5690	284	5	trihardianingsih	trihardianingsih	PROPN
easat-5690	284	6	,	,	PUNCT
easat-5690	284	7	a.	a.	NOUN
easat-5690	284	8	sunyoto	sunyoto	PROPN
easat-5690	284	9	,	,	PUNCT
easat-5690	284	10	and	and	CCONJ
easat-5690	284	11	t.	t.	PROPN
easat-5690	284	12	hidayat	hidayat	PROPN
easat-5690	284	13	,	,	PUNCT
easat-5690	284	14	"	"	PUNCT
easat-5690	284	15	classification	classification	NOUN
easat-5690	284	16	of	of	ADP
easat-5690	284	17	tea	tea	NOUN
easat-5690	284	18	leaf	leaf	NOUN
easat-5690	284	19	diseases	disease	NOUN
easat-5690	284	20	based	base	VERB
easat-5690	284	21	on	on	ADP
easat-5690	284	22	resnet-50	resnet-50	PROPN
easat-5690	284	23	and	and	CCONJ
easat-5690	284	24	inception	inception	PROPN
easat-5690	284	25	v3	v3	PROPN
easat-5690	284	26	,	,	PUNCT
easat-5690	284	27	"	"	PUNCT
easat-5690	284	28	sinkron	sinkron	NOUN
easat-5690	284	29	:	:	PUNCT
easat-5690	284	30	jurnal	jurnal	ADJ
easat-5690	284	31	dan	dan	PROPN
easat-5690	284	32	penelitian	penelitian	PROPN
easat-5690	284	33	teknik	teknik	PROPN
easat-5690	284	34	informatika	informatika	PROPN
easat-5690	284	35	,	,	PUNCT
easat-5690	284	36	vol	vol	NOUN
easat-5690	284	37	.	.	PROPN
easat-5690	284	38	8	8	NUM
easat-5690	284	39	,	,	PUNCT
easat-5690	284	40	no	no	INTJ
easat-5690	284	41	.	.	NOUN
easat-5690	284	42	3	3	NUM
easat-5690	284	43	,	,	PUNCT
easat-5690	284	44	pp	pp	ADJ
easat-5690	284	45	.	.	PUNCT
easat-5690	285	1	1564	1564	NUM
easat-5690	285	2	-	-	SYM
easat-5690	285	3	1573	1573	NUM
easat-5690	285	4	,	,	PUNCT
easat-5690	285	5	2023	2023	NUM
easat-5690	285	6	.	.	PUNCT
easat-5690	286	1	https://doi.org/10.33395/sinkron.v8i3.12604	https://doi.org/10.33395/sinkron.v8i3.12604	PROPN
easat-5690	287	1	[	[	X
easat-5690	287	2	8	8	NUM
easat-5690	287	3	]	]	PUNCT
easat-5690	287	4	s.	s.	PROPN
easat-5690	287	5	bhowmik	bhowmik	PROPN
easat-5690	287	6	,	,	PUNCT
easat-5690	287	7	a.	a.	PROPN
easat-5690	287	8	k.	k.	PROPN
easat-5690	287	9	talukdar	talukdar	PROPN
easat-5690	287	10	,	,	PUNCT
easat-5690	287	11	and	and	CCONJ
easat-5690	287	12	k.	k.	PROPN
easat-5690	287	13	k.	k.	PROPN
easat-5690	287	14	sarma	sarma	PROPN
easat-5690	287	15	,	,	PUNCT
easat-5690	287	16	"	"	PUNCT
easat-5690	287	17	detection	detection	NOUN
easat-5690	287	18	of	of	ADP
easat-5690	287	19	disease	disease	NOUN
easat-5690	287	20	in	in	ADP
easat-5690	287	21	tea	tea	NOUN
easat-5690	287	22	leaves	leave	NOUN
easat-5690	287	23	using	use	VERB
easat-5690	287	24	convolution	convolution	NOUN
easat-5690	287	25	neural	neural	ADJ
easat-5690	287	26	network	network	NOUN
easat-5690	287	27	,	,	PUNCT
easat-5690	287	28	"	"	PUNCT
easat-5690	287	29	in	in	ADP
easat-5690	287	30	2020	2020	NUM
easat-5690	287	31	advanced	advanced	ADJ
easat-5690	287	32	communication	communication	NOUN
easat-5690	287	33	technologies	technology	NOUN
easat-5690	287	34	and	and	CCONJ
easat-5690	287	35	signal	signal	NOUN
easat-5690	287	36	processing	processing	NOUN
easat-5690	287	37	(	(	PUNCT
easat-5690	287	38	acts	act	NOUN
easat-5690	287	39	)	)	PUNCT
easat-5690	287	40	,	,	PUNCT
easat-5690	287	41	2020	2020	NUM
easat-5690	287	42	:	:	PUNCT
easat-5690	288	1	ieee	ieee	NOUN
easat-5690	288	2	,	,	PUNCT
easat-5690	288	3	pp	pp	ADJ
easat-5690	288	4	.	.	PUNCT
easat-5690	289	1	1	1	NUM
easat-5690	289	2	-	-	SYM
easat-5690	289	3	6	6	NUM
easat-5690	289	4	,	,	PUNCT
easat-5690	289	5	doi	doi	NOUN
easat-5690	289	6	:	:	PUNCT
easat-5690	289	7	https://doi.org/10.1109/acts49415.2020.9350413	https://doi.org/10.1109/acts49415.2020.9350413	ADJ
easat-5690	289	8	.	.	PUNCT
easat-5690	290	1	[	[	X
easat-5690	290	2	9	9	NUM
easat-5690	290	3	]	]	X
easat-5690	290	4	y.	y.	PROPN
easat-5690	290	5	raju	raju	PROPN
easat-5690	290	6	and	and	CCONJ
easat-5690	290	7	m.	m.	PROPN
easat-5690	290	8	narayana	narayana	PROPN
easat-5690	290	9	,	,	PUNCT
easat-5690	290	10	"	"	PUNCT
easat-5690	290	11	satellite	satellite	NOUN
easat-5690	290	12	image	image	NOUN
easat-5690	290	13	segmentation	segmentation	NOUN
easat-5690	290	14	using	use	VERB
easat-5690	290	15	unet++	unet++	PROPN
easat-5690	290	16	with	with	ADP
easat-5690	290	17	vgg19	vgg19	PROPN
easat-5690	290	18	deep	deep	ADJ
easat-5690	290	19	learning	learning	NOUN
easat-5690	290	20	model	model	NOUN
easat-5690	290	21	,	,	PUNCT
easat-5690	290	22	"	"	PUNCT
easat-5690	290	23	edelweiss	edelweiss	PROPN
easat-5690	290	24	applied	apply	VERB
easat-5690	290	25	science	science	NOUN
easat-5690	290	26	and	and	CCONJ
easat-5690	290	27	technology	technology	NOUN
easat-5690	290	28	,	,	PUNCT
easat-5690	290	29	vol	vol	NOUN
easat-5690	290	30	.	.	PROPN
easat-5690	290	31	8	8	NUM
easat-5690	290	32	,	,	PUNCT
easat-5690	290	33	no	no	INTJ
easat-5690	290	34	.	.	NOUN
easat-5690	290	35	6	6	NUM
easat-5690	290	36	,	,	PUNCT
easat-5690	290	37	pp	pp	ADJ
easat-5690	290	38	.	.	PUNCT
easat-5690	291	1	5707	5707	NUM
easat-5690	291	2	-	-	SYM
easat-5690	291	3	5722	5722	NUM
easat-5690	291	4	,	,	PUNCT
easat-5690	291	5	2024	2024	NUM
easat-5690	291	6	.	.	PUNCT
easat-5690	292	1	https://doi.org/10.55214/25768484.v8i6.3242	https://doi.org/10.55214/25768484.v8i6.3242	NOUN
easat-5690	292	2	[	[	X
easat-5690	292	3	10	10	NUM
easat-5690	292	4	]	]	X
easat-5690	292	5	l.	l.	PROPN
easat-5690	292	6	c.	c.	PROPN
easat-5690	292	7	jaffrin	jaffrin	PROPN
easat-5690	292	8	and	and	CCONJ
easat-5690	292	9	j.	j.	PROPN
easat-5690	292	10	visumathi	visumathi	PROPN
easat-5690	292	11	,	,	PUNCT
easat-5690	292	12	"	"	PUNCT
easat-5690	292	13	impact	impact	NOUN
easat-5690	292	14	of	of	ADP
easat-5690	292	15	feature	feature	NOUN
easat-5690	292	16	selection	selection	NOUN
easat-5690	292	17	techniques	technique	NOUN
easat-5690	292	18	on	on	ADP
easat-5690	292	19	machine	machine	NOUN
easat-5690	292	20	learning	learning	NOUN
easat-5690	292	21	and	and	CCONJ
easat-5690	292	22	deep	deep	ADJ
easat-5690	292	23	learning	learning	NOUN
easat-5690	292	24	techniques	technique	NOUN
easat-5690	292	25	for	for	ADP
easat-5690	292	26	cardiovascular	cardiovascular	ADJ
easat-5690	292	27	disease	disease	NOUN
easat-5690	292	28	prediction	prediction	NOUN
easat-5690	292	29	-	-	PUNCT
easat-5690	292	30	an	an	DET
easat-5690	292	31	analysis	analysis	NOUN
easat-5690	292	32	,	,	PUNCT
easat-5690	292	33	"	"	PUNCT
easat-5690	292	34	edelweiss	edelweiss	PROPN
easat-5690	292	35	applied	apply	VERB
easat-5690	292	36	science	science	NOUN
easat-5690	292	37	and	and	CCONJ
easat-5690	292	38	technology	technology	NOUN
easat-5690	292	39	,	,	PUNCT
easat-5690	292	40	vol	vol	NOUN
easat-5690	292	41	.	.	PROPN
easat-5690	292	42	8	8	NUM
easat-5690	292	43	,	,	PUNCT
easat-5690	292	44	no	no	INTJ
easat-5690	292	45	.	.	NOUN
easat-5690	292	46	5	5	NUM
easat-5690	292	47	,	,	PUNCT
easat-5690	292	48	pp	pp	ADJ
easat-5690	292	49	.	.	PUNCT
easat-5690	292	50	1454	1454	NUM
easat-5690	292	51	-	-	SYM
easat-5690	292	52	1471	1471	NUM
easat-5690	292	53	,	,	PUNCT
easat-5690	292	54	2024	2024	NUM
easat-5690	292	55	.	.	PUNCT
easat-5690	293	1	https://doi.org/10.55214/25768484.v8i5.1848	https://doi.org/10.55214/25768484.v8i5.1848	NOUN
easat-5690	294	1	[	[	X
easat-5690	294	2	11	11	NUM
easat-5690	294	3	]	]	PUNCT
easat-5690	294	4	a.	a.	NOUN
easat-5690	294	5	m.	m.	NOUN
easat-5690	294	6	ridha	ridha	PROPN
easat-5690	294	7	,	,	PUNCT
easat-5690	294	8	m.	m.	NOUN
easat-5690	294	9	j.	j.	PROPN
easat-5690	294	10	mohammed	mohammed	PROPN
easat-5690	294	11	,	,	PUNCT
easat-5690	294	12	h.	h.	PROPN
easat-5690	294	13	a.	a.	PROPN
easat-5690	294	14	saber	saber	PROPN
easat-5690	294	15	,	,	PUNCT
easat-5690	294	16	m.	m.	NOUN
easat-5690	294	17	h.	h.	PROPN
easat-5690	294	18	chyad	chyad	PROPN
easat-5690	294	19	,	,	PUNCT
easat-5690	294	20	and	and	CCONJ
easat-5690	294	21	m.	m.	PROPN
easat-5690	294	22	h.	h.	PROPN
easat-5690	294	23	abdulameer	abdulameer	PROPN
easat-5690	294	24	,	,	PUNCT
easat-5690	294	25	"	"	PUNCT
easat-5690	294	26	advanced	advanced	ADJ
easat-5690	294	27	deep	deep	ADJ
easat-5690	294	28	learning	learning	NOUN
easat-5690	294	29	approaches	approach	NOUN
easat-5690	294	30	for	for	ADP
easat-5690	294	31	early	early	ADJ
easat-5690	294	32	detection	detection	NOUN
easat-5690	294	33	and	and	CCONJ
easat-5690	294	34	localization	localization	NOUN
easat-5690	294	35	of	of	ADP
easat-5690	294	36	ocular	ocular	ADJ
easat-5690	294	37	diseases	disease	NOUN
easat-5690	294	38	,	,	PUNCT
easat-5690	294	39	"	"	PUNCT
easat-5690	294	40	edelweiss	edelweiss	PROPN
easat-5690	294	41	applied	apply	VERB
easat-5690	294	42	science	science	NOUN
easat-5690	294	43	and	and	CCONJ
easat-5690	294	44	technology	technology	NOUN
easat-5690	294	45	,	,	PUNCT
easat-5690	294	46	vol	vol	NOUN
easat-5690	294	47	.	.	PROPN
easat-5690	294	48	8	8	NUM
easat-5690	294	49	,	,	PUNCT
easat-5690	294	50	no	no	INTJ
easat-5690	294	51	.	.	NOUN
easat-5690	294	52	6	6	NUM
easat-5690	294	53	,	,	PUNCT
easat-5690	294	54	pp	pp	ADJ
easat-5690	294	55	.	.	PUNCT
easat-5690	295	1	3708	3708	NUM
easat-5690	295	2	-	-	SYM
easat-5690	295	3	3721	3721	NUM
easat-5690	295	4	,	,	PUNCT
easat-5690	295	5	2024	2024	NUM
easat-5690	295	6	.	.	PUNCT
easat-5690	296	1	https://doi.org/10.55214/25768484.v8i6.2813	https://doi.org/10.55214/25768484.v8i6.2813	PROPN
easat-5690	296	2	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-5690	296	3	https://doi.org/10.1109/spac46244.2018.8965555	https://doi.org/10.1109/spac46244.2018.8965555	PROPN
easat-5690	296	4	https://doi.org/10.5755/j02.eie.33866	https://doi.org/10.5755/j02.eie.33866	PROPN
easat-5690	297	1	https://doi.org/10.1109/cspa.2018.8368703	https://doi.org/10.1109/cspa.2018.8368703	PROPN
easat-5690	297	2	https://doi.org/10.1038/s41598-022-06181-z	https://doi.org/10.1038/s41598-022-06181-z	PROPN
easat-5690	297	3	https://doi.org/10.3390/sym11030343	https://doi.org/10.3390/sym11030343	NOUN
easat-5690	297	4	https://doi.org/10.1109/iccci50826.2021.9402225	https://doi.org/10.1109/iccci50826.2021.9402225	PRON
easat-5690	298	1	https://doi.org/10.33395/sinkron.v8i3.12604	https://doi.org/10.33395/sinkron.v8i3.12604	VERB
easat-5690	298	2	https://doi.org/10.1109/acts49415.2020.9350413	https://doi.org/10.1109/acts49415.2020.9350413	PROPN
easat-5690	298	3	https://doi.org/10.55214/25768484.v8i6.3242	https://doi.org/10.55214/25768484.v8i6.3242	NOUN
easat-5690	298	4	https://doi.org/10.55214/25768484.v8i5.1848	https://doi.org/10.55214/25768484.v8i5.1848	PROPN
easat-5690	298	5	https://doi.org/10.55214/25768484.v8i6.2813	https://doi.org/10.55214/25768484.v8i6.2813	PROPN
easat-5690	298	6	1799	1799	NUM
easat-5690	298	7	edelweiss	edelweiss	PROPN
easat-5690	298	8	applied	apply	VERB
easat-5690	298	9	science	science	NOUN
easat-5690	298	10	and	and	CCONJ
easat-5690	298	11	technology	technology	NOUN
easat-5690	298	12	issn	issn	PROPN
easat-5690	298	13	:	:	PUNCT
easat-5690	298	14	2576	2576	NUM
easat-5690	298	15	-	-	SYM
easat-5690	298	16	8484	8484	NUM
easat-5690	298	17	vol	vol	NOUN
easat-5690	298	18	.	.	PROPN
easat-5690	299	1	9	9	NUM
easat-5690	299	2	,	,	PUNCT
easat-5690	299	3	no	no	INTJ
easat-5690	299	4	.	.	NOUN
easat-5690	300	1	3	3	NUM
easat-5690	300	2	:	:	SYM
easat-5690	300	3	1777	1777	NUM
easat-5690	300	4	-	-	SYM
easat-5690	300	5	1794	1794	NUM
easat-5690	300	6	,	,	PUNCT
easat-5690	300	7	2025	2025	NUM
easat-5690	300	8	doi	doi	NOUN
easat-5690	300	9	:	:	PUNCT
easat-5690	300	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	300	11	©	©	PROPN
easat-5690	300	12	2025	2025	NUM
easat-5690	300	13	by	by	ADP
easat-5690	300	14	the	the	DET
easat-5690	300	15	authors	author	NOUN
easat-5690	300	16	;	;	PUNCT
easat-5690	300	17	licensee	licensee	PROPN
easat-5690	300	18	learning	learning	NOUN
easat-5690	300	19	gate	gate	NOUN
easat-5690	301	1	[	[	X
easat-5690	301	2	12	12	NUM
easat-5690	301	3	]	]	PUNCT
easat-5690	301	4	k.	k.	PROPN
easat-5690	302	1	p.	p.	PROPN
easat-5690	302	2	ferentinos	ferentinos	PROPN
easat-5690	302	3	,	,	PUNCT
easat-5690	302	4	"	"	PUNCT
easat-5690	302	5	deep	deep	ADJ
easat-5690	302	6	learning	learning	NOUN
easat-5690	302	7	models	model	NOUN
easat-5690	302	8	for	for	ADP
easat-5690	302	9	plant	plant	NOUN
easat-5690	302	10	disease	disease	NOUN
easat-5690	302	11	detection	detection	NOUN
easat-5690	302	12	and	and	CCONJ
easat-5690	302	13	diagnosis	diagnosis	NOUN
easat-5690	302	14	,	,	PUNCT
easat-5690	302	15	"	"	PUNCT
easat-5690	302	16	computers	computer	NOUN
easat-5690	302	17	and	and	CCONJ
easat-5690	302	18	electronics	electronic	NOUN
easat-5690	302	19	in	in	ADP
easat-5690	302	20	agriculture	agriculture	NOUN
easat-5690	302	21	,	,	PUNCT
easat-5690	302	22	vol	vol	NOUN
easat-5690	302	23	.	.	PROPN
easat-5690	303	1	145	145	NUM
easat-5690	303	2	,	,	PUNCT
easat-5690	303	3	pp	pp	ADJ
easat-5690	303	4	.	.	PUNCT
easat-5690	304	1	311	311	NUM
easat-5690	304	2	-	-	SYM
easat-5690	304	3	318	318	NUM
easat-5690	304	4	,	,	PUNCT
easat-5690	304	5	2018	2018	NUM
easat-5690	304	6	.	.	PUNCT
easat-5690	305	1	https://doi.org/10.1016/j.compag.2018.01.009	https://doi.org/10.1016/j.compag.2018.01.009	NUM
easat-5690	305	2	[	[	X
easat-5690	305	3	13	13	NUM
easat-5690	305	4	]	]	PUNCT
easat-5690	305	5	m.	m.	NOUN
easat-5690	305	6	h.	h.	PROPN
easat-5690	305	7	kamrul	kamrul	PROPN
easat-5690	305	8	,	,	PUNCT
easat-5690	305	9	m.	m.	PROPN
easat-5690	305	10	rahman	rahman	PROPN
easat-5690	305	11	,	,	PUNCT
easat-5690	305	12	m.	m.	PROPN
easat-5690	305	13	r.	r.	PROPN
easat-5690	305	14	i.	i.	PROPN
easat-5690	305	15	robin	robin	PROPN
easat-5690	305	16	,	,	PUNCT
easat-5690	305	17	m.	m.	PROPN
easat-5690	305	18	s.	s.	PROPN
easat-5690	305	19	hossain	hossain	PROPN
easat-5690	305	20	,	,	PUNCT
easat-5690	305	21	m.	m.	PROPN
easat-5690	305	22	h.	h.	PROPN
easat-5690	305	23	hasan	hasan	PROPN
easat-5690	305	24	,	,	PUNCT
easat-5690	305	25	and	and	CCONJ
easat-5690	305	26	p.	p.	NOUN
easat-5690	305	27	paul	paul	PROPN
easat-5690	305	28	,	,	PUNCT
easat-5690	305	29	"	"	PUNCT
easat-5690	305	30	a	a	DET
easat-5690	305	31	deep	deep	ADJ
easat-5690	305	32	learning	learning	NOUN
easat-5690	305	33	based	base	VERB
easat-5690	305	34	approach	approach	NOUN
easat-5690	305	35	on	on	ADP
easat-5690	305	36	categorization	categorization	NOUN
easat-5690	305	37	of	of	ADP
easat-5690	305	38	tea	tea	NOUN
easat-5690	305	39	leaf	leaf	NOUN
easat-5690	305	40	,	,	PUNCT
easat-5690	305	41	"	"	PUNCT
easat-5690	305	42	in	in	ADP
easat-5690	305	43	proceedings	proceeding	NOUN
easat-5690	305	44	of	of	ADP
easat-5690	305	45	the	the	DET
easat-5690	305	46	international	international	ADJ
easat-5690	305	47	conference	conference	NOUN
easat-5690	305	48	on	on	ADP
easat-5690	305	49	computing	compute	VERB
easat-5690	305	50	advancements	advancement	NOUN
easat-5690	305	51	,	,	PUNCT
easat-5690	305	52	2020	2020	NUM
easat-5690	305	53	,	,	PUNCT
easat-5690	305	54	pp	pp	ADV
easat-5690	305	55	.	.	PUNCT
easat-5690	306	1	1	1	NUM
easat-5690	306	2	-	-	SYM
easat-5690	306	3	8	8	NUM
easat-5690	306	4	,	,	PUNCT
easat-5690	306	5	doi	doi	NOUN
easat-5690	306	6	:	:	PUNCT
easat-5690	306	7	https://doi.org/10.1145/3377049.3377122	https://doi.org/10.1145/3377049.3377122	NOUN
easat-5690	306	8	.	.	PUNCT
easat-5690	307	1	[	[	X
easat-5690	307	2	14	14	NUM
easat-5690	307	3	]	]	PUNCT
easat-5690	307	4	s.-h	s.-h	NOUN
easat-5690	307	5	.	.	PUNCT
easat-5690	308	1	lee	lee	PROPN
easat-5690	308	2	,	,	PUNCT
easat-5690	308	3	c.-c	c.-c	PROPN
easat-5690	308	4	.	.	PUNCT
easat-5690	309	1	wu	wu	PROPN
easat-5690	309	2	,	,	PUNCT
easat-5690	309	3	and	and	CCONJ
easat-5690	309	4	s.-f	s.-f	PROPN
easat-5690	309	5	.	.	PUNCT
easat-5690	310	1	chen	chen	PROPN
easat-5690	310	2	,	,	PUNCT
easat-5690	310	3	"	"	PUNCT
easat-5690	310	4	development	development	NOUN
easat-5690	310	5	of	of	ADP
easat-5690	310	6	image	image	NOUN
easat-5690	310	7	recognition	recognition	NOUN
easat-5690	310	8	and	and	CCONJ
easat-5690	310	9	classification	classification	NOUN
easat-5690	310	10	algorithm	algorithm	NOUN
easat-5690	310	11	for	for	ADP
easat-5690	310	12	tea	tea	NOUN
easat-5690	310	13	leaf	leaf	NOUN
easat-5690	310	14	diseases	disease	NOUN
easat-5690	310	15	using	use	VERB
easat-5690	310	16	convolutional	convolutional	ADJ
easat-5690	310	17	neural	neural	ADJ
easat-5690	310	18	network	network	NOUN
easat-5690	310	19	,	,	PUNCT
easat-5690	310	20	"	"	PUNCT
easat-5690	310	21	in	in	ADP
easat-5690	310	22	2018	2018	NUM
easat-5690	310	23	asabe	asabe	PRON
easat-5690	310	24	annual	annual	ADJ
easat-5690	310	25	international	international	ADJ
easat-5690	310	26	meeting	meeting	NOUN
easat-5690	310	27	,	,	PUNCT
easat-5690	310	28	2018	2018	NUM
easat-5690	310	29	:	:	PUNCT
easat-5690	310	30	american	american	ADJ
easat-5690	310	31	society	society	NOUN
easat-5690	310	32	of	of	ADP
easat-5690	310	33	agricultural	agricultural	ADJ
easat-5690	310	34	and	and	CCONJ
easat-5690	310	35	biological	biological	ADJ
easat-5690	310	36	engineers	engineer	NOUN
easat-5690	310	37	,	,	PUNCT
easat-5690	310	38	p.	p.	NOUN
easat-5690	310	39	1	1	NUM
easat-5690	310	40	,	,	PUNCT
easat-5690	310	41	doi	doi	NOUN
easat-5690	310	42	:	:	PUNCT
easat-5690	310	43	https://doi.org/10.13031/aim.201801254	https://doi.org/10.13031/aim.201801254	ADJ
easat-5690	310	44	.	.	PUNCT
easat-5690	311	1	[	[	X
easat-5690	311	2	15	15	NUM
easat-5690	311	3	]	]	X
easat-5690	311	4	y.	y.	PROPN
easat-5690	311	5	zhang	zhang	PROPN
easat-5690	311	6	,	,	PUNCT
easat-5690	311	7	c.	c.	PROPN
easat-5690	311	8	song	song	PROPN
easat-5690	311	9	,	,	PUNCT
easat-5690	311	10	and	and	CCONJ
easat-5690	311	11	d.	d.	PROPN
easat-5690	311	12	zhang	zhang	PROPN
easat-5690	311	13	,	,	PUNCT
easat-5690	311	14	"	"	PUNCT
easat-5690	311	15	deep	deep	ADJ
easat-5690	311	16	learning	learning	NOUN
easat-5690	311	17	-	-	PUNCT
easat-5690	311	18	based	base	VERB
easat-5690	311	19	object	object	NOUN
easat-5690	311	20	detection	detection	NOUN
easat-5690	311	21	improvement	improvement	NOUN
easat-5690	311	22	for	for	ADP
easat-5690	311	23	tomato	tomato	NOUN
easat-5690	311	24	disease	disease	NOUN
easat-5690	311	25	,	,	PUNCT
easat-5690	311	26	"	"	PUNCT
easat-5690	311	27	ieee	ieee	NOUN
easat-5690	311	28	access	access	NOUN
easat-5690	311	29	,	,	PUNCT
easat-5690	311	30	vol	vol	NOUN
easat-5690	311	31	.	.	PROPN
easat-5690	311	32	8	8	NUM
easat-5690	311	33	,	,	PUNCT
easat-5690	311	34	pp	pp	ADJ
easat-5690	311	35	.	.	PUNCT
easat-5690	312	1	56607	56607	NUM
easat-5690	312	2	-	-	SYM
easat-5690	312	3	56614	56614	NUM
easat-5690	312	4	,	,	PUNCT
easat-5690	312	5	2020	2020	NUM
easat-5690	312	6	.	.	PUNCT
easat-5690	313	1	https://doi.org/10.1109/access.2020.2982456	https://doi.org/10.1109/access.2020.2982456	PROPN
easat-5690	314	1	[	[	X
easat-5690	314	2	16	16	NUM
easat-5690	314	3	]	]	X
easat-5690	314	4	g.	g.	PROPN
easat-5690	314	5	hu	hu	PROPN
easat-5690	314	6	,	,	PUNCT
easat-5690	314	7	h.	h.	PROPN
easat-5690	314	8	wang	wang	PROPN
easat-5690	314	9	,	,	PUNCT
easat-5690	314	10	y.	y.	PROPN
easat-5690	314	11	zhang	zhang	PROPN
easat-5690	314	12	,	,	PUNCT
easat-5690	314	13	and	and	CCONJ
easat-5690	314	14	m.	m.	PROPN
easat-5690	314	15	wan	wan	PROPN
easat-5690	314	16	,	,	PUNCT
easat-5690	314	17	"	"	PUNCT
easat-5690	314	18	detection	detection	NOUN
easat-5690	314	19	and	and	CCONJ
easat-5690	314	20	severity	severity	NOUN
easat-5690	314	21	analysis	analysis	NOUN
easat-5690	314	22	of	of	ADP
easat-5690	314	23	tea	tea	NOUN
easat-5690	314	24	leaf	leaf	NOUN
easat-5690	314	25	blight	blight	NOUN
easat-5690	314	26	based	base	VERB
easat-5690	314	27	on	on	ADP
easat-5690	314	28	deep	deep	ADJ
easat-5690	314	29	learning	learning	NOUN
easat-5690	314	30	,	,	PUNCT
easat-5690	314	31	"	"	PUNCT
easat-5690	314	32	computers	computer	NOUN
easat-5690	314	33	&	&	CCONJ
easat-5690	314	34	electrical	electrical	ADJ
easat-5690	314	35	engineering	engineering	NOUN
easat-5690	314	36	,	,	PUNCT
easat-5690	314	37	vol	vol	NOUN
easat-5690	314	38	.	.	PROPN
easat-5690	314	39	90	90	NUM
easat-5690	314	40	,	,	PUNCT
easat-5690	314	41	p.	p.	NOUN
easat-5690	314	42	107023	107023	NUM
easat-5690	314	43	,	,	PUNCT
easat-5690	314	44	2021	2021	NUM
easat-5690	314	45	.	.	PUNCT
easat-5690	315	1	https://doi.org/10.1016/j.compeleceng.2021.107023	https://doi.org/10.1016/j.compeleceng.2021.107023	NOUN
easat-5690	316	1	[	[	X
easat-5690	316	2	17	17	NUM
easat-5690	316	3	]	]	PUNCT
easat-5690	316	4	l.	l.	PROPN
easat-5690	316	5	wang	wang	PROPN
easat-5690	316	6	and	and	CCONJ
easat-5690	316	7	w.	w.	PROPN
easat-5690	316	8	q.	q.	PROPN
easat-5690	316	9	yan	yan	PROPN
easat-5690	316	10	,	,	PUNCT
easat-5690	316	11	"	"	PUNCT
easat-5690	316	12	tree	tree	NOUN
easat-5690	316	13	leaves	leave	VERB
easat-5690	316	14	detection	detection	NOUN
easat-5690	316	15	based	base	VERB
easat-5690	316	16	on	on	ADP
easat-5690	316	17	deep	deep	ADJ
easat-5690	316	18	learning	learning	NOUN
easat-5690	316	19	,	,	PUNCT
easat-5690	316	20	"	"	PUNCT
easat-5690	316	21	in	in	ADP
easat-5690	316	22	geometry	geometry	NOUN
easat-5690	316	23	and	and	CCONJ
easat-5690	316	24	vision	vision	NOUN
easat-5690	316	25	:	:	PUNCT
easat-5690	316	26	first	first	ADJ
easat-5690	316	27	international	international	ADJ
easat-5690	316	28	symposium	symposium	NOUN
easat-5690	316	29	,	,	PUNCT
easat-5690	316	30	isgv	isgv	NOUN
easat-5690	316	31	2021	2021	NUM
easat-5690	316	32	,	,	PUNCT
easat-5690	316	33	auckland	auckland	PROPN
easat-5690	316	34	,	,	PUNCT
easat-5690	316	35	new	new	PROPN
easat-5690	316	36	zealand	zealand	PROPN
easat-5690	316	37	,	,	PUNCT
easat-5690	316	38	january	january	PROPN
easat-5690	316	39	28	28	NUM
easat-5690	316	40	-	-	SYM
easat-5690	316	41	29	29	NUM
easat-5690	316	42	,	,	PUNCT
easat-5690	316	43	2021	2021	NUM
easat-5690	316	44	,	,	PUNCT
easat-5690	316	45	revised	revise	VERB
easat-5690	316	46	selected	select	VERB
easat-5690	316	47	papers	paper	NOUN
easat-5690	316	48	1	1	NUM
easat-5690	316	49	,	,	PUNCT
easat-5690	316	50	2021	2021	NUM
easat-5690	316	51	:	:	PUNCT
easat-5690	316	52	springer	springer	NOUN
easat-5690	316	53	,	,	PUNCT
easat-5690	316	54	pp	pp	ADP
easat-5690	316	55	.	.	PUNCT
easat-5690	317	1	2638	2638	NUM
easat-5690	317	2	,	,	PUNCT
easat-5690	317	3	doi	doi	NOUN
easat-5690	317	4	:	:	PUNCT
easat-5690	317	5	https://doi.org/10.1007/978-3-030-72073-5_3	https://doi.org/10.1007/978-3-030-72073-5_3	X
easat-5690	317	6	.	.	PUNCT
easat-5690	318	1	[	[	X
easat-5690	318	2	18	18	NUM
easat-5690	318	3	]	]	PUNCT
easat-5690	318	4	j.	j.	PROPN
easat-5690	318	5	lin	lin	PROPN
easat-5690	318	6	,	,	PUNCT
easat-5690	318	7	d.	d.	PROPN
easat-5690	318	8	bai	bai	PROPN
easat-5690	318	9	,	,	PUNCT
easat-5690	318	10	r.	r.	PROPN
easat-5690	318	11	xu	xu	PROPN
easat-5690	318	12	,	,	PUNCT
easat-5690	318	13	and	and	CCONJ
easat-5690	318	14	h.	h.	PROPN
easat-5690	318	15	lin	lin	PROPN
easat-5690	318	16	,	,	PUNCT
easat-5690	318	17	"	"	PUNCT
easat-5690	318	18	tsba	tsba	NOUN
easat-5690	318	19	-	-	PUNCT
easat-5690	318	20	yolo	yolo	NOUN
easat-5690	318	21	:	:	PUNCT
easat-5690	318	22	an	an	DET
easat-5690	318	23	improved	improved	ADJ
easat-5690	318	24	tea	tea	NOUN
easat-5690	318	25	diseases	disease	NOUN
easat-5690	318	26	detection	detection	NOUN
easat-5690	318	27	model	model	NOUN
easat-5690	318	28	based	base	VERB
easat-5690	318	29	on	on	ADP
easat-5690	318	30	attention	attention	NOUN
easat-5690	318	31	mechanisms	mechanism	NOUN
easat-5690	318	32	and	and	CCONJ
easat-5690	318	33	feature	feature	NOUN
easat-5690	318	34	fusion	fusion	NOUN
easat-5690	318	35	,	,	PUNCT
easat-5690	318	36	"	"	PUNCT
easat-5690	318	37	forests	forest	NOUN
easat-5690	318	38	,	,	PUNCT
easat-5690	318	39	vol	vol	NOUN
easat-5690	318	40	.	.	PROPN
easat-5690	318	41	14	14	NUM
easat-5690	318	42	,	,	PUNCT
easat-5690	318	43	no	no	INTJ
easat-5690	318	44	.	.	NOUN
easat-5690	318	45	3	3	NUM
easat-5690	318	46	,	,	PUNCT
easat-5690	318	47	p.	p.	NOUN
easat-5690	318	48	619	619	NUM
easat-5690	318	49	,	,	PUNCT
easat-5690	318	50	2023	2023	NUM
easat-5690	318	51	.	.	PUNCT
easat-5690	319	1	https://doi.org/10.3390/f14030619	https://doi.org/10.3390/f14030619	X
easat-5690	320	1	[	[	X
easat-5690	320	2	19	19	NUM
easat-5690	320	3	]	]	PUNCT
easat-5690	320	4	z.	z.	PROPN
easat-5690	320	5	xue	xue	PROPN
easat-5690	320	6	,	,	PUNCT
easat-5690	320	7	r.	r.	PROPN
easat-5690	320	8	xu	xu	PROPN
easat-5690	320	9	,	,	PUNCT
easat-5690	320	10	d.	d.	PROPN
easat-5690	320	11	bai	bai	PROPN
easat-5690	320	12	,	,	PUNCT
easat-5690	320	13	and	and	CCONJ
easat-5690	320	14	h.	h.	PROPN
easat-5690	320	15	lin	lin	PROPN
easat-5690	320	16	,	,	PUNCT
easat-5690	320	17	"	"	PUNCT
easat-5690	320	18	yolo	yolo	PROPN
easat-5690	320	19	-	-	PUNCT
easat-5690	320	20	tea	tea	NOUN
easat-5690	320	21	:	:	PUNCT
easat-5690	320	22	a	a	DET
easat-5690	320	23	tea	tea	NOUN
easat-5690	320	24	disease	disease	NOUN
easat-5690	320	25	detection	detection	NOUN
easat-5690	320	26	model	model	NOUN
easat-5690	320	27	improved	improve	VERB
easat-5690	320	28	by	by	ADP
easat-5690	320	29	yolov5	yolov5	NOUN
easat-5690	320	30	,	,	PUNCT
easat-5690	320	31	"	"	PUNCT
easat-5690	320	32	forests	forest	NOUN
easat-5690	320	33	,	,	PUNCT
easat-5690	320	34	vol	vol	NOUN
easat-5690	320	35	.	.	PROPN
easat-5690	320	36	14	14	NUM
easat-5690	320	37	,	,	PUNCT
easat-5690	320	38	no	no	INTJ
easat-5690	320	39	.	.	NOUN
easat-5690	320	40	2	2	NUM
easat-5690	320	41	,	,	PUNCT
easat-5690	320	42	p.	p.	NOUN
easat-5690	320	43	415	415	NUM
easat-5690	320	44	,	,	PUNCT
easat-5690	320	45	2023	2023	NUM
easat-5690	320	46	.	.	PUNCT
easat-5690	321	1	https://doi.org/10.3390/f14020415	https://doi.org/10.3390/f14020415	PRON
easat-5690	322	1	[	[	SYM
easat-5690	322	2	20	20	NUM
easat-5690	322	3	]	]	X
easat-5690	322	4	y.	y.	PROPN
easat-5690	322	5	li	li	PROPN
easat-5690	322	6	,	,	PUNCT
easat-5690	322	7	a.	a.	PROPN
easat-5690	322	8	li	li	PROPN
easat-5690	322	9	,	,	PUNCT
easat-5690	322	10	x.	x.	PROPN
easat-5690	322	11	li	li	PROPN
easat-5690	322	12	,	,	PUNCT
easat-5690	322	13	and	and	CCONJ
easat-5690	322	14	d.	d.	PROPN
easat-5690	322	15	liang	liang	PROPN
easat-5690	322	16	,	,	PUNCT
easat-5690	322	17	"	"	PUNCT
easat-5690	322	18	detection	detection	NOUN
easat-5690	322	19	and	and	CCONJ
easat-5690	322	20	identification	identification	NOUN
easat-5690	322	21	of	of	ADP
easat-5690	322	22	peach	peach	NOUN
easat-5690	322	23	leaf	leaf	NOUN
easat-5690	322	24	diseases	disease	NOUN
easat-5690	322	25	based	base	VERB
easat-5690	322	26	on	on	ADP
easat-5690	322	27	yolo	yolo	ADJ
easat-5690	322	28	v5	v5	PROPN
easat-5690	322	29	improved	improve	VERB
easat-5690	322	30	model	model	NOUN
easat-5690	322	31	,	,	PUNCT
easat-5690	322	32	"	"	PUNCT
easat-5690	322	33	in	in	ADP
easat-5690	322	34	proceedings	proceeding	NOUN
easat-5690	322	35	of	of	ADP
easat-5690	322	36	the	the	DET
easat-5690	322	37	5th	5th	ADJ
easat-5690	322	38	international	international	ADJ
easat-5690	322	39	conference	conference	NOUN
easat-5690	322	40	on	on	ADP
easat-5690	322	41	control	control	NOUN
easat-5690	322	42	and	and	CCONJ
easat-5690	322	43	computer	computer	NOUN
easat-5690	322	44	vision	vision	NOUN
easat-5690	322	45	,	,	PUNCT
easat-5690	322	46	2022	2022	NUM
easat-5690	322	47	,	,	PUNCT
easat-5690	322	48	pp	pp	ADJ
easat-5690	322	49	.	.	PUNCT
easat-5690	323	1	79	79	NUM
easat-5690	323	2	-	-	SYM
easat-5690	323	3	84	84	NUM
easat-5690	323	4	,	,	PUNCT
easat-5690	323	5	doi	doi	NOUN
easat-5690	323	6	:	:	PUNCT
easat-5690	323	7	https://doi.org/10.1145/3561613.3561626	https://doi.org/10.1145/3561613.3561626	NOUN
easat-5690	323	8	.	.	PUNCT
easat-5690	324	1	[	[	X
easat-5690	324	2	21	21	NUM
easat-5690	324	3	]	]	X
easat-5690	324	4	b.	b.	PROPN
easat-5690	324	5	yan	yan	PROPN
easat-5690	324	6	,	,	PUNCT
easat-5690	324	7	p.	p.	PROPN
easat-5690	324	8	fan	fan	PROPN
easat-5690	324	9	,	,	PUNCT
easat-5690	324	10	x.	x.	PROPN
easat-5690	324	11	lei	lei	PROPN
easat-5690	324	12	,	,	PUNCT
easat-5690	324	13	z.	z.	PROPN
easat-5690	324	14	liu	liu	PROPN
easat-5690	324	15	,	,	PUNCT
easat-5690	324	16	and	and	CCONJ
easat-5690	324	17	f.	f.	PROPN
easat-5690	324	18	yang	yang	PROPN
easat-5690	324	19	,	,	PUNCT
easat-5690	324	20	"	"	PUNCT
easat-5690	324	21	a	a	DET
easat-5690	324	22	real	real	ADJ
easat-5690	324	23	-	-	PUNCT
easat-5690	324	24	time	time	NOUN
easat-5690	324	25	apple	apple	NOUN
easat-5690	324	26	targets	target	NOUN
easat-5690	324	27	detection	detection	NOUN
easat-5690	324	28	method	method	NOUN
easat-5690	324	29	for	for	ADP
easat-5690	324	30	picking	pick	VERB
easat-5690	324	31	robot	robot	NOUN
easat-5690	324	32	based	base	VERB
easat-5690	324	33	on	on	ADP
easat-5690	324	34	improved	improved	ADJ
easat-5690	324	35	yolov5	yolov5	NOUN
easat-5690	324	36	,	,	PUNCT
easat-5690	324	37	"	"	PUNCT
easat-5690	324	38	remote	remote	ADJ
easat-5690	324	39	sensing	sensing	NOUN
easat-5690	324	40	,	,	PUNCT
easat-5690	324	41	vol	vol	NOUN
easat-5690	324	42	.	.	PROPN
easat-5690	324	43	13	13	NUM
easat-5690	324	44	,	,	PUNCT
easat-5690	324	45	no	no	INTJ
easat-5690	324	46	.	.	NOUN
easat-5690	324	47	9	9	NUM
easat-5690	324	48	,	,	PUNCT
easat-5690	324	49	p.	p.	NOUN
easat-5690	324	50	1619	1619	NUM
easat-5690	324	51	,	,	PUNCT
easat-5690	324	52	2021	2021	NUM
easat-5690	324	53	.	.	PUNCT
easat-5690	325	1	https://doi.org/10.3390/rs13091619	https://doi.org/10.3390/rs13091619	X
easat-5690	326	1	[	[	X
easat-5690	326	2	22	22	NUM
easat-5690	326	3	]	]	PUNCT
easat-5690	326	4	z.	z.	PROPN
easat-5690	326	5	chen	chen	PROPN
easat-5690	326	6	et	et	PROPN
easat-5690	326	7	al	al	PROPN
easat-5690	326	8	.	.	PROPN
easat-5690	326	9	,	,	PUNCT
easat-5690	326	10	"	"	PUNCT
easat-5690	326	11	plant	plant	NOUN
easat-5690	326	12	disease	disease	NOUN
easat-5690	326	13	recognition	recognition	NOUN
easat-5690	326	14	model	model	NOUN
easat-5690	326	15	based	base	VERB
easat-5690	326	16	on	on	ADP
easat-5690	326	17	improved	improved	ADJ
easat-5690	326	18	yolov5	yolov5	NOUN
easat-5690	326	19	,	,	PUNCT
easat-5690	326	20	"	"	PUNCT
easat-5690	326	21	agronomy	agronomy	NOUN
easat-5690	326	22	,	,	PUNCT
easat-5690	326	23	vol	vol	NOUN
easat-5690	326	24	.	.	PROPN
easat-5690	326	25	12	12	NUM
easat-5690	326	26	,	,	PUNCT
easat-5690	326	27	no	no	INTJ
easat-5690	326	28	.	.	NOUN
easat-5690	326	29	2	2	NUM
easat-5690	326	30	,	,	PUNCT
easat-5690	326	31	p.	p.	NOUN
easat-5690	326	32	365	365	NUM
easat-5690	326	33	,	,	PUNCT
easat-5690	326	34	2022	2022	NUM
easat-5690	326	35	.	.	PUNCT
easat-5690	327	1	https://doi.org/10.3390/agronomy12020365	https://doi.org/10.3390/agronomy12020365	PROPN
easat-5690	327	2	[	[	X
easat-5690	327	3	23	23	X
easat-5690	327	4	]	]	X
easat-5690	327	5	y.	y.	PROPN
easat-5690	327	6	wang	wang	PROPN
easat-5690	327	7	,	,	PUNCT
easat-5690	327	8	r.	r.	PROPN
easat-5690	327	9	xu	xu	PROPN
easat-5690	327	10	,	,	PUNCT
easat-5690	327	11	d.	d.	PROPN
easat-5690	327	12	bai	bai	PROPN
easat-5690	327	13	,	,	PUNCT
easat-5690	327	14	and	and	CCONJ
easat-5690	327	15	h.	h.	PROPN
easat-5690	327	16	lin	lin	PROPN
easat-5690	327	17	,	,	PUNCT
easat-5690	327	18	"	"	PUNCT
easat-5690	327	19	integrated	integrated	ADJ
easat-5690	327	20	learning	learning	NOUN
easat-5690	327	21	-	-	PUNCT
easat-5690	327	22	based	base	VERB
easat-5690	327	23	pest	pest	NOUN
easat-5690	327	24	and	and	CCONJ
easat-5690	327	25	disease	disease	NOUN
easat-5690	327	26	detection	detection	NOUN
easat-5690	327	27	method	method	NOUN
easat-5690	327	28	for	for	ADP
easat-5690	327	29	tea	tea	NOUN
easat-5690	327	30	leaves	leave	NOUN
easat-5690	327	31	,	,	PUNCT
easat-5690	327	32	"	"	PUNCT
easat-5690	327	33	forests	forest	NOUN
easat-5690	327	34	,	,	PUNCT
easat-5690	327	35	vol	vol	NOUN
easat-5690	327	36	.	.	PROPN
easat-5690	327	37	14	14	NUM
easat-5690	327	38	,	,	PUNCT
easat-5690	327	39	no	no	INTJ
easat-5690	327	40	.	.	NOUN
easat-5690	327	41	5	5	NUM
easat-5690	327	42	,	,	PUNCT
easat-5690	327	43	p.	p.	NOUN
easat-5690	327	44	1012	1012	NUM
easat-5690	327	45	,	,	PUNCT
easat-5690	327	46	2023	2023	NUM
easat-5690	327	47	.	.	PUNCT
easat-5690	328	1	https://doi.org/10.3390/f14051012	https://doi.org/10.3390/f14051012	NUM
easat-5690	328	2	[	[	X
easat-5690	328	3	24	24	NUM
easat-5690	328	4	]	]	PUNCT
easat-5690	328	5	j.	j.	PROPN
easat-5690	328	6	qi	qi	PROPN
easat-5690	328	7	et	et	PROPN
easat-5690	328	8	al	al	PROPN
easat-5690	328	9	.	.	PROPN
easat-5690	328	10	,	,	PUNCT
easat-5690	328	11	"	"	PUNCT
easat-5690	328	12	an	an	DET
easat-5690	328	13	improved	improved	ADJ
easat-5690	328	14	yolov5	yolov5	NOUN
easat-5690	328	15	model	model	NOUN
easat-5690	328	16	based	base	VERB
easat-5690	328	17	on	on	ADP
easat-5690	328	18	visual	visual	ADJ
easat-5690	328	19	attention	attention	NOUN
easat-5690	328	20	mechanism	mechanism	NOUN
easat-5690	328	21	:	:	PUNCT
easat-5690	328	22	application	application	NOUN
easat-5690	328	23	to	to	ADP
easat-5690	328	24	recognition	recognition	NOUN
easat-5690	328	25	of	of	ADP
easat-5690	328	26	tomato	tomato	NOUN
easat-5690	328	27	virus	virus	NOUN
easat-5690	328	28	disease	disease	NOUN
easat-5690	328	29	,	,	PUNCT
easat-5690	328	30	"	"	PUNCT
easat-5690	328	31	computers	computer	NOUN
easat-5690	328	32	and	and	CCONJ
easat-5690	328	33	electronics	electronic	NOUN
easat-5690	328	34	in	in	ADP
easat-5690	328	35	agriculture	agriculture	NOUN
easat-5690	328	36	,	,	PUNCT
easat-5690	328	37	vol	vol	NOUN
easat-5690	328	38	.	.	PROPN
easat-5690	328	39	194	194	NUM
easat-5690	328	40	,	,	PUNCT
easat-5690	328	41	p.	p.	NOUN
easat-5690	328	42	106780	106780	NUM
easat-5690	328	43	,	,	PUNCT
easat-5690	328	44	2022	2022	NUM
easat-5690	328	45	.	.	PUNCT
easat-5690	329	1	https://doi.org/10.1016/j.compag.2022.106780	https://doi.org/10.1016/j.compag.2022.106780	NOUN
easat-5690	329	2	[	[	X
easat-5690	329	3	25	25	NUM
easat-5690	329	4	]	]	PUNCT
easat-5690	329	5	r.	r.	PROPN
easat-5690	329	6	li	li	PROPN
easat-5690	329	7	and	and	CCONJ
easat-5690	329	8	y.	y.	PROPN
easat-5690	329	9	wu	wu	PROPN
easat-5690	329	10	,	,	PUNCT
easat-5690	329	11	"	"	PUNCT
easat-5690	329	12	improved	improve	VERB
easat-5690	329	13	yolo	yolo	ADJ
easat-5690	329	14	v5	v5	NOUN
easat-5690	329	15	wheat	wheat	NOUN
easat-5690	329	16	ear	ear	NOUN
easat-5690	329	17	detection	detection	NOUN
easat-5690	329	18	algorithm	algorithm	NOUN
easat-5690	329	19	based	base	VERB
easat-5690	329	20	on	on	ADP
easat-5690	329	21	attention	attention	NOUN
easat-5690	329	22	mechanism	mechanism	NOUN
easat-5690	329	23	,	,	PUNCT
easat-5690	329	24	"	"	PUNCT
easat-5690	329	25	electronics	electronic	NOUN
easat-5690	329	26	,	,	PUNCT
easat-5690	329	27	vol	vol	NOUN
easat-5690	329	28	.	.	PROPN
easat-5690	330	1	11	11	NUM
easat-5690	330	2	,	,	PUNCT
easat-5690	330	3	no	no	INTJ
easat-5690	330	4	.	.	NOUN
easat-5690	330	5	11	11	NUM
easat-5690	330	6	,	,	PUNCT
easat-5690	330	7	p.	p.	NOUN
easat-5690	330	8	1673	1673	NUM
easat-5690	330	9	,	,	PUNCT
easat-5690	330	10	2022	2022	NUM
easat-5690	330	11	.	.	PUNCT
easat-5690	331	1	https://doi.org/10.3390/electronics11111673	https://doi.org/10.3390/electronics11111673	PROPN
easat-5690	332	1	[	[	X
easat-5690	332	2	26	26	NUM
easat-5690	332	3	]	]	PUNCT
easat-5690	332	4	l.	l.	PROPN
easat-5690	332	5	wang	wang	PROPN
easat-5690	332	6	,	,	PUNCT
easat-5690	332	7	y.	y.	PROPN
easat-5690	332	8	zhao	zhao	PROPN
easat-5690	332	9	,	,	PUNCT
easat-5690	332	10	z.	z.	PROPN
easat-5690	332	11	xiong	xiong	PROPN
easat-5690	332	12	,	,	PUNCT
easat-5690	332	13	s.	s.	PROPN
easat-5690	332	14	wang	wang	PROPN
easat-5690	332	15	,	,	PUNCT
easat-5690	332	16	y.	y.	PROPN
easat-5690	332	17	li	li	PROPN
easat-5690	332	18	,	,	PUNCT
easat-5690	332	19	and	and	CCONJ
easat-5690	332	20	y.	y.	PROPN
easat-5690	332	21	lan	lan	PROPN
easat-5690	332	22	,	,	PUNCT
easat-5690	332	23	"	"	PUNCT
easat-5690	332	24	fast	fast	ADJ
easat-5690	332	25	and	and	CCONJ
easat-5690	332	26	precise	precise	ADJ
easat-5690	332	27	detection	detection	NOUN
easat-5690	332	28	of	of	ADP
easat-5690	332	29	litchi	litchi	PROPN
easat-5690	332	30	fruits	fruit	NOUN
easat-5690	332	31	for	for	ADP
easat-5690	332	32	yield	yield	NOUN
easat-5690	332	33	estimation	estimation	NOUN
easat-5690	332	34	based	base	VERB
easat-5690	332	35	on	on	ADP
easat-5690	332	36	the	the	DET
easat-5690	332	37	improved	improved	ADJ
easat-5690	332	38	yolov5	yolov5	NOUN
easat-5690	332	39	model	model	NOUN
easat-5690	332	40	,	,	PUNCT
easat-5690	332	41	"	"	PUNCT
easat-5690	332	42	frontiers	frontier	NOUN
easat-5690	332	43	in	in	ADP
easat-5690	332	44	plant	plant	NOUN
easat-5690	332	45	science	science	NOUN
easat-5690	332	46	,	,	PUNCT
easat-5690	332	47	vol	vol	NOUN
easat-5690	332	48	.	.	PROPN
easat-5690	332	49	13	13	NUM
easat-5690	332	50	,	,	PUNCT
easat-5690	332	51	p.	p.	NOUN
easat-5690	332	52	965425	965425	NUM
easat-5690	332	53	,	,	PUNCT
easat-5690	332	54	2022	2022	NUM
easat-5690	332	55	.	.	PUNCT
easat-5690	333	1	https://doi.org/10.3389/fpls.2022.965425	https://doi.org/10.3389/fpls.2022.965425	PROPN
easat-5690	334	1	[	[	X
easat-5690	334	2	27	27	NUM
easat-5690	334	3	]	]	PUNCT
easat-5690	334	4	z.	z.	PROPN
easat-5690	334	5	ye	ye	PRON
easat-5690	334	6	et	et	PROPN
easat-5690	334	7	al	al	PROPN
easat-5690	334	8	.	.	PROPN
easat-5690	334	9	,	,	PUNCT
easat-5690	334	10	"	"	PUNCT
easat-5690	334	11	recognition	recognition	NOUN
easat-5690	334	12	of	of	ADP
easat-5690	334	13	terminal	terminal	ADJ
easat-5690	334	14	buds	bud	NOUN
easat-5690	334	15	of	of	ADP
easat-5690	334	16	densely	densely	ADV
easat-5690	334	17	-	-	PUNCT
easat-5690	334	18	planted	plant	VERB
easat-5690	334	19	chinese	chinese	ADJ
easat-5690	334	20	fir	fir	NOUN
easat-5690	334	21	seedlings	seedling	NOUN
easat-5690	334	22	using	use	VERB
easat-5690	334	23	improved	improved	ADJ
easat-5690	334	24	yolov5	yolov5	NOUN
easat-5690	334	25	by	by	ADP
easat-5690	334	26	integrating	integrate	VERB
easat-5690	334	27	attention	attention	NOUN
easat-5690	334	28	mechanism	mechanism	NOUN
easat-5690	334	29	,	,	PUNCT
easat-5690	334	30	"	"	PUNCT
easat-5690	334	31	frontiers	frontier	NOUN
easat-5690	334	32	in	in	ADP
easat-5690	334	33	plant	plant	NOUN
easat-5690	334	34	science	science	NOUN
easat-5690	334	35	,	,	PUNCT
easat-5690	334	36	vol	vol	NOUN
easat-5690	334	37	.	.	PROPN
easat-5690	334	38	13	13	NUM
easat-5690	334	39	,	,	PUNCT
easat-5690	334	40	p.	p.	NOUN
easat-5690	334	41	991929	991929	NUM
easat-5690	334	42	,	,	PUNCT
easat-5690	334	43	2022	2022	NUM
easat-5690	334	44	.	.	PUNCT
easat-5690	335	1	https://doi.org/10.3389/fpls.2022.991929	https://doi.org/10.3389/fpls.2022.991929	PROPN
easat-5690	336	1	[	[	X
easat-5690	336	2	28	28	NUM
easat-5690	336	3	]	]	X
easat-5690	336	4	h.	h.	PROPN
easat-5690	336	5	zang	zang	PROPN
easat-5690	336	6	et	et	PROPN
easat-5690	336	7	al	al	PROPN
easat-5690	336	8	.	.	PROPN
easat-5690	336	9	,	,	PUNCT
easat-5690	336	10	"	"	PUNCT
easat-5690	336	11	detection	detection	NOUN
easat-5690	336	12	method	method	NOUN
easat-5690	336	13	of	of	ADP
easat-5690	336	14	wheat	wheat	NOUN
easat-5690	336	15	spike	spike	NOUN
easat-5690	336	16	improved	improve	VERB
easat-5690	336	17	yolov5s	yolov5s	PROPN
easat-5690	336	18	based	base	VERB
easat-5690	336	19	on	on	ADP
easat-5690	336	20	the	the	DET
easat-5690	336	21	attention	attention	NOUN
easat-5690	336	22	mechanism	mechanism	NOUN
easat-5690	336	23	,	,	PUNCT
easat-5690	336	24	"	"	PUNCT
easat-5690	336	25	frontiers	frontier	NOUN
easat-5690	336	26	in	in	ADP
easat-5690	336	27	plant	plant	NOUN
easat-5690	336	28	science	science	NOUN
easat-5690	336	29	,	,	PUNCT
easat-5690	336	30	vol	vol	NOUN
easat-5690	336	31	.	.	PROPN
easat-5690	336	32	13	13	NUM
easat-5690	336	33	,	,	PUNCT
easat-5690	336	34	p.	p.	NOUN
easat-5690	336	35	993244	993244	NUM
easat-5690	336	36	,	,	PUNCT
easat-5690	336	37	2022	2022	NUM
easat-5690	336	38	.	.	PUNCT
easat-5690	337	1	https://doi.org/10.3389/fpls.2022.993244	https://doi.org/10.3389/fpls.2022.993244	PROPN
easat-5690	338	1	[	[	X
easat-5690	338	2	29	29	NUM
easat-5690	338	3	]	]	PUNCT
easat-5690	338	4	x.	x.	NOUN
easat-5690	338	5	song	song	PROPN
easat-5690	338	6	et	et	PROPN
easat-5690	338	7	al	al	PROPN
easat-5690	338	8	.	.	PROPN
easat-5690	338	9	,	,	PUNCT
easat-5690	338	10	"	"	PUNCT
easat-5690	338	11	real	real	ADJ
easat-5690	338	12	-	-	PUNCT
easat-5690	338	13	time	time	NOUN
easat-5690	338	14	determination	determination	NOUN
easat-5690	338	15	of	of	ADP
easat-5690	338	16	flowering	flower	VERB
easat-5690	338	17	period	period	NOUN
easat-5690	338	18	for	for	ADP
easat-5690	338	19	field	field	NOUN
easat-5690	338	20	wheat	wheat	NOUN
easat-5690	338	21	based	base	VERB
easat-5690	338	22	on	on	ADP
easat-5690	338	23	improved	improved	ADJ
easat-5690	338	24	yolov5s	yolov5s	PROPN
easat-5690	338	25	model	model	NOUN
easat-5690	338	26	,	,	PUNCT
easat-5690	338	27	"	"	PUNCT
easat-5690	338	28	frontiers	frontier	NOUN
easat-5690	338	29	in	in	ADP
easat-5690	338	30	plant	plant	NOUN
easat-5690	338	31	science	science	NOUN
easat-5690	338	32	,	,	PUNCT
easat-5690	338	33	vol	vol	NOUN
easat-5690	338	34	.	.	PROPN
easat-5690	338	35	13	13	NUM
easat-5690	338	36	,	,	PUNCT
easat-5690	338	37	p.	p.	NOUN
easat-5690	338	38	1025663	1025663	NUM
easat-5690	338	39	,	,	PUNCT
easat-5690	338	40	2023	2023	NUM
easat-5690	338	41	.	.	PUNCT
easat-5690	339	1	https://doi.org/10.3389/fpls.2022.1025663	https://doi.org/10.3389/fpls.2022.1025663	PROPN
easat-5690	340	1	[	[	X
easat-5690	340	2	30	30	NUM
easat-5690	340	3	]	]	X
easat-5690	340	4	y.	y.	PROPN
easat-5690	340	5	zhao	zhao	PROPN
easat-5690	340	6	,	,	PUNCT
easat-5690	340	7	y.	y.	PROPN
easat-5690	340	8	yang	yang	PROPN
easat-5690	340	9	,	,	PUNCT
easat-5690	340	10	x.	x.	PROPN
easat-5690	340	11	xu	xu	PROPN
easat-5690	340	12	,	,	PUNCT
easat-5690	340	13	and	and	CCONJ
easat-5690	340	14	c.	c.	PROPN
easat-5690	340	15	sun	sun	PROPN
easat-5690	340	16	,	,	PUNCT
easat-5690	340	17	"	"	PUNCT
easat-5690	340	18	precision	precision	NOUN
easat-5690	340	19	detection	detection	NOUN
easat-5690	340	20	of	of	ADP
easat-5690	340	21	crop	crop	NOUN
easat-5690	340	22	diseases	disease	NOUN
easat-5690	340	23	based	base	VERB
easat-5690	340	24	on	on	ADP
easat-5690	340	25	improved	improved	ADJ
easat-5690	340	26	yolov5	yolov5	NOUN
easat-5690	340	27	model	model	NOUN
easat-5690	340	28	,	,	PUNCT
easat-5690	340	29	"	"	PUNCT
easat-5690	340	30	frontiers	frontier	NOUN
easat-5690	340	31	in	in	ADP
easat-5690	340	32	plant	plant	NOUN
easat-5690	340	33	science	science	NOUN
easat-5690	340	34	,	,	PUNCT
easat-5690	340	35	vol	vol	NOUN
easat-5690	340	36	.	.	PROPN
easat-5690	340	37	13	13	NUM
easat-5690	340	38	,	,	PUNCT
easat-5690	340	39	p.	p.	NOUN
easat-5690	340	40	1066835	1066835	NUM
easat-5690	340	41	,	,	PUNCT
easat-5690	340	42	2023	2023	NUM
easat-5690	340	43	.	.	PUNCT
easat-5690	341	1	https://doi.org/10.3389/fpls.2022.1066835	https://doi.org/10.3389/fpls.2022.1066835	PROPN
easat-5690	341	2	[	[	X
easat-5690	341	3	31	31	NUM
easat-5690	341	4	]	]	PUNCT
easat-5690	341	5	a.	a.	PROPN
easat-5690	341	6	wang	wang	PROPN
easat-5690	341	7	,	,	PUNCT
easat-5690	341	8	t.	t.	PROPN
easat-5690	341	9	peng	peng	PROPN
easat-5690	341	10	,	,	PUNCT
easat-5690	341	11	h.	h.	PROPN
easat-5690	341	12	cao	cao	PROPN
easat-5690	341	13	,	,	PUNCT
easat-5690	341	14	y.	y.	PROPN
easat-5690	341	15	xu	xu	PROPN
easat-5690	341	16	,	,	PUNCT
easat-5690	341	17	x.	x.	PROPN
easat-5690	341	18	wei	wei	PROPN
easat-5690	341	19	,	,	PUNCT
easat-5690	341	20	and	and	CCONJ
easat-5690	341	21	b.	b.	PROPN
easat-5690	341	22	cui	cui	PROPN
easat-5690	341	23	,	,	PUNCT
easat-5690	341	24	"	"	PUNCT
easat-5690	341	25	tia	tia	PROPN
easat-5690	341	26	-	-	PUNCT
easat-5690	341	27	yolov5	yolov5	PROPN
easat-5690	341	28	:	:	PUNCT
easat-5690	341	29	an	an	DET
easat-5690	341	30	improved	improved	ADJ
easat-5690	341	31	yolov5	yolov5	NOUN
easat-5690	341	32	network	network	NOUN
easat-5690	341	33	for	for	ADP
easat-5690	341	34	real	real	ADJ
easat-5690	341	35	-	-	PUNCT
easat-5690	341	36	time	time	NOUN
easat-5690	341	37	detection	detection	NOUN
easat-5690	341	38	of	of	ADP
easat-5690	341	39	crop	crop	NOUN
easat-5690	341	40	and	and	CCONJ
easat-5690	341	41	weed	weed	NOUN
easat-5690	341	42	in	in	ADP
easat-5690	341	43	the	the	DET
easat-5690	341	44	field	field	NOUN
easat-5690	341	45	,	,	PUNCT
easat-5690	341	46	"	"	PUNCT
easat-5690	341	47	frontiers	frontier	NOUN
easat-5690	341	48	in	in	ADP
easat-5690	341	49	plant	plant	NOUN
easat-5690	341	50	science	science	NOUN
easat-5690	341	51	,	,	PUNCT
easat-5690	341	52	vol	vol	NOUN
easat-5690	341	53	.	.	PROPN
easat-5690	341	54	13	13	NUM
easat-5690	341	55	,	,	PUNCT
easat-5690	341	56	p.	p.	NOUN
easat-5690	341	57	1091655	1091655	NUM
easat-5690	341	58	,	,	PUNCT
easat-5690	341	59	2022	2022	NUM
easat-5690	341	60	.	.	PUNCT
easat-5690	342	1	https://doi.org/10.3389/fpls.2022.1091655	https://doi.org/10.3389/fpls.2022.1091655	PROPN
easat-5690	343	1	[	[	X
easat-5690	343	2	32	32	NUM
easat-5690	343	3	]	]	PUNCT
easat-5690	343	4	h.	h.	PROPN
easat-5690	343	5	he	he	PRON
easat-5690	343	6	,	,	PUNCT
easat-5690	343	7	x.	x.	PROPN
easat-5690	343	8	ma	ma	PROPN
easat-5690	343	9	,	,	PUNCT
easat-5690	343	10	h.	h.	PROPN
easat-5690	343	11	guan	guan	PROPN
easat-5690	343	12	,	,	PUNCT
easat-5690	343	13	f.	f.	PROPN
easat-5690	343	14	wang	wang	PROPN
easat-5690	343	15	,	,	PUNCT
easat-5690	343	16	and	and	CCONJ
easat-5690	343	17	p.	p.	PROPN
easat-5690	343	18	shen	shen	NOUN
easat-5690	343	19	,	,	PUNCT
easat-5690	343	20	"	"	PUNCT
easat-5690	343	21	recognition	recognition	NOUN
easat-5690	343	22	of	of	ADP
easat-5690	343	23	soybean	soybean	NOUN
easat-5690	343	24	pods	pod	NOUN
easat-5690	343	25	and	and	CCONJ
easat-5690	343	26	yield	yield	NOUN
easat-5690	343	27	prediction	prediction	NOUN
easat-5690	343	28	based	base	VERB
easat-5690	343	29	on	on	ADP
easat-5690	343	30	improved	improved	ADJ
easat-5690	343	31	deep	deep	ADJ
easat-5690	343	32	learning	learning	NOUN
easat-5690	343	33	model	model	NOUN
easat-5690	343	34	,	,	PUNCT
easat-5690	343	35	"	"	PUNCT
easat-5690	343	36	frontiers	frontier	NOUN
easat-5690	343	37	in	in	ADP
easat-5690	343	38	plant	plant	NOUN
easat-5690	343	39	science	science	NOUN
easat-5690	343	40	,	,	PUNCT
easat-5690	343	41	vol	vol	NOUN
easat-5690	343	42	.	.	PROPN
easat-5690	343	43	13	13	NUM
easat-5690	343	44	,	,	PUNCT
easat-5690	343	45	p.	p.	NOUN
easat-5690	343	46	1096619	1096619	NUM
easat-5690	343	47	,	,	PUNCT
easat-5690	343	48	2023	2023	NUM
easat-5690	343	49	.	.	PUNCT
easat-5690	344	1	https://doi.org/10.3389/fpls.2022.1096619	https://doi.org/10.3389/fpls.2022.1096619	X
easat-5690	345	1	[	[	X
easat-5690	345	2	33	33	NUM
easat-5690	345	3	]	]	PUNCT
easat-5690	345	4	s.	s.	PROPN
easat-5690	345	5	zhu	zhu	PROPN
easat-5690	345	6	,	,	PUNCT
easat-5690	345	7	w.	w.	PROPN
easat-5690	345	8	ma	ma	PROPN
easat-5690	345	9	,	,	PUNCT
easat-5690	345	10	j.	j.	PROPN
easat-5690	345	11	wang	wang	PROPN
easat-5690	345	12	,	,	PUNCT
easat-5690	345	13	m.	m.	PROPN
easat-5690	345	14	yang	yang	PROPN
easat-5690	345	15	,	,	PUNCT
easat-5690	345	16	y.	y.	PROPN
easat-5690	345	17	wang	wang	PROPN
easat-5690	345	18	,	,	PUNCT
easat-5690	345	19	and	and	CCONJ
easat-5690	345	20	c.	c.	PROPN
easat-5690	345	21	wang	wang	PROPN
easat-5690	345	22	,	,	PUNCT
easat-5690	345	23	"	"	PUNCT
easat-5690	345	24	eadd	eadd	NOUN
easat-5690	345	25	-	-	PUNCT
easat-5690	345	26	yolo	yolo	NOUN
easat-5690	345	27	:	:	PUNCT
easat-5690	345	28	an	an	DET
easat-5690	345	29	efficient	efficient	ADJ
easat-5690	345	30	and	and	CCONJ
easat-5690	345	31	accurate	accurate	ADJ
easat-5690	345	32	disease	disease	NOUN
easat-5690	345	33	detector	detector	NOUN
easat-5690	345	34	for	for	ADP
easat-5690	345	35	apple	apple	NOUN
easat-5690	345	36	leaf	leaf	NOUN
easat-5690	345	37	using	use	VERB
easat-5690	345	38	improved	improve	VERB
easat-5690	345	39	lightweight	lightweight	ADJ
easat-5690	345	40	yolov5	yolov5	NOUN
easat-5690	345	41	,	,	PUNCT
easat-5690	345	42	"	"	PUNCT
easat-5690	345	43	frontiers	frontier	NOUN
easat-5690	345	44	in	in	ADP
easat-5690	345	45	plant	plant	NOUN
easat-5690	345	46	science	science	NOUN
easat-5690	345	47	,	,	PUNCT
easat-5690	345	48	vol	vol	NOUN
easat-5690	345	49	.	.	PROPN
easat-5690	345	50	14	14	NUM
easat-5690	345	51	,	,	PUNCT
easat-5690	345	52	p.	p.	NOUN
easat-5690	345	53	1120724	1120724	NUM
easat-5690	345	54	,	,	PUNCT
easat-5690	345	55	2023	2023	NUM
easat-5690	345	56	.	.	PUNCT
easat-5690	346	1	https://doi.org/10.3389/fpls.2023.1120724	https://doi.org/10.3389/fpls.2023.1120724	PROPN
easat-5690	346	2	[	[	X
easat-5690	346	3	34	34	NUM
easat-5690	346	4	]	]	PUNCT
easat-5690	346	5	j.	j.	PROPN
easat-5690	346	6	lin	lin	PROPN
easat-5690	346	7	et	et	PROPN
easat-5690	346	8	al	al	PROPN
easat-5690	346	9	.	.	PROPN
easat-5690	346	10	,	,	PUNCT
easat-5690	346	11	"	"	PUNCT
easat-5690	346	12	improved	improved	ADJ
easat-5690	346	13	yolox	yolox	NOUN
easat-5690	346	14	-	-	PUNCT
easat-5690	346	15	tiny	tiny	ADJ
easat-5690	346	16	network	network	NOUN
easat-5690	346	17	for	for	ADP
easat-5690	346	18	detection	detection	NOUN
easat-5690	346	19	of	of	ADP
easat-5690	346	20	tobacco	tobacco	NOUN
easat-5690	346	21	brown	brown	ADJ
easat-5690	346	22	spot	spot	NOUN
easat-5690	346	23	disease	disease	NOUN
easat-5690	346	24	,	,	PUNCT
easat-5690	346	25	"	"	PUNCT
easat-5690	346	26	frontiers	frontier	NOUN
easat-5690	346	27	in	in	ADP
easat-5690	346	28	plant	plant	NOUN
easat-5690	346	29	science	science	NOUN
easat-5690	346	30	,	,	PUNCT
easat-5690	346	31	vol	vol	NOUN
easat-5690	346	32	.	.	PROPN
easat-5690	346	33	14	14	NUM
easat-5690	346	34	,	,	PUNCT
easat-5690	346	35	p.	p.	NOUN
easat-5690	346	36	1135105	1135105	NUM
easat-5690	346	37	,	,	PUNCT
easat-5690	346	38	2023	2023	NUM
easat-5690	346	39	.	.	PUNCT
easat-5690	347	1	https://doi.org/10.3389/fpls.2023.1135105	https://doi.org/10.3389/fpls.2023.1135105	PROPN
easat-5690	347	2	[	[	X
easat-5690	347	3	35	35	NUM
easat-5690	347	4	]	]	X
easat-5690	347	5	o.	o.	PROPN
easat-5690	347	6	m.	m.	PROPN
easat-5690	347	7	lawal	lawal	PROPN
easat-5690	347	8	,	,	PUNCT
easat-5690	347	9	s.	s.	PROPN
easat-5690	347	10	zhu	zhu	PROPN
easat-5690	347	11	,	,	PUNCT
easat-5690	347	12	and	and	CCONJ
easat-5690	347	13	k.	k.	PROPN
easat-5690	347	14	cheng	cheng	PROPN
easat-5690	347	15	,	,	PUNCT
easat-5690	347	16	"	"	PUNCT
easat-5690	347	17	an	an	DET
easat-5690	347	18	improved	improved	ADJ
easat-5690	347	19	yolov5s	yolov5s	NOUN
easat-5690	347	20	model	model	NOUN
easat-5690	347	21	using	use	VERB
easat-5690	347	22	feature	feature	NOUN
easat-5690	347	23	concatenation	concatenation	NOUN
easat-5690	347	24	with	with	ADP
easat-5690	347	25	attention	attention	NOUN
easat-5690	347	26	mechanism	mechanism	NOUN
easat-5690	347	27	for	for	ADP
easat-5690	347	28	real	real	ADJ
easat-5690	347	29	-	-	PUNCT
easat-5690	347	30	time	time	NOUN
easat-5690	347	31	fruit	fruit	NOUN
easat-5690	347	32	detection	detection	NOUN
easat-5690	347	33	and	and	CCONJ
easat-5690	347	34	counting	counting	NOUN
easat-5690	347	35	,	,	PUNCT
easat-5690	347	36	"	"	PUNCT
easat-5690	347	37	frontiers	frontier	NOUN
easat-5690	347	38	in	in	ADP
easat-5690	347	39	plant	plant	NOUN
easat-5690	347	40	science	science	NOUN
easat-5690	347	41	,	,	PUNCT
easat-5690	347	42	vol	vol	NOUN
easat-5690	347	43	.	.	PROPN
easat-5690	347	44	14	14	NUM
easat-5690	347	45	,	,	PUNCT
easat-5690	347	46	2023	2023	NUM
easat-5690	347	47	.	.	PUNCT
easat-5690	348	1	https://doi.org/10.3389/fpls.2023.1153505	https://doi.org/10.3389/fpls.2023.1153505	PROPN
easat-5690	348	2	[	[	X
easat-5690	348	3	36	36	NUM
easat-5690	348	4	]	]	X
easat-5690	348	5	d.	d.	PROPN
easat-5690	348	6	tzutalin	tzutalin	PROPN
easat-5690	348	7	,	,	PUNCT
easat-5690	348	8	"	"	PUNCT
easat-5690	348	9	tzutalin	tzutalin	PROPN
easat-5690	348	10	/	/	SYM
easat-5690	348	11	labelimg	labelimg	NOUN
easat-5690	348	12	,	,	PUNCT
easat-5690	348	13	"	"	PUNCT
easat-5690	348	14	github	github	NOUN
easat-5690	348	15	,	,	PUNCT
easat-5690	348	16	2015	2015	NUM
easat-5690	348	17	.	.	PUNCT
easat-5690	349	1	https://doi.org/10.1016/j.compag.2018.01.009	https://doi.org/10.1016/j.compag.2018.01.009	X
easat-5690	349	2	https://doi.org/10.1145/3377049.3377122	https://doi.org/10.1145/3377049.3377122	NOUN
easat-5690	349	3	https://doi.org/10.13031/aim.201801254	https://doi.org/10.13031/aim.201801254	PROPN
easat-5690	349	4	https://doi.org/10.1109/access.2020.2982456	https://doi.org/10.1109/access.2020.2982456	PROPN
easat-5690	349	5	https://doi.org/10.1016/j.compeleceng.2021.107023	https://doi.org/10.1016/j.compeleceng.2021.107023	NOUN
easat-5690	349	6	https://doi.org/10.1007/978-3-030-72073-5_3	https://doi.org/10.1007/978-3-030-72073-5_3	ADP
easat-5690	349	7	https://doi.org/10.3390/f14030619	https://doi.org/10.3390/f14030619	NUM
easat-5690	349	8	https://doi.org/10.3390/f14020415	https://doi.org/10.3390/f14020415	PRON
easat-5690	349	9	https://doi.org/10.1145/3561613.3561626	https://doi.org/10.1145/3561613.3561626	NOUN
easat-5690	349	10	https://doi.org/10.3390/rs13091619	https://doi.org/10.3390/rs13091619	X
easat-5690	349	11	https://doi.org/10.3390/agronomy12020365	https://doi.org/10.3390/agronomy12020365	PROPN
easat-5690	349	12	https://doi.org/10.3390/f14051012	https://doi.org/10.3390/f14051012	PROPN
easat-5690	349	13	https://doi.org/10.1016/j.compag.2022.106780	https://doi.org/10.1016/j.compag.2022.106780	NOUN
easat-5690	349	14	https://doi.org/10.3390/electronics11111673	https://doi.org/10.3390/electronics11111673	PROPN
easat-5690	349	15	https://doi.org/10.3389/fpls.2022.965425	https://doi.org/10.3389/fpls.2022.965425	PROPN
easat-5690	349	16	https://doi.org/10.3389/fpls.2022.991929	https://doi.org/10.3389/fpls.2022.991929	PROPN
easat-5690	349	17	https://doi.org/10.3389/fpls.2022.993244	https://doi.org/10.3389/fpls.2022.993244	PROPN
easat-5690	349	18	https://doi.org/10.3389/fpls.2022.1025663	https://doi.org/10.3389/fpls.2022.1025663	PROPN
easat-5690	349	19	https://doi.org/10.3389/fpls.2022.1066835	https://doi.org/10.3389/fpls.2022.1066835	PROPN
easat-5690	349	20	https://doi.org/10.3389/fpls.2022.1091655	https://doi.org/10.3389/fpls.2022.1091655	PROPN
easat-5690	349	21	https://doi.org/10.3389/fpls.2022.1096619	https://doi.org/10.3389/fpls.2022.1096619	PROPN
easat-5690	350	1	https://doi.org/10.3389/fpls.2023.1120724	https://doi.org/10.3389/fpls.2023.1120724	PROPN
easat-5690	350	2	https://doi.org/10.3389/fpls.2023.1135105	https://doi.org/10.3389/fpls.2023.1135105	PROPN
easat-5690	350	3	https://doi.org/10.3389/fpls.2023.1153505	https://doi.org/10.3389/fpls.2023.1153505	PROPN
easat-5690	350	4	1800	1800	NUM
easat-5690	350	5	edelweiss	edelweiss	PROPN
easat-5690	350	6	applied	apply	VERB
easat-5690	350	7	science	science	NOUN
easat-5690	350	8	and	and	CCONJ
easat-5690	350	9	technology	technology	NOUN
easat-5690	350	10	issn	issn	PROPN
easat-5690	350	11	:	:	PUNCT
easat-5690	350	12	2576	2576	NUM
easat-5690	350	13	-	-	SYM
easat-5690	350	14	8484	8484	NUM
easat-5690	350	15	vol	vol	NOUN
easat-5690	350	16	.	.	PROPN
easat-5690	351	1	9	9	NUM
easat-5690	351	2	,	,	PUNCT
easat-5690	351	3	no	no	INTJ
easat-5690	351	4	.	.	NOUN
easat-5690	352	1	3	3	NUM
easat-5690	352	2	:	:	SYM
easat-5690	352	3	1777	1777	NUM
easat-5690	352	4	-	-	SYM
easat-5690	352	5	1794	1794	NUM
easat-5690	352	6	,	,	PUNCT
easat-5690	352	7	2025	2025	NUM
easat-5690	352	8	doi	doi	NOUN
easat-5690	352	9	:	:	PUNCT
easat-5690	352	10	10.55214/25768484.v9i3.5690	10.55214/25768484.v9i3.5690	NUM
easat-5690	352	11	©	©	PROPN
easat-5690	352	12	2025	2025	NUM
easat-5690	352	13	by	by	ADP
easat-5690	352	14	the	the	DET
easat-5690	352	15	authors	author	NOUN
easat-5690	352	16	;	;	PUNCT
easat-5690	352	17	licensee	licensee	PROPN
easat-5690	352	18	learning	learning	NOUN
easat-5690	352	19	gate	gate	NOUN
easat-5690	353	1	[	[	X
easat-5690	353	2	37	37	NUM
easat-5690	353	3	]	]	PUNCT
easat-5690	353	4	j.	j.	PROPN
easat-5690	353	5	qian	qian	PROPN
easat-5690	353	6	and	and	CCONJ
easat-5690	353	7	h.	h.	PROPN
easat-5690	353	8	lin	lin	PROPN
easat-5690	353	9	,	,	PUNCT
easat-5690	353	10	"	"	PUNCT
easat-5690	353	11	a	a	DET
easat-5690	353	12	forest	forest	NOUN
easat-5690	353	13	fire	fire	NOUN
easat-5690	353	14	identification	identification	NOUN
easat-5690	353	15	system	system	NOUN
easat-5690	353	16	based	base	VERB
easat-5690	353	17	on	on	ADP
easat-5690	353	18	weighted	weight	VERB
easat-5690	353	19	fusion	fusion	NOUN
easat-5690	353	20	algorithm	algorithm	NOUN
easat-5690	353	21	,	,	PUNCT
easat-5690	353	22	"	"	PUNCT
easat-5690	353	23	forests	forest	NOUN
easat-5690	353	24	,	,	PUNCT
easat-5690	353	25	vol	vol	NOUN
easat-5690	353	26	.	.	PROPN
easat-5690	353	27	13	13	NUM
easat-5690	353	28	,	,	PUNCT
easat-5690	353	29	no	no	INTJ
easat-5690	353	30	.	.	NOUN
easat-5690	353	31	8	8	NUM
easat-5690	353	32	,	,	PUNCT
easat-5690	353	33	p.	p.	NOUN
easat-5690	353	34	1301	1301	NUM
easat-5690	353	35	,	,	PUNCT
easat-5690	353	36	2022	2022	NUM
easat-5690	354	1	.	.	PUNCT
easat-5690	355	1	https://doi.org/10.3390/f13081301	https://doi.org/10.3390/f13081301	PRON
easat-5690	355	2	[	[	X
easat-5690	355	3	38	38	NUM
easat-5690	355	4	]	]	PUNCT
easat-5690	355	5	k.	k.	PROPN
easat-5690	355	6	s.	s.	PROPN
easat-5690	355	7	kouassi	kouassi	PROPN
easat-5690	355	8	,	,	PUNCT
easat-5690	355	9	m.	m.	PROPN
easat-5690	355	10	diarra	diarra	PROPN
easat-5690	355	11	,	,	PUNCT
easat-5690	355	12	e.	e.	PROPN
easat-5690	355	13	kouassi	kouassi	PROPN
easat-5690	355	14	hilaire	hilaire	PROPN
easat-5690	355	15	,	,	PUNCT
easat-5690	355	16	and	and	CCONJ
easat-5690	355	17	k.	k.	PROPN
easat-5690	355	18	b.	b.	PROPN
easat-5690	355	19	jean	jean	PROPN
easat-5690	355	20	-	-	PUNCT
easat-5690	355	21	claude	claude	PROPN
easat-5690	355	22	,	,	PUNCT
easat-5690	355	23	"	"	PUNCT
easat-5690	355	24	detection	detection	NOUN
easat-5690	355	25	of	of	ADP
easat-5690	355	26	cocoa	cocoa	NOUN
easat-5690	355	27	pod	pod	NOUN
easat-5690	355	28	diseases	disease	NOUN
easat-5690	355	29	using	use	VERB
easat-5690	355	30	a	a	DET
easat-5690	355	31	hybrid	hybrid	ADJ
easat-5690	355	32	feature	feature	NOUN
easat-5690	355	33	extractor	extractor	NOUN
easat-5690	355	34	combining	combine	VERB
easat-5690	355	35	cnn	cnn	PROPN
easat-5690	355	36	and	and	CCONJ
easat-5690	355	37	vision	vision	NOUN
easat-5690	355	38	transformer	transformer	NOUN
easat-5690	355	39	with	with	ADP
easat-5690	355	40	dual	dual	ADJ
easat-5690	355	41	classifier	classifier	NOUN
easat-5690	355	42	,	,	PUNCT
easat-5690	355	43	"	"	PUNCT
easat-5690	355	44	edelweiss	edelweiss	PROPN
easat-5690	355	45	applied	apply	VERB
easat-5690	355	46	science	science	NOUN
easat-5690	355	47	and	and	CCONJ
easat-5690	355	48	technology	technology	NOUN
easat-5690	355	49	,	,	PUNCT
easat-5690	355	50	vol	vol	NOUN
easat-5690	355	51	.	.	PROPN
easat-5690	355	52	9	9	NUM
easat-5690	355	53	,	,	PUNCT
easat-5690	355	54	no	no	INTJ
easat-5690	355	55	.	.	NOUN
easat-5690	355	56	1	1	NUM
easat-5690	355	57	,	,	PUNCT
easat-5690	355	58	pp	pp	ADJ
easat-5690	355	59	.	.	PUNCT
easat-5690	356	1	668	668	NUM
easat-5690	356	2	-	-	SYM
easat-5690	356	3	681	681	NUM
easat-5690	356	4	,	,	PUNCT
easat-5690	356	5	2025	2025	NUM
easat-5690	356	6	.	.	PUNCT
easat-5690	357	1	https://doi.org/10.55214/25768484.v9i1.4209	https://doi.org/10.55214/25768484.v9i1.4209	NOUN
easat-5690	358	1	[	[	X
easat-5690	358	2	39	39	NUM
easat-5690	358	3	]	]	PUNCT
easat-5690	358	4	s.	s.	PROPN
easat-5690	358	5	ozbay	ozbay	VERB
easat-5690	358	6	and	and	CCONJ
easat-5690	358	7	o.	o.	PROPN
easat-5690	358	8	tunc	tunc	PROPN
easat-5690	358	9	,	,	PUNCT
easat-5690	358	10	"	"	PUNCT
easat-5690	358	11	deep	deep	ADJ
easat-5690	358	12	learning	learning	NOUN
easat-5690	358	13	in	in	ADP
easat-5690	358	14	analysing	analyse	VERB
easat-5690	358	15	paranasal	paranasal	ADJ
easat-5690	358	16	sinuses	sinus	NOUN
easat-5690	358	17	,	,	PUNCT
easat-5690	358	18	"	"	PUNCT
easat-5690	358	19	elektronika	elektronika	PROPN
easat-5690	358	20	ir	ir	PROPN
easat-5690	358	21	elektrotechnika	elektrotechnika	NOUN
easat-5690	358	22	,	,	PUNCT
easat-5690	358	23	vol	vol	NOUN
easat-5690	358	24	.	.	PROPN
easat-5690	358	25	28	28	NUM
easat-5690	358	26	,	,	PUNCT
easat-5690	358	27	no	no	INTJ
easat-5690	358	28	.	.	NOUN
easat-5690	358	29	3	3	NUM
easat-5690	358	30	,	,	PUNCT
easat-5690	358	31	pp	pp	ADJ
easat-5690	358	32	.	.	PUNCT
easat-5690	359	1	65	65	NUM
easat-5690	359	2	-	-	SYM
easat-5690	359	3	70	70	NUM
easat-5690	359	4	,	,	PUNCT
easat-5690	359	5	2022	2022	NUM
easat-5690	359	6	.	.	PUNCT
easat-5690	360	1	https://doi.org/10.5755/j02.eie.31133	https://doi.org/10.5755/j02.eie.31133	PROPN
easat-5690	360	2	https://doi.org/10.3390/f13081301	https://doi.org/10.3390/f13081301	NUM
easat-5690	360	3	https://doi.org/10.55214/25768484.v9i1.4209	https://doi.org/10.55214/25768484.v9i1.4209	NOUN
easat-5690	360	4	https://doi.org/10.5755/j02.eie.31133	https://doi.org/10.5755/j02.eie.31133	PROPN