id	sid	tid	token	lemma	pos
ajst-16789	1	1	academic	academic	ADJ
ajst-16789	1	2	journal	journal	NOUN
ajst-16789	1	3	of	of	ADP
ajst-16789	1	4	science	science	NOUN
ajst-16789	1	5	and	and	CCONJ
ajst-16789	1	6	technology	technology	NOUN
ajst-16789	1	7	issn	issn	NOUN
ajst-16789	1	8	:	:	PUNCT
ajst-16789	1	9	2771	2771	NUM
ajst-16789	1	10	-	-	SYM
ajst-16789	1	11	3032	3032	NUM
ajst-16789	1	12	|	|	NOUN
ajst-16789	1	13	vol	vol	NOUN
ajst-16789	1	14	.	.	PROPN
ajst-16789	2	1	9	9	NUM
ajst-16789	2	2	,	,	PUNCT
ajst-16789	2	3	no	no	INTJ
ajst-16789	2	4	.	.	NOUN
ajst-16789	2	5	1	1	NUM
ajst-16789	2	6	,	,	PUNCT
ajst-16789	2	7	2024	2024	NUM
ajst-16789	2	8	198	198	NUM
ajst-16789	2	9	the	the	DET
ajst-16789	2	10	study	study	NOUN
ajst-16789	2	11	on	on	ADP
ajst-16789	2	12	recognition	recognition	NOUN
ajst-16789	2	13	and	and	CCONJ
ajst-16789	2	14	detection	detection	NOUN
ajst-16789	2	15	of	of	ADP
ajst-16789	2	16	express	express	ADJ
ajst-16789	2	17	package	package	NOUN
ajst-16789	2	18	grabbing	grabbing	NOUN
ajst-16789	2	19	based	base	VERB
ajst-16789	2	20	on	on	ADP
ajst-16789	2	21	machine	machine	NOUN
ajst-16789	2	22	vision	vision	PROPN
ajst-16789	2	23	weixiang	weixiang	PROPN
ajst-16789	2	24	qi	qi	PROPN
ajst-16789	2	25	,	,	PUNCT
ajst-16789	2	26	xiangguo	xiangguo	PROPN
ajst-16789	2	27	sun	sun	PROPN
ajst-16789	2	28	school	school	NOUN
ajst-16789	2	29	of	of	ADP
ajst-16789	2	30	mechanical	mechanical	ADJ
ajst-16789	2	31	engineering	engineering	NOUN
ajst-16789	2	32	,	,	PUNCT
ajst-16789	2	33	sichuan	sichuan	PROPN
ajst-16789	2	34	university	university	PROPN
ajst-16789	2	35	of	of	ADP
ajst-16789	2	36	science	science	PROPN
ajst-16789	2	37	&	&	CCONJ
ajst-16789	2	38	engineering	engineering	PROPN
ajst-16789	2	39	,	,	PUNCT
ajst-16789	2	40	yibin	yibin	PROPN
ajst-16789	2	41	,	,	PUNCT
ajst-16789	2	42	644002	644002	NUM
ajst-16789	2	43	,	,	PUNCT
ajst-16789	2	44	sichuan	sichuan	PROPN
ajst-16789	2	45	,	,	PUNCT
ajst-16789	2	46	china	china	PROPN
ajst-16789	2	47	abstract	abstract	PROPN
ajst-16789	2	48	:	:	PUNCT
ajst-16789	2	49	this	this	DET
ajst-16789	2	50	research	research	NOUN
ajst-16789	2	51	addresses	address	VERB
ajst-16789	2	52	the	the	DET
ajst-16789	2	53	automated	automate	VERB
ajst-16789	2	54	supply	supply	NOUN
ajst-16789	2	55	of	of	ADP
ajst-16789	2	56	parcels	parcel	NOUN
ajst-16789	2	57	in	in	ADP
ajst-16789	2	58	logistics	logistic	NOUN
ajst-16789	2	59	by	by	ADP
ajst-16789	2	60	employing	employ	VERB
ajst-16789	2	61	a	a	DET
ajst-16789	2	62	robotic	robotic	ADJ
ajst-16789	2	63	arm	arm	NOUN
ajst-16789	2	64	to	to	PART
ajst-16789	2	65	replace	replace	VERB
ajst-16789	2	66	manual	manual	ADJ
ajst-16789	2	67	intervention	intervention	NOUN
ajst-16789	2	68	.	.	PUNCT
ajst-16789	3	1	the	the	DET
ajst-16789	3	2	process	process	NOUN
ajst-16789	3	3	involves	involve	VERB
ajst-16789	3	4	the	the	DET
ajst-16789	3	5	suction	suction	NOUN
ajst-16789	3	6	-	-	PUNCT
ajst-16789	3	7	based	base	VERB
ajst-16789	3	8	retrieval	retrieval	NOUN
ajst-16789	3	9	of	of	ADP
ajst-16789	3	10	disorderedly	disorderedly	ADV
ajst-16789	3	11	stacked	stack	VERB
ajst-16789	3	12	parcels	parcel	NOUN
ajst-16789	3	13	and	and	CCONJ
ajst-16789	3	14	their	their	PRON
ajst-16789	3	15	placement	placement	NOUN
ajst-16789	3	16	onto	onto	ADP
ajst-16789	3	17	a	a	DET
ajst-16789	3	18	conveyor	conveyor	NOUN
ajst-16789	3	19	belt	belt	NOUN
ajst-16789	3	20	,	,	PUNCT
ajst-16789	3	21	presenting	present	VERB
ajst-16789	3	22	an	an	DET
ajst-16789	3	23	economically	economically	ADV
ajst-16789	3	24	significant	significant	ADJ
ajst-16789	3	25	engineering	engineering	NOUN
ajst-16789	3	26	challenge	challenge	NOUN
ajst-16789	3	27	.	.	PUNCT
ajst-16789	4	1	this	this	DET
ajst-16789	4	2	paper	paper	NOUN
ajst-16789	4	3	designs	design	VERB
ajst-16789	4	4	a	a	DET
ajst-16789	4	5	visual	visual	ADJ
ajst-16789	4	6	recognition	recognition	NOUN
ajst-16789	4	7	system	system	NOUN
ajst-16789	4	8	to	to	PART
ajst-16789	4	9	assist	assist	VERB
ajst-16789	4	10	the	the	DET
ajst-16789	4	11	robotic	robotic	ADJ
ajst-16789	4	12	arm	arm	NOUN
ajst-16789	4	13	in	in	ADP
ajst-16789	4	14	identifying	identify	VERB
ajst-16789	4	15	the	the	DET
ajst-16789	4	16	grasping	grasp	VERB
ajst-16789	4	17	surface	surface	NOUN
ajst-16789	4	18	of	of	ADP
ajst-16789	4	19	disorderedly	disorderedly	ADV
ajst-16789	4	20	stacked	stack	VERB
ajst-16789	4	21	parcels	parcel	NOUN
ajst-16789	4	22	.	.	PUNCT
ajst-16789	5	1	leveraging	leverage	VERB
ajst-16789	5	2	transfer	transfer	NOUN
ajst-16789	5	3	learning	learning	NOUN
ajst-16789	5	4	,	,	PUNCT
ajst-16789	5	5	a	a	DET
ajst-16789	5	6	pretrained	pretraine	VERB
ajst-16789	5	7	model	model	NOUN
ajst-16789	5	8	constructs	construct	VERB
ajst-16789	5	9	a	a	DET
ajst-16789	5	10	multimodal	multimodal	ADJ
ajst-16789	5	11	network	network	NOUN
ajst-16789	5	12	framework	framework	NOUN
ajst-16789	5	13	.	.	PUNCT
ajst-16789	6	1	pre	pre	VERB
ajst-16789	6	2	-	-	ADJ
ajst-16789	6	3	trained	train	VERB
ajst-16789	6	4	models	model	NOUN
ajst-16789	6	5	of	of	ADP
ajst-16789	6	6	resnet	resnet	NOUN
ajst-16789	6	7	169	169	NUM
ajst-16789	6	8	,	,	PUNCT
ajst-16789	6	9	densenet121	densenet121	PROPN
ajst-16789	6	10	,	,	PUNCT
ajst-16789	6	11	resnet101	resnet101	PROPN
ajst-16789	6	12	,	,	PUNCT
ajst-16789	6	13	and	and	CCONJ
ajst-16789	6	14	resnet50	resnet50	NOUN
ajst-16789	6	15	serve	serve	VERB
ajst-16789	6	16	as	as	ADP
ajst-16789	6	17	backbone	backbone	NOUN
ajst-16789	6	18	networks	network	NOUN
ajst-16789	6	19	to	to	PART
ajst-16789	6	20	train	train	VERB
ajst-16789	6	21	and	and	CCONJ
ajst-16789	6	22	test	test	VERB
ajst-16789	6	23	on	on	ADP
ajst-16789	6	24	custom	custom	NOUN
ajst-16789	6	25	parcel	parcel	NOUN
ajst-16789	6	26	datasets	dataset	NOUN
ajst-16789	6	27	.	.	PUNCT
ajst-16789	7	1	post	post	VERB
ajst-16789	7	2	comparative	comparative	ADJ
ajst-16789	7	3	testing	testing	NOUN
ajst-16789	7	4	of	of	ADP
ajst-16789	7	5	model	model	NOUN
ajst-16789	7	6	performance	performance	NOUN
ajst-16789	7	7	,	,	PUNCT
ajst-16789	7	8	densenet169	densenet169	PROPN
ajst-16789	7	9	is	be	AUX
ajst-16789	7	10	chosen	choose	VERB
ajst-16789	7	11	to	to	PART
ajst-16789	7	12	construct	construct	VERB
ajst-16789	7	13	the	the	DET
ajst-16789	7	14	visual	visual	ADJ
ajst-16789	7	15	recognition	recognition	NOUN
ajst-16789	7	16	network	network	NOUN
ajst-16789	7	17	.	.	PUNCT
ajst-16789	8	1	specifically	specifically	ADV
ajst-16789	8	2	,	,	PUNCT
ajst-16789	8	3	the	the	DET
ajst-16789	8	4	rgb	rgb	PROPN
ajst-16789	8	5	and	and	CCONJ
ajst-16789	8	6	depth	depth	NOUN
ajst-16789	8	7	data	datum	NOUN
ajst-16789	8	8	of	of	ADP
ajst-16789	8	9	parcels	parcel	NOUN
ajst-16789	8	10	are	be	AUX
ajst-16789	8	11	separately	separately	ADV
ajst-16789	8	12	fed	feed	VERB
ajst-16789	8	13	into	into	ADP
ajst-16789	8	14	densenet169	densenet169	PROPN
ajst-16789	8	15	for	for	ADP
ajst-16789	8	16	feature	feature	NOUN
ajst-16789	8	17	extraction	extraction	NOUN
ajst-16789	8	18	.	.	PUNCT
ajst-16789	9	1	post	post	ADJ
ajst-16789	9	2	-	-	ADJ
ajst-16789	9	3	feature	feature	ADJ
ajst-16789	9	4	extraction	extraction	NOUN
ajst-16789	9	5	,	,	PUNCT
ajst-16789	9	6	multimodal	multimodal	ADJ
ajst-16789	9	7	fusion	fusion	NOUN
ajst-16789	9	8	is	be	AUX
ajst-16789	9	9	applied	apply	VERB
ajst-16789	9	10	,	,	PUNCT
ajst-16789	9	11	integrating	integrate	VERB
ajst-16789	9	12	a	a	DET
ajst-16789	9	13	lightweight	lightweight	ADJ
ajst-16789	9	14	attention	attention	NOUN
ajst-16789	9	15	mechanism	mechanism	NOUN
ajst-16789	9	16	,	,	PUNCT
ajst-16789	9	17	cbam	cbam	NOUN
ajst-16789	9	18	,	,	PUNCT
ajst-16789	9	19	to	to	PART
ajst-16789	9	20	enhance	enhance	VERB
ajst-16789	9	21	semantic	semantic	ADJ
ajst-16789	9	22	segmentation	segmentation	NOUN
ajst-16789	9	23	accuracy	accuracy	NOUN
ajst-16789	9	24	.	.	PUNCT
ajst-16789	10	1	subsequently	subsequently	ADV
ajst-16789	10	2	,	,	PUNCT
ajst-16789	10	3	post	post	ADJ
ajst-16789	10	4	-	-	ADJ
ajst-16789	10	5	processing	processing	NOUN
ajst-16789	10	6	of	of	ADP
ajst-16789	10	7	image	image	NOUN
ajst-16789	10	8	features	feature	NOUN
ajst-16789	10	9	filters	filter	VERB
ajst-16789	10	10	out	out	ADP
ajst-16789	10	11	the	the	DET
ajst-16789	10	12	background	background	NOUN
ajst-16789	10	13	,	,	PUNCT
ajst-16789	10	14	achieving	achieve	VERB
ajst-16789	10	15	precise	precise	ADJ
ajst-16789	10	16	identification	identification	NOUN
ajst-16789	10	17	of	of	ADP
ajst-16789	10	18	the	the	DET
ajst-16789	10	19	parcel	parcel	NOUN
ajst-16789	10	20	grasping	grasp	VERB
ajst-16789	10	21	region	region	NOUN
ajst-16789	10	22	.	.	PUNCT
ajst-16789	11	1	ultimately	ultimately	ADV
ajst-16789	11	2	,	,	PUNCT
ajst-16789	11	3	the	the	DET
ajst-16789	11	4	constructed	construct	VERB
ajst-16789	11	5	network	network	NOUN
ajst-16789	11	6	achieves	achieve	VERB
ajst-16789	11	7	an	an	DET
ajst-16789	11	8	average	average	ADJ
ajst-16789	11	9	top1	top1	NOUN
ajst-16789	11	10	true	true	ADJ
ajst-16789	11	11	class	class	NOUN
ajst-16789	11	12	accuracy	accuracy	NOUN
ajst-16789	11	13	of	of	ADP
ajst-16789	11	14	95.86	95.86	NUM
ajst-16789	11	15	%	%	NOUN
ajst-16789	11	16	on	on	ADP
ajst-16789	11	17	the	the	DET
ajst-16789	11	18	test	test	NOUN
ajst-16789	11	19	set	set	NOUN
ajst-16789	11	20	.	.	PUNCT
ajst-16789	12	1	thus	thus	ADV
ajst-16789	12	2	,	,	PUNCT
ajst-16789	12	3	the	the	DET
ajst-16789	12	4	designed	design	VERB
ajst-16789	12	5	parcel	parcel	PROPN
ajst-16789	12	6	visual	visual	ADJ
ajst-16789	12	7	recognition	recognition	NOUN
ajst-16789	12	8	system	system	NOUN
ajst-16789	12	9	based	base	VERB
ajst-16789	12	10	on	on	ADP
ajst-16789	12	11	this	this	DET
ajst-16789	12	12	methodology	methodology	NOUN
ajst-16789	12	13	exhibits	exhibit	VERB
ajst-16789	12	14	robustness	robustness	NOUN
ajst-16789	12	15	,	,	PUNCT
ajst-16789	12	16	meeting	meet	VERB
ajst-16789	12	17	the	the	DET
ajst-16789	12	18	demand	demand	NOUN
ajst-16789	12	19	for	for	ADP
ajst-16789	12	20	autonomous	autonomous	PROPN
ajst-16789	12	21	parcel	parcel	PROPN
ajst-16789	12	22	retrieval	retrieval	NOUN
ajst-16789	12	23	by	by	ADP
ajst-16789	12	24	the	the	DET
ajst-16789	12	25	robotic	robotic	ADJ
ajst-16789	12	26	arm	arm	NOUN
ajst-16789	12	27	.	.	PUNCT
ajst-16789	13	1	it	it	PRON
ajst-16789	13	2	offers	offer	VERB
ajst-16789	13	3	significant	significant	ADJ
ajst-16789	13	4	support	support	NOUN
ajst-16789	13	5	towards	towards	ADP
ajst-16789	13	6	resolving	resolve	VERB
ajst-16789	13	7	challenges	challenge	NOUN
ajst-16789	13	8	in	in	ADP
ajst-16789	13	9	automated	automate	VERB
ajst-16789	13	10	parcel	parcel	NOUN
ajst-16789	13	11	supply	supply	NOUN
ajst-16789	13	12	within	within	ADP
ajst-16789	13	13	logistics	logistic	NOUN
ajst-16789	13	14	.	.	PUNCT
ajst-16789	14	1	keywords	keyword	NOUN
ajst-16789	14	2	:	:	PUNCT
ajst-16789	14	3	resnet	resnet	NOUN
ajst-16789	14	4	,	,	PUNCT
ajst-16789	14	5	densenet	densenet	NOUN
ajst-16789	14	6	,	,	PUNCT
ajst-16789	14	7	deep	deep	ADJ
ajst-16789	14	8	learning	learning	NOUN
ajst-16789	14	9	,	,	PUNCT
ajst-16789	14	10	convolutional	convolutional	ADJ
ajst-16789	14	11	neural	neural	ADJ
ajst-16789	14	12	networks	network	NOUN
ajst-16789	14	13	,	,	PUNCT
ajst-16789	14	14	courier	courier	NOUN
ajst-16789	14	15	parcels	parcel	NOUN
ajst-16789	14	16	,	,	PUNCT
ajst-16789	14	17	visual	visual	ADJ
ajst-16789	14	18	recognition	recognition	NOUN
ajst-16789	14	19	of	of	ADP
ajst-16789	14	20	gripping	grip	VERB
ajst-16789	14	21	surfaces	surface	NOUN
ajst-16789	14	22	.	.	PUNCT
ajst-16789	15	1	1	1	X
ajst-16789	15	2	.	.	X
ajst-16789	15	3	introduction	introduction	NOUN
ajst-16789	15	4	in	in	ADP
ajst-16789	15	5	recent	recent	ADJ
ajst-16789	15	6	years	year	NOUN
ajst-16789	15	7	,	,	PUNCT
ajst-16789	15	8	with	with	ADP
ajst-16789	15	9	the	the	DET
ajst-16789	15	10	rapid	rapid	ADJ
ajst-16789	15	11	development	development	NOUN
ajst-16789	15	12	of	of	ADP
ajst-16789	15	13	e	e	NOUN
ajst-16789	15	14	-	-	NOUN
ajst-16789	15	15	commerce	commerce	NOUN
ajst-16789	15	16	and	and	CCONJ
ajst-16789	15	17	internet	internet	NOUN
ajst-16789	15	18	technology	technology	NOUN
ajst-16789	15	19	,	,	PUNCT
ajst-16789	15	20	the	the	DET
ajst-16789	15	21	volume	volume	NOUN
ajst-16789	15	22	of	of	ADP
ajst-16789	15	23	express	express	ADJ
ajst-16789	15	24	parcels	parcel	NOUN
ajst-16789	15	25	has	have	AUX
ajst-16789	15	26	continued	continue	VERB
ajst-16789	15	27	to	to	PART
ajst-16789	15	28	soar	soar	VERB
ajst-16789	15	29	,	,	PUNCT
ajst-16789	15	30	leading	lead	VERB
ajst-16789	15	31	to	to	ADP
ajst-16789	15	32	an	an	DET
ajst-16789	15	33	urgent	urgent	ADJ
ajst-16789	15	34	demand	demand	NOUN
ajst-16789	15	35	for	for	ADP
ajst-16789	15	36	rapid	rapid	ADJ
ajst-16789	15	37	and	and	CCONJ
ajst-16789	15	38	efficient	efficient	ADJ
ajst-16789	15	39	sorting	sorting	NOUN
ajst-16789	15	40	[	[	X
ajst-16789	15	41	1	1	NUM
ajst-16789	15	42	]	]	PUNCT
ajst-16789	15	43	.	.	PUNCT
ajst-16789	16	1	in	in	ADP
ajst-16789	16	2	the	the	DET
ajst-16789	16	3	realization	realization	NOUN
ajst-16789	16	4	of	of	ADP
ajst-16789	16	5	automated	automate	VERB
ajst-16789	16	6	supply	supply	NOUN
ajst-16789	16	7	of	of	ADP
ajst-16789	16	8	packages	package	NOUN
ajst-16789	16	9	in	in	ADP
ajst-16789	16	10	the	the	DET
ajst-16789	16	11	logistics	logistic	NOUN
ajst-16789	16	12	front	front	ADJ
ajst-16789	16	13	end	end	NOUN
ajst-16789	16	14	,	,	PUNCT
ajst-16789	16	15	the	the	DET
ajst-16789	16	16	identification	identification	NOUN
ajst-16789	16	17	and	and	CCONJ
ajst-16789	16	18	detection	detection	NOUN
ajst-16789	16	19	of	of	ADP
ajst-16789	16	20	express	express	ADJ
ajst-16789	16	21	parcels	parcel	NOUN
ajst-16789	16	22	have	have	AUX
ajst-16789	16	23	become	become	VERB
ajst-16789	16	24	a	a	DET
ajst-16789	16	25	critical	critical	ADJ
ajst-16789	16	26	link	link	NOUN
ajst-16789	16	27	.	.	PUNCT
ajst-16789	17	1	the	the	DET
ajst-16789	17	2	rapid	rapid	ADJ
ajst-16789	17	3	advancement	advancement	NOUN
ajst-16789	17	4	of	of	ADP
ajst-16789	17	5	deep	deep	ADJ
ajst-16789	17	6	learning	learning	NOUN
ajst-16789	17	7	has	have	AUX
ajst-16789	17	8	further	far	ADV
ajst-16789	17	9	propelled	propel	VERB
ajst-16789	17	10	the	the	DET
ajst-16789	17	11	progress	progress	NOUN
ajst-16789	17	12	of	of	ADP
ajst-16789	17	13	target	target	NOUN
ajst-16789	17	14	detection	detection	NOUN
ajst-16789	17	15	technology	technology	NOUN
ajst-16789	17	16	[	[	X
ajst-16789	17	17	2	2	NUM
ajst-16789	17	18	-	-	SYM
ajst-16789	17	19	4	4	NUM
ajst-16789	17	20	]	]	PUNCT
ajst-16789	17	21	,	,	PUNCT
ajst-16789	17	22	offering	offer	VERB
ajst-16789	17	23	more	more	ADJ
ajst-16789	17	24	promising	promising	ADJ
ajst-16789	17	25	prospects	prospect	NOUN
ajst-16789	17	26	for	for	ADP
ajst-16789	17	27	logistics	logistic	NOUN
ajst-16789	17	28	automation	automation	NOUN
ajst-16789	17	29	[	[	X
ajst-16789	17	30	5	5	NUM
ajst-16789	17	31	]	]	PUNCT
ajst-16789	17	32	.	.	PUNCT
ajst-16789	18	1	he	he	PRON
ajst-16789	18	2	k	k	PROPN
ajst-16789	18	3	et	et	PROPN
ajst-16789	18	4	al	al	PROPN
ajst-16789	18	5	.	.	PUNCT
ajst-16789	19	1	[	[	X
ajst-16789	19	2	6	6	NUM
ajst-16789	19	3	]	]	PUNCT
ajst-16789	19	4	introduced	introduce	VERB
ajst-16789	19	5	the	the	DET
ajst-16789	19	6	concept	concept	NOUN
ajst-16789	19	7	of	of	ADP
ajst-16789	19	8	residual	residual	ADJ
ajst-16789	19	9	expression	expression	NOUN
ajst-16789	19	10	into	into	ADP
ajst-16789	19	11	the	the	DET
ajst-16789	19	12	architecture	architecture	NOUN
ajst-16789	19	13	of	of	ADP
ajst-16789	19	14	convolutional	convolutional	ADJ
ajst-16789	19	15	neural	neural	ADJ
ajst-16789	19	16	networks	network	NOUN
ajst-16789	19	17	,	,	PUNCT
ajst-16789	19	18	proposing	propose	VERB
ajst-16789	19	19	a	a	DET
ajst-16789	19	20	deep	deep	ADJ
ajst-16789	19	21	residual	residual	ADJ
ajst-16789	19	22	convolutional	convolutional	ADJ
ajst-16789	19	23	neural	neural	ADJ
ajst-16789	19	24	network	network	NOUN
ajst-16789	19	25	framework	framework	NOUN
ajst-16789	19	26	to	to	PART
ajst-16789	19	27	alleviate	alleviate	VERB
ajst-16789	19	28	the	the	DET
ajst-16789	19	29	difficulty	difficulty	NOUN
ajst-16789	19	30	of	of	ADP
ajst-16789	19	31	training	train	VERB
ajst-16789	19	32	deep	deep	ADJ
ajst-16789	19	33	neural	neural	ADJ
ajst-16789	19	34	networks	network	NOUN
ajst-16789	19	35	.	.	PUNCT
ajst-16789	20	1	the	the	DET
ajst-16789	20	2	residual	residual	ADJ
ajst-16789	20	3	network	network	NOUN
ajst-16789	20	4	enhances	enhance	VERB
ajst-16789	20	5	classification	classification	NOUN
ajst-16789	20	6	accuracy	accuracy	NOUN
ajst-16789	20	7	by	by	ADP
ajst-16789	20	8	increasing	increase	VERB
ajst-16789	20	9	network	network	NOUN
ajst-16789	20	10	depth	depth	NOUN
ajst-16789	20	11	,	,	PUNCT
ajst-16789	20	12	effectively	effectively	ADV
ajst-16789	20	13	avoiding	avoid	VERB
ajst-16789	20	14	the	the	DET
ajst-16789	20	15	problem	problem	NOUN
ajst-16789	20	16	of	of	ADP
ajst-16789	20	17	decreased	decrease	VERB
ajst-16789	20	18	accuracy	accuracy	NOUN
ajst-16789	20	19	after	after	ADP
ajst-16789	20	20	reaching	reach	VERB
ajst-16789	20	21	a	a	DET
ajst-16789	20	22	certain	certain	ADJ
ajst-16789	20	23	number	number	NOUN
ajst-16789	20	24	of	of	ADP
ajst-16789	20	25	layers	layer	NOUN
ajst-16789	20	26	.	.	PUNCT
ajst-16789	21	1	huang	huang	PROPN
ajst-16789	21	2	et	et	PROPN
ajst-16789	21	3	al	al	PROPN
ajst-16789	21	4	.	.	PUNCT
ajst-16789	22	1	[	[	X
ajst-16789	22	2	7	7	X
ajst-16789	22	3	]	]	PUNCT
ajst-16789	22	4	proposed	propose	VERB
ajst-16789	22	5	the	the	DET
ajst-16789	22	6	densenet	densenet	NOUN
ajst-16789	22	7	network	network	NOUN
ajst-16789	22	8	architecture	architecture	NOUN
ajst-16789	22	9	,	,	PUNCT
ajst-16789	22	10	where	where	SCONJ
ajst-16789	22	11	each	each	DET
ajst-16789	22	12	layer	layer	NOUN
ajst-16789	22	13	's	's	PART
ajst-16789	22	14	input	input	NOUN
ajst-16789	22	15	is	be	AUX
ajst-16789	22	16	a	a	DET
ajst-16789	22	17	combination	combination	NOUN
ajst-16789	22	18	of	of	ADP
ajst-16789	22	19	feature	feature	NOUN
ajst-16789	22	20	maps	map	NOUN
ajst-16789	22	21	from	from	ADP
ajst-16789	22	22	preceding	precede	VERB
ajst-16789	22	23	layers	layer	NOUN
ajst-16789	22	24	,	,	PUNCT
ajst-16789	22	25	and	and	CCONJ
ajst-16789	22	26	the	the	DET
ajst-16789	22	27	feature	feature	NOUN
ajst-16789	22	28	maps	map	NOUN
ajst-16789	22	29	of	of	ADP
ajst-16789	22	30	that	that	DET
ajst-16789	22	31	layer	layer	NOUN
ajst-16789	22	32	are	be	AUX
ajst-16789	22	33	used	use	VERB
ajst-16789	22	34	as	as	ADP
ajst-16789	22	35	inputs	input	NOUN
ajst-16789	22	36	for	for	ADP
ajst-16789	22	37	all	all	DET
ajst-16789	22	38	subsequent	subsequent	ADJ
ajst-16789	22	39	layers	layer	NOUN
ajst-16789	22	40	.	.	PUNCT
ajst-16789	23	1	this	this	DET
ajst-16789	23	2	architecture	architecture	NOUN
ajst-16789	23	3	mitigates	mitigate	VERB
ajst-16789	23	4	the	the	DET
ajst-16789	23	5	vanishing	vanish	VERB
ajst-16789	23	6	gradient	gradient	ADJ
ajst-16789	23	7	problem	problem	NOUN
ajst-16789	23	8	to	to	ADP
ajst-16789	23	9	some	some	DET
ajst-16789	23	10	extent	extent	NOUN
ajst-16789	23	11	through	through	ADP
ajst-16789	23	12	feature	feature	NOUN
ajst-16789	23	13	reuse	reuse	NOUN
ajst-16789	23	14	and	and	CCONJ
ajst-16789	23	15	skip	skip	ADJ
ajst-16789	23	16	connections	connection	NOUN
ajst-16789	23	17	,	,	PUNCT
ajst-16789	23	18	enhancing	enhance	VERB
ajst-16789	23	19	feature	feature	NOUN
ajst-16789	23	20	propagation	propagation	NOUN
ajst-16789	23	21	,	,	PUNCT
ajst-16789	23	22	and	and	CCONJ
ajst-16789	23	23	significantly	significantly	ADV
ajst-16789	23	24	reducing	reduce	VERB
ajst-16789	23	25	the	the	DET
ajst-16789	23	26	number	number	NOUN
ajst-16789	23	27	of	of	ADP
ajst-16789	23	28	parameters	parameter	NOUN
ajst-16789	23	29	.	.	PUNCT
ajst-16789	24	1	to	to	PART
ajst-16789	24	2	enhance	enhance	VERB
ajst-16789	24	3	the	the	DET
ajst-16789	24	4	robustness	robustness	NOUN
ajst-16789	24	5	of	of	ADP
ajst-16789	24	6	target	target	NOUN
ajst-16789	24	7	recognition	recognition	NOUN
ajst-16789	24	8	algorithms	algorithm	NOUN
ajst-16789	24	9	,	,	PUNCT
ajst-16789	24	10	eitel	eitel	NOUN
ajst-16789	24	11	et	et	PROPN
ajst-16789	24	12	al	al	PROPN
ajst-16789	24	13	.	.	PUNCT
ajst-16789	25	1	[	[	X
ajst-16789	25	2	8	8	NUM
ajst-16789	25	3	]	]	PUNCT
ajst-16789	25	4	proposed	propose	VERB
ajst-16789	25	5	a	a	DET
ajst-16789	25	6	multi	multi	ADJ
ajst-16789	25	7	-	-	ADJ
ajst-16789	25	8	modal	modal	ADJ
ajst-16789	25	9	target	target	NOUN
ajst-16789	25	10	recognition	recognition	NOUN
ajst-16789	25	11	network	network	NOUN
ajst-16789	25	12	architecture	architecture	NOUN
ajst-16789	25	13	consisting	consist	VERB
ajst-16789	25	14	of	of	ADP
ajst-16789	25	15	two	two	NUM
ajst-16789	25	16	relatively	relatively	ADV
ajst-16789	25	17	independent	independent	ADJ
ajst-16789	25	18	neural	neural	ADJ
ajst-16789	25	19	network	network	NOUN
ajst-16789	25	20	streams	stream	NOUN
ajst-16789	25	21	.	.	PUNCT
ajst-16789	26	1	each	each	DET
ajst-16789	26	2	network	network	NOUN
ajst-16789	26	3	stream	stream	NOUN
ajst-16789	26	4	primarily	primarily	ADV
ajst-16789	26	5	processes	process	VERB
ajst-16789	26	6	a	a	DET
ajst-16789	26	7	specific	specific	ADJ
ajst-16789	26	8	modality	modality	NOUN
ajst-16789	26	9	of	of	ADP
ajst-16789	26	10	information	information	NOUN
ajst-16789	26	11	and	and	CCONJ
ajst-16789	26	12	integrates	integrate	VERB
ajst-16789	26	13	different	different	ADJ
ajst-16789	26	14	modality	modality	NOUN
ajst-16789	26	15	features	feature	VERB
ajst-16789	26	16	through	through	ADP
ajst-16789	26	17	a	a	DET
ajst-16789	26	18	late	late	ADJ
ajst-16789	26	19	fusion	fusion	NOUN
ajst-16789	26	20	approach	approach	NOUN
ajst-16789	26	21	.	.	PUNCT
ajst-16789	27	1	during	during	ADP
ajst-16789	27	2	the	the	DET
ajst-16789	27	3	training	training	NOUN
ajst-16789	27	4	of	of	ADP
ajst-16789	27	5	this	this	DET
ajst-16789	27	6	network	network	NOUN
ajst-16789	27	7	,	,	PUNCT
ajst-16789	27	8	depth	depth	NOUN
ajst-16789	27	9	data	datum	NOUN
ajst-16789	27	10	is	be	AUX
ajst-16789	27	11	rendered	render	VERB
ajst-16789	27	12	into	into	ADP
ajst-16789	27	13	three	three	NUM
ajst-16789	27	14	-	-	PUNCT
ajst-16789	27	15	channel	channel	NOUN
ajst-16789	27	16	color	color	NOUN
ajst-16789	27	17	images	image	NOUN
ajst-16789	27	18	,	,	PUNCT
ajst-16789	27	19	enabling	enable	VERB
ajst-16789	27	20	model	model	NOUN
ajst-16789	27	21	training	training	NOUN
ajst-16789	27	22	without	without	ADP
ajst-16789	27	23	the	the	DET
ajst-16789	27	24	need	need	NOUN
ajst-16789	27	25	for	for	ADP
ajst-16789	27	26	large	large	ADJ
ajst-16789	27	27	-	-	PUNCT
ajst-16789	27	28	scale	scale	NOUN
ajst-16789	27	29	depth	depth	NOUN
ajst-16789	27	30	datasets	dataset	NOUN
ajst-16789	27	31	.	.	PUNCT
ajst-16789	28	1	simultaneously	simultaneously	ADV
ajst-16789	28	2	,	,	PUNCT
ajst-16789	28	3	augmenting	augment	VERB
ajst-16789	28	4	data	datum	NOUN
ajst-16789	28	5	is	be	AUX
ajst-16789	28	6	achieved	achieve	VERB
ajst-16789	28	7	by	by	ADP
ajst-16789	28	8	corrupting	corrupt	VERB
ajst-16789	28	9	depth	depth	NOUN
ajst-16789	28	10	images	image	NOUN
ajst-16789	28	11	with	with	ADP
ajst-16789	28	12	noise	noise	NOUN
ajst-16789	28	13	patterns	pattern	NOUN
ajst-16789	28	14	from	from	ADP
ajst-16789	28	15	real	real	ADJ
ajst-16789	28	16	-	-	PUNCT
ajst-16789	28	17	world	world	NOUN
ajst-16789	28	18	scenarios	scenario	NOUN
ajst-16789	28	19	,	,	PUNCT
ajst-16789	28	20	thus	thus	ADV
ajst-16789	28	21	improving	improve	VERB
ajst-16789	28	22	the	the	DET
ajst-16789	28	23	robustness	robustness	NOUN
ajst-16789	28	24	of	of	ADP
ajst-16789	28	25	deep	deep	ADJ
ajst-16789	28	26	neural	neural	ADJ
ajst-16789	28	27	networks	network	NOUN
ajst-16789	28	28	.	.	PUNCT
ajst-16789	29	1	this	this	DET
ajst-16789	29	2	approach	approach	NOUN
ajst-16789	29	3	demonstrated	demonstrate	VERB
ajst-16789	29	4	improved	improved	ADJ
ajst-16789	29	5	performance	performance	NOUN
ajst-16789	29	6	on	on	ADP
ajst-16789	29	7	multimodal	multimodal	NOUN
ajst-16789	29	8	datasets	dataset	NOUN
ajst-16789	29	9	and	and	CCONJ
ajst-16789	29	10	showcased	showcase	VERB
ajst-16789	29	11	recognition	recognition	NOUN
ajst-16789	29	12	capabilities	capability	NOUN
ajst-16789	29	13	in	in	ADP
ajst-16789	29	14	actual	actual	ADJ
ajst-16789	29	15	noisy	noisy	ADJ
ajst-16789	29	16	environments	environment	NOUN
ajst-16789	29	17	.	.	PUNCT
ajst-16789	30	1	due	due	ADP
ajst-16789	30	2	to	to	ADP
ajst-16789	30	3	the	the	DET
ajst-16789	30	4	capability	capability	NOUN
ajst-16789	30	5	of	of	ADP
ajst-16789	30	6	capturing	capture	VERB
ajst-16789	30	7	both	both	PRON
ajst-16789	30	8	rgb	rgb	PROPN
ajst-16789	30	9	images	image	NOUN
ajst-16789	30	10	and	and	CCONJ
ajst-16789	30	11	depth	depth	NOUN
ajst-16789	30	12	images	image	NOUN
ajst-16789	30	13	of	of	ADP
ajst-16789	30	14	objects	object	NOUN
ajst-16789	30	15	simultaneously	simultaneously	ADV
ajst-16789	30	16	[	[	X
ajst-16789	30	17	9	9	NUM
ajst-16789	30	18	]	]	PUNCT
ajst-16789	30	19	,	,	PUNCT
ajst-16789	30	20	rgbd	rgbd	NOUN
ajst-16789	30	21	cameras	camera	NOUN
ajst-16789	30	22	enable	enable	VERB
ajst-16789	30	23	the	the	DET
ajst-16789	30	24	acquisition	acquisition	NOUN
ajst-16789	30	25	of	of	ADP
ajst-16789	30	26	three	three	NUM
ajst-16789	30	27	-	-	PUNCT
ajst-16789	30	28	dimensional	dimensional	ADJ
ajst-16789	30	29	positional	positional	ADJ
ajst-16789	30	30	information	information	NOUN
ajst-16789	30	31	of	of	ADP
ajst-16789	30	32	objects	object	NOUN
ajst-16789	30	33	[	[	X
ajst-16789	30	34	10,11	10,11	NOUN
ajst-16789	30	35	]	]	X
ajst-16789	30	36	.	.	PUNCT
ajst-16789	31	1	consequently	consequently	ADV
ajst-16789	31	2	,	,	PUNCT
ajst-16789	31	3	rgbd	rgbd	NOUN
ajst-16789	31	4	cameras	camera	NOUN
ajst-16789	31	5	are	be	AUX
ajst-16789	31	6	increasingly	increasingly	ADV
ajst-16789	31	7	being	be	AUX
ajst-16789	31	8	applied	apply	VERB
ajst-16789	31	9	in	in	ADP
ajst-16789	31	10	the	the	DET
ajst-16789	31	11	field	field	NOUN
ajst-16789	31	12	of	of	ADP
ajst-16789	31	13	robotic	robotic	ADJ
ajst-16789	31	14	arm	arm	NOUN
ajst-16789	31	15	visual	visual	ADJ
ajst-16789	31	16	grasping	grasping	NOUN
ajst-16789	31	17	.	.	PUNCT
ajst-16789	32	1	dealing	deal	VERB
ajst-16789	32	2	with	with	ADP
ajst-16789	32	3	cluttered	cluttered	ADJ
ajst-16789	32	4	and	and	CCONJ
ajst-16789	32	5	disorganized	disorganized	ADJ
ajst-16789	32	6	items	item	NOUN
ajst-16789	32	7	becomes	become	VERB
ajst-16789	32	8	challenging	challenge	VERB
ajst-16789	32	9	when	when	SCONJ
ajst-16789	32	10	inferring	infer	VERB
ajst-16789	32	11	their	their	PRON
ajst-16789	32	12	relative	relative	ADJ
ajst-16789	32	13	poses	pose	NOUN
ajst-16789	32	14	from	from	ADP
ajst-16789	32	15	textured	textured	ADJ
ajst-16789	32	16	color	color	NOUN
ajst-16789	32	17	images	image	NOUN
ajst-16789	32	18	due	due	ADJ
ajst-16789	32	19	to	to	ADP
ajst-16789	32	20	factors	factor	NOUN
ajst-16789	32	21	such	such	ADJ
ajst-16789	32	22	as	as	ADP
ajst-16789	32	23	sensor	sensor	NOUN
ajst-16789	32	24	noise	noise	NOUN
ajst-16789	32	25	,	,	PUNCT
ajst-16789	32	26	occlusion	occlusion	NOUN
ajst-16789	32	27	,	,	PUNCT
ajst-16789	32	28	and	and	CCONJ
ajst-16789	32	29	obstruction	obstruction	NOUN
ajst-16789	32	30	[	[	X
ajst-16789	32	31	12,13	12,13	NUM
ajst-16789	32	32	]	]	PUNCT
ajst-16789	32	33	.	.	PUNCT
ajst-16789	33	1	depth	depth	NOUN
ajst-16789	33	2	image	image	NOUN
ajst-16789	33	3	data	datum	NOUN
ajst-16789	33	4	contains	contain	VERB
ajst-16789	33	5	rich	rich	ADJ
ajst-16789	33	6	geometric	geometric	ADJ
ajst-16789	33	7	information	information	NOUN
ajst-16789	33	8	.	.	PUNCT
ajst-16789	34	1	hence	hence	ADV
ajst-16789	34	2	,	,	PUNCT
ajst-16789	34	3	when	when	SCONJ
ajst-16789	34	4	picking	pick	VERB
ajst-16789	34	5	up	up	ADP
ajst-16789	34	6	objects	object	NOUN
ajst-16789	34	7	with	with	ADP
ajst-16789	34	8	planar	planar	ADJ
ajst-16789	34	9	surfaces	surface	NOUN
ajst-16789	34	10	in	in	ADP
ajst-16789	34	11	crowded	crowded	ADJ
ajst-16789	34	12	environments	environment	NOUN
ajst-16789	34	13	,	,	PUNCT
ajst-16789	34	14	using	use	VERB
ajst-16789	34	15	depth	depth	NOUN
ajst-16789	34	16	images	image	NOUN
ajst-16789	34	17	is	be	AUX
ajst-16789	34	18	more	more	ADV
ajst-16789	34	19	suitable	suitable	ADJ
ajst-16789	34	20	.	.	PUNCT
ajst-16789	35	1	this	this	DET
ajst-16789	35	2	approach	approach	NOUN
ajst-16789	35	3	reduces	reduce	VERB
ajst-16789	35	4	reliance	reliance	NOUN
ajst-16789	35	5	on	on	ADP
ajst-16789	35	6	target	target	NOUN
ajst-16789	35	7	image	image	NOUN
ajst-16789	35	8	and	and	CCONJ
ajst-16789	35	9	texture	texture	ADJ
ajst-16789	35	10	feature	feature	NOUN
ajst-16789	35	11	extraction	extraction	NOUN
ajst-16789	35	12	,	,	PUNCT
ajst-16789	35	13	thereby	thereby	ADV
ajst-16789	35	14	enhancing	enhance	VERB
ajst-16789	35	15	the	the	DET
ajst-16789	35	16	robustness	robustness	NOUN
ajst-16789	35	17	of	of	ADP
ajst-16789	35	18	visual	visual	ADJ
ajst-16789	35	19	recognition	recognition	NOUN
ajst-16789	35	20	systems	system	NOUN
ajst-16789	35	21	.	.	PUNCT
ajst-16789	36	1	zeng	zeng	PROPN
ajst-16789	36	2	et	et	PROPN
ajst-16789	36	3	al	al	PROPN
ajst-16789	36	4	.	.	PUNCT
ajst-16789	37	1	[	[	X
ajst-16789	37	2	14	14	NUM
ajst-16789	37	3	]	]	PUNCT
ajst-16789	37	4	proposed	propose	VERB
ajst-16789	37	5	an	an	DET
ajst-16789	37	6	'	'	PUNCT
ajst-16789	37	7	grasp	grasp	NOUN
ajst-16789	37	8	-	-	PUNCT
ajst-16789	37	9	first	first	ADJ
ajst-16789	37	10	-	-	PUNCT
ajst-16789	37	11	recognize	recognize	VERB
ajst-16789	37	12	-	-	PUNCT
ajst-16789	37	13	later	later	NOUN
ajst-16789	37	14	'	'	PUNCT
ajst-16789	37	15	approach	approach	NOUN
ajst-16789	37	16	to	to	PART
ajst-16789	37	17	address	address	VERB
ajst-16789	37	18	the	the	DET
ajst-16789	37	19	grasping	grasp	VERB
ajst-16789	37	20	problem	problem	NOUN
ajst-16789	37	21	of	of	ADP
ajst-16789	37	22	unknown	unknown	ADJ
ajst-16789	37	23	objects	object	NOUN
ajst-16789	37	24	in	in	ADP
ajst-16789	37	25	cluttered	cluttered	ADJ
ajst-16789	37	26	scenes	scene	NOUN
ajst-16789	37	27	,	,	PUNCT
ajst-16789	37	28	significantly	significantly	ADV
ajst-16789	37	29	reducing	reduce	VERB
ajst-16789	37	30	the	the	DET
ajst-16789	37	31	complexity	complexity	NOUN
ajst-16789	37	32	of	of	ADP
ajst-16789	37	33	recognition	recognition	NOUN
ajst-16789	37	34	.	.	PUNCT
ajst-16789	38	1	they	they	PRON
ajst-16789	38	2	utilized	utilize	VERB
ajst-16789	38	3	a	a	DET
ajst-16789	38	4	full	full	ADJ
ajst-16789	38	5	convolutional	convolutional	ADJ
ajst-16789	38	6	residual	residual	ADJ
ajst-16789	38	7	network	network	NOUN
ajst-16789	38	8	,	,	PUNCT
ajst-16789	38	9	resnet101	resnet101	PROPN
ajst-16789	38	10	,	,	PUNCT
ajst-16789	38	11	to	to	PART
ajst-16789	38	12	process	process	VERB
ajst-16789	38	13	multi	multi	ADJ
ajst-16789	38	14	-	-	ADJ
ajst-16789	38	15	modal	modal	ADJ
ajst-16789	38	16	information	information	NOUN
ajst-16789	38	17	from	from	ADP
ajst-16789	38	18	rgb	rgb	PROPN
ajst-16789	38	19	images	image	NOUN
ajst-16789	38	20	and	and	CCONJ
ajst-16789	38	21	depth	depth	NOUN
ajst-16789	38	22	maps	map	NOUN
ajst-16789	38	23	.	.	PUNCT
ajst-16789	39	1	this	this	DET
ajst-16789	39	2	network	network	NOUN
ajst-16789	39	3	outputted	output	VERB
ajst-16789	39	4	labels	label	NOUN
ajst-16789	39	5	for	for	ADP
ajst-16789	39	6	grasping	grasp	VERB
ajst-16789	39	7	regions	region	NOUN
ajst-16789	39	8	.	.	PUNCT
ajst-16789	40	1	simultaneously	simultaneously	ADV
ajst-16789	40	2	,	,	PUNCT
ajst-16789	40	3	the	the	DET
ajst-16789	40	4	depth	depth	NOUN
ajst-16789	40	5	map	map	NOUN
ajst-16789	40	6	was	be	AUX
ajst-16789	40	7	transformed	transform	VERB
ajst-16789	40	8	into	into	ADP
ajst-16789	40	9	a	a	DET
ajst-16789	40	10	3d	3d	NUM
ajst-16789	40	11	point	point	NOUN
ajst-16789	40	12	cloud	cloud	NOUN
ajst-16789	40	13	,	,	PUNCT
ajst-16789	40	14	enabling	enable	VERB
ajst-16789	40	15	the	the	DET
ajst-16789	40	16	calculation	calculation	NOUN
ajst-16789	40	17	of	of	ADP
ajst-16789	40	18	surface	surface	ADJ
ajst-16789	40	19	normal	normal	ADJ
ajst-16789	40	20	vectors	vector	NOUN
ajst-16789	40	21	for	for	ADP
ajst-16789	40	22	grasping	grasp	VERB
ajst-16789	40	23	areas	area	NOUN
ajst-16789	40	24	,	,	PUNCT
ajst-16789	40	25	thereby	thereby	ADV
ajst-16789	40	26	providing	provide	VERB
ajst-16789	40	27	information	information	NOUN
ajst-16789	40	28	about	about	ADP
ajst-16789	40	29	grasp	grasp	NOUN
ajst-16789	40	30	angles	angle	NOUN
ajst-16789	40	31	.	.	PUNCT
ajst-16789	41	1	199	199	NUM
ajst-16789	41	2	2	2	NUM
ajst-16789	41	3	.	.	PUNCT
ajst-16789	42	1	the	the	DET
ajst-16789	42	2	collection	collection	NOUN
ajst-16789	42	3	and	and	CCONJ
ajst-16789	42	4	annotation	annotation	NOUN
ajst-16789	42	5	of	of	ADP
ajst-16789	42	6	datasets	dataset	NOUN
ajst-16789	42	7	given	give	VERB
ajst-16789	42	8	the	the	DET
ajst-16789	42	9	absence	absence	NOUN
ajst-16789	42	10	of	of	ADP
ajst-16789	42	11	publicly	publicly	ADV
ajst-16789	42	12	available	available	ADJ
ajst-16789	42	13	datasets	dataset	NOUN
ajst-16789	42	14	for	for	ADP
ajst-16789	42	15	express	express	ADJ
ajst-16789	42	16	parcels	parcel	NOUN
ajst-16789	42	17	,	,	PUNCT
ajst-16789	42	18	we	we	PRON
ajst-16789	42	19	independently	independently	ADV
ajst-16789	42	20	collected	collect	VERB
ajst-16789	42	21	images	image	NOUN
ajst-16789	42	22	of	of	ADP
ajst-16789	42	23	15	15	NUM
ajst-16789	42	24	different	different	ADJ
ajst-16789	42	25	types	type	NOUN
ajst-16789	42	26	of	of	ADP
ajst-16789	42	27	express	express	ADJ
ajst-16789	42	28	parcels	parcel	NOUN
ajst-16789	42	29	using	use	VERB
ajst-16789	42	30	an	an	DET
ajst-16789	42	31	rgbd	rgbd	NOUN
ajst-16789	42	32	camera	camera	NOUN
ajst-16789	42	33	,	,	PUNCT
ajst-16789	42	34	comprising	comprise	VERB
ajst-16789	42	35	8	8	NUM
ajst-16789	42	36	types	type	NOUN
ajst-16789	42	37	of	of	ADP
ajst-16789	42	38	cardboard	cardboard	NOUN
ajst-16789	42	39	boxes	box	NOUN
ajst-16789	42	40	,	,	PUNCT
ajst-16789	42	41	2	2	NUM
ajst-16789	42	42	types	type	NOUN
ajst-16789	42	43	of	of	ADP
ajst-16789	42	44	bubble	bubble	NOUN
ajst-16789	42	45	-	-	PUNCT
ajst-16789	42	46	wrapped	wrap	VERB
ajst-16789	42	47	flat	flat	ADJ
ajst-16789	42	48	items	item	NOUN
ajst-16789	42	49	,	,	PUNCT
ajst-16789	42	50	and	and	CCONJ
ajst-16789	42	51	5	5	NUM
ajst-16789	42	52	types	type	NOUN
ajst-16789	42	53	of	of	ADP
ajst-16789	42	54	soft	soft	ADJ
ajst-16789	42	55	pouches	pouch	NOUN
ajst-16789	42	56	(	(	PUNCT
ajst-16789	42	57	refer	refer	VERB
ajst-16789	42	58	to	to	PART
ajst-16789	42	59	figure	figure	VERB
ajst-16789	42	60	1	1	NUM
ajst-16789	42	61	)	)	PUNCT
ajst-16789	42	62	.	.	PUNCT
ajst-16789	43	1	throughout	throughout	ADP
ajst-16789	43	2	this	this	DET
ajst-16789	43	3	process	process	NOUN
ajst-16789	43	4	,	,	PUNCT
ajst-16789	43	5	we	we	PRON
ajst-16789	43	6	considered	consider	VERB
ajst-16789	43	7	combinations	combination	NOUN
ajst-16789	43	8	of	of	ADP
ajst-16789	43	9	parcels	parcel	NOUN
ajst-16789	43	10	with	with	ADP
ajst-16789	43	11	varying	vary	VERB
ajst-16789	43	12	quantities	quantity	NOUN
ajst-16789	43	13	,	,	PUNCT
ajst-16789	43	14	types	type	NOUN
ajst-16789	43	15	,	,	PUNCT
ajst-16789	43	16	and	and	CCONJ
ajst-16789	43	17	orientations	orientation	NOUN
ajst-16789	43	18	,	,	PUNCT
ajst-16789	43	19	amassing	amass	VERB
ajst-16789	43	20	a	a	DET
ajst-16789	43	21	total	total	NOUN
ajst-16789	43	22	of	of	ADP
ajst-16789	43	23	713	713	NUM
ajst-16789	43	24	images	image	NOUN
ajst-16789	43	25	.	.	PUNCT
ajst-16789	44	1	these	these	DET
ajst-16789	44	2	collected	collect	VERB
ajst-16789	44	3	images	image	NOUN
ajst-16789	44	4	were	be	AUX
ajst-16789	44	5	annotated	annotate	VERB
ajst-16789	44	6	.	.	PUNCT
ajst-16789	45	1	subsequently	subsequently	ADV
ajst-16789	45	2	,	,	PUNCT
ajst-16789	45	3	we	we	PRON
ajst-16789	45	4	augmented	augment	VERB
ajst-16789	45	5	the	the	DET
ajst-16789	45	6	dataset	dataset	NOUN
ajst-16789	45	7	through	through	ADP
ajst-16789	45	8	operations	operation	NOUN
ajst-16789	45	9	such	such	ADJ
ajst-16789	45	10	as	as	ADP
ajst-16789	45	11	rotation	rotation	NOUN
ajst-16789	45	12	and	and	CCONJ
ajst-16789	45	13	mirroring	mirroring	NOUN
ajst-16789	45	14	,	,	PUNCT
ajst-16789	45	15	resulting	result	VERB
ajst-16789	45	16	in	in	ADP
ajst-16789	45	17	an	an	DET
ajst-16789	45	18	expanded	expand	VERB
ajst-16789	45	19	dataset	dataset	NOUN
ajst-16789	45	20	containing	contain	VERB
ajst-16789	45	21	2,842	2,842	NUM
ajst-16789	45	22	images	image	NOUN
ajst-16789	45	23	.	.	PUNCT
ajst-16789	46	1	figure	figure	NOUN
ajst-16789	46	2	1	1	NUM
ajst-16789	46	3	.	.	PUNCT
ajst-16789	46	4	different	different	ADJ
ajst-16789	46	5	types	type	NOUN
ajst-16789	46	6	of	of	ADP
ajst-16789	46	7	packages	package	NOUN
ajst-16789	46	8	the	the	DET
ajst-16789	46	9	annotated	annotate	VERB
ajst-16789	46	10	images	image	NOUN
ajst-16789	46	11	were	be	AUX
ajst-16789	46	12	generated	generate	VERB
ajst-16789	46	13	based	base	VERB
ajst-16789	46	14	on	on	ADP
ajst-16789	46	15	the	the	DET
ajst-16789	46	16	collected	collect	VERB
ajst-16789	46	17	rgb	rgb	PROPN
ajst-16789	46	18	images	image	NOUN
ajst-16789	46	19	(	(	PUNCT
ajst-16789	46	20	detailed	detailed	ADJ
ajst-16789	46	21	in	in	ADP
ajst-16789	46	22	figures	figure	NOUN
ajst-16789	46	23	2	2	NUM
ajst-16789	46	24	and	and	CCONJ
ajst-16789	46	25	3	3	NUM
ajst-16789	46	26	)	)	PUNCT
ajst-16789	46	27	.	.	PUNCT
ajst-16789	47	1	unlike	unlike	ADP
ajst-16789	47	2	the	the	DET
ajst-16789	47	3	conventional	conventional	ADJ
ajst-16789	47	4	coordinate	coordinate	NOUN
ajst-16789	47	5	-	-	PUNCT
ajst-16789	47	6	based	base	VERB
ajst-16789	47	7	annotation	annotation	NOUN
ajst-16789	47	8	methods	method	NOUN
ajst-16789	47	9	used	use	VERB
ajst-16789	47	10	in	in	ADP
ajst-16789	47	11	traditional	traditional	ADJ
ajst-16789	47	12	object	object	NOUN
ajst-16789	47	13	detection	detection	NOUN
ajst-16789	47	14	,	,	PUNCT
ajst-16789	47	15	the	the	DET
ajst-16789	47	16	labeling	labeling	NOUN
ajst-16789	47	17	results	result	VERB
ajst-16789	47	18	in	in	ADP
ajst-16789	47	19	this	this	DET
ajst-16789	47	20	study	study	NOUN
ajst-16789	47	21	's	's	PART
ajst-16789	47	22	model	model	NOUN
ajst-16789	47	23	are	be	AUX
ajst-16789	47	24	presented	present	VERB
ajst-16789	47	25	in	in	ADP
ajst-16789	47	26	grayscale	grayscale	NOUN
ajst-16789	47	27	format	format	NOUN
ajst-16789	47	28	.	.	PUNCT
ajst-16789	48	1	three	three	NUM
ajst-16789	48	2	colors	color	NOUN
ajst-16789	48	3	(	(	PUNCT
ajst-16789	48	4	black	black	ADJ
ajst-16789	48	5	,	,	PUNCT
ajst-16789	48	6	gray	gray	ADJ
ajst-16789	48	7	,	,	PUNCT
ajst-16789	48	8	white	white	ADJ
ajst-16789	48	9	)	)	PUNCT
ajst-16789	48	10	represent	represent	VERB
ajst-16789	48	11	three	three	NUM
ajst-16789	48	12	labels	label	NOUN
ajst-16789	48	13	(	(	PUNCT
ajst-16789	48	14	non	non	ADJ
ajst-16789	48	15	-	-	ADJ
ajst-16789	48	16	graspable	graspable	ADJ
ajst-16789	48	17	area	area	NOUN
ajst-16789	48	18	,	,	PUNCT
ajst-16789	48	19	graspable	graspable	ADJ
ajst-16789	48	20	area	area	NOUN
ajst-16789	48	21	,	,	PUNCT
ajst-16789	48	22	background	background	NOUN
ajst-16789	48	23	area	area	NOUN
ajst-16789	48	24	)	)	PUNCT
ajst-16789	48	25	.	.	PUNCT
ajst-16789	49	1	typically	typically	ADV
ajst-16789	49	2	,	,	PUNCT
ajst-16789	49	3	the	the	DET
ajst-16789	49	4	nongraspable	nongraspable	ADJ
ajst-16789	49	5	area	area	NOUN
ajst-16789	49	6	refers	refer	VERB
ajst-16789	49	7	to	to	ADP
ajst-16789	49	8	parcels	parcel	NOUN
ajst-16789	49	9	unsuitable	unsuitable	ADJ
ajst-16789	49	10	for	for	ADP
ajst-16789	49	11	grasping	grasp	VERB
ajst-16789	49	12	,	,	PUNCT
ajst-16789	49	13	such	such	ADJ
ajst-16789	49	14	as	as	ADP
ajst-16789	49	15	those	those	PRON
ajst-16789	49	16	mostly	mostly	ADV
ajst-16789	49	17	covered	cover	VERB
ajst-16789	49	18	by	by	ADP
ajst-16789	49	19	other	other	ADJ
ajst-16789	49	20	parcels	parcel	NOUN
ajst-16789	49	21	or	or	CCONJ
ajst-16789	49	22	positioned	position	VERB
ajst-16789	49	23	at	at	ADP
ajst-16789	49	24	the	the	DET
ajst-16789	49	25	edges	edge	NOUN
ajst-16789	49	26	.	.	PUNCT
ajst-16789	50	1	figure	figure	NOUN
ajst-16789	50	2	2	2	NUM
ajst-16789	50	3	.	.	NUM
ajst-16789	50	4	captured	capture	VERB
ajst-16789	50	5	rgb	rgb	PROPN
ajst-16789	50	6	image	image	NOUN
ajst-16789	50	7	figure	figure	NOUN
ajst-16789	50	8	3	3	NUM
ajst-16789	50	9	.	.	PUNCT
ajst-16789	50	10	annotated	annotate	VERB
ajst-16789	50	11	image	image	NOUN
ajst-16789	50	12	200	200	NUM
ajst-16789	50	13	3	3	NUM
ajst-16789	50	14	.	.	PUNCT
ajst-16789	51	1	the	the	DET
ajst-16789	51	2	model	model	NOUN
ajst-16789	51	3	design	design	NOUN
ajst-16789	51	4	in	in	ADP
ajst-16789	51	5	this	this	DET
ajst-16789	51	6	study	study	NOUN
ajst-16789	51	7	,	,	PUNCT
ajst-16789	51	8	for	for	ADP
ajst-16789	51	9	the	the	DET
ajst-16789	51	10	suction	suction	NOUN
ajst-16789	51	11	cup	cup	NOUN
ajst-16789	51	12	grasp	grasp	VERB
ajst-16789	51	13	surface	surface	NOUN
ajst-16789	51	14	recognition	recognition	NOUN
ajst-16789	51	15	of	of	ADP
ajst-16789	51	16	express	express	ADJ
ajst-16789	51	17	parcels	parcel	NOUN
ajst-16789	51	18	,	,	PUNCT
ajst-16789	51	19	we	we	PRON
ajst-16789	51	20	propose	propose	VERB
ajst-16789	51	21	a	a	DET
ajst-16789	51	22	detection	detection	NOUN
ajst-16789	51	23	and	and	CCONJ
ajst-16789	51	24	identification	identification	NOUN
ajst-16789	51	25	network	network	NOUN
ajst-16789	51	26	called	call	VERB
ajst-16789	51	27	parcelnet	parcelnet	NOUN
ajst-16789	51	28	.	.	PUNCT
ajst-16789	52	1	the	the	DET
ajst-16789	52	2	structure	structure	NOUN
ajst-16789	52	3	of	of	ADP
ajst-16789	52	4	the	the	DET
ajst-16789	52	5	network	network	NOUN
ajst-16789	52	6	is	be	AUX
ajst-16789	52	7	illustrated	illustrate	VERB
ajst-16789	52	8	in	in	ADP
ajst-16789	52	9	figure	figure	NOUN
ajst-16789	52	10	4	4	NUM
ajst-16789	52	11	.	.	PUNCT
ajst-16789	53	1	r	r	NOUN
ajst-16789	53	2	g	g	PROPN
ajst-16789	53	3	b	b	PROPN
ajst-16789	53	4	 	 	SPACE
ajst-16789	53	5	i	i	PRON
ajst-16789	53	6	m	m	VERB
ajst-16789	53	7	a	a	DET
ajst-16789	53	8	g	g	NOUN
ajst-16789	53	9	e	e	NOUN
ajst-16789	53	10	d	d	X
ajst-16789	53	11	e	e	X
ajst-16789	53	12	p	p	X
ajst-16789	53	13	th	th	INTJ
ajst-16789	53	14	 	 	SPACE
ajst-16789	53	15	i	i	PRON
ajst-16789	53	16	m	m	VERB
ajst-16789	53	17	a	a	DET
ajst-16789	53	18	ge	ge	PROPN
ajst-16789	53	19	r	r	PROPN
ajst-16789	53	20	e	e	X
ajst-16789	53	21	se	se	X
ajst-16789	53	22	n	n	PROPN
ajst-16789	53	23	e	e	PROPN
ajst-16789	53	24	t	t	PROPN
ajst-16789	53	25	（	（	PUNCT
ajst-16789	53	26	d	d	X
ajst-16789	53	27	e	e	X
ajst-16789	53	28	n	n	X
ajst-16789	53	29	se	se	X
ajst-16789	53	30	  	  	SPACE
ajst-16789	53	31	n	n	PROPN
ajst-16789	53	32	e	e	PROPN
ajst-16789	53	33	t	t	PROPN
ajst-16789	53	34	）	）	PROPN
ajst-16789	53	35	b	b	PROPN
ajst-16789	53	36	a	a	DET
ajst-16789	53	37	ckb	ckb	NOUN
ajst-16789	53	38	o	o	NOUN
ajst-16789	53	39	n	n	PROPN
ajst-16789	53	40	e	e	PROPN
ajst-16789	53	41	a	a	PROPN
ajst-16789	53	42	d	d	X
ajst-16789	53	43	d	d	NOUN
ajst-16789	53	44	sp	sp	ADP
ajst-16789	53	45	a	a	DET
ajst-16789	53	46	tial	tial	NOUN
ajst-16789	53	47	 	 	SPACE
ajst-16789	53	48	c	c	NOUN
ajst-16789	54	1	o	o	NOUN
ajst-16789	54	2	n	n	CCONJ
ajst-16789	54	3	vo	vo	X
ajst-16789	54	4	lu	lu	NOUN
ajst-16789	54	5	tio	tio	NOUN
ajst-16789	54	6	n	n	CCONJ
ajst-16789	54	7	r	r	NOUN
ajst-16789	54	8	e	e	X
ajst-16789	54	9	se	se	X
ajst-16789	54	10	n	n	PROPN
ajst-16789	54	11	e	e	PROPN
ajst-16789	54	12	t	t	PROPN
ajst-16789	54	13	（	（	PUNCT
ajst-16789	55	1	d	d	X
ajst-16789	55	2	e	e	X
ajst-16789	55	3	n	n	X
ajst-16789	55	4	se	se	X
ajst-16789	55	5	  	  	SPACE
ajst-16789	55	6	n	n	PROPN
ajst-16789	55	7	e	e	PROPN
ajst-16789	55	8	t	t	PROPN
ajst-16789	55	9	）	）	PROPN
ajst-16789	55	10	b	b	PROPN
ajst-16789	55	11	a	a	DET
ajst-16789	55	12	ckb	ckb	NOUN
ajst-16789	55	13	o	o	NOUN
ajst-16789	55	14	n	n	ADP
ajst-16789	55	15	e	e	X
ajst-16789	55	16	sp	sp	ADP
ajst-16789	55	17	a	a	DET
ajst-16789	55	18	tial	tial	ADJ
ajst-16789	55	19	 	 	SPACE
ajst-16789	55	20	c	c	NOUN
ajst-16789	56	1	o	o	NOUN
ajst-16789	56	2	n	n	CCONJ
ajst-16789	56	3	vo	vo	X
ajst-16789	56	4	lu	lu	PROPN
ajst-16789	56	5	tio	tio	NOUN
ajst-16789	56	6	n	n	PROPN
ajst-16789	56	7	sp	sp	ADP
ajst-16789	56	8	a	a	DET
ajst-16789	56	9	tial	tial	NOUN
ajst-16789	56	10	 	 	SPACE
ajst-16789	56	11	c	c	NOUN
ajst-16789	57	1	o	o	NOUN
ajst-16789	57	2	n	n	CCONJ
ajst-16789	57	3	vo	vo	X
ajst-16789	57	4	lu	lu	NOUN
ajst-16789	58	1	tio	tio	NOUN
ajst-16789	59	1	n	n	ADV
ajst-16789	60	1	u	u	NOUN
ajst-16789	60	2	p	p	PROPN
ajst-16789	60	3	sam	sam	PROPN
ajst-16789	60	4	p	p	PROPN
ajst-16789	60	5	lin	lin	PROPN
ajst-16789	61	1	g	g	PROPN
ajst-16789	61	2	p	p	PROPN
ajst-16789	61	3	ad	ad	NOUN
ajst-16789	61	4	in	in	ADP
ajst-16789	61	5	g	g	PROPN
ajst-16789	61	6	b	b	PROPN
ajst-16789	61	7	a	a	DET
ajst-16789	61	8	ckg	ckg	NOUN
ajst-16789	61	9	ro	ro	X
ajst-16789	61	10	u	u	PROPN
ajst-16789	61	11	n	n	PROPN
ajst-16789	61	12	d	d	NOUN
ajst-16789	61	13	filte	filte	NOUN
ajst-16789	62	1	r	r	NOUN
ajst-16789	62	2	r	r	NOUN
ajst-16789	62	3	e	e	X
ajst-16789	62	4	su	su	PROPN
ajst-16789	62	5	lt	lt	PRON
ajst-16789	62	6	figure	figure	NOUN
ajst-16789	62	7	4	4	NUM
ajst-16789	62	8	.	.	PUNCT
ajst-16789	62	9	parcel	parcel	NOUN
ajst-16789	62	10	recognition	recognition	PROPN
ajst-16789	62	11	network	network	NOUN
ajst-16789	62	12	parcelnet	parcelnet	NOUN
ajst-16789	62	13	model	model	NOUN
ajst-16789	62	14	architecture	architecture	NOUN
ajst-16789	62	15	the	the	DET
ajst-16789	62	16	visual	visual	ADJ
ajst-16789	62	17	recognition	recognition	NOUN
ajst-16789	62	18	network	network	NOUN
ajst-16789	62	19	model	model	NOUN
ajst-16789	62	20	adopts	adopt	VERB
ajst-16789	62	21	a	a	DET
ajst-16789	62	22	step	step	NOUN
ajst-16789	62	23	-	-	PUNCT
ajst-16789	62	24	bystep	bystep	NOUN
ajst-16789	62	25	processing	processing	NOUN
ajst-16789	62	26	approach	approach	NOUN
ajst-16789	62	27	to	to	PART
ajst-16789	62	28	handle	handle	VERB
ajst-16789	62	29	the	the	DET
ajst-16789	62	30	collected	collect	VERB
ajst-16789	62	31	express	express	PROPN
ajst-16789	62	32	parcel	parcel	NOUN
ajst-16789	62	33	data	datum	NOUN
ajst-16789	62	34	.	.	PUNCT
ajst-16789	63	1	initially	initially	ADV
ajst-16789	63	2	,	,	PUNCT
ajst-16789	63	3	rgb	rgb	PROPN
ajst-16789	63	4	and	and	CCONJ
ajst-16789	63	5	depth	depth	NOUN
ajst-16789	63	6	data	datum	NOUN
ajst-16789	63	7	are	be	AUX
ajst-16789	63	8	separately	separately	ADV
ajst-16789	63	9	fed	feed	VERB
ajst-16789	63	10	into	into	ADP
ajst-16789	63	11	selected	select	VERB
ajst-16789	63	12	pre	pre	ADJ
ajst-16789	63	13	-	-	ADJ
ajst-16789	63	14	trained	train	VERB
ajst-16789	63	15	residual	residual	ADJ
ajst-16789	63	16	networks	network	NOUN
ajst-16789	63	17	(	(	PUNCT
ajst-16789	63	18	resnet	resnet	NOUN
ajst-16789	63	19	)	)	PUNCT
ajst-16789	63	20	or	or	CCONJ
ajst-16789	63	21	densely	densely	ADV
ajst-16789	63	22	connected	connected	ADJ
ajst-16789	63	23	networks	network	NOUN
ajst-16789	63	24	(	(	PUNCT
ajst-16789	63	25	densenet	densenet	NOUN
ajst-16789	63	26	)	)	PUNCT
ajst-16789	63	27	to	to	PART
ajst-16789	63	28	extract	extract	VERB
ajst-16789	63	29	feature	feature	NOUN
ajst-16789	63	30	information	information	NOUN
ajst-16789	63	31	from	from	ADP
ajst-16789	63	32	the	the	DET
ajst-16789	63	33	images	image	NOUN
ajst-16789	63	34	.	.	PUNCT
ajst-16789	64	1	subsequently	subsequently	ADV
ajst-16789	64	2	,	,	PUNCT
ajst-16789	64	3	these	these	PRON
ajst-16789	64	4	extracted	extract	VERB
ajst-16789	64	5	features	feature	NOUN
ajst-16789	64	6	undergo	undergo	VERB
ajst-16789	64	7	multimodal	multimodal	ADJ
ajst-16789	64	8	fusion	fusion	NOUN
ajst-16789	64	9	to	to	PART
ajst-16789	64	10	obtain	obtain	VERB
ajst-16789	64	11	a	a	DET
ajst-16789	64	12	more	more	ADV
ajst-16789	64	13	comprehensive	comprehensive	ADJ
ajst-16789	64	14	and	and	CCONJ
ajst-16789	64	15	enriched	enriched	ADJ
ajst-16789	64	16	information	information	NOUN
ajst-16789	64	17	representation	representation	NOUN
ajst-16789	64	18	.	.	PUNCT
ajst-16789	65	1	the	the	DET
ajst-16789	65	2	specific	specific	ADJ
ajst-16789	65	3	process	process	NOUN
ajst-16789	65	4	involves	involve	VERB
ajst-16789	65	5	multimodal	multimodal	ADJ
ajst-16789	65	6	concatenation	concatenation	NOUN
ajst-16789	65	7	followed	follow	VERB
ajst-16789	65	8	by	by	ADP
ajst-16789	65	9	a	a	DET
ajst-16789	65	10	series	series	NOUN
ajst-16789	65	11	of	of	ADP
ajst-16789	65	12	convolutional	convolutional	ADJ
ajst-16789	65	13	operations	operation	NOUN
ajst-16789	65	14	to	to	PART
ajst-16789	65	15	fuse	fuse	VERB
ajst-16789	65	16	the	the	DET
ajst-16789	65	17	concatenated	concatenate	VERB
ajst-16789	65	18	features	feature	NOUN
ajst-16789	65	19	,	,	PUNCT
ajst-16789	65	20	including	include	VERB
ajst-16789	65	21	adjustments	adjustment	NOUN
ajst-16789	65	22	in	in	ADP
ajst-16789	65	23	channel	channel	NOUN
ajst-16789	65	24	numbers	number	NOUN
ajst-16789	65	25	,	,	PUNCT
ajst-16789	65	26	upsampling	upsampling	NOUN
ajst-16789	65	27	,	,	PUNCT
ajst-16789	65	28	padding	padding	NOUN
ajst-16789	65	29	,	,	PUNCT
ajst-16789	65	30	and	and	CCONJ
ajst-16789	65	31	other	other	ADJ
ajst-16789	65	32	operations	operation	NOUN
ajst-16789	65	33	,	,	PUNCT
ajst-16789	65	34	completing	complete	VERB
ajst-16789	65	35	the	the	DET
ajst-16789	65	36	feature	feature	NOUN
ajst-16789	65	37	extraction	extraction	NOUN
ajst-16789	65	38	process	process	NOUN
ajst-16789	65	39	.	.	PUNCT
ajst-16789	66	1	once	once	ADV
ajst-16789	66	2	the	the	DET
ajst-16789	66	3	feature	feature	NOUN
ajst-16789	66	4	extraction	extraction	NOUN
ajst-16789	66	5	is	be	AUX
ajst-16789	66	6	completed	complete	VERB
ajst-16789	66	7	,	,	PUNCT
ajst-16789	66	8	a	a	DET
ajst-16789	66	9	background	background	NOUN
ajst-16789	66	10	filtering	filtering	NOUN
ajst-16789	66	11	is	be	AUX
ajst-16789	66	12	applied	apply	VERB
ajst-16789	66	13	to	to	ADP
ajst-16789	66	14	the	the	DET
ajst-16789	66	15	grasp	grasp	NOUN
ajst-16789	66	16	surface	surface	NOUN
ajst-16789	66	17	features	feature	NOUN
ajst-16789	66	18	,	,	PUNCT
ajst-16789	66	19	yielding	yield	VERB
ajst-16789	66	20	the	the	DET
ajst-16789	66	21	foreground	foreground	NOUN
ajst-16789	66	22	grasp	grasp	NOUN
ajst-16789	66	23	surfaces	surface	NOUN
ajst-16789	66	24	.	.	PUNCT
ajst-16789	67	1	this	this	DET
ajst-16789	67	2	process	process	NOUN
ajst-16789	67	3	effectively	effectively	ADV
ajst-16789	67	4	integrates	integrate	VERB
ajst-16789	67	5	the	the	DET
ajst-16789	67	6	multimodal	multimodal	ADJ
ajst-16789	67	7	information	information	NOUN
ajst-16789	67	8	of	of	ADP
ajst-16789	67	9	the	the	DET
ajst-16789	67	10	data	datum	NOUN
ajst-16789	67	11	and	and	CCONJ
ajst-16789	67	12	extracts	extract	NOUN
ajst-16789	67	13	the	the	DET
ajst-16789	67	14	target	target	NOUN
ajst-16789	67	15	grasp	grasp	NOUN
ajst-16789	67	16	surface	surface	NOUN
ajst-16789	67	17	features	feature	NOUN
ajst-16789	67	18	that	that	PRON
ajst-16789	67	19	meet	meet	VERB
ajst-16789	67	20	the	the	DET
ajst-16789	67	21	grasping	grasp	VERB
ajst-16789	67	22	requirements	requirement	NOUN
ajst-16789	67	23	.	.	PUNCT
ajst-16789	68	1	in	in	ADP
ajst-16789	68	2	the	the	DET
ajst-16789	68	3	rapidly	rapidly	ADV
ajst-16789	68	4	evolving	evolve	VERB
ajst-16789	68	5	field	field	NOUN
ajst-16789	68	6	of	of	ADP
ajst-16789	68	7	deep	deep	ADJ
ajst-16789	68	8	learning	learning	NOUN
ajst-16789	68	9	,	,	PUNCT
ajst-16789	68	10	transfer	transfer	NOUN
ajst-16789	68	11	learning	learning	NOUN
ajst-16789	68	12	plays	play	VERB
ajst-16789	68	13	a	a	DET
ajst-16789	68	14	crucial	crucial	ADJ
ajst-16789	68	15	role	role	NOUN
ajst-16789	68	16	in	in	ADP
ajst-16789	68	17	two	two	NUM
ajst-16789	68	18	main	main	ADJ
ajst-16789	68	19	aspects	aspect	NOUN
ajst-16789	68	20	:	:	PUNCT
ajst-16789	68	21	feature	feature	NOUN
ajst-16789	68	22	extraction	extraction	NOUN
ajst-16789	68	23	and	and	CCONJ
ajst-16789	68	24	fine	fine	ADV
ajst-16789	68	25	-	-	PUNCT
ajst-16789	68	26	tuning	tuning	NOUN
ajst-16789	68	27	.	.	PUNCT
ajst-16789	69	1	in	in	ADP
ajst-16789	69	2	feature	feature	NOUN
ajst-16789	69	3	extraction	extraction	NOUN
ajst-16789	69	4	,	,	PUNCT
ajst-16789	69	5	we	we	PRON
ajst-16789	69	6	leverage	leverage	VERB
ajst-16789	69	7	knowledge	knowledge	NOUN
ajst-16789	69	8	acquired	acquire	VERB
ajst-16789	69	9	by	by	ADP
ajst-16789	69	10	pre	pre	ADJ
ajst-16789	69	11	-	-	ADJ
ajst-16789	69	12	trained	train	VERB
ajst-16789	69	13	models	model	NOUN
ajst-16789	69	14	on	on	ADP
ajst-16789	69	15	standard	standard	ADJ
ajst-16789	69	16	datasets	dataset	NOUN
ajst-16789	69	17	(	(	PUNCT
ajst-16789	69	18	such	such	ADJ
ajst-16789	69	19	as	as	ADP
ajst-16789	69	20	imagenet	imagenet	NOUN
ajst-16789	69	21	)	)	PUNCT
ajst-16789	69	22	but	but	CCONJ
ajst-16789	69	23	excluding	exclude	VERB
ajst-16789	69	24	the	the	DET
ajst-16789	69	25	top	top	ADJ
ajst-16789	69	26	-	-	PUNCT
ajst-16789	69	27	level	level	NOUN
ajst-16789	69	28	classification	classification	NOUN
ajst-16789	69	29	structure	structure	NOUN
ajst-16789	69	30	.	.	PUNCT
ajst-16789	70	1	subsequently	subsequently	ADV
ajst-16789	70	2	,	,	PUNCT
ajst-16789	70	3	a	a	DET
ajst-16789	70	4	new	new	ADJ
ajst-16789	70	5	classifier	classifier	NOUN
ajst-16789	70	6	is	be	AUX
ajst-16789	70	7	built	build	VERB
ajst-16789	70	8	on	on	ADP
ajst-16789	70	9	this	this	DET
ajst-16789	70	10	foundation	foundation	NOUN
ajst-16789	70	11	to	to	PART
ajst-16789	70	12	perform	perform	VERB
ajst-16789	70	13	specific	specific	ADJ
ajst-16789	70	14	classification	classification	NOUN
ajst-16789	70	15	tasks	task	NOUN
ajst-16789	70	16	.	.	PUNCT
ajst-16789	71	1	removing	remove	VERB
ajst-16789	71	2	the	the	DET
ajst-16789	71	3	top	top	ADJ
ajst-16789	71	4	-	-	PUNCT
ajst-16789	71	5	level	level	NOUN
ajst-16789	71	6	classifier	classifier	NOUN
ajst-16789	71	7	from	from	ADP
ajst-16789	71	8	the	the	DET
ajst-16789	71	9	pre	pre	ADJ
ajst-16789	71	10	-	-	ADJ
ajst-16789	71	11	trained	train	VERB
ajst-16789	71	12	model	model	NOUN
ajst-16789	71	13	renders	render	VERB
ajst-16789	71	14	it	it	PRON
ajst-16789	71	15	as	as	ADP
ajst-16789	71	16	a	a	DET
ajst-16789	71	17	universal	universal	ADJ
ajst-16789	71	18	feature	feature	NOUN
ajst-16789	71	19	extractor	extractor	NOUN
ajst-16789	71	20	,	,	PUNCT
ajst-16789	71	21	capable	capable	ADJ
ajst-16789	71	22	of	of	ADP
ajst-16789	71	23	extracting	extract	VERB
ajst-16789	71	24	effective	effective	ADJ
ajst-16789	71	25	feature	feature	NOUN
ajst-16789	71	26	information	information	NOUN
ajst-16789	71	27	from	from	ADP
ajst-16789	71	28	new	new	ADJ
ajst-16789	71	29	datasets	dataset	NOUN
ajst-16789	71	30	.	.	PUNCT
ajst-16789	72	1	meanwhile	meanwhile	ADV
ajst-16789	72	2	,	,	PUNCT
ajst-16789	72	3	in	in	ADP
ajst-16789	72	4	fine	fine	ADV
ajst-16789	72	5	-	-	PUNCT
ajst-16789	72	6	tuning	tuning	NOUN
ajst-16789	72	7	,	,	PUNCT
ajst-16789	72	8	the	the	DET
ajst-16789	72	9	weights	weight	NOUN
ajst-16789	72	10	of	of	ADP
ajst-16789	72	11	the	the	DET
ajst-16789	72	12	pre	pre	ADJ
ajst-16789	72	13	-	-	ADJ
ajst-16789	72	14	trained	train	VERB
ajst-16789	72	15	model	model	NOUN
ajst-16789	72	16	are	be	AUX
ajst-16789	72	17	considered	consider	VERB
ajst-16789	72	18	initial	initial	ADJ
ajst-16789	72	19	values	value	NOUN
ajst-16789	72	20	for	for	ADP
ajst-16789	72	21	the	the	DET
ajst-16789	72	22	new	new	ADJ
ajst-16789	72	23	task	task	NOUN
ajst-16789	72	24	and	and	CCONJ
ajst-16789	72	25	are	be	AUX
ajst-16789	72	26	adjusted	adjust	VERB
ajst-16789	72	27	and	and	CCONJ
ajst-16789	72	28	updated	update	VERB
ajst-16789	72	29	during	during	ADP
ajst-16789	72	30	training	training	NOUN
ajst-16789	72	31	.	.	PUNCT
ajst-16789	73	1	this	this	DET
ajst-16789	73	2	approach	approach	NOUN
ajst-16789	73	3	aims	aim	VERB
ajst-16789	73	4	to	to	ADP
ajst-16789	73	5	fine	fine	ADJ
ajst-16789	73	6	-	-	PUNCT
ajst-16789	73	7	tune	tune	NOUN
ajst-16789	73	8	the	the	DET
ajst-16789	73	9	general	general	ADJ
ajst-16789	73	10	feature	feature	NOUN
ajst-16789	73	11	maps	map	NOUN
ajst-16789	73	12	towards	towards	ADP
ajst-16789	73	13	specific	specific	ADJ
ajst-16789	73	14	features	feature	NOUN
ajst-16789	73	15	relevant	relevant	ADJ
ajst-16789	73	16	to	to	ADP
ajst-16789	73	17	the	the	DET
ajst-16789	73	18	new	new	ADJ
ajst-16789	73	19	dataset	dataset	NOUN
ajst-16789	73	20	,	,	PUNCT
ajst-16789	73	21	enabling	enable	VERB
ajst-16789	73	22	the	the	DET
ajst-16789	73	23	adaptation	adaptation	NOUN
ajst-16789	73	24	of	of	ADP
ajst-16789	73	25	universal	universal	ADJ
ajst-16789	73	26	functionalities	functionality	NOUN
ajst-16789	73	27	to	to	ADP
ajst-16789	73	28	specific	specific	ADJ
ajst-16789	73	29	tasks	task	NOUN
ajst-16789	73	30	without	without	ADP
ajst-16789	73	31	overriding	override	VERB
ajst-16789	73	32	general	general	ADJ
ajst-16789	73	33	learning	learning	NOUN
ajst-16789	73	34	.	.	PUNCT
ajst-16789	74	1	the	the	DET
ajst-16789	74	2	transfer	transfer	NOUN
ajst-16789	74	3	learning	learning	NOUN
ajst-16789	74	4	method	method	NOUN
ajst-16789	74	5	employed	employ	VERB
ajst-16789	74	6	in	in	ADP
ajst-16789	74	7	the	the	DET
ajst-16789	74	8	study	study	NOUN
ajst-16789	74	9	is	be	AUX
ajst-16789	74	10	the	the	DET
ajst-16789	74	11	second	second	ADJ
ajst-16789	74	12	approach	approach	NOUN
ajst-16789	74	13	,	,	PUNCT
ajst-16789	74	14	utilizing	utilize	VERB
ajst-16789	74	15	pre	pre	ADJ
ajst-16789	74	16	-	-	ADJ
ajst-16789	74	17	trained	train	VERB
ajst-16789	74	18	weights	weight	NOUN
ajst-16789	74	19	from	from	ADP
ajst-16789	74	20	imagenet	imagenet	NOUN
ajst-16789	74	21	to	to	PART
ajst-16789	74	22	compensate	compensate	VERB
ajst-16789	74	23	for	for	ADP
ajst-16789	74	24	the	the	DET
ajst-16789	74	25	relatively	relatively	ADV
ajst-16789	74	26	smaller	small	ADJ
ajst-16789	74	27	training	training	NOUN
ajst-16789	74	28	dataset	dataset	NOUN
ajst-16789	74	29	.	.	PUNCT
ajst-16789	75	1	through	through	ADP
ajst-16789	75	2	transfer	transfer	NOUN
ajst-16789	75	3	learning	learning	NOUN
ajst-16789	75	4	,	,	PUNCT
ajst-16789	75	5	it	it	PRON
ajst-16789	75	6	helps	help	VERB
ajst-16789	75	7	prevent	prevent	VERB
ajst-16789	75	8	model	model	NOUN
ajst-16789	75	9	overfitting	overfitting	NOUN
ajst-16789	75	10	due	due	ADP
ajst-16789	75	11	to	to	ADP
ajst-16789	75	12	the	the	DET
ajst-16789	75	13	smaller	small	ADJ
ajst-16789	75	14	dataset	dataset	NOUN
ajst-16789	75	15	size	size	NOUN
ajst-16789	75	16	.	.	PUNCT
ajst-16789	76	1	in	in	ADP
ajst-16789	76	2	our	our	PRON
ajst-16789	76	3	study	study	NOUN
ajst-16789	76	4	,	,	PUNCT
ajst-16789	76	5	for	for	ADP
ajst-16789	76	6	the	the	DET
ajst-16789	76	7	backbone	backbone	NOUN
ajst-16789	76	8	depicted	depict	VERB
ajst-16789	76	9	in	in	ADP
ajst-16789	76	10	figure	figure	NOUN
ajst-16789	76	11	4	4	NUM
ajst-16789	76	12	,	,	PUNCT
ajst-16789	76	13	we	we	PRON
ajst-16789	76	14	introduced	introduce	VERB
ajst-16789	76	15	several	several	ADJ
ajst-16789	76	16	mainstream	mainstream	NOUN
ajst-16789	76	17	models	model	NOUN
ajst-16789	76	18	pre	pre	VERB
ajst-16789	76	19	-	-	VERB
ajst-16789	76	20	trained	train	VERB
ajst-16789	76	21	on	on	ADP
ajst-16789	76	22	imagenet	imagenet	NOUN
ajst-16789	76	23	,	,	PUNCT
ajst-16789	76	24	including	include	VERB
ajst-16789	76	25	resnet50	resnet50	NOUN
ajst-16789	76	26	,	,	PUNCT
ajst-16789	76	27	resnet101	resnet101	PROPN
ajst-16789	76	28	,	,	PUNCT
ajst-16789	76	29	densenet121	densenet121	PROPN
ajst-16789	76	30	,	,	PUNCT
ajst-16789	76	31	and	and	CCONJ
ajst-16789	76	32	densenet169	densenet169	PROPN
ajst-16789	76	33	.	.	PUNCT
ajst-16789	77	1	to	to	PART
ajst-16789	77	2	clearly	clearly	ADV
ajst-16789	77	3	distinguish	distinguish	VERB
ajst-16789	77	4	between	between	ADP
ajst-16789	77	5	these	these	DET
ajst-16789	77	6	models	model	NOUN
ajst-16789	77	7	,	,	PUNCT
ajst-16789	77	8	we	we	PRON
ajst-16789	77	9	assigned	assign	VERB
ajst-16789	77	10	unique	unique	ADJ
ajst-16789	77	11	names	name	NOUN
ajst-16789	77	12	to	to	ADP
ajst-16789	77	13	the	the	DET
ajst-16789	77	14	networks	network	NOUN
ajst-16789	77	15	built	build	VERB
ajst-16789	77	16	based	base	VERB
ajst-16789	77	17	on	on	ADP
ajst-16789	77	18	different	different	ADJ
ajst-16789	77	19	pre	pre	ADJ
ajst-16789	77	20	-	-	ADJ
ajst-16789	77	21	trained	train	VERB
ajst-16789	77	22	models	model	NOUN
ajst-16789	77	23	.	.	PUNCT
ajst-16789	78	1	for	for	ADP
ajst-16789	78	2	instance	instance	NOUN
ajst-16789	78	3	,	,	PUNCT
ajst-16789	78	4	the	the	DET
ajst-16789	78	5	network	network	NOUN
ajst-16789	78	6	constructed	construct	VERB
ajst-16789	78	7	using	use	VERB
ajst-16789	78	8	resnet50	resnet50	NOUN
ajst-16789	78	9	is	be	AUX
ajst-16789	78	10	referred	refer	VERB
ajst-16789	78	11	to	to	ADP
ajst-16789	78	12	as	as	ADP
ajst-16789	78	13	parcelnet(r50	parcelnet(r50	VERB
ajst-16789	78	14	)	)	PUNCT
ajst-16789	78	15	,	,	PUNCT
ajst-16789	78	16	the	the	DET
ajst-16789	78	17	one	one	NOUN
ajst-16789	78	18	corresponding	correspond	VERB
ajst-16789	78	19	to	to	ADP
ajst-16789	78	20	resnet101	resnet101	PROPN
ajst-16789	78	21	is	be	AUX
ajst-16789	78	22	named	name	VERB
ajst-16789	78	23	parcelnet(r101	parcelnet(r101	PROPN
ajst-16789	78	24	)	)	PUNCT
ajst-16789	78	25	,	,	PUNCT
ajst-16789	78	26	the	the	DET
ajst-16789	78	27	network	network	NOUN
ajst-16789	78	28	based	base	VERB
ajst-16789	78	29	on	on	ADP
ajst-16789	78	30	densenet121	densenet121	PROPN
ajst-16789	78	31	is	be	AUX
ajst-16789	78	32	labeled	label	VERB
ajst-16789	78	33	parcelnet(d121	parcelnet(d121	NUM
ajst-16789	78	34	)	)	PUNCT
ajst-16789	78	35	,	,	PUNCT
ajst-16789	78	36	and	and	CCONJ
ajst-16789	78	37	the	the	DET
ajst-16789	78	38	network	network	NOUN
ajst-16789	78	39	derived	derive	VERB
ajst-16789	78	40	from	from	ADP
ajst-16789	78	41	densenet169	densenet169	PROPN
ajst-16789	78	42	is	be	AUX
ajst-16789	78	43	designated	designate	VERB
ajst-16789	78	44	as	as	ADP
ajst-16789	78	45	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	78	46	)	)	PUNCT
ajst-16789	78	47	.	.	PUNCT
ajst-16789	79	1	this	this	DET
ajst-16789	79	2	naming	name	VERB
ajst-16789	79	3	convention	convention	NOUN
ajst-16789	79	4	helps	help	VERB
ajst-16789	79	5	accurately	accurately	ADV
ajst-16789	79	6	identify	identify	VERB
ajst-16789	79	7	and	and	CCONJ
ajst-16789	79	8	reference	reference	VERB
ajst-16789	79	9	the	the	DET
ajst-16789	79	10	source	source	NOUN
ajst-16789	79	11	and	and	CCONJ
ajst-16789	79	12	specific	specific	ADJ
ajst-16789	79	13	attributes	attribute	NOUN
ajst-16789	79	14	of	of	ADP
ajst-16789	79	15	the	the	DET
ajst-16789	79	16	utilized	utilize	VERB
ajst-16789	79	17	models	model	NOUN
ajst-16789	79	18	.	.	PUNCT
ajst-16789	80	1	the	the	DET
ajst-16789	80	2	parameter	parameter	PROPN
ajst-16789	80	3	comparison	comparison	NOUN
ajst-16789	80	4	results	result	VERB
ajst-16789	80	5	for	for	ADP
ajst-16789	80	6	these	these	DET
ajst-16789	80	7	network	network	NOUN
ajst-16789	80	8	models	model	NOUN
ajst-16789	80	9	are	be	AUX
ajst-16789	80	10	shown	show	VERB
ajst-16789	80	11	in	in	ADP
ajst-16789	80	12	table	table	NOUN
ajst-16789	80	13	1	1	NUM
ajst-16789	80	14	.	.	PUNCT
ajst-16789	80	15	table	table	NOUN
ajst-16789	80	16	1	1	NUM
ajst-16789	80	17	.	.	PUNCT
ajst-16789	80	18	comparison	comparison	NOUN
ajst-16789	80	19	of	of	ADP
ajst-16789	80	20	model	model	NOUN
ajst-16789	80	21	parameters	parameter	NOUN
ajst-16789	80	22	net	net	ADJ
ajst-16789	80	23	parameter	parameter	NOUN
ajst-16789	80	24	amount	amount	NOUN
ajst-16789	80	25	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	80	26	)	)	PUNCT
ajst-16789	80	27	14254979	14254979	NUM
ajst-16789	80	28	parcelnet(d121	parcelnet(d121	NOUN
ajst-16789	80	29	)	)	PUNCT
ajst-16789	80	30	8068995	8068995	NUM
ajst-16789	80	31	parcelnet(r101	parcelnet(r101	NOUN
ajst-16789	80	32	)	)	PUNCT
ajst-16789	80	33	44663875	44663875	NUM
ajst-16789	80	34	parcelnet(r50	parcelnet(r50	VERB
ajst-16789	80	35	)	)	PUNCT
ajst-16789	80	36	25671747	25671747	NUM
ajst-16789	80	37	the	the	DET
ajst-16789	80	38	table	table	NOUN
ajst-16789	80	39	clearly	clearly	ADV
ajst-16789	80	40	demonstrates	demonstrate	VERB
ajst-16789	80	41	that	that	SCONJ
ajst-16789	80	42	the	the	DET
ajst-16789	80	43	express	express	ADJ
ajst-16789	80	44	parcel	parcel	NOUN
ajst-16789	80	45	detection	detection	NOUN
ajst-16789	80	46	and	and	CCONJ
ajst-16789	80	47	recognition	recognition	NOUN
ajst-16789	80	48	network	network	NOUN
ajst-16789	80	49	model	model	NOUN
ajst-16789	80	50	constructed	construct	VERB
ajst-16789	80	51	using	use	VERB
ajst-16789	80	52	densenet	densenet	NOUN
ajst-16789	80	53	has	have	VERB
ajst-16789	80	54	fewer	few	ADJ
ajst-16789	80	55	parameters	parameter	NOUN
ajst-16789	80	56	compared	compare	VERB
ajst-16789	80	57	to	to	ADP
ajst-16789	80	58	the	the	DET
ajst-16789	80	59	one	one	NUM
ajst-16789	80	60	built	build	VERB
ajst-16789	80	61	using	use	VERB
ajst-16789	80	62	resnet	resnet	NOUN
ajst-16789	80	63	.	.	PUNCT
ajst-16789	81	1	this	this	DET
ajst-16789	81	2	disparity	disparity	NOUN
ajst-16789	81	3	arises	arise	VERB
ajst-16789	81	4	from	from	ADP
ajst-16789	81	5	the	the	DET
ajst-16789	81	6	skip	skip	ADJ
ajst-16789	81	7	connections	connection	NOUN
ajst-16789	81	8	employed	employ	VERB
ajst-16789	81	9	by	by	ADP
ajst-16789	81	10	resnet	resnet	NOUN
ajst-16789	81	11	,	,	PUNCT
ajst-16789	81	12	which	which	PRON
ajst-16789	81	13	increase	increase	VERB
ajst-16789	81	14	the	the	DET
ajst-16789	81	15	overall	overall	ADJ
ajst-16789	81	16	parameter	parameter	NOUN
ajst-16789	81	17	count	count	NOUN
ajst-16789	81	18	.	.	PUNCT
ajst-16789	82	1	these	these	DET
ajst-16789	82	2	connections	connection	NOUN
ajst-16789	82	3	necessitate	necessitate	ADJ
ajst-16789	82	4	additional	additional	ADJ
ajst-16789	82	5	weights	weight	NOUN
ajst-16789	82	6	to	to	PART
ajst-16789	82	7	learn	learn	VERB
ajst-16789	82	8	residuals	residual	NOUN
ajst-16789	82	9	for	for	ADP
ajst-16789	82	10	gradient	gradient	ADJ
ajst-16789	82	11	propagation	propagation	NOUN
ajst-16789	82	12	and	and	CCONJ
ajst-16789	82	13	alleviation	alleviation	NOUN
ajst-16789	82	14	of	of	ADP
ajst-16789	82	15	gradient	gradient	ADJ
ajst-16789	82	16	vanishing	vanish	VERB
ajst-16789	82	17	issues	issue	NOUN
ajst-16789	82	18	.	.	PUNCT
ajst-16789	83	1	in	in	ADP
ajst-16789	83	2	contrast	contrast	NOUN
ajst-16789	83	3	,	,	PUNCT
ajst-16789	83	4	densenet	densenet	NOUN
ajst-16789	83	5	utilizes	utilize	VERB
ajst-16789	83	6	a	a	DET
ajst-16789	83	7	dense	dense	ADJ
ajst-16789	83	8	connectivity	connectivity	NOUN
ajst-16789	83	9	structure	structure	NOUN
ajst-16789	83	10	,	,	PUNCT
ajst-16789	83	11	directly	directly	ADV
ajst-16789	83	12	linking	link	VERB
ajst-16789	83	13	each	each	DET
ajst-16789	83	14	layer	layer	NOUN
ajst-16789	83	15	with	with	ADP
ajst-16789	83	16	all	all	DET
ajst-16789	83	17	other	other	ADJ
ajst-16789	83	18	layers	layer	NOUN
ajst-16789	83	19	by	by	ADP
ajst-16789	83	20	concatenating	concatenate	VERB
ajst-16789	83	21	outputs	output	NOUN
ajst-16789	83	22	from	from	ADP
ajst-16789	83	23	preceding	precede	VERB
ajst-16789	83	24	layers	layer	NOUN
ajst-16789	83	25	as	as	ADP
ajst-16789	83	26	inputs	input	NOUN
ajst-16789	83	27	to	to	ADP
ajst-16789	83	28	subsequent	subsequent	ADJ
ajst-16789	83	29	layers	layer	NOUN
ajst-16789	83	30	.	.	PUNCT
ajst-16789	84	1	this	this	PRON
ajst-16789	84	2	facilitates	facilitate	VERB
ajst-16789	84	3	more	more	ADV
ajst-16789	84	4	efficient	efficient	ADJ
ajst-16789	84	5	parameter	parameter	NOUN
ajst-16789	84	6	sharing	sharing	NOUN
ajst-16789	84	7	and	and	CCONJ
ajst-16789	84	8	reuse	reuse	NOUN
ajst-16789	84	9	,	,	PUNCT
ajst-16789	84	10	allowing	allow	VERB
ajst-16789	84	11	densenet	densenet	NOUN
ajst-16789	84	12	to	to	PART
ajst-16789	84	13	effectively	effectively	ADV
ajst-16789	84	14	utilize	utilize	VERB
ajst-16789	84	15	parameters	parameter	NOUN
ajst-16789	84	16	to	to	ADP
ajst-16789	84	17	a	a	DET
ajst-16789	84	18	greater	great	ADJ
ajst-16789	84	19	extent	extent	NOUN
ajst-16789	84	20	.	.	PUNCT
ajst-16789	85	1	this	this	DET
ajst-16789	85	2	design	design	NOUN
ajst-16789	85	3	enables	enable	VERB
ajst-16789	85	4	densenet	densenet	NOUN
ajst-16789	85	5	to	to	PART
ajst-16789	85	6	achieve	achieve	VERB
ajst-16789	85	7	greater	great	ADJ
ajst-16789	85	8	computational	computational	ADJ
ajst-16789	85	9	efficiency	efficiency	NOUN
ajst-16789	85	10	despite	despite	SCONJ
ajst-16789	85	11	having	have	VERB
ajst-16789	85	12	several	several	ADJ
ajst-16789	85	13	times	time	NOUN
ajst-16789	85	14	more	more	ADJ
ajst-16789	85	15	layers	layer	NOUN
ajst-16789	85	16	than	than	SCONJ
ajst-16789	85	17	resnet	resnet	NOUN
ajst-16789	85	18	.	.	PUNCT
ajst-16789	86	1	although	although	SCONJ
ajst-16789	86	2	densenet	densenet	NOUN
ajst-16789	86	3	boasts	boast	VERB
ajst-16789	86	4	multiple	multiple	ADJ
ajst-16789	86	5	layers	layer	NOUN
ajst-16789	86	6	compared	compare	VERB
ajst-16789	86	7	to	to	ADP
ajst-16789	86	8	resnet	resnet	NOUN
ajst-16789	86	9	,	,	PUNCT
ajst-16789	86	10	it	it	PRON
ajst-16789	86	11	exhibits	exhibit	VERB
ajst-16789	86	12	lower	low	ADJ
ajst-16789	86	13	computational	computational	ADJ
ajst-16789	86	14	overhead	overhead	NOUN
ajst-16789	86	15	.	.	PUNCT
ajst-16789	87	1	4	4	X
ajst-16789	87	2	.	.	PUNCT
ajst-16789	87	3	model	model	NOUN
ajst-16789	87	4	training	training	NOUN
ajst-16789	87	5	and	and	CCONJ
ajst-16789	87	6	testing	testing	NOUN
ajst-16789	87	7	utilizing	utilize	VERB
ajst-16789	87	8	the	the	DET
ajst-16789	87	9	previously	previously	ADV
ajst-16789	87	10	annotated	annotate	VERB
ajst-16789	87	11	2,842	2,842	NUM
ajst-16789	87	12	images	image	NOUN
ajst-16789	87	13	along	along	ADP
ajst-16789	87	14	with	with	ADP
ajst-16789	87	15	their	their	PRON
ajst-16789	87	16	corresponding	corresponding	ADJ
ajst-16789	87	17	labels	label	NOUN
ajst-16789	87	18	,	,	PUNCT
ajst-16789	87	19	we	we	PRON
ajst-16789	87	20	randomly	randomly	ADV
ajst-16789	87	21	partitioned	partition	VERB
ajst-16789	87	22	the	the	DET
ajst-16789	87	23	data	datum	NOUN
ajst-16789	87	24	into	into	ADP
ajst-16789	87	25	training	training	NOUN
ajst-16789	87	26	and	and	CCONJ
ajst-16789	87	27	test	test	NOUN
ajst-16789	87	28	sets	set	NOUN
ajst-16789	87	29	in	in	ADP
ajst-16789	87	30	a	a	DET
ajst-16789	87	31	ratio	ratio	NOUN
ajst-16789	87	32	of	of	ADP
ajst-16789	87	33	8:2	8:2	NUM
ajst-16789	87	34	.	.	PUNCT
ajst-16789	88	1	subsequently	subsequently	ADV
ajst-16789	88	2	,	,	PUNCT
ajst-16789	88	3	these	these	DET
ajst-16789	88	4	201	201	NUM
ajst-16789	88	5	datasets	dataset	NOUN
ajst-16789	88	6	were	be	AUX
ajst-16789	88	7	fed	feed	VERB
ajst-16789	88	8	into	into	ADP
ajst-16789	88	9	the	the	DET
ajst-16789	88	10	model	model	NOUN
ajst-16789	88	11	for	for	ADP
ajst-16789	88	12	training	training	NOUN
ajst-16789	88	13	,	,	PUNCT
ajst-16789	88	14	followed	follow	VERB
ajst-16789	88	15	by	by	ADP
ajst-16789	88	16	testing	test	VERB
ajst-16789	88	17	post	post	NOUN
ajst-16789	88	18	-	-	NOUN
ajst-16789	88	19	training	training	NOUN
ajst-16789	88	20	to	to	PART
ajst-16789	88	21	evaluate	evaluate	VERB
ajst-16789	88	22	the	the	DET
ajst-16789	88	23	performance	performance	NOUN
ajst-16789	88	24	of	of	ADP
ajst-16789	88	25	each	each	DET
ajst-16789	88	26	model	model	NOUN
ajst-16789	88	27	.	.	PUNCT
ajst-16789	89	1	4.1	4.1	NUM
ajst-16789	89	2	.	.	PUNCT
ajst-16789	89	3	model	model	NOUN
ajst-16789	89	4	training	training	NOUN
ajst-16789	89	5	in	in	ADP
ajst-16789	89	6	this	this	DET
ajst-16789	89	7	study	study	NOUN
ajst-16789	89	8	's	's	PART
ajst-16789	89	9	experiments	experiment	NOUN
ajst-16789	89	10	,	,	PUNCT
ajst-16789	89	11	we	we	PRON
ajst-16789	89	12	utilized	utilize	VERB
ajst-16789	89	13	the	the	DET
ajst-16789	89	14	intel(r	intel(r	PROPN
ajst-16789	89	15	)	)	PUNCT
ajst-16789	89	16	xeon(r	xeon(r	NOUN
ajst-16789	89	17	)	)	PUNCT
ajst-16789	89	18	gold	gold	NOUN
ajst-16789	89	19	6234	6234	NUM
ajst-16789	89	20	processor	processor	NOUN
ajst-16789	89	21	and	and	CCONJ
ajst-16789	89	22	executed	execute	VERB
ajst-16789	89	23	operations	operation	NOUN
ajst-16789	89	24	on	on	ADP
ajst-16789	89	25	an	an	DET
ajst-16789	89	26	nvidia	nvidia	PROPN
ajst-16789	89	27	gtx	gtx	ADJ
ajst-16789	89	28	2080	2080	NUM
ajst-16789	89	29	ti	ti	X
ajst-16789	89	30	gpu	gpu	PROPN
ajst-16789	89	31	.	.	PUNCT
ajst-16789	90	1	the	the	DET
ajst-16789	90	2	chosen	choose	VERB
ajst-16789	90	3	deep	deep	ADJ
ajst-16789	90	4	learning	learning	NOUN
ajst-16789	90	5	framework	framework	NOUN
ajst-16789	90	6	was	be	AUX
ajst-16789	90	7	pytorch	pytorch	NOUN
ajst-16789	90	8	1.12	1.12	NUM
ajst-16789	90	9	,	,	PUNCT
ajst-16789	90	10	with	with	ADP
ajst-16789	90	11	cuda	cuda	PROPN
ajst-16789	90	12	version	version	NOUN
ajst-16789	90	13	12.0	12.0	NUM
ajst-16789	90	14	,	,	PUNCT
ajst-16789	90	15	and	and	CCONJ
ajst-16789	90	16	the	the	DET
ajst-16789	90	17	model	model	NOUN
ajst-16789	90	18	training	training	NOUN
ajst-16789	90	19	was	be	AUX
ajst-16789	90	20	conducted	conduct	VERB
ajst-16789	90	21	using	use	VERB
ajst-16789	90	22	python	python	NOUN
ajst-16789	90	23	version	version	NOUN
ajst-16789	90	24	3.8	3.8	NUM
ajst-16789	90	25	.	.	PUNCT
ajst-16789	91	1	the	the	DET
ajst-16789	91	2	cross	cross	ADJ
ajst-16789	91	3	-	-	ADJ
ajst-16789	91	4	entropy	entropy	ADJ
ajst-16789	91	5	loss	loss	NOUN
ajst-16789	91	6	function	function	NOUN
ajst-16789	91	7	,	,	PUNCT
ajst-16789	91	8	sometimes	sometimes	ADV
ajst-16789	91	9	referred	refer	VERB
ajst-16789	91	10	to	to	ADP
ajst-16789	91	11	as	as	ADP
ajst-16789	91	12	pixel	pixel	ADJ
ajst-16789	91	13	-	-	ADJ
ajst-16789	91	14	wise	wise	ADJ
ajst-16789	91	15	cross	cross	ADJ
ajst-16789	91	16	-	-	ADJ
ajst-16789	91	17	entropy	entropy	ADJ
ajst-16789	91	18	loss	loss	NOUN
ajst-16789	91	19	function	function	NOUN
ajst-16789	91	20	,	,	PUNCT
ajst-16789	91	21	was	be	AUX
ajst-16789	91	22	employed	employ	VERB
ajst-16789	91	23	in	in	ADP
ajst-16789	91	24	our	our	PRON
ajst-16789	91	25	approach	approach	NOUN
ajst-16789	91	26	.	.	PUNCT
ajst-16789	92	1	the	the	DET
ajst-16789	92	2	primary	primary	ADJ
ajst-16789	92	3	objective	objective	NOUN
ajst-16789	92	4	of	of	ADP
ajst-16789	92	5	this	this	DET
ajst-16789	92	6	loss	loss	NOUN
ajst-16789	92	7	function	function	NOUN
ajst-16789	92	8	is	be	AUX
ajst-16789	92	9	to	to	PART
ajst-16789	92	10	compare	compare	VERB
ajst-16789	92	11	the	the	DET
ajst-16789	92	12	model	model	NOUN
ajst-16789	92	13	's	's	PART
ajst-16789	92	14	predicted	predict	VERB
ajst-16789	92	15	class	class	NOUN
ajst-16789	92	16	probabilities	probability	NOUN
ajst-16789	92	17	for	for	ADP
ajst-16789	92	18	each	each	DET
ajst-16789	92	19	pixel	pixel	NOUN
ajst-16789	92	20	position	position	NOUN
ajst-16789	92	21	with	with	ADP
ajst-16789	92	22	the	the	DET
ajst-16789	92	23	actual	actual	ADJ
ajst-16789	92	24	pixel	pixel	ADJ
ajst-16789	92	25	-	-	PUNCT
ajst-16789	92	26	level	level	NOUN
ajst-16789	92	27	labels	label	NOUN
ajst-16789	92	28	,	,	PUNCT
ajst-16789	92	29	ensuring	ensure	VERB
ajst-16789	92	30	accurate	accurate	ADJ
ajst-16789	92	31	pixel	pixel	ADJ
ajst-16789	92	32	-	-	ADJ
ajst-16789	92	33	wise	wise	ADJ
ajst-16789	92	34	classification	classification	NOUN
ajst-16789	92	35	by	by	ADP
ajst-16789	92	36	the	the	DET
ajst-16789	92	37	model	model	NOUN
ajst-16789	92	38	.	.	PUNCT
ajst-16789	93	1	the	the	DET
ajst-16789	93	2	mathematical	mathematical	ADJ
ajst-16789	93	3	expression	expression	NOUN
ajst-16789	93	4	for	for	ADP
ajst-16789	93	5	the	the	DET
ajst-16789	93	6	cross	cross	ADJ
ajst-16789	93	7	-	-	ADJ
ajst-16789	93	8	entropy	entropy	ADJ
ajst-16789	93	9	loss	loss	NOUN
ajst-16789	93	10	function	function	NOUN
ajst-16789	93	11	is	be	AUX
ajst-16789	93	12	as	as	SCONJ
ajst-16789	93	13	follows	follow	VERB
ajst-16789	93	14	:	:	PUNCT
ajst-16789	93	15	ce	ce	PROPN
ajst-16789	93	16	(	(	PUNCT
ajst-16789	93	17	,	,	PUNCT
ajst-16789	93	18	)	)	PUNCT
ajst-16789	93	19	log	log	NOUN
ajst-16789	93	20	(	(	PUNCT
ajst-16789	93	21	)	)	PUNCT
ajst-16789	93	22	i	i	PRON
ajst-16789	93	23	i	i	PRON
ajst-16789	94	1	i	i	PRON
ajst-16789	94	2	p	p	VERB
ajst-16789	94	3	q	q	X
ajst-16789	94	4	p	p	X
ajst-16789	94	5	q	q	NOUN
ajst-16789	94	6			PROPN
ajst-16789	94	7	in	in	ADP
ajst-16789	94	8	this	this	DET
ajst-16789	94	9	study	study	NOUN
ajst-16789	94	10	,	,	PUNCT
ajst-16789	94	11	we	we	PRON
ajst-16789	94	12	utilized	utilize	VERB
ajst-16789	94	13	hyperparameters	hyperparameter	NOUN
ajst-16789	94	14	set	set	VERB
ajst-16789	94	15	at	at	ADP
ajst-16789	94	16	a	a	DET
ajst-16789	94	17	learning	learning	NOUN
ajst-16789	94	18	rate	rate	NOUN
ajst-16789	94	19	of	of	ADP
ajst-16789	94	20	0.001	0.001	NUM
ajst-16789	94	21	,	,	PUNCT
ajst-16789	94	22	an	an	DET
ajst-16789	94	23	assumed	assumed	ADJ
ajst-16789	94	24	training	training	NOUN
ajst-16789	94	25	cycle	cycle	NOUN
ajst-16789	94	26	of	of	ADP
ajst-16789	94	27	200	200	NUM
ajst-16789	94	28	epochs	epoch	NOUN
ajst-16789	94	29	,	,	PUNCT
ajst-16789	94	30	and	and	CCONJ
ajst-16789	94	31	a	a	DET
ajst-16789	94	32	weight	weight	NOUN
ajst-16789	94	33	decay	decay	NOUN
ajst-16789	94	34	coefficient	coefficient	NOUN
ajst-16789	94	35	of	of	ADP
ajst-16789	94	36	1e-4	1e-4	PROPN
ajst-16789	94	37	.	.	PUNCT
ajst-16789	94	38	to	to	PART
ajst-16789	94	39	prevent	prevent	VERB
ajst-16789	94	40	model	model	NOUN
ajst-16789	94	41	overfitting	overfitting	NOUN
ajst-16789	94	42	,	,	PUNCT
ajst-16789	94	43	an	an	DET
ajst-16789	94	44	early	early	ADJ
ajst-16789	94	45	stopping	stopping	NOUN
ajst-16789	94	46	strategy	strategy	NOUN
ajst-16789	94	47	was	be	AUX
ajst-16789	94	48	implemented	implement	VERB
ajst-16789	94	49	.	.	PUNCT
ajst-16789	95	1	we	we	PRON
ajst-16789	95	2	trained	train	VERB
ajst-16789	95	3	the	the	DET
ajst-16789	95	4	training	training	NOUN
ajst-16789	95	5	dataset	dataset	NOUN
ajst-16789	95	6	using	use	VERB
ajst-16789	95	7	parcelnet(r50	parcelnet(r50	NOUN
ajst-16789	95	8	)	)	PUNCT
ajst-16789	95	9	,	,	PUNCT
ajst-16789	95	10	parcelnet(r101	parcelnet(r101	PROPN
ajst-16789	95	11	)	)	PUNCT
ajst-16789	95	12	,	,	PUNCT
ajst-16789	95	13	parcelnet(d121	parcelnet(d121	NOUN
ajst-16789	95	14	)	)	PUNCT
ajst-16789	95	15	,	,	PUNCT
ajst-16789	95	16	and	and	CCONJ
ajst-16789	95	17	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	95	18	)	)	PUNCT
ajst-16789	95	19	.	.	PUNCT
ajst-16789	96	1	after	after	ADP
ajst-16789	96	2	150	150	NUM
ajst-16789	96	3	training	training	NOUN
ajst-16789	96	4	batches	batch	NOUN
ajst-16789	96	5	,	,	PUNCT
ajst-16789	96	6	the	the	DET
ajst-16789	96	7	training	training	NOUN
ajst-16789	96	8	loss	loss	NOUN
ajst-16789	96	9	for	for	ADP
ajst-16789	96	10	all	all	DET
ajst-16789	96	11	models	model	NOUN
ajst-16789	96	12	dropped	drop	VERB
ajst-16789	96	13	to	to	ADP
ajst-16789	96	14	1	1	NUM
ajst-16789	96	15	%	%	NOUN
ajst-16789	96	16	and	and	CCONJ
ajst-16789	96	17	stabilized	stabilize	VERB
ajst-16789	96	18	.	.	PUNCT
ajst-16789	97	1	notably	notably	ADV
ajst-16789	97	2	,	,	PUNCT
ajst-16789	97	3	densenet169	densenet169	PROPN
ajst-16789	97	4	exhibited	exhibit	VERB
ajst-16789	97	5	the	the	DET
ajst-16789	97	6	lowest	low	ADJ
ajst-16789	97	7	loss	loss	NOUN
ajst-16789	97	8	rate	rate	NOUN
ajst-16789	97	9	,	,	PUNCT
ajst-16789	97	10	decreasing	decrease	VERB
ajst-16789	97	11	to	to	ADP
ajst-16789	97	12	0.5	0.5	NUM
ajst-16789	97	13	%	%	NOUN
ajst-16789	97	14	(	(	PUNCT
ajst-16789	97	15	refer	refer	VERB
ajst-16789	97	16	to	to	PART
ajst-16789	97	17	figure	figure	VERB
ajst-16789	97	18	5	5	NUM
ajst-16789	97	19	)	)	PUNCT
ajst-16789	97	20	.	.	PUNCT
ajst-16789	98	1	subsequently	subsequently	ADV
ajst-16789	98	2	,	,	PUNCT
ajst-16789	98	3	we	we	PRON
ajst-16789	98	4	tested	test	VERB
ajst-16789	98	5	these	these	DET
ajst-16789	98	6	trained	train	VERB
ajst-16789	98	7	models	model	NOUN
ajst-16789	98	8	on	on	ADP
ajst-16789	98	9	a	a	DET
ajst-16789	98	10	test	test	NOUN
ajst-16789	98	11	set	set	NOUN
ajst-16789	98	12	containing	contain	VERB
ajst-16789	98	13	527	527	NUM
ajst-16789	98	14	images	image	NOUN
ajst-16789	98	15	,	,	PUNCT
ajst-16789	98	16	obtaining	obtain	VERB
ajst-16789	98	17	the	the	DET
ajst-16789	98	18	top1	top1	PROPN
ajst-16789	98	19	graspable	graspable	ADJ
ajst-16789	98	20	surface	surface	NOUN
ajst-16789	98	21	detection	detection	NOUN
ajst-16789	98	22	accuracy	accuracy	NOUN
ajst-16789	98	23	for	for	ADP
ajst-16789	98	24	each	each	DET
ajst-16789	98	25	model	model	NOUN
ajst-16789	98	26	,	,	PUNCT
ajst-16789	98	27	as	as	SCONJ
ajst-16789	98	28	presented	present	VERB
ajst-16789	98	29	in	in	ADP
ajst-16789	98	30	table	table	NOUN
ajst-16789	98	31	2	2	NUM
ajst-16789	98	32	.	.	PUNCT
ajst-16789	99	1	these	these	DET
ajst-16789	99	2	results	result	NOUN
ajst-16789	99	3	offer	offer	VERB
ajst-16789	99	4	significant	significant	ADJ
ajst-16789	99	5	empirical	empirical	ADJ
ajst-16789	99	6	evidence	evidence	NOUN
ajst-16789	99	7	supporting	support	VERB
ajst-16789	99	8	our	our	PRON
ajst-16789	99	9	choices	choice	NOUN
ajst-16789	99	10	.	.	PUNCT
ajst-16789	100	1	figure	figure	NOUN
ajst-16789	100	2	5	5	NUM
ajst-16789	100	3	.	.	PUNCT
ajst-16789	101	1	training	training	NOUN
ajst-16789	101	2	loss	loss	NOUN
ajst-16789	101	3	graph	graph	NOUN
ajst-16789	101	4	4.2	4.2	NUM
ajst-16789	101	5	.	.	PUNCT
ajst-16789	102	1	testing	testing	NOUN
ajst-16789	102	2	and	and	CCONJ
ajst-16789	102	3	evaluation	evaluation	NOUN
ajst-16789	102	4	in	in	ADP
ajst-16789	102	5	evaluating	evaluate	VERB
ajst-16789	102	6	the	the	DET
ajst-16789	102	7	model	model	NOUN
ajst-16789	102	8	's	's	PART
ajst-16789	102	9	performance	performance	NOUN
ajst-16789	102	10	,	,	PUNCT
ajst-16789	102	11	various	various	ADJ
ajst-16789	102	12	metrics	metric	NOUN
ajst-16789	102	13	come	come	VERB
ajst-16789	102	14	into	into	ADP
ajst-16789	102	15	play	play	NOUN
ajst-16789	102	16	,	,	PUNCT
ajst-16789	102	17	including	include	VERB
ajst-16789	102	18	overall	overall	ADJ
ajst-16789	102	19	classification	classification	NOUN
ajst-16789	102	20	accuracy	accuracy	NOUN
ajst-16789	102	21	,	,	PUNCT
ajst-16789	102	22	recall	recall	NOUN
ajst-16789	102	23	(	(	PUNCT
ajst-16789	102	24	also	also	ADV
ajst-16789	102	25	known	know	VERB
ajst-16789	102	26	as	as	ADP
ajst-16789	102	27	sensitivity	sensitivity	NOUN
ajst-16789	102	28	)	)	PUNCT
ajst-16789	102	29	,	,	PUNCT
ajst-16789	102	30	and	and	CCONJ
ajst-16789	102	31	precision	precision	NOUN
ajst-16789	102	32	.	.	PUNCT
ajst-16789	103	1	in	in	ADP
ajst-16789	103	2	the	the	DET
ajst-16789	103	3	context	context	NOUN
ajst-16789	103	4	of	of	ADP
ajst-16789	103	5	recognizing	recognize	VERB
ajst-16789	103	6	and	and	CCONJ
ajst-16789	103	7	grasping	grasp	VERB
ajst-16789	103	8	express	express	ADJ
ajst-16789	103	9	parcels	parcel	NOUN
ajst-16789	103	10	with	with	ADP
ajst-16789	103	11	a	a	DET
ajst-16789	103	12	robotic	robotic	ADJ
ajst-16789	103	13	arm	arm	NOUN
ajst-16789	103	14	,	,	PUNCT
ajst-16789	103	15	as	as	SCONJ
ajst-16789	103	16	the	the	DET
ajst-16789	103	17	arm	arm	NOUN
ajst-16789	103	18	typically	typically	ADV
ajst-16789	103	19	handles	handle	VERB
ajst-16789	103	20	one	one	NUM
ajst-16789	103	21	parcel	parcel	NOUN
ajst-16789	103	22	at	at	ADP
ajst-16789	103	23	a	a	DET
ajst-16789	103	24	time	time	NOUN
ajst-16789	103	25	,	,	PUNCT
ajst-16789	103	26	the	the	DET
ajst-16789	103	27	robustness	robustness	NOUN
ajst-16789	103	28	of	of	ADP
ajst-16789	103	29	the	the	DET
ajst-16789	103	30	object	object	NOUN
ajst-16789	103	31	detection	detection	NOUN
ajst-16789	103	32	network	network	NOUN
ajst-16789	103	33	lies	lie	VERB
ajst-16789	103	34	in	in	ADP
ajst-16789	103	35	its	its	PRON
ajst-16789	103	36	continuous	continuous	ADJ
ajst-16789	103	37	ability	ability	NOUN
ajst-16789	103	38	to	to	PART
ajst-16789	103	39	correctly	correctly	ADV
ajst-16789	103	40	detect	detect	VERB
ajst-16789	103	41	at	at	ADV
ajst-16789	103	42	least	least	ADV
ajst-16789	103	43	one	one	NUM
ajst-16789	103	44	graspable	graspable	ADJ
ajst-16789	103	45	surface	surface	NOUN
ajst-16789	103	46	.	.	PUNCT
ajst-16789	104	1	reflecting	reflect	VERB
ajst-16789	104	2	this	this	PRON
ajst-16789	104	3	,	,	PUNCT
ajst-16789	104	4	our	our	PRON
ajst-16789	104	5	evaluation	evaluation	NOUN
ajst-16789	104	6	metric	metric	NOUN
ajst-16789	104	7	is	be	AUX
ajst-16789	104	8	the	the	DET
ajst-16789	104	9	precision	precision	NOUN
ajst-16789	104	10	of	of	ADP
ajst-16789	104	11	inference	inference	NOUN
ajst-16789	104	12	proposals	proposal	NOUN
ajst-16789	104	13	against	against	ADP
ajst-16789	104	14	manually	manually	ADV
ajst-16789	104	15	annotated	annotate	VERB
ajst-16789	104	16	accuracy	accuracy	NOUN
ajst-16789	104	17	.	.	PUNCT
ajst-16789	105	1	if	if	SCONJ
ajst-16789	105	2	the	the	DET
ajst-16789	105	3	center	center	NOUN
ajst-16789	105	4	pixel	pixel	NOUN
ajst-16789	105	5	of	of	ADP
ajst-16789	105	6	a	a	DET
ajst-16789	105	7	proposal	proposal	NOUN
ajst-16789	105	8	is	be	AUX
ajst-16789	105	9	manually	manually	ADV
ajst-16789	105	10	labeled	label	VERB
ajst-16789	105	11	as	as	ADP
ajst-16789	105	12	a	a	DET
ajst-16789	105	13	suction	suction	NOUN
ajst-16789	105	14	-	-	PUNCT
ajst-16789	105	15	cup	cup	NOUN
ajst-16789	105	16	graspable	graspable	ADJ
ajst-16789	105	17	area	area	NOUN
ajst-16789	105	18	,	,	PUNCT
ajst-16789	105	19	it	it	PRON
ajst-16789	105	20	's	be	AUX
ajst-16789	105	21	considered	consider	VERB
ajst-16789	105	22	a	a	DET
ajst-16789	105	23	true	true	ADJ
ajst-16789	105	24	positive	positive	ADJ
ajst-16789	105	25	(	(	PUNCT
ajst-16789	105	26	tp	tp	NOUN
ajst-16789	105	27	)	)	PUNCT
ajst-16789	105	28	.	.	PUNCT
ajst-16789	106	1	if	if	SCONJ
ajst-16789	106	2	labeled	label	VERB
ajst-16789	106	3	as	as	ADP
ajst-16789	106	4	a	a	DET
ajst-16789	106	5	non	non	ADJ
ajst-16789	106	6	-	-	ADJ
ajst-16789	106	7	graspable	graspable	ADJ
ajst-16789	106	8	area	area	NOUN
ajst-16789	106	9	but	but	CCONJ
ajst-16789	106	10	detected	detect	VERB
ajst-16789	106	11	as	as	ADP
ajst-16789	106	12	graspable	graspable	ADJ
ajst-16789	106	13	,	,	PUNCT
ajst-16789	106	14	it	it	PRON
ajst-16789	106	15	's	be	AUX
ajst-16789	106	16	a	a	DET
ajst-16789	106	17	false	false	ADJ
ajst-16789	106	18	positive	positive	ADJ
ajst-16789	106	19	(	(	PUNCT
ajst-16789	106	20	fp	fp	NOUN
ajst-16789	106	21	)	)	PUNCT
ajst-16789	106	22	.	.	PUNCT
ajst-16789	107	1	among	among	ADP
ajst-16789	107	2	the	the	DET
ajst-16789	107	3	detected	detect	VERB
ajst-16789	107	4	graspable	graspable	ADJ
ajst-16789	107	5	regions	region	NOUN
ajst-16789	107	6	,	,	PUNCT
ajst-16789	107	7	the	the	DET
ajst-16789	107	8	one	one	NOUN
ajst-16789	107	9	with	with	ADP
ajst-16789	107	10	the	the	DET
ajst-16789	107	11	largest	large	ADJ
ajst-16789	107	12	area	area	NOUN
ajst-16789	107	13	is	be	AUX
ajst-16789	107	14	designated	designate	VERB
ajst-16789	107	15	as	as	ADP
ajst-16789	107	16	top1	top1	PROPN
ajst-16789	107	17	.	.	PUNCT
ajst-16789	108	1	therefore	therefore	ADV
ajst-16789	108	2	,	,	PUNCT
ajst-16789	108	3	we	we	PRON
ajst-16789	108	4	focus	focus	VERB
ajst-16789	108	5	on	on	ADP
ajst-16789	108	6	the	the	DET
ajst-16789	108	7	precision	precision	NOUN
ajst-16789	108	8	of	of	ADP
ajst-16789	108	9	top1	top1	PROPN
ajst-16789	108	10	graspable	graspable	ADJ
ajst-16789	108	11	areas	area	NOUN
ajst-16789	108	12	as	as	ADP
ajst-16789	108	13	the	the	DET
ajst-16789	108	14	performance	performance	NOUN
ajst-16789	108	15	metric	metric	ADJ
ajst-16789	108	16	,	,	PUNCT
ajst-16789	108	17	utilizing	utilize	VERB
ajst-16789	108	18	top1	top1	PROPN
ajst-16789	108	19	true	true	ADJ
ajst-16789	108	20	positive	positive	ADJ
ajst-16789	108	21	accuracy	accuracy	NOUN
ajst-16789	108	22	to	to	PART
ajst-16789	108	23	assessre	assessre	VERB
ajst-16789	108	24	cognition	cognition	NOUN
ajst-16789	108	25	performance	performance	NOUN
ajst-16789	108	26	,	,	PUNCT
ajst-16789	108	27	expressedas	expresseda	NOUN
ajst-16789	108	28	:	:	PUNCT
ajst-16789	108	29			NOUN
ajst-16789	108	30	1	1	ADJ
ajst-16789	108	31	top	top	NOUN
ajst-16789	108	32	tp	tp	ADP
ajst-16789	108	33	precision	precision	NOUN
ajst-16789	108	34	tp	tp	ADP
ajst-16789	108	35	fp	fp	PROPN
ajst-16789	108	36			PROPN
ajst-16789	108	37			PUNCT
ajst-16789	108	38	after	after	ADP
ajst-16789	108	39	completing	complete	VERB
ajst-16789	108	40	model	model	NOUN
ajst-16789	108	41	training	training	NOUN
ajst-16789	108	42	,	,	PUNCT
ajst-16789	108	43	we	we	PRON
ajst-16789	108	44	evaluated	evaluate	VERB
ajst-16789	108	45	the	the	DET
ajst-16789	108	46	models	model	NOUN
ajst-16789	108	47	using	use	VERB
ajst-16789	108	48	a	a	DET
ajst-16789	108	49	test	test	NOUN
ajst-16789	108	50	set	set	NOUN
ajst-16789	108	51	comprising	comprise	VERB
ajst-16789	108	52	568	568	NUM
ajst-16789	108	53	images	image	NOUN
ajst-16789	108	54	to	to	PART
ajst-16789	108	55	validate	validate	VERB
ajst-16789	108	56	their	their	PRON
ajst-16789	108	57	accuracy	accuracy	NOUN
ajst-16789	108	58	.	.	PUNCT
ajst-16789	109	1	below	below	ADV
ajst-16789	109	2	are	be	AUX
ajst-16789	109	3	the	the	DET
ajst-16789	109	4	average	average	ADJ
ajst-16789	109	5	accuracies	accuracy	NOUN
ajst-16789	109	6	from	from	ADP
ajst-16789	109	7	the	the	DET
ajst-16789	109	8	tests	test	NOUN
ajst-16789	109	9	conducted	conduct	VERB
ajst-16789	109	10	for	for	ADP
ajst-16789	109	11	each	each	DET
ajst-16789	109	12	model	model	NOUN
ajst-16789	109	13	,	,	PUNCT
ajst-16789	109	14	as	as	SCONJ
ajst-16789	109	15	detailed	detailed	ADJ
ajst-16789	109	16	in	in	ADP
ajst-16789	109	17	table	table	NOUN
ajst-16789	109	18	2	2	NUM
ajst-16789	109	19	.	.	PUNCT
ajst-16789	109	20	table	table	NOUN
ajst-16789	109	21	2	2	NUM
ajst-16789	109	22	.	.	PUNCT
ajst-16789	109	23	comparison	comparison	NOUN
ajst-16789	109	24	of	of	ADP
ajst-16789	109	25	test	test	NOUN
ajst-16789	109	26	set	set	VERB
ajst-16789	109	27	prediction	prediction	NOUN
ajst-16789	109	28	net	net	ADJ
ajst-16789	109	29	top1	top1	PROPN
ajst-16789	109	30	parcelnet(d169	parcelnet(d169	PROPN
ajst-16789	109	31	)	)	PUNCT
ajst-16789	109	32	93.39	93.39	NUM
ajst-16789	109	33	%	%	NOUN
ajst-16789	109	34	parcelnet(d121	parcelnet(d121	NOUN
ajst-16789	109	35	)	)	PUNCT
ajst-16789	109	36	88.21	88.21	NUM
ajst-16789	109	37	%	%	NOUN
ajst-16789	109	38	parcelnet(r101	parcelnet(r101	NOUN
ajst-16789	109	39	)	)	PUNCT
ajst-16789	109	40	90.56	90.56	NUM
ajst-16789	109	41	%	%	NOUN
ajst-16789	109	42	parcelnet(r50	parcelnet(r50	VERB
ajst-16789	109	43	)	)	PUNCT
ajst-16789	109	44	82.32	82.32	NUM
ajst-16789	109	45	%	%	NOUN
ajst-16789	109	46	the	the	DET
ajst-16789	109	47	experimental	experimental	ADJ
ajst-16789	109	48	results	result	NOUN
ajst-16789	109	49	indicate	indicate	VERB
ajst-16789	109	50	that	that	DET
ajst-16789	109	51	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	109	52	)	)	PUNCT
ajst-16789	109	53	exhibits	exhibit	VERB
ajst-16789	109	54	the	the	DET
ajst-16789	109	55	best	good	ADJ
ajst-16789	109	56	performance	performance	NOUN
ajst-16789	109	57	,	,	PUNCT
ajst-16789	109	58	achieving	achieve	VERB
ajst-16789	109	59	a	a	DET
ajst-16789	109	60	top1	top1	PROPN
ajst-16789	109	61	accuracy	accuracy	NOUN
ajst-16789	109	62	of	of	ADP
ajst-16789	109	63	93.39	93.39	NUM
ajst-16789	109	64	%	%	NOUN
ajst-16789	109	65	.	.	PUNCT
ajst-16789	110	1	following	follow	VERB
ajst-16789	110	2	closely	closely	ADV
ajst-16789	110	3	is	be	AUX
ajst-16789	110	4	parcelnet(r101	parcelnet(r101	PROPN
ajst-16789	110	5	)	)	PUNCT
ajst-16789	110	6	with	with	ADP
ajst-16789	110	7	a	a	DET
ajst-16789	110	8	top1	top1	NOUN
ajst-16789	110	9	accuracy	accuracy	NOUN
ajst-16789	110	10	of	of	ADP
ajst-16789	110	11	90.56	90.56	NUM
ajst-16789	110	12	%	%	NOUN
ajst-16789	110	13	.	.	PUNCT
ajst-16789	111	1	notably	notably	ADV
ajst-16789	111	2	,	,	PUNCT
ajst-16789	111	3	parcelnet(d169	parcelnet(d169	PROPN
ajst-16789	111	4	)	)	PUNCT
ajst-16789	111	5	comprises	comprise	VERB
ajst-16789	111	6	14,254,979	14,254,979	NUM
ajst-16789	111	7	parameters	parameter	NOUN
ajst-16789	111	8	,	,	PUNCT
ajst-16789	111	9	while	while	SCONJ
ajst-16789	111	10	parcelnet(r101	parcelnet(r101	NOUN
ajst-16789	111	11	)	)	PUNCT
ajst-16789	111	12	consists	consist	VERB
ajst-16789	111	13	of	of	ADP
ajst-16789	111	14	44,663,875	44,663,875	NUM
ajst-16789	111	15	parameters	parameter	NOUN
ajst-16789	111	16	.	.	PUNCT
ajst-16789	112	1	considering	consider	VERB
ajst-16789	112	2	the	the	DET
ajst-16789	112	3	trade	trade	NOUN
ajst-16789	112	4	-	-	PUNCT
ajst-16789	112	5	off	off	NOUN
ajst-16789	112	6	between	between	ADP
ajst-16789	112	7	model	model	NOUN
ajst-16789	112	8	performance	performance	NOUN
ajst-16789	112	9	and	and	CCONJ
ajst-16789	112	10	parameter	parameter	NOUN
ajst-16789	112	11	count	count	NOUN
ajst-16789	112	12	,	,	PUNCT
ajst-16789	112	13	the	the	DET
ajst-16789	112	14	detection	detection	NOUN
ajst-16789	112	15	and	and	CCONJ
ajst-16789	112	16	recognition	recognition	NOUN
ajst-16789	112	17	network	network	NOUN
ajst-16789	112	18	constructed	construct	VERB
ajst-16789	112	19	based	base	VERB
ajst-16789	112	20	on	on	ADP
ajst-16789	112	21	densenet169	densenet169	PROPN
ajst-16789	112	22	,	,	PUNCT
ajst-16789	112	23	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	112	24	)	)	PUNCT
ajst-16789	112	25	,	,	PUNCT
ajst-16789	112	26	was	be	AUX
ajst-16789	112	27	ultimately	ultimately	ADV
ajst-16789	112	28	selected	select	VERB
ajst-16789	112	29	.	.	PUNCT
ajst-16789	113	1	5	5	X
ajst-16789	113	2	.	.	PUNCT
ajst-16789	113	3	model	model	NOUN
ajst-16789	113	4	optimization	optimization	NOUN
ajst-16789	113	5	in	in	ADP
ajst-16789	113	6	order	order	NOUN
ajst-16789	113	7	to	to	PART
ajst-16789	113	8	further	far	ADV
ajst-16789	113	9	improve	improve	VERB
ajst-16789	113	10	the	the	DET
ajst-16789	113	11	system	system	NOUN
ajst-16789	113	12	's	's	PART
ajst-16789	113	13	robustness	robustness	NOUN
ajst-16789	113	14	and	and	CCONJ
ajst-16789	113	15	optimize	optimize	VERB
ajst-16789	113	16	the	the	DET
ajst-16789	113	17	accuracy	accuracy	NOUN
ajst-16789	113	18	of	of	ADP
ajst-16789	113	19	semantic	semantic	ADJ
ajst-16789	113	20	segmentation	segmentation	NOUN
ajst-16789	113	21	,	,	PUNCT
ajst-16789	113	22	we	we	PRON
ajst-16789	113	23	introduced	introduce	VERB
ajst-16789	113	24	cbam	cbam	NOUN
ajst-16789	113	25	(	(	PUNCT
ajst-16789	113	26	convolutional	convolutional	ADJ
ajst-16789	113	27	block	block	NOUN
ajst-16789	113	28	attention	attention	NOUN
ajst-16789	113	29	module	module	NOUN
ajst-16789	113	30	)	)	PUNCT
ajst-16789	114	1	[	[	X
ajst-16789	114	2	15	15	NUM
ajst-16789	114	3	]	]	X
ajst-16789	114	4	,	,	PUNCT
ajst-16789	114	5	which	which	PRON
ajst-16789	114	6	is	be	AUX
ajst-16789	114	7	a	a	DET
ajst-16789	114	8	lightweight	lightweight	ADJ
ajst-16789	114	9	attention	attention	NOUN
ajst-16789	114	10	mechanism	mechanism	NOUN
ajst-16789	114	11	.	.	PUNCT
ajst-16789	115	1	we	we	PRON
ajst-16789	115	2	compared	compare	VERB
ajst-16789	115	3	the	the	DET
ajst-16789	115	4	model	model	NOUN
ajst-16789	115	5	parameters	parameter	NOUN
ajst-16789	115	6	before	before	ADV
ajst-16789	115	7	and	and	CCONJ
ajst-16789	115	8	after	after	ADP
ajst-16789	115	9	introducing	introduce	VERB
ajst-16789	115	10	cbam	cbam	NOUN
ajst-16789	115	11	(	(	PUNCT
ajst-16789	115	12	detailed	detail	VERB
ajst-16789	115	13	in	in	ADP
ajst-16789	115	14	table	table	NOUN
ajst-16789	115	15	3	3	NUM
ajst-16789	115	16	)	)	PUNCT
ajst-16789	115	17	.	.	PUNCT
ajst-16789	116	1	table	table	NOUN
ajst-16789	117	1	3	3	NUM
ajst-16789	117	2	.	.	PUNCT
ajst-16789	117	3	comparison	comparison	NOUN
ajst-16789	117	4	of	of	ADP
ajst-16789	117	5	model	model	NOUN
ajst-16789	117	6	parameters	parameter	NOUN
ajst-16789	117	7	net	net	ADJ
ajst-16789	117	8	parameter	parameter	NOUN
ajst-16789	117	9	amount	amount	NOUN
ajst-16789	117	10	parcelnet(d169)+cbam	parcelnet(d169)+cbam	NOUN
ajst-16789	117	11	14289991	14289991	NUM
ajst-16789	117	12	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	117	13	)	)	PUNCT
ajst-16789	117	14	14254979	14254979	NUM
ajst-16789	117	15	cbam	cbam	NOUN
ajst-16789	117	16	,	,	PUNCT
ajst-16789	117	17	as	as	ADP
ajst-16789	117	18	a	a	DET
ajst-16789	117	19	lightweight	lightweight	ADJ
ajst-16789	117	20	convolutional	convolutional	ADJ
ajst-16789	117	21	attention	attention	NOUN
ajst-16789	117	22	module	module	NOUN
ajst-16789	117	23	,	,	PUNCT
ajst-16789	117	24	integrates	integrate	NOUN
ajst-16789	117	25	channel	channel	NOUN
ajst-16789	117	26	and	and	CCONJ
ajst-16789	117	27	spatial	spatial	ADJ
ajst-16789	117	28	attention	attention	NOUN
ajst-16789	117	29	mechanisms	mechanism	NOUN
ajst-16789	117	30	without	without	ADP
ajst-16789	117	31	significantly	significantly	ADV
ajst-16789	117	32	increasing	increase	VERB
ajst-16789	117	33	the	the	DET
ajst-16789	117	34	model	model	NOUN
ajst-16789	117	35	parameters	parameter	NOUN
ajst-16789	117	36	(	(	PUNCT
ajst-16789	117	37	refer	refer	VERB
ajst-16789	117	38	to	to	PART
ajst-16789	117	39	figure	figure	VERB
ajst-16789	117	40	6	6	NUM
ajst-16789	117	41	)	)	PUNCT
ajst-16789	117	42	.	.	PUNCT
ajst-16789	118	1	its	its	PRON
ajst-16789	118	2	implementation	implementation	NOUN
ajst-16789	118	3	involves	involve	VERB
ajst-16789	118	4	several	several	ADJ
ajst-16789	118	5	steps	step	NOUN
ajst-16789	118	6	:	:	PUNCT
ajst-16789	118	7	firstly	firstly	ADV
ajst-16789	118	8	,	,	PUNCT
ajst-16789	118	9	global	global	ADJ
ajst-16789	118	10	maxpool	maxpool	NOUN
ajst-16789	118	11	and	and	CCONJ
ajst-16789	118	12	avgpool	avgpool	PROPN
ajst-16789	118	13	operations	operation	NOUN
ajst-16789	118	14	are	be	AUX
ajst-16789	118	15	applied	apply	VERB
ajst-16789	118	16	to	to	ADP
ajst-16789	118	17	the	the	DET
ajst-16789	118	18	feature	feature	NOUN
ajst-16789	118	19	maps	map	NOUN
ajst-16789	118	20	to	to	PART
ajst-16789	118	21	obtain	obtain	VERB
ajst-16789	118	22	two	two	NUM
ajst-16789	118	23	distinct	distinct	ADJ
ajst-16789	118	24	feature	feature	NOUN
ajst-16789	118	25	vectors	vector	NOUN
ajst-16789	118	26	.	.	PUNCT
ajst-16789	119	1	subsequently	subsequently	ADV
ajst-16789	119	2	,	,	PUNCT
ajst-16789	119	3	these	these	DET
ajst-16789	119	4	two	two	NUM
ajst-16789	119	5	feature	feature	NOUN
ajst-16789	119	6	vectors	vector	NOUN
ajst-16789	119	7	pass	pass	VERB
ajst-16789	119	8	through	through	ADP
ajst-16789	119	9	a	a	DET
ajst-16789	119	10	shared	share	VERB
ajst-16789	119	11	perceptron	perceptron	NOUN
ajst-16789	119	12	for	for	ADP
ajst-16789	119	13	processing	processing	NOUN
ajst-16789	119	14	and	and	CCONJ
ajst-16789	119	15	are	be	AUX
ajst-16789	119	16	merged	merge	VERB
ajst-16789	119	17	into	into	ADP
ajst-16789	119	18	a	a	DET
ajst-16789	119	19	completely	completely	ADV
ajst-16789	119	20	new	new	ADJ
ajst-16789	119	21	feature	feature	NOUN
ajst-16789	119	22	vector	vector	NOUN
ajst-16789	119	23	.	.	PUNCT
ajst-16789	120	1	finally	finally	ADV
ajst-16789	120	2	,	,	PUNCT
ajst-16789	120	3	a	a	DET
ajst-16789	120	4	sigmoid	sigmoid	NOUN
ajst-16789	120	5	operation	operation	NOUN
ajst-16789	120	6	is	be	AUX
ajst-16789	120	7	performed	perform	VERB
ajst-16789	120	8	on	on	ADP
ajst-16789	120	9	this	this	DET
ajst-16789	120	10	new	new	ADJ
ajst-16789	120	11	feature	feature	NOUN
ajst-16789	120	12	vector	vector	NOUN
ajst-16789	120	13	to	to	PART
ajst-16789	120	14	obtain	obtain	VERB
ajst-16789	120	15	attention	attention	NOUN
ajst-16789	120	16	weights	weight	NOUN
ajst-16789	120	17	in	in	ADP
ajst-16789	120	18	the	the	DET
ajst-16789	120	19	channel	channel	NOUN
ajst-16789	120	20	domain	domain	NOUN
ajst-16789	120	21	.	.	PUNCT
ajst-16789	121	1	this	this	DET
ajst-16789	121	2	approach	approach	NOUN
ajst-16789	121	3	by	by	ADP
ajst-16789	121	4	cbam	cbam	NOUN
ajst-16789	121	5	,	,	PUNCT
ajst-16789	121	6	through	through	ADP
ajst-16789	121	7	extraction	extraction	NOUN
ajst-16789	121	8	and	and	CCONJ
ajst-16789	121	9	fusion	fusion	NOUN
ajst-16789	121	10	of	of	ADP
ajst-16789	121	11	different	different	ADJ
ajst-16789	121	12	features	feature	NOUN
ajst-16789	121	13	,	,	PUNCT
ajst-16789	121	14	effectively	effectively	ADV
ajst-16789	121	15	enhances	enhance	VERB
ajst-16789	121	16	the	the	DET
ajst-16789	121	17	performance	performance	NOUN
ajst-16789	121	18	of	of	ADP
ajst-16789	121	19	deep	deep	ADJ
ajst-16789	121	20	learning	learning	NOUN
ajst-16789	121	21	models	model	NOUN
ajst-16789	121	22	.	.	PUNCT
ajst-16789	122	1	202	202	NUM
ajst-16789	122	2	figure	figure	NOUN
ajst-16789	122	3	6	6	NUM
ajst-16789	122	4	.	.	PUNCT
ajst-16789	122	5	schematic	schematic	ADJ
ajst-16789	122	6	diagram	diagram	NOUN
ajst-16789	122	7	of	of	ADP
ajst-16789	122	8	cbam	cbam	NOUN
ajst-16789	122	9	attention	attention	NOUN
ajst-16789	122	10	module	module	NOUN
ajst-16789	122	11	after	after	ADP
ajst-16789	122	12	the	the	DET
ajst-16789	122	13	multimodal	multimodal	ADJ
ajst-16789	122	14	concatenation	concatenation	NOUN
ajst-16789	122	15	in	in	ADP
ajst-16789	122	16	the	the	DET
ajst-16789	122	17	model	model	NOUN
ajst-16789	122	18	,	,	PUNCT
ajst-16789	122	19	we	we	PRON
ajst-16789	122	20	further	far	ADV
ajst-16789	122	21	optimize	optimize	VERB
ajst-16789	122	22	it	it	PRON
ajst-16789	122	23	by	by	ADP
ajst-16789	122	24	inserting	insert	VERB
ajst-16789	122	25	the	the	DET
ajst-16789	122	26	convolutional	convolutional	ADJ
ajst-16789	122	27	block	block	NOUN
ajst-16789	122	28	attention	attention	NOUN
ajst-16789	122	29	module	module	NOUN
ajst-16789	122	30	(	(	PUNCT
ajst-16789	122	31	cbam	cbam	NOUN
ajst-16789	122	32	)	)	PUNCT
ajst-16789	122	33	attention	attention	NOUN
ajst-16789	122	34	mechanism	mechanism	NOUN
ajst-16789	122	35	between	between	ADP
ajst-16789	122	36	convolutional	convolutional	ADJ
ajst-16789	122	37	layers	layer	NOUN
ajst-16789	122	38	.	.	PUNCT
ajst-16789	123	1	the	the	DET
ajst-16789	123	2	specific	specific	ADJ
ajst-16789	123	3	structure	structure	NOUN
ajst-16789	123	4	is	be	AUX
ajst-16789	123	5	illustrated	illustrate	VERB
ajst-16789	123	6	in	in	ADP
ajst-16789	123	7	figure	figure	NOUN
ajst-16789	123	8	7	7	NUM
ajst-16789	123	9	.	.	PUNCT
ajst-16789	124	1	r	r	NOUN
ajst-16789	124	2	g	g	PROPN
ajst-16789	124	3	b	b	PROPN
ajst-16789	124	4	 	 	SPACE
ajst-16789	124	5	i	i	PRON
ajst-16789	124	6	m	m	VERB
ajst-16789	124	7	a	a	PRON
ajst-16789	124	8	g	g	NOUN
ajst-16789	124	9	e	e	NOUN
ajst-16789	124	10	d	d	X
ajst-16789	124	11	e	e	X
ajst-16789	124	12	p	p	X
ajst-16789	124	13	th	th	INTJ
ajst-16789	124	14	 	 	SPACE
ajst-16789	124	15	i	i	PRON
ajst-16789	124	16	m	m	VERB
ajst-16789	124	17	a	a	DET
ajst-16789	124	18	ge	ge	PROPN
ajst-16789	124	19	d	d	PROPN
ajst-16789	124	20	e	e	PROPN
ajst-16789	124	21	n	n	X
ajst-16789	124	22	se	se	X
ajst-16789	124	23	n	n	PROPN
ajst-16789	124	24	e	e	NOUN
ajst-16789	124	25	t1	t1	NOUN
ajst-16789	124	26	6	6	NUM
ajst-16789	124	27	9	9	NUM
ajst-16789	124	28	b	b	PROPN
ajst-16789	125	1	a	a	DET
ajst-16789	125	2	ckb	ckb	NOUN
ajst-16789	125	3	o	o	NOUN
ajst-16789	125	4	n	n	PROPN
ajst-16789	125	5	e	e	PROPN
ajst-16789	125	6	a	a	PROPN
ajst-16789	125	7	d	d	X
ajst-16789	125	8	d	d	NOUN
ajst-16789	125	9	sp	sp	ADP
ajst-16789	125	10	a	a	DET
ajst-16789	125	11	tial	tial	NOUN
ajst-16789	125	12	 	 	SPACE
ajst-16789	125	13	c	c	NOUN
ajst-16789	126	1	o	o	NOUN
ajst-16789	126	2	n	n	CCONJ
ajst-16789	126	3	vo	vo	X
ajst-16789	126	4	lu	lu	NOUN
ajst-16789	126	5	tio	tio	NOUN
ajst-16789	126	6	n	n	PROPN
ajst-16789	126	7	d	d	PROPN
ajst-16789	126	8	e	e	X
ajst-16789	126	9	n	n	X
ajst-16789	126	10	se	se	X
ajst-16789	126	11	n	n	PROPN
ajst-16789	126	12	e	e	NOUN
ajst-16789	126	13	t1	t1	NOUN
ajst-16789	126	14	6	6	NUM
ajst-16789	126	15	9	9	NUM
ajst-16789	126	16	b	b	PROPN
ajst-16789	126	17	a	a	DET
ajst-16789	126	18	ckb	ckb	NOUN
ajst-16789	126	19	o	o	NOUN
ajst-16789	126	20	n	n	NOUN
ajst-16789	126	21	e	e	NOUN
ajst-16789	126	22	c	c	PROPN
ajst-16789	126	23	b	b	PROPN
ajst-16789	126	24	a	a	DET
ajst-16789	126	25	m	m	NOUN
ajst-16789	126	26	sp	sp	ADP
ajst-16789	126	27	a	a	DET
ajst-16789	126	28	tial	tial	NOUN
ajst-16789	126	29	 	 	SPACE
ajst-16789	126	30	c	c	NOUN
ajst-16789	127	1	o	o	NOUN
ajst-16789	127	2	n	n	CCONJ
ajst-16789	127	3	vo	vo	X
ajst-16789	127	4	lu	lu	NOUN
ajst-16789	128	1	tio	tio	NOUN
ajst-16789	129	1	n	n	ADV
ajst-16789	130	1	u	u	NOUN
ajst-16789	130	2	p	p	PROPN
ajst-16789	130	3	sam	sam	PROPN
ajst-16789	130	4	p	p	PROPN
ajst-16789	130	5	lin	lin	PROPN
ajst-16789	131	1	g	g	PROPN
ajst-16789	131	2	p	p	PROPN
ajst-16789	131	3	ad	ad	NOUN
ajst-16789	131	4	in	in	ADP
ajst-16789	131	5	g	g	PROPN
ajst-16789	131	6	b	b	PROPN
ajst-16789	131	7	a	a	DET
ajst-16789	131	8	ckg	ckg	NOUN
ajst-16789	131	9	ro	ro	X
ajst-16789	131	10	u	u	PROPN
ajst-16789	131	11	n	n	PROPN
ajst-16789	131	12	d	d	NOUN
ajst-16789	131	13	filte	filte	NOUN
ajst-16789	132	1	r	r	NOUN
ajst-16789	132	2	r	r	NOUN
ajst-16789	132	3	e	e	X
ajst-16789	132	4	su	su	PROPN
ajst-16789	132	5	lt	lt	PRON
ajst-16789	132	6	c	c	PROPN
ajst-16789	132	7	b	b	PROPN
ajst-16789	132	8	a	a	DET
ajst-16789	132	9	m	m	NOUN
ajst-16789	132	10	sp	sp	ADP
ajst-16789	132	11	a	a	DET
ajst-16789	132	12	tial	tial	NOUN
ajst-16789	132	13	 	 	SPACE
ajst-16789	132	14	c	c	NOUN
ajst-16789	132	15	o	o	NOUN
ajst-16789	132	16	n	n	CCONJ
ajst-16789	132	17	vo	vo	X
ajst-16789	132	18	lu	lu	NOUN
ajst-16789	132	19	tio	tio	NOUN
ajst-16789	132	20	n	n	PRON
ajst-16789	132	21	figure	figure	VERB
ajst-16789	132	22	7	7	NUM
ajst-16789	132	23	.	.	PUNCT
ajst-16789	133	1	parcelnet(d169)+cbam	parcelnet(d169)+cbam	NOUN
ajst-16789	133	2	after	after	ADP
ajst-16789	133	3	integrating	integrate	VERB
ajst-16789	133	4	the	the	DET
ajst-16789	133	5	attention	attention	NOUN
ajst-16789	133	6	mechanism	mechanism	NOUN
ajst-16789	133	7	into	into	ADP
ajst-16789	133	8	the	the	DET
ajst-16789	133	9	model	model	NOUN
ajst-16789	133	10	,	,	PUNCT
ajst-16789	133	11	we	we	PRON
ajst-16789	133	12	employed	employ	VERB
ajst-16789	133	13	the	the	DET
ajst-16789	133	14	previous	previous	ADJ
ajst-16789	133	15	training	training	NOUN
ajst-16789	133	16	approach	approach	NOUN
ajst-16789	133	17	and	and	CCONJ
ajst-16789	133	18	conducted	conduct	VERB
ajst-16789	133	19	training	training	NOUN
ajst-16789	133	20	for	for	ADP
ajst-16789	133	21	150	150	NUM
ajst-16789	133	22	epochs	epoch	NOUN
ajst-16789	133	23	on	on	ADP
ajst-16789	133	24	the	the	DET
ajst-16789	133	25	training	training	NOUN
ajst-16789	133	26	dataset	dataset	NOUN
ajst-16789	133	27	.	.	PUNCT
ajst-16789	134	1	during	during	ADP
ajst-16789	134	2	this	this	DET
ajst-16789	134	3	process	process	NOUN
ajst-16789	134	4	,	,	PUNCT
ajst-16789	134	5	we	we	PRON
ajst-16789	134	6	observed	observe	VERB
ajst-16789	134	7	the	the	DET
ajst-16789	134	8	training	training	NOUN
ajst-16789	134	9	loss	loss	NOUN
ajst-16789	134	10	function	function	NOUN
ajst-16789	134	11	graph	graph	NOUN
ajst-16789	134	12	*	*	PUNCT
ajst-16789	134	13	(	(	PUNCT
ajst-16789	134	14	refer	refer	VERB
ajst-16789	134	15	to	to	PART
ajst-16789	134	16	figure	figure	VERB
ajst-16789	134	17	8)	8)	NUM
ajst-16789	134	18	and	and	CCONJ
ajst-16789	134	19	noted	note	VERB
ajst-16789	134	20	that	that	SCONJ
ajst-16789	134	21	the	the	DET
ajst-16789	134	22	training	training	NOUN
ajst-16789	134	23	loss	loss	NOUN
ajst-16789	134	24	also	also	ADV
ajst-16789	134	25	decreased	decrease	VERB
ajst-16789	134	26	to	to	ADP
ajst-16789	134	27	a	a	DET
ajst-16789	134	28	level	level	NOUN
ajst-16789	134	29	of	of	ADP
ajst-16789	134	30	0.5	0.5	NUM
ajst-16789	134	31	%	%	NOUN
ajst-16789	134	32	and	and	CCONJ
ajst-16789	134	33	remained	remain	VERB
ajst-16789	134	34	stable	stable	ADJ
ajst-16789	134	35	.	.	PUNCT
ajst-16789	135	1	figure	figure	NOUN
ajst-16789	135	2	8	8	NUM
ajst-16789	135	3	.	.	PUNCT
ajst-16789	136	1	training	training	NOUN
ajst-16789	136	2	loss	loss	NOUN
ajst-16789	136	3	function	function	NOUN
ajst-16789	136	4	graph	graph	NOUN
ajst-16789	136	5	*	*	PUNCT
ajst-16789	136	6	after	after	ADP
ajst-16789	136	7	completing	complete	VERB
ajst-16789	136	8	the	the	DET
ajst-16789	136	9	model	model	NOUN
ajst-16789	136	10	training	training	NOUN
ajst-16789	136	11	,	,	PUNCT
ajst-16789	136	12	we	we	PRON
ajst-16789	136	13	tested	test	VERB
ajst-16789	136	14	its	its	PRON
ajst-16789	136	15	performance	performance	NOUN
ajst-16789	136	16	using	use	VERB
ajst-16789	136	17	a	a	DET
ajst-16789	136	18	test	test	NOUN
ajst-16789	136	19	set	set	NOUN
ajst-16789	136	20	containing	contain	VERB
ajst-16789	136	21	568	568	NUM
ajst-16789	136	22	images	image	NOUN
ajst-16789	136	23	,	,	PUNCT
ajst-16789	136	24	with	with	ADP
ajst-16789	136	25	detailed	detailed	ADJ
ajst-16789	136	26	results	result	NOUN
ajst-16789	136	27	provided	provide	VERB
ajst-16789	136	28	in	in	ADP
ajst-16789	136	29	table	table	NOUN
ajst-16789	136	30	4	4	NUM
ajst-16789	136	31	.	.	PUNCT
ajst-16789	136	32	upon	upon	SCONJ
ajst-16789	136	33	integrating	integrate	VERB
ajst-16789	136	34	the	the	DET
ajst-16789	136	35	cbam	cbam	NOUN
ajst-16789	136	36	module	module	NOUN
ajst-16789	136	37	,	,	PUNCT
ajst-16789	136	38	parcelnet	parcelnet	NOUN
ajst-16789	136	39	(	(	PUNCT
ajst-16789	136	40	d169	d169	PROPN
ajst-16789	136	41	)	)	PUNCT
ajst-16789	136	42	achieved	achieve	VERB
ajst-16789	136	43	a	a	DET
ajst-16789	136	44	top1	top1	PROPN
ajst-16789	136	45	accuracy	accuracy	NOUN
ajst-16789	136	46	of	of	ADP
ajst-16789	136	47	95.86	95.86	NUM
ajst-16789	136	48	%	%	NOUN
ajst-16789	136	49	.	.	PUNCT
ajst-16789	137	1	this	this	PRON
ajst-16789	137	2	represents	represent	VERB
ajst-16789	137	3	a	a	DET
ajst-16789	137	4	gain	gain	NOUN
ajst-16789	137	5	of	of	ADP
ajst-16789	137	6	over	over	ADP
ajst-16789	137	7	2	2	NUM
ajst-16789	137	8	%	%	NOUN
ajst-16789	137	9	in	in	ADP
ajst-16789	137	10	accuracy	accuracy	NOUN
ajst-16789	137	11	compared	compare	VERB
ajst-16789	137	12	to	to	ADP
ajst-16789	137	13	the	the	DET
ajst-16789	137	14	model	model	NOUN
ajst-16789	137	15	without	without	ADP
ajst-16789	137	16	the	the	DET
ajst-16789	137	17	cbam	cbam	NOUN
ajst-16789	137	18	module	module	NOUN
ajst-16789	137	19	.	.	PUNCT
ajst-16789	138	1	we	we	PRON
ajst-16789	138	2	can	can	AUX
ajst-16789	138	3	distinctly	distinctly	ADV
ajst-16789	138	4	observe	observe	VERB
ajst-16789	138	5	the	the	DET
ajst-16789	138	6	impact	impact	NOUN
ajst-16789	138	7	of	of	ADP
ajst-16789	138	8	cbam	cbam	NOUN
ajst-16789	138	9	on	on	ADP
ajst-16789	138	10	the	the	DET
ajst-16789	138	11	model	model	NOUN
ajst-16789	138	12	's	's	PART
ajst-16789	138	13	performance	performance	NOUN
ajst-16789	138	14	.	.	PUNCT
ajst-16789	139	1	the	the	DET
ajst-16789	139	2	introduction	introduction	NOUN
ajst-16789	139	3	of	of	ADP
ajst-16789	139	4	this	this	DET
ajst-16789	139	5	attention	attention	NOUN
ajst-16789	139	6	mechanism	mechanism	NOUN
ajst-16789	139	7	helps	help	VERB
ajst-16789	139	8	the	the	DET
ajst-16789	139	9	model	model	NOUN
ajst-16789	139	10	better	well	ADJ
ajst-16789	139	11	capture	capture	VERB
ajst-16789	139	12	and	and	CCONJ
ajst-16789	139	13	utilize	utilize	VERB
ajst-16789	139	14	contextual	contextual	ADJ
ajst-16789	139	15	information	information	NOUN
ajst-16789	139	16	within	within	ADP
ajst-16789	139	17	images	image	NOUN
ajst-16789	139	18	,	,	PUNCT
ajst-16789	139	19	thereby	thereby	ADV
ajst-16789	139	20	enhancing	enhance	VERB
ajst-16789	139	21	the	the	DET
ajst-16789	139	22	accuracy	accuracy	NOUN
ajst-16789	139	23	of	of	ADP
ajst-16789	139	24	semantic	semantic	ADJ
ajst-16789	139	25	segmentation	segmentation	NOUN
ajst-16789	139	26	.	.	PUNCT
ajst-16789	140	1	consequently	consequently	ADV
ajst-16789	140	2	,	,	PUNCT
ajst-16789	140	3	it	it	PRON
ajst-16789	140	4	improves	improve	VERB
ajst-16789	140	5	the	the	DET
ajst-16789	140	6	detection	detection	NOUN
ajst-16789	140	7	accuracy	accuracy	NOUN
ajst-16789	140	8	of	of	ADP
ajst-16789	140	9	parcelnet	parcelnet	NOUN
ajst-16789	140	10	(	(	PUNCT
ajst-16789	140	11	d169	d169	PROPN
ajst-16789	140	12	)	)	PUNCT
ajst-16789	140	13	.	.	PUNCT
ajst-16789	141	1	table	table	NOUN
ajst-16789	141	2	4	4	NUM
ajst-16789	141	3	.	.	PUNCT
ajst-16789	141	4	comparison	comparison	NOUN
ajst-16789	141	5	of	of	ADP
ajst-16789	141	6	test	test	NOUN
ajst-16789	141	7	set	set	VERB
ajst-16789	141	8	prediction	prediction	NOUN
ajst-16789	141	9	net	net	ADJ
ajst-16789	141	10	top1	top1	PROPN
ajst-16789	141	11	parcelnet(d169)+cbam	parcelnet(d169)+cbam	NOUN
ajst-16789	141	12	95.86	95.86	NUM
ajst-16789	141	13	%	%	NOUN
ajst-16789	141	14	parcelnet(d169	parcelnet(d169	NOUN
ajst-16789	141	15	)	)	PUNCT
ajst-16789	141	16	93.39	93.39	NUM
ajst-16789	141	17	%	%	NOUN
ajst-16789	141	18	6	6	NUM
ajst-16789	141	19	.	.	PUNCT
ajst-16789	141	20	conclusion	conclusion	NOUN
ajst-16789	141	21	in	in	ADP
ajst-16789	141	22	this	this	DET
ajst-16789	141	23	study	study	NOUN
ajst-16789	141	24	focusing	focus	VERB
ajst-16789	141	25	on	on	ADP
ajst-16789	141	26	unstructured	unstructured	ADJ
ajst-16789	141	27	package	package	NOUN
ajst-16789	141	28	grab	grab	PROPN
ajst-16789	141	29	area	area	NOUN
ajst-16789	141	30	recognition	recognition	NOUN
ajst-16789	141	31	,	,	PUNCT
ajst-16789	141	32	we	we	PRON
ajst-16789	141	33	constructed	construct	VERB
ajst-16789	141	34	a	a	DET
ajst-16789	141	35	multimodal	multimodal	NOUN
ajst-16789	141	36	recognition	recognition	NOUN
ajst-16789	141	37	network	network	NOUN
ajst-16789	141	38	using	use	VERB
ajst-16789	141	39	different	different	ADJ
ajst-16789	141	40	pre	pre	ADJ
ajst-16789	141	41	-	-	ADJ
ajst-16789	141	42	trained	train	VERB
ajst-16789	141	43	deep	deep	ADJ
ajst-16789	141	44	learning	learning	NOUN
ajst-16789	141	45	models	model	NOUN
ajst-16789	141	46	.	.	PUNCT
ajst-16789	142	1	through	through	ADP
ajst-16789	142	2	a	a	DET
ajst-16789	142	3	series	series	NOUN
ajst-16789	142	4	of	of	ADP
ajst-16789	142	5	training	training	NOUN
ajst-16789	142	6	and	and	CCONJ
ajst-16789	142	7	testing	testing	NOUN
ajst-16789	142	8	experiments	experiment	NOUN
ajst-16789	142	9	and	and	CCONJ
ajst-16789	142	10	203	203	NUM
ajst-16789	142	11	considering	consider	VERB
ajst-16789	142	12	comprehensive	comprehensive	ADJ
ajst-16789	142	13	results	result	NOUN
ajst-16789	142	14	,	,	PUNCT
ajst-16789	142	15	we	we	PRON
ajst-16789	142	16	selected	select	VERB
ajst-16789	142	17	parcelnet	parcelnet	NOUN
ajst-16789	142	18	(	(	PUNCT
ajst-16789	142	19	d169	d169	PROPN
ajst-16789	142	20	)	)	PUNCT
ajst-16789	142	21	built	build	VERB
ajst-16789	142	22	upon	upon	SCONJ
ajst-16789	142	23	densenet	densenet	NOUN
ajst-16789	142	24	for	for	ADP
ajst-16789	142	25	our	our	PRON
ajst-16789	142	26	approach	approach	NOUN
ajst-16789	142	27	.	.	PUNCT
ajst-16789	143	1	concurrently	concurrently	ADV
ajst-16789	143	2	,	,	PUNCT
ajst-16789	143	3	we	we	PRON
ajst-16789	143	4	optimized	optimize	VERB
ajst-16789	143	5	the	the	DET
ajst-16789	143	6	network	network	NOUN
ajst-16789	143	7	by	by	ADP
ajst-16789	143	8	introducing	introduce	VERB
ajst-16789	143	9	cbam	cbam	NOUN
ajst-16789	143	10	,	,	PUNCT
ajst-16789	143	11	a	a	DET
ajst-16789	143	12	lightweight	lightweight	ADJ
ajst-16789	143	13	attention	attention	NOUN
ajst-16789	143	14	mechanism	mechanism	NOUN
ajst-16789	143	15	,	,	PUNCT
ajst-16789	143	16	resulting	result	VERB
ajst-16789	143	17	in	in	ADP
ajst-16789	143	18	a	a	DET
ajst-16789	143	19	significant	significant	ADJ
ajst-16789	143	20	improvement	improvement	NOUN
ajst-16789	143	21	in	in	ADP
ajst-16789	143	22	recognition	recognition	NOUN
ajst-16789	143	23	accuracy	accuracy	NOUN
ajst-16789	143	24	,	,	PUNCT
ajst-16789	143	25	elevating	elevate	VERB
ajst-16789	143	26	its	its	PRON
ajst-16789	143	27	top1	top1	NOUN
ajst-16789	143	28	accuracy	accuracy	NOUN
ajst-16789	143	29	to	to	ADP
ajst-16789	143	30	95.86	95.86	NUM
ajst-16789	143	31	%	%	NOUN
ajst-16789	143	32	.	.	PUNCT
ajst-16789	144	1	this	this	DET
ajst-16789	144	2	enhancement	enhancement	NOUN
ajst-16789	144	3	greatly	greatly	ADV
ajst-16789	144	4	boosted	boost	VERB
ajst-16789	144	5	the	the	DET
ajst-16789	144	6	network	network	NOUN
ajst-16789	144	7	's	's	PART
ajst-16789	144	8	robustness	robustness	NOUN
ajst-16789	144	9	,	,	PUNCT
ajst-16789	144	10	providing	provide	VERB
ajst-16789	144	11	a	a	DET
ajst-16789	144	12	significant	significant	ADJ
ajst-16789	144	13	solution	solution	NOUN
ajst-16789	144	14	to	to	ADP
ajst-16789	144	15	unstructured	unstructured	ADJ
ajst-16789	144	16	package	package	NOUN
ajst-16789	144	17	grab	grab	PROPN
ajst-16789	144	18	area	area	NOUN
ajst-16789	144	19	recognition	recognition	NOUN
ajst-16789	144	20	.	.	PUNCT
ajst-16789	145	1	references	reference	NOUN
ajst-16789	145	2	[	[	X
ajst-16789	145	3	1	1	NUM
ajst-16789	145	4	]	]	X
ajst-16789	145	5	bai	bai	PROPN
ajst-16789	145	6	wenjie	wenjie	PROPN
ajst-16789	145	7	.	.	PUNCT
ajst-16789	146	1	research	research	NOUN
ajst-16789	146	2	on	on	ADP
ajst-16789	146	3	logistics	logistic	NOUN
ajst-16789	146	4	sorting	sorting	NOUN
ajst-16789	146	5	and	and	CCONJ
ajst-16789	146	6	planning	planning	NOUN
ajst-16789	146	7	system	system	NOUN
ajst-16789	146	8	based	base	VERB
ajst-16789	146	9	on	on	ADP
ajst-16789	146	10	deep	deep	ADJ
ajst-16789	146	11	learning	learning	NOUN
ajst-16789	146	12	of	of	ADP
ajst-16789	146	13	express	express	ADJ
ajst-16789	146	14	delivery	delivery	NOUN
ajst-16789	146	15	waybill	waybill	NOUN
ajst-16789	147	1	[	[	X
ajst-16789	147	2	d	d	X
ajst-16789	147	3	]	]	X
ajst-16789	147	4	.	.	PUNCT
ajst-16789	148	1	anhui	anhui	PROPN
ajst-16789	148	2	university	university	PROPN
ajst-16789	148	3	of	of	ADP
ajst-16789	148	4	science	science	NOUN
ajst-16789	148	5	and	and	CCONJ
ajst-16789	148	6	technology	technology	NOUN
ajst-16789	148	7	,	,	PUNCT
ajst-16789	148	8	2021	2021	NUM
ajst-16789	148	9	.	.	PUNCT
ajst-16789	149	1	[	[	X
ajst-16789	149	2	2	2	X
ajst-16789	149	3	]	]	PUNCT
ajst-16789	149	4	wang	wang	PROPN
ajst-16789	149	5	c	c	PROPN
ajst-16789	149	6	,	,	PUNCT
ajst-16789	149	7	bai	bai	PROPN
ajst-16789	149	8	x	x	PROPN
ajst-16789	149	9	,	,	PUNCT
ajst-16789	149	10	wang	wang	PROPN
ajst-16789	149	11	x	x	PROPN
ajst-16789	149	12	,	,	PUNCT
ajst-16789	149	13	et	et	PROPN
ajst-16789	149	14	al	al	PROPN
ajst-16789	149	15	.	.	PROPN
ajst-16789	149	16	self	self	NOUN
ajst-16789	149	17	-	-	PUNCT
ajst-16789	149	18	supervised	supervise	VERB
ajst-16789	149	19	multiscale	multiscale	ADJ
ajst-16789	149	20	adversarial	adversarial	ADJ
ajst-16789	149	21	regression	regression	NOUN
ajst-16789	149	22	network	network	NOUN
ajst-16789	149	23	for	for	ADP
ajst-16789	149	24	stereo	stereo	ADJ
ajst-16789	149	25	disparity	disparity	NOUN
ajst-16789	149	26	estimation[j	estimation[j	PROPN
ajst-16789	149	27	]	]	PUNCT
ajst-16789	149	28	.	.	PUNCT
ajst-16789	150	1	ieee	ieee	NOUN
ajst-16789	150	2	transactions	transaction	NOUN
ajst-16789	150	3	on	on	ADP
ajst-16789	150	4	cybernetics	cybernetic	NOUN
ajst-16789	150	5	,	,	PUNCT
ajst-16789	150	6	2020	2020	NUM
ajst-16789	150	7	,	,	PUNCT
ajst-16789	150	8	51(10	51(10	NUM
ajst-16789	150	9	):	):	PUNCT
ajst-16789	150	10	4770	4770	NUM
ajst-16789	150	11	-	-	SYM
ajst-16789	150	12	4783	4783	NUM
ajst-16789	150	13	.	.	PUNCT
ajst-16789	151	1	[	[	X
ajst-16789	151	2	3	3	X
ajst-16789	151	3	]	]	X
ajst-16789	151	4	zhang	zhang	PROPN
ajst-16789	151	5	j	j	PROPN
ajst-16789	151	6	,	,	PUNCT
ajst-16789	151	7	xie	xie	PROPN
ajst-16789	151	8	z	z	PROPN
ajst-16789	151	9	,	,	PUNCT
ajst-16789	151	10	sun	sun	PROPN
ajst-16789	151	11	j	j	PROPN
ajst-16789	151	12	,	,	PUNCT
ajst-16789	151	13	et	et	PROPN
ajst-16789	151	14	al	al	PROPN
ajst-16789	151	15	.	.	PUNCT
ajst-16789	152	1	a	a	DET
ajst-16789	152	2	cascaded	cascade	VERB
ajst-16789	152	3	r	r	NOUN
ajst-16789	152	4	-	-	PUNCT
ajst-16789	152	5	cnn	cnn	PROPN
ajst-16789	152	6	with	with	ADP
ajst-16789	152	7	multiscale	multiscale	ADJ
ajst-16789	152	8	attention	attention	NOUN
ajst-16789	152	9	and	and	CCONJ
ajst-16789	152	10	imbalanced	imbalanced	ADJ
ajst-16789	152	11	samples	sample	NOUN
ajst-16789	152	12	for	for	ADP
ajst-16789	152	13	traffic	traffic	NOUN
ajst-16789	152	14	sign	sign	NOUN
ajst-16789	152	15	detection[j	detection[j	PROPN
ajst-16789	152	16	]	]	PUNCT
ajst-16789	152	17	.	.	PUNCT
ajst-16789	153	1	ieee	ieee	NOUN
ajst-16789	153	2	access	access	NOUN
ajst-16789	153	3	,	,	PUNCT
ajst-16789	153	4	2020	2020	NUM
ajst-16789	153	5	,	,	PUNCT
ajst-16789	153	6	8	8	NUM
ajst-16789	153	7	:	:	SYM
ajst-16789	153	8	29742	29742	NUM
ajst-16789	153	9	-	-	SYM
ajst-16789	153	10	29754	29754	NUM
ajst-16789	153	11	.	.	PUNCT
ajst-16789	154	1	[	[	X
ajst-16789	154	2	4	4	NUM
ajst-16789	154	3	]	]	X
ajst-16789	154	4	alasmari	alasmari	PROPN
ajst-16789	154	5	n	n	CCONJ
ajst-16789	154	6	,	,	PUNCT
ajst-16789	154	7	alohali	alohali	PROPN
ajst-16789	154	8	m	m	PROPN
ajst-16789	154	9	a	a	PROPN
ajst-16789	154	10	,	,	PUNCT
ajst-16789	154	11	khalid	khalid	PROPN
ajst-16789	154	12	m	m	PROPN
ajst-16789	154	13	,	,	PUNCT
ajst-16789	154	14	et	et	PROPN
ajst-16789	154	15	al	al	PROPN
ajst-16789	154	16	.	.	PROPN
ajst-16789	154	17	improved	improve	VERB
ajst-16789	154	18	metaheuristics	metaheuristic	NOUN
ajst-16789	154	19	with	with	ADP
ajst-16789	154	20	deep	deep	ADJ
ajst-16789	154	21	learning	learning	NOUN
ajst-16789	154	22	based	base	VERB
ajst-16789	154	23	object	object	NOUN
ajst-16789	154	24	detector	detector	NOUN
ajst-16789	154	25	for	for	ADP
ajst-16789	154	26	intelligent	intelligent	ADJ
ajst-16789	154	27	control	control	NOUN
ajst-16789	154	28	in	in	ADP
ajst-16789	154	29	autonomous	autonomous	ADJ
ajst-16789	154	30	vehicles[j	vehicles[j	NOUN
ajst-16789	154	31	]	]	PUNCT
ajst-16789	154	32	.	.	PUNCT
ajst-16789	155	1	computers	computer	NOUN
ajst-16789	155	2	,	,	PUNCT
ajst-16789	155	3	2023	2023	NUM
ajst-16789	155	4	,	,	PUNCT
ajst-16789	155	5	108	108	NUM
ajst-16789	155	6	:	:	SYM
ajst-16789	155	7	108718	108718	NUM
ajst-16789	155	8	.	.	PUNCT
ajst-16789	156	1	[	[	X
ajst-16789	156	2	5	5	NUM
ajst-16789	156	3	]	]	PUNCT
ajst-16789	156	4	localization	localization	NOUN
ajst-16789	156	5	algorithm	algorithm	NOUN
ajst-16789	156	6	and	and	CCONJ
ajst-16789	156	7	system	system	NOUN
ajst-16789	156	8	design	design	NOUN
ajst-16789	156	9	for	for	ADP
ajst-16789	156	10	vision	vision	NOUN
ajst-16789	156	11	-	-	PUNCT
ajst-16789	156	12	based	base	VERB
ajst-16789	156	13	navigation	navigation	NOUN
ajst-16789	156	14	agv	agv	PROPN
ajst-16789	156	15	journal	journal	PROPN
ajst-16789	156	16	of	of	ADP
ajst-16789	156	17	qingdao	qingdao	PROPN
ajst-16789	156	18	university(natural	university(natural	PROPN
ajst-16789	156	19	science	science	NOUN
ajst-16789	156	20	edition	edition	NOUN
ajst-16789	156	21	)	)	PUNCT
ajst-16789	156	22	,	,	PUNCT
ajst-16789	156	23	2022	2022	NUM
ajst-16789	156	24	,	,	PUNCT
ajst-16789	156	25	35	35	NUM
ajst-16789	156	26	:	:	SYM
ajst-16789	156	27	83	83	NUM
ajst-16789	156	28	-	-	SYM
ajst-16789	156	29	91	91	NUM
ajst-16789	156	30	.	.	PUNCT
ajst-16789	157	1	[	[	X
ajst-16789	157	2	6	6	NUM
ajst-16789	157	3	]	]	PUNCT
ajst-16789	157	4	he	he	PRON
ajst-16789	157	5	k	k	PROPN
ajst-16789	157	6	,	,	PUNCT
ajst-16789	157	7	zhang	zhang	PROPN
ajst-16789	157	8	x	x	PROPN
ajst-16789	157	9	,	,	PUNCT
ajst-16789	157	10	ren	ren	PROPN
ajst-16789	157	11	s	s	PROPN
ajst-16789	157	12	,	,	PUNCT
ajst-16789	157	13	et	et	PROPN
ajst-16789	157	14	al	al	PROPN
ajst-16789	157	15	.	.	PUNCT
ajst-16789	158	1	deep	deep	ADJ
ajst-16789	158	2	residual	residual	ADJ
ajst-16789	158	3	learning	learning	NOUN
ajst-16789	158	4	for	for	ADP
ajst-16789	158	5	image	image	NOUN
ajst-16789	158	6	recognition[c	recognition[c	PROPN
ajst-16789	158	7	]	]	PUNCT
ajst-16789	158	8	.	.	PUNCT
ajst-16789	159	1	proceedings	proceeding	NOUN
ajst-16789	159	2	of	of	ADP
ajst-16789	159	3	the	the	DET
ajst-16789	159	4	ieee	ieee	NOUN
ajst-16789	159	5	conference	conference	NOUN
ajst-16789	159	6	on	on	ADP
ajst-16789	159	7	computer	computer	NOUN
ajst-16789	159	8	vision	vision	NOUN
ajst-16789	159	9	and	and	CCONJ
ajst-16789	159	10	pattern	pattern	NOUN
ajst-16789	159	11	recognition	recognition	NOUN
ajst-16789	159	12	,	,	PUNCT
ajst-16789	159	13	2016	2016	NUM
ajst-16789	159	14	:	:	PUNCT
ajst-16789	159	15	770	770	NUM
ajst-16789	159	16	-	-	SYM
ajst-16789	159	17	778	778	NUM
ajst-16789	159	18	.	.	PUNCT
ajst-16789	160	1	[	[	X
ajst-16789	160	2	7	7	X
ajst-16789	160	3	]	]	X
ajst-16789	160	4	huang	huang	PROPN
ajst-16789	160	5	g	g	PROPN
ajst-16789	160	6	,	,	PUNCT
ajst-16789	160	7	liu	liu	PROPN
ajst-16789	160	8	z	z	PROPN
ajst-16789	160	9	,	,	PUNCT
ajst-16789	160	10	van	van	PROPN
ajst-16789	160	11	der	der	NOUN
ajst-16789	160	12	maaten	maaten	VERB
ajst-16789	160	13	l	l	NOUN
ajst-16789	160	14	,	,	PUNCT
ajst-16789	160	15	et	et	PROPN
ajst-16789	160	16	al	al	PROPN
ajst-16789	160	17	.	.	PROPN
ajst-16789	161	1	densely	densely	ADV
ajst-16789	161	2	connected	connect	VERB
ajst-16789	161	3	convolutional	convolutional	ADJ
ajst-16789	161	4	networks[c	networks[c	PROPN
ajst-16789	161	5	]	]	PUNCT
ajst-16789	161	6	.	.	PUNCT
ajst-16789	162	1	proceedings	proceeding	NOUN
ajst-16789	162	2	of	of	ADP
ajst-16789	162	3	the	the	DET
ajst-16789	162	4	ieee	ieee	NOUN
ajst-16789	162	5	conference	conference	NOUN
ajst-16789	162	6	on	on	ADP
ajst-16789	162	7	computer	computer	NOUN
ajst-16789	162	8	vision	vision	NOUN
ajst-16789	162	9	and	and	CCONJ
ajst-16789	162	10	pattern	pattern	NOUN
ajst-16789	162	11	recognition	recognition	NOUN
ajst-16789	162	12	,	,	PUNCT
ajst-16789	162	13	2017	2017	NUM
ajst-16789	162	14	:	:	SYM
ajst-16789	162	15	4700	4700	NUM
ajst-16789	162	16	-	-	SYM
ajst-16789	162	17	4708	4708	NUM
ajst-16789	162	18	.	.	PUNCT
ajst-16789	163	1	[	[	X
ajst-16789	163	2	8	8	NUM
ajst-16789	163	3	]	]	X
ajst-16789	163	4	eitel	eitel	NOUN
ajst-16789	163	5	a	a	NOUN
ajst-16789	163	6	,	,	PUNCT
ajst-16789	163	7	springenberg	springenberg	PROPN
ajst-16789	163	8	j	j	PROPN
ajst-16789	163	9	t	t	PROPN
ajst-16789	163	10	,	,	PUNCT
ajst-16789	163	11	spinello	spinello	PROPN
ajst-16789	163	12	l	l	NOUN
ajst-16789	163	13	,	,	PUNCT
ajst-16789	163	14	et	et	PROPN
ajst-16789	163	15	al	al	PROPN
ajst-16789	163	16	.	.	PUNCT
ajst-16789	163	17	multimodal	multimodal	ADJ
ajst-16789	163	18	deep	deep	ADJ
ajst-16789	163	19	learning	learning	NOUN
ajst-16789	163	20	for	for	ADP
ajst-16789	163	21	robust	robust	ADJ
ajst-16789	163	22	rgb	rgb	PROPN
ajst-16789	163	23	-	-	PUNCT
ajst-16789	163	24	d	d	PROPN
ajst-16789	163	25	object	object	NOUN
ajst-16789	163	26	recognition[c	recognition[c	NOUN
ajst-16789	163	27	]	]	PUNCT
ajst-16789	163	28	.	.	PUNCT
ajst-16789	164	1	2015	2015	NUM
ajst-16789	164	2	ieee	ieee	NOUN
ajst-16789	164	3	/	/	SYM
ajst-16789	164	4	rsj	rsj	NOUN
ajst-16789	164	5	international	international	ADJ
ajst-16789	164	6	conference	conference	NOUN
ajst-16789	164	7	on	on	ADP
ajst-16789	164	8	intelligent	intelligent	ADJ
ajst-16789	164	9	robots	robot	NOUN
ajst-16789	164	10	and	and	CCONJ
ajst-16789	164	11	systems	system	NOUN
ajst-16789	164	12	(	(	PUNCT
ajst-16789	164	13	iros	iro	NOUN
ajst-16789	164	14	)	)	PUNCT
ajst-16789	164	15	,	,	PUNCT
ajst-16789	164	16	2015	2015	NUM
ajst-16789	164	17	:	:	PUNCT
ajst-16789	164	18	681	681	NUM
ajst-16789	164	19	-	-	NOUN
ajst-16789	164	20	687	687	NUM
ajst-16789	164	21	.	.	PUNCT
ajst-16789	165	1	[	[	X
ajst-16789	165	2	9	9	NUM
ajst-16789	165	3	]	]	X
ajst-16789	165	4	wang	wang	PROPN
ajst-16789	165	5	y	y	PROPN
ajst-16789	165	6	,	,	PUNCT
ajst-16789	165	7	wang	wang	PROPN
ajst-16789	165	8	c	c	PROPN
ajst-16789	165	9	,	,	PUNCT
ajst-16789	165	10	long	long	ADV
ajst-16789	165	11	p	p	X
ajst-16789	165	12	,	,	PUNCT
ajst-16789	165	13	et	et	PROPN
ajst-16789	165	14	al	al	PROPN
ajst-16789	165	15	.	.	PUNCT
ajst-16789	166	1	recent	recent	ADJ
ajst-16789	166	2	advances	advance	NOUN
ajst-16789	166	3	in	in	ADP
ajst-16789	166	4	3d	3d	NUM
ajst-16789	166	5	object	object	NOUN
ajst-16789	166	6	detection	detection	NOUN
ajst-16789	166	7	based	base	VERB
ajst-16789	166	8	on	on	ADP
ajst-16789	166	9	rgb	rgb	PROPN
ajst-16789	166	10	-	-	PROPN
ajst-16789	166	11	d	d	PROPN
ajst-16789	166	12	:	:	PUNCT
ajst-16789	166	13	a	a	DET
ajst-16789	166	14	survey[j	survey[j	PROPN
ajst-16789	166	15	]	]	PUNCT
ajst-16789	166	16	.	.	PUNCT
ajst-16789	167	1	displays	display	NOUN
ajst-16789	167	2	,	,	PUNCT
ajst-16789	167	3	2021	2021	NUM
ajst-16789	167	4	,	,	PUNCT
ajst-16789	167	5	70	70	NUM
ajst-16789	167	6	:	:	SYM
ajst-16789	167	7	102077	102077	NUM
ajst-16789	167	8	.	.	PUNCT
ajst-16789	168	1	[	[	X
ajst-16789	168	2	10	10	NUM
ajst-16789	168	3	]	]	X
ajst-16789	168	4	rahman	rahman	PROPN
ajst-16789	168	5	m	m	PROPN
ajst-16789	168	6	m	m	PROPN
ajst-16789	168	7	,	,	PUNCT
ajst-16789	168	8	tan	tan	PROPN
ajst-16789	168	9	y	y	PROPN
ajst-16789	168	10	,	,	PUNCT
ajst-16789	168	11	xue	xue	PROPN
ajst-16789	168	12	j	j	PROPN
ajst-16789	168	13	,	,	PUNCT
ajst-16789	168	14	et	et	PROPN
ajst-16789	168	15	al	al	PROPN
ajst-16789	168	16	.	.	PROPN
ajst-16789	168	17	notice	notice	NOUN
ajst-16789	168	18	of	of	ADP
ajst-16789	168	19	violation	violation	NOUN
ajst-16789	168	20	of	of	ADP
ajst-16789	168	21	ieee	ieee	NOUN
ajst-16789	168	22	publication	publication	NOUN
ajst-16789	168	23	principles	principle	NOUN
ajst-16789	168	24	:	:	PUNCT
ajst-16789	168	25	recent	recent	ADJ
ajst-16789	168	26	advances	advance	NOUN
ajst-16789	168	27	in	in	ADP
ajst-16789	168	28	3d	3d	NUM
ajst-16789	168	29	object	object	NOUN
ajst-16789	168	30	detection	detection	NOUN
ajst-16789	168	31	in	in	ADP
ajst-16789	168	32	the	the	DET
ajst-16789	168	33	era	era	NOUN
ajst-16789	168	34	of	of	ADP
ajst-16789	168	35	deep	deep	ADJ
ajst-16789	168	36	neural	neural	ADJ
ajst-16789	168	37	networks	network	NOUN
ajst-16789	168	38	:	:	PUNCT
ajst-16789	168	39	a	a	DET
ajst-16789	168	40	survey[j	survey[j	PROPN
ajst-16789	168	41	]	]	PUNCT
ajst-16789	168	42	.	.	PUNCT
ajst-16789	169	1	ieee	ieee	NOUN
ajst-16789	169	2	transactions	transaction	NOUN
ajst-16789	169	3	on	on	ADP
ajst-16789	169	4	image	image	NOUN
ajst-16789	169	5	processing	processing	NOUN
ajst-16789	169	6	,	,	PUNCT
ajst-16789	169	7	2019	2019	NUM
ajst-16789	169	8	,	,	PUNCT
ajst-16789	169	9	29	29	NUM
ajst-16789	169	10	:	:	SYM
ajst-16789	169	11	2947	2947	NUM
ajst-16789	169	12	-	-	SYM
ajst-16789	169	13	2962	2962	NUM
ajst-16789	169	14	.	.	PUNCT
ajst-16789	170	1	[	[	X
ajst-16789	170	2	11	11	NUM
ajst-16789	170	3	]	]	PUNCT
ajst-16789	170	4	arnold	arnold	PROPN
ajst-16789	170	5	e	e	PROPN
ajst-16789	170	6	,	,	PUNCT
ajst-16789	170	7	al	al	PROPN
ajst-16789	170	8	-	-	PUNCT
ajst-16789	170	9	jarrah	jarrah	NOUN
ajst-16789	170	10	o	o	NOUN
ajst-16789	170	11	y	y	PROPN
ajst-16789	170	12	,	,	PUNCT
ajst-16789	170	13	dianati	dianati	PROPN
ajst-16789	170	14	m	m	PROPN
ajst-16789	170	15	,	,	PUNCT
ajst-16789	170	16	et	et	PROPN
ajst-16789	170	17	al	al	PROPN
ajst-16789	170	18	.	.	PUNCT
ajst-16789	171	1	a	a	DET
ajst-16789	171	2	survey	survey	NOUN
ajst-16789	171	3	on	on	ADP
ajst-16789	171	4	3d	3d	NUM
ajst-16789	171	5	object	object	NOUN
ajst-16789	171	6	detection	detection	NOUN
ajst-16789	171	7	methods	method	NOUN
ajst-16789	171	8	for	for	ADP
ajst-16789	171	9	autonomous	autonomous	ADJ
ajst-16789	171	10	driving	drive	VERB
ajst-16789	171	11	applications[j	applications[j	NOUN
ajst-16789	171	12	]	]	PUNCT
ajst-16789	171	13	.	.	PUNCT
ajst-16789	172	1	ieee	ieee	NOUN
ajst-16789	172	2	transactions	transaction	NOUN
ajst-16789	172	3	on	on	ADP
ajst-16789	172	4	intelligent	intelligent	ADJ
ajst-16789	172	5	transportation	transportation	NOUN
ajst-16789	172	6	systems	system	NOUN
ajst-16789	172	7	,	,	PUNCT
ajst-16789	172	8	2019	2019	NUM
ajst-16789	172	9	,	,	PUNCT
ajst-16789	172	10	20(10	20(10	NUM
ajst-16789	172	11	):	):	PUNCT
ajst-16789	172	12	3782	3782	NUM
ajst-16789	172	13	-	-	SYM
ajst-16789	172	14	3795	3795	NUM
ajst-16789	172	15	.	.	PUNCT
ajst-16789	173	1	[	[	X
ajst-16789	173	2	12	12	NUM
ajst-16789	173	3	]	]	X
ajst-16789	173	4	kruglyak	kruglyak	PROPN
ajst-16789	173	5	l	l	PROPN
ajst-16789	173	6	,	,	PUNCT
ajst-16789	173	7	lander	lander	PROPN
ajst-16789	173	8	e	e	PROPN
ajst-16789	173	9	s.	s.	PROPN
ajst-16789	173	10	complete	complete	PROPN
ajst-16789	173	11	multipoint	multipoint	PROPN
ajst-16789	173	12	sib	sib	PROPN
ajst-16789	173	13	-	-	PUNCT
ajst-16789	173	14	pair	pair	NOUN
ajst-16789	173	15	analysis	analysis	NOUN
ajst-16789	173	16	of	of	ADP
ajst-16789	173	17	qualitative	qualitative	NOUN
ajst-16789	173	18	and	and	CCONJ
ajst-16789	173	19	quantitative	quantitative	ADJ
ajst-16789	173	20	traits[j	traits[j	NOUN
ajst-16789	173	21	]	]	PUNCT
ajst-16789	173	22	.	.	PUNCT
ajst-16789	174	1	american	american	PROPN
ajst-16789	174	2	journal	journal	PROPN
ajst-16789	174	3	of	of	ADP
ajst-16789	174	4	human	human	ADJ
ajst-16789	174	5	genetics	genetic	NOUN
ajst-16789	174	6	,	,	PUNCT
ajst-16789	174	7	1995	1995	NUM
ajst-16789	174	8	,	,	PUNCT
ajst-16789	174	9	57(2	57(2	NUM
ajst-16789	174	10	):	):	PUNCT
ajst-16789	174	11	439	439	NUM
ajst-16789	174	12	.	.	PUNCT
ajst-16789	175	1	[	[	X
ajst-16789	175	2	13	13	NUM
ajst-16789	175	3	]	]	PUNCT
ajst-16789	175	4	vidal	vidal	PROPN
ajst-16789	175	5	j	j	PROPN
ajst-16789	175	6	,	,	PUNCT
ajst-16789	175	7	lin	lin	PROPN
ajst-16789	175	8	c	c	PROPN
ajst-16789	175	9	-	-	PUNCT
ajst-16789	175	10	y	y	PROPN
ajst-16789	175	11	,	,	PUNCT
ajst-16789	175	12	martí	martí	PROPN
ajst-16789	175	13	r.	r.	PROPN
ajst-16789	175	14	6d	6d	PROPN
ajst-16789	175	15	pose	pose	NOUN
ajst-16789	175	16	estimation	estimation	NOUN
ajst-16789	175	17	using	use	VERB
ajst-16789	175	18	an	an	DET
ajst-16789	175	19	improved	improved	ADJ
ajst-16789	175	20	method	method	NOUN
ajst-16789	175	21	based	base	VERB
ajst-16789	175	22	on	on	ADP
ajst-16789	175	23	point	point	NOUN
ajst-16789	175	24	pair	pair	NOUN
ajst-16789	175	25	features[c	features[c	PROPN
ajst-16789	175	26	]	]	PUNCT
ajst-16789	175	27	.	.	PUNCT
ajst-16789	175	28	2018	2018	NUM
ajst-16789	175	29	4th	4th	ADJ
ajst-16789	175	30	international	international	ADJ
ajst-16789	175	31	conference	conference	NOUN
ajst-16789	175	32	on	on	ADP
ajst-16789	175	33	control	control	NOUN
ajst-16789	175	34	,	,	PUNCT
ajst-16789	175	35	automation	automation	NOUN
ajst-16789	175	36	and	and	CCONJ
ajst-16789	175	37	robotics	robotic	NOUN
ajst-16789	175	38	(	(	PUNCT
ajst-16789	175	39	iccar	iccar	NOUN
ajst-16789	175	40	)	)	PUNCT
ajst-16789	175	41	,	,	PUNCT
ajst-16789	175	42	2018	2018	NUM
ajst-16789	175	43	:	:	PUNCT
ajst-16789	175	44	405	405	NUM
ajst-16789	175	45	-	-	SYM
ajst-16789	175	46	409	409	NUM
ajst-16789	175	47	.	.	PUNCT
ajst-16789	176	1	[	[	X
ajst-16789	176	2	14	14	NUM
ajst-16789	176	3	]	]	X
ajst-16789	176	4	zeng	zeng	PROPN
ajst-16789	176	5	a	a	X
ajst-16789	176	6	,	,	PUNCT
ajst-16789	176	7	song	song	NOUN
ajst-16789	176	8	s	s	PROPN
ajst-16789	176	9	,	,	PUNCT
ajst-16789	176	10	yu	yu	PROPN
ajst-16789	176	11	k	k	PROPN
ajst-16789	176	12	-	-	PUNCT
ajst-16789	176	13	t	t	PROPN
ajst-16789	176	14	,	,	PUNCT
ajst-16789	176	15	et	et	PROPN
ajst-16789	176	16	al	al	PROPN
ajst-16789	176	17	.	.	PROPN
ajst-16789	176	18	robotic	robotic	PROPN
ajst-16789	176	19	pick	pick	VERB
ajst-16789	176	20	-	-	PUNCT
ajst-16789	176	21	and	and	CCONJ
ajst-16789	176	22	-	-	PUNCT
ajst-16789	176	23	place	place	NOUN
ajst-16789	176	24	of	of	ADP
ajst-16789	176	25	novel	novel	ADJ
ajst-16789	176	26	objects	object	NOUN
ajst-16789	176	27	in	in	ADP
ajst-16789	176	28	clutter	clutter	NOUN
ajst-16789	176	29	with	with	ADP
ajst-16789	176	30	multi	multi	ADJ
ajst-16789	176	31	-	-	ADJ
ajst-16789	176	32	affordance	affordance	ADJ
ajst-16789	176	33	grasping	grasp	VERB
ajst-16789	176	34	and	and	CCONJ
ajst-16789	176	35	crossdomain	crossdomain	ADJ
ajst-16789	176	36	image	image	NOUN
ajst-16789	176	37	matching[j	matching[j	NOUN
ajst-16789	176	38	]	]	PUNCT
ajst-16789	176	39	.	.	PUNCT
ajst-16789	177	1	international	international	ADJ
ajst-16789	177	2	journal	journal	PROPN
ajst-16789	177	3	of	of	ADP
ajst-16789	177	4	robotics	robotic	NOUN
ajst-16789	177	5	research	research	NOUN
ajst-16789	177	6	,	,	PUNCT
ajst-16789	177	7	2022	2022	NUM
ajst-16789	177	8	,	,	PUNCT
ajst-16789	177	9	41(7	41(7	NUM
ajst-16789	177	10	):	):	PUNCT
ajst-16789	177	11	690	690	NUM
ajst-16789	177	12	-	-	SYM
ajst-16789	177	13	705	705	NUM
ajst-16789	177	14	.	.	PUNCT
ajst-16789	178	1	[	[	X
ajst-16789	178	2	15	15	NUM
ajst-16789	178	3	]	]	X
ajst-16789	178	4	woo	woo	NOUN
ajst-16789	178	5	s	s	PROPN
ajst-16789	178	6	,	,	PUNCT
ajst-16789	178	7	park	park	PROPN
ajst-16789	178	8	j	j	PROPN
ajst-16789	178	9	,	,	PUNCT
ajst-16789	178	10	lee	lee	PROPN
ajst-16789	178	11	j	j	PROPN
ajst-16789	178	12	-	-	PROPN
ajst-16789	178	13	y	y	PROPN
ajst-16789	178	14	,	,	PUNCT
ajst-16789	178	15	et	et	PROPN
ajst-16789	178	16	al	al	PROPN
ajst-16789	178	17	.	.	PUNCT
ajst-16789	178	18	cbam	cbam	NOUN
ajst-16789	178	19	:	:	PUNCT
ajst-16789	178	20	convolutional	convolutional	ADJ
ajst-16789	178	21	block	block	NOUN
ajst-16789	178	22	attention	attention	NOUN
ajst-16789	178	23	module[c	module[c	NOUN
ajst-16789	178	24	]	]	PUNCT
ajst-16789	178	25	.	.	PUNCT
ajst-16789	179	1	proceedings	proceeding	NOUN
ajst-16789	179	2	of	of	ADP
ajst-16789	179	3	the	the	DET
ajst-16789	179	4	european	european	PROPN
ajst-16789	179	5	conference	conference	PROPN
ajst-16789	179	6	on	on	ADP
ajst-16789	179	7	computer	computer	NOUN
ajst-16789	179	8	vision	vision	NOUN
ajst-16789	179	9	(	(	PUNCT
ajst-16789	179	10	eccv	eccv	ADV
ajst-16789	179	11	)	)	PUNCT
ajst-16789	179	12	,	,	PUNCT
ajst-16789	179	13	2018	2018	NUM
ajst-16789	179	14	:	:	PUNCT
ajst-16789	179	15	3	3	NUM
ajst-16789	179	16	-	-	SYM
ajst-16789	179	17	19	19	NUM
ajst-16789	179	18	.	.	PUNCT