id	sid	tid	token	lemma	pos
ajst-390	1	1	academic	academic	ADJ
ajst-390	1	2	journal	journal	NOUN
ajst-390	1	3	of	of	ADP
ajst-390	1	4	science	science	NOUN
ajst-390	1	5	and	and	CCONJ
ajst-390	1	6	technology	technology	NOUN
ajst-390	1	7	issn	issn	NOUN
ajst-390	1	8	:	:	PUNCT
ajst-390	1	9	2771	2771	NUM
ajst-390	1	10	-	-	SYM
ajst-390	1	11	3032	3032	NUM
ajst-390	1	12	|	|	NOUN
ajst-390	1	13	vol	vol	NOUN
ajst-390	1	14	.	.	PROPN
ajst-390	2	1	1	1	NUM
ajst-390	2	2	,	,	PUNCT
ajst-390	2	3	no	no	INTJ
ajst-390	2	4	.	.	NOUN
ajst-390	2	5	3	3	NUM
ajst-390	2	6	,	,	PUNCT
ajst-390	2	7	2022	2022	NUM
ajst-390	2	8	1	1	NUM
ajst-390	2	9	research	research	NOUN
ajst-390	2	10	of	of	ADP
ajst-390	2	11	speech	speech	NOUN
ajst-390	2	12	style	style	NOUN
ajst-390	2	13	transfer	transfer	NOUN
ajst-390	2	14	based	base	VERB
ajst-390	2	15	on	on	ADP
ajst-390	2	16	neural	neural	ADJ
ajst-390	2	17	network	network	NOUN
ajst-390	2	18	ziyi	ziyi	PROPN
ajst-390	2	19	sun	sun	PROPN
ajst-390	2	20	school	school	PROPN
ajst-390	2	21	of	of	ADP
ajst-390	2	22	economics	economics	PROPN
ajst-390	2	23	,	,	PUNCT
ajst-390	2	24	jinan	jinan	PROPN
ajst-390	2	25	university	university	PROPN
ajst-390	2	26	,	,	PUNCT
ajst-390	2	27	guangzhou	guangzhou	PROPN
ajst-390	2	28	,	,	PUNCT
ajst-390	2	29	510632	510632	NUM
ajst-390	2	30	,	,	PUNCT
ajst-390	2	31	china	china	PROPN
ajst-390	2	32	abstract	abstract	NOUN
ajst-390	2	33	:	:	PUNCT
ajst-390	2	34	this	this	DET
ajst-390	2	35	paper	paper	NOUN
ajst-390	2	36	draws	draw	VERB
ajst-390	2	37	inspiration	inspiration	NOUN
ajst-390	2	38	from	from	ADP
ajst-390	2	39	image	image	NOUN
ajst-390	2	40	style	style	NOUN
ajst-390	2	41	transfer	transfer	NOUN
ajst-390	2	42	model	model	NOUN
ajst-390	2	43	neural	neural	ADJ
ajst-390	2	44	style	style	NOUN
ajst-390	2	45	transfer	transfer	NOUN
ajst-390	2	46	,	,	PUNCT
ajst-390	2	47	which	which	PRON
ajst-390	2	48	leads	lead	VERB
ajst-390	2	49	to	to	ADP
ajst-390	2	50	the	the	DET
ajst-390	2	51	research	research	NOUN
ajst-390	2	52	topic	topic	NOUN
ajst-390	2	53	of	of	ADP
ajst-390	2	54	speech	speech	NOUN
ajst-390	2	55	style	style	NOUN
ajst-390	2	56	transfer	transfer	NOUN
ajst-390	2	57	based	base	VERB
ajst-390	2	58	on	on	ADP
ajst-390	2	59	neural	neural	ADJ
ajst-390	2	60	network	network	NOUN
ajst-390	2	61	.	.	PUNCT
ajst-390	3	1	first	first	ADV
ajst-390	3	2	,	,	PUNCT
ajst-390	3	3	the	the	DET
ajst-390	3	4	article	article	NOUN
ajst-390	3	5	describes	describe	VERB
ajst-390	3	6	the	the	DET
ajst-390	3	7	extraction	extraction	NOUN
ajst-390	3	8	process	process	NOUN
ajst-390	3	9	on	on	ADP
ajst-390	3	10	2d	2d	NUM
ajst-390	3	11	spectrogram	spectrogram	NOUN
ajst-390	3	12	of	of	ADP
ajst-390	3	13	speech	speech	NOUN
ajst-390	3	14	signal	signal	NOUN
ajst-390	3	15	.	.	PUNCT
ajst-390	4	1	then	then	ADV
ajst-390	4	2	,	,	PUNCT
ajst-390	4	3	the	the	DET
ajst-390	4	4	speech	speech	NOUN
ajst-390	4	5	style	style	NOUN
ajst-390	4	6	transfer	transfer	NOUN
ajst-390	4	7	based	base	VERB
ajst-390	4	8	on	on	ADP
ajst-390	4	9	convolutional	convolutional	ADJ
ajst-390	4	10	neural	neural	ADJ
ajst-390	4	11	network	network	NOUN
ajst-390	4	12	is	be	AUX
ajst-390	4	13	constructed	construct	VERB
ajst-390	4	14	.	.	PUNCT
ajst-390	5	1	keywords	keyword	NOUN
ajst-390	5	2	:	:	PUNCT
ajst-390	5	3	image	image	NOUN
ajst-390	5	4	style	style	NOUN
ajst-390	5	5	transfer	transfer	NOUN
ajst-390	5	6	,	,	PUNCT
ajst-390	5	7	speech	speech	NOUN
ajst-390	5	8	style	style	NOUN
ajst-390	5	9	transfer	transfer	NOUN
ajst-390	5	10	,	,	PUNCT
ajst-390	5	11	convolutional	convolutional	ADJ
ajst-390	5	12	neural	neural	ADJ
ajst-390	5	13	network	network	NOUN
ajst-390	5	14	,	,	PUNCT
ajst-390	5	15	2d	2d	NUM
ajst-390	5	16	spectrogram	spectrogram	NOUN
ajst-390	5	17	.	.	PUNCT
ajst-390	6	1	1	1	X
ajst-390	6	2	.	.	X
ajst-390	6	3	introduction	introduction	NOUN
ajst-390	6	4	voice	voice	NOUN
ajst-390	6	5	conversion	conversion	NOUN
ajst-390	6	6	(	(	PUNCT
ajst-390	6	7	vc	vc	NOUN
ajst-390	6	8	)	)	PUNCT
ajst-390	7	1	[	[	X
ajst-390	7	2	1	1	X
ajst-390	7	3	]	]	PUNCT
ajst-390	7	4	refers	refer	VERB
ajst-390	7	5	to	to	ADP
ajst-390	7	6	the	the	DET
ajst-390	7	7	conversion	conversion	NOUN
ajst-390	7	8	of	of	ADP
ajst-390	7	9	the	the	DET
ajst-390	7	10	phonetic	phonetic	ADJ
ajst-390	7	11	style	style	NOUN
ajst-390	7	12	features	feature	NOUN
ajst-390	7	13	of	of	ADP
ajst-390	7	14	the	the	DET
ajst-390	7	15	source	source	NOUN
ajst-390	7	16	speaker	speaker	NOUN
ajst-390	7	17	to	to	ADP
ajst-390	7	18	those	those	PRON
ajst-390	7	19	of	of	ADP
ajst-390	7	20	the	the	DET
ajst-390	7	21	target	target	NOUN
ajst-390	7	22	speaker	speaker	NOUN
ajst-390	7	23	,	,	PUNCT
ajst-390	7	24	while	while	SCONJ
ajst-390	7	25	keeping	keep	VERB
ajst-390	7	26	the	the	DET
ajst-390	7	27	semantic	semantic	ADJ
ajst-390	7	28	information	information	NOUN
ajst-390	7	29	of	of	ADP
ajst-390	7	30	the	the	DET
ajst-390	7	31	source	source	NOUN
ajst-390	7	32	speaker	speaker	NOUN
ajst-390	7	33	unchanged	unchanged	ADJ
ajst-390	7	34	.	.	PUNCT
ajst-390	8	1	actually	actually	ADV
ajst-390	8	2	,	,	PUNCT
ajst-390	8	3	speech	speech	NOUN
ajst-390	8	4	style	style	NOUN
ajst-390	8	5	transfer	transfer	NOUN
ajst-390	8	6	can	can	AUX
ajst-390	8	7	be	be	AUX
ajst-390	8	8	applied	apply	VERB
ajst-390	8	9	to	to	ADP
ajst-390	8	10	communication	communication	NOUN
ajst-390	8	11	,	,	PUNCT
ajst-390	8	12	medical	medical	ADJ
ajst-390	8	13	care	care	NOUN
ajst-390	8	14	,	,	PUNCT
ajst-390	8	15	entertainment	entertainment	NOUN
ajst-390	8	16	and	and	CCONJ
ajst-390	8	17	other	other	ADJ
ajst-390	8	18	fields	field	NOUN
ajst-390	8	19	:	:	PUNCT
ajst-390	8	20	in	in	ADP
ajst-390	8	21	the	the	DET
ajst-390	8	22	text	text	NOUN
ajst-390	8	23	-	-	PUNCT
ajst-390	8	24	to	to	ADP
ajst-390	8	25	-	-	PUNCT
ajst-390	8	26	speech	speech	NOUN
ajst-390	8	27	(	(	PUNCT
ajst-390	8	28	tts	tts	PROPN
ajst-390	8	29	)	)	PUNCT
ajst-390	8	30	model	model	NOUN
ajst-390	8	31	[	[	X
ajst-390	8	32	2	2	NUM
ajst-390	8	33	]	]	PUNCT
ajst-390	8	34	,	,	PUNCT
ajst-390	8	35	the	the	DET
ajst-390	8	36	synthesized	synthesize	VERB
ajst-390	8	37	speech	speech	NOUN
ajst-390	8	38	might	might	AUX
ajst-390	8	39	sound	sound	VERB
ajst-390	8	40	more	more	ADV
ajst-390	8	41	like	like	ADP
ajst-390	8	42	the	the	DET
ajst-390	8	43	voice	voice	NOUN
ajst-390	8	44	delivered	deliver	VERB
ajst-390	8	45	by	by	ADP
ajst-390	8	46	a	a	DET
ajst-390	8	47	real	real	ADJ
ajst-390	8	48	person	person	NOUN
ajst-390	8	49	if	if	SCONJ
ajst-390	8	50	it	it	PRON
ajst-390	8	51	is	be	AUX
ajst-390	8	52	processed	process	VERB
ajst-390	8	53	by	by	ADP
ajst-390	8	54	speech	speech	NOUN
ajst-390	8	55	style	style	NOUN
ajst-390	8	56	transfer	transfer	NOUN
ajst-390	8	57	at	at	ADP
ajst-390	8	58	the	the	DET
ajst-390	8	59	meantime	meantime	NOUN
ajst-390	8	60	;	;	PUNCT
ajst-390	8	61	for	for	ADP
ajst-390	8	62	the	the	DET
ajst-390	8	63	sake	sake	NOUN
ajst-390	8	64	of	of	ADP
ajst-390	8	65	confidentiality	confidentiality	NOUN
ajst-390	8	66	and	and	CCONJ
ajst-390	8	67	security	security	NOUN
ajst-390	8	68	,	,	PUNCT
ajst-390	8	69	speech	speech	NOUN
ajst-390	8	70	style	style	NOUN
ajst-390	8	71	transfer	transfer	NOUN
ajst-390	8	72	technology	technology	NOUN
ajst-390	8	73	can	can	AUX
ajst-390	8	74	be	be	AUX
ajst-390	8	75	adopted	adopt	VERB
ajst-390	8	76	to	to	PART
ajst-390	8	77	change	change	VERB
ajst-390	8	78	the	the	DET
ajst-390	8	79	style	style	NOUN
ajst-390	8	80	characteristics	characteristic	NOUN
ajst-390	8	81	of	of	ADP
ajst-390	8	82	speaker	speaker	NOUN
ajst-390	8	83	's	's	PART
ajst-390	8	84	voice	voice	NOUN
ajst-390	9	1	[	[	X
ajst-390	9	2	3	3	NUM
ajst-390	9	3	]	]	PUNCT
ajst-390	9	4	;	;	PUNCT
ajst-390	9	5	in	in	ADP
ajst-390	9	6	the	the	DET
ajst-390	9	7	medical	medical	ADJ
ajst-390	9	8	field	field	NOUN
ajst-390	9	9	,	,	PUNCT
ajst-390	9	10	the	the	DET
ajst-390	9	11	speech	speech	NOUN
ajst-390	9	12	signal	signal	NOUN
ajst-390	9	13	phonated	phonate	VERB
ajst-390	9	14	by	by	ADP
ajst-390	9	15	patients	patient	NOUN
ajst-390	9	16	with	with	ADP
ajst-390	9	17	damaged	damage	VERB
ajst-390	9	18	larynx	larynx	NOUN
ajst-390	9	19	can	can	AUX
ajst-390	9	20	be	be	AUX
ajst-390	9	21	repaired	repair	VERB
ajst-390	9	22	with	with	ADP
ajst-390	9	23	the	the	DET
ajst-390	9	24	help	help	NOUN
ajst-390	9	25	of	of	ADP
ajst-390	9	26	voice	voice	NOUN
ajst-390	9	27	style	style	NOUN
ajst-390	9	28	transfer	transfer	NOUN
ajst-390	9	29	technology	technology	NOUN
ajst-390	9	30	[	[	X
ajst-390	9	31	4	4	NUM
ajst-390	9	32	]	]	PUNCT
ajst-390	9	33	;	;	PUNCT
ajst-390	9	34	in	in	ADP
ajst-390	9	35	the	the	DET
ajst-390	9	36	film	film	NOUN
ajst-390	9	37	dubbing	dub	VERB
ajst-390	9	38	field	field	NOUN
ajst-390	9	39	,	,	PUNCT
ajst-390	9	40	especially	especially	ADV
ajst-390	9	41	,	,	PUNCT
ajst-390	9	42	when	when	SCONJ
ajst-390	9	43	another	another	DET
ajst-390	9	44	language	language	NOUN
ajst-390	9	45	is	be	AUX
ajst-390	9	46	used	use	VERB
ajst-390	9	47	in	in	ADP
ajst-390	9	48	film	film	NOUN
ajst-390	9	49	dubbing	dubbing	NOUN
ajst-390	9	50	,	,	PUNCT
ajst-390	9	51	speech	speech	NOUN
ajst-390	9	52	style	style	NOUN
ajst-390	9	53	transfer	transfer	NOUN
ajst-390	9	54	technology	technology	NOUN
ajst-390	9	55	can	can	AUX
ajst-390	9	56	make	make	VERB
ajst-390	9	57	the	the	DET
ajst-390	9	58	voice	voice	NOUN
ajst-390	9	59	style	style	NOUN
ajst-390	9	60	of	of	ADP
ajst-390	9	61	the	the	DET
ajst-390	9	62	dubbing	dubbing	NOUN
ajst-390	9	63	actor	actor	NOUN
ajst-390	9	64	the	the	DET
ajst-390	9	65	same	same	ADJ
ajst-390	9	66	as	as	ADP
ajst-390	9	67	that	that	PRON
ajst-390	9	68	of	of	ADP
ajst-390	9	69	the	the	DET
ajst-390	9	70	film	film	NOUN
ajst-390	9	71	actor	actor	NOUN
ajst-390	9	72	,	,	PUNCT
ajst-390	9	73	and	and	CCONJ
ajst-390	9	74	the	the	DET
ajst-390	9	75	ideal	ideal	ADJ
ajst-390	9	76	dubbing	dubbing	NOUN
ajst-390	9	77	effect	effect	NOUN
ajst-390	9	78	is	be	AUX
ajst-390	9	79	achieved	achieve	VERB
ajst-390	9	80	finally[5	finally[5	ADV
ajst-390	9	81	]	]	PUNCT
ajst-390	9	82	.	.	PUNCT
ajst-390	10	1	it	it	PRON
ajst-390	10	2	can	can	AUX
ajst-390	10	3	be	be	AUX
ajst-390	10	4	seen	see	VERB
ajst-390	10	5	from	from	ADP
ajst-390	10	6	the	the	DET
ajst-390	10	7	aforementioned	aforementioned	ADJ
ajst-390	10	8	examples	example	NOUN
ajst-390	10	9	that	that	PRON
ajst-390	10	10	speech	speech	NOUN
ajst-390	10	11	style	style	NOUN
ajst-390	10	12	transfer	transfer	NOUN
ajst-390	10	13	technology	technology	NOUN
ajst-390	10	14	,	,	PUNCT
ajst-390	10	15	as	as	ADP
ajst-390	10	16	a	a	DET
ajst-390	10	17	subject	subject	NOUN
ajst-390	10	18	with	with	ADP
ajst-390	10	19	strong	strong	ADJ
ajst-390	10	20	interdisciplinarity	interdisciplinarity	NOUN
ajst-390	10	21	,	,	PUNCT
ajst-390	10	22	has	have	VERB
ajst-390	10	23	extremely	extremely	ADV
ajst-390	10	24	important	important	ADJ
ajst-390	10	25	research	research	NOUN
ajst-390	10	26	role	role	NOUN
ajst-390	10	27	and	and	CCONJ
ajst-390	10	28	value	value	NOUN
ajst-390	10	29	,	,	PUNCT
ajst-390	10	30	and	and	CCONJ
ajst-390	10	31	attracts	attract	VERB
ajst-390	10	32	numerous	numerous	ADJ
ajst-390	10	33	researchers	researcher	NOUN
ajst-390	10	34	to	to	PART
ajst-390	10	35	explore	explore	VERB
ajst-390	10	36	.	.	PUNCT
ajst-390	11	1	in	in	ADP
ajst-390	11	2	1988	1988	NUM
ajst-390	11	3	,	,	PUNCT
ajst-390	11	4	abe	abe	PROPN
ajst-390	11	5	’s	’s	PART
ajst-390	11	6	group	group	NOUN
ajst-390	11	7	[	[	X
ajst-390	11	8	6	6	NUM
ajst-390	11	9	]	]	PUNCT
ajst-390	11	10	firstly	firstly	ADV
ajst-390	11	11	proposed	propose	VERB
ajst-390	11	12	speech	speech	NOUN
ajst-390	11	13	style	style	NOUN
ajst-390	11	14	transfer	transfer	NOUN
ajst-390	11	15	based	base	VERB
ajst-390	11	16	on	on	ADP
ajst-390	11	17	vector	vector	NOUN
ajst-390	11	18	quantization	quantization	NOUN
ajst-390	11	19	(	(	PUNCT
ajst-390	11	20	vq	vq	NOUN
ajst-390	11	21	)	)	PUNCT
ajst-390	11	22	and	and	CCONJ
ajst-390	11	23	codebook	codebook	NOUN
ajst-390	11	24	mapping	mapping	NOUN
ajst-390	11	25	.	.	PUNCT
ajst-390	12	1	the	the	DET
ajst-390	12	2	spectral	spectral	ADJ
ajst-390	12	3	envelope	envelope	NOUN
ajst-390	12	4	characteristic	characteristic	ADJ
ajst-390	12	5	parameters	parameter	NOUN
ajst-390	12	6	of	of	ADP
ajst-390	12	7	primitive	primitive	ADJ
ajst-390	12	8	and	and	CCONJ
ajst-390	12	9	stylized	stylized	ADJ
ajst-390	12	10	speeches	speech	NOUN
ajst-390	12	11	are	be	AUX
ajst-390	12	12	divided	divide	VERB
ajst-390	12	13	into	into	ADP
ajst-390	12	14	a	a	DET
ajst-390	12	15	series	series	NOUN
ajst-390	12	16	of	of	ADP
ajst-390	12	17	codebooks	codebook	NOUN
ajst-390	12	18	by	by	ADP
ajst-390	12	19	vector	vector	NOUN
ajst-390	12	20	quantization	quantization	NOUN
ajst-390	12	21	and	and	CCONJ
ajst-390	12	22	speech	speech	NOUN
ajst-390	12	23	style	style	NOUN
ajst-390	12	24	transfer	transfer	NOUN
ajst-390	12	25	is	be	AUX
ajst-390	12	26	realized	realize	VERB
ajst-390	12	27	by	by	ADP
ajst-390	12	28	establishing	establish	VERB
ajst-390	12	29	the	the	DET
ajst-390	12	30	mapping	mapping	NOUN
ajst-390	12	31	relationship	relationship	NOUN
ajst-390	12	32	between	between	ADP
ajst-390	12	33	the	the	DET
ajst-390	12	34	codebooks	codebook	NOUN
ajst-390	12	35	.	.	PUNCT
ajst-390	13	1	although	although	SCONJ
ajst-390	13	2	this	this	DET
ajst-390	13	3	method	method	NOUN
ajst-390	13	4	is	be	AUX
ajst-390	13	5	simple	simple	ADJ
ajst-390	13	6	and	and	CCONJ
ajst-390	13	7	easy	easy	ADJ
ajst-390	13	8	to	to	PART
ajst-390	13	9	implement	implement	VERB
ajst-390	13	10	,	,	PUNCT
ajst-390	13	11	the	the	DET
ajst-390	13	12	quality	quality	NOUN
ajst-390	13	13	of	of	ADP
ajst-390	13	14	transformed	transform	VERB
ajst-390	13	15	voice	voice	NOUN
ajst-390	13	16	is	be	AUX
ajst-390	13	17	poor	poor	ADJ
ajst-390	13	18	,	,	PUNCT
ajst-390	13	19	discontinued	discontinued	ADJ
ajst-390	13	20	and	and	CCONJ
ajst-390	13	21	intermittent	intermittent	ADJ
ajst-390	13	22	.	.	PUNCT
ajst-390	14	1	in	in	ADP
ajst-390	14	2	1992	1992	NUM
ajst-390	14	3	,	,	PUNCT
ajst-390	14	4	savic[7	savic[7	PROPN
ajst-390	14	5	]	]	X
ajst-390	14	6	et	et	PROPN
ajst-390	14	7	al	al	PROPN
ajst-390	14	8	.	.	PROPN
ajst-390	14	9	improved	improve	VERB
ajst-390	14	10	the	the	DET
ajst-390	14	11	codebooks	codebook	NOUN
ajst-390	14	12	mapping	mapping	NOUN
ajst-390	14	13	into	into	ADP
ajst-390	14	14	a	a	DET
ajst-390	14	15	neural	neural	ADJ
ajst-390	14	16	network	network	NOUN
ajst-390	14	17	on	on	ADP
ajst-390	14	18	the	the	DET
ajst-390	14	19	basis	basis	NOUN
ajst-390	14	20	of	of	ADP
ajst-390	14	21	abe	abe	PROPN
ajst-390	14	22	's	's	PART
ajst-390	14	23	research	research	NOUN
ajst-390	14	24	,	,	PUNCT
ajst-390	14	25	which	which	PRON
ajst-390	14	26	greatly	greatly	ADV
ajst-390	14	27	enhanced	enhance	VERB
ajst-390	14	28	the	the	DET
ajst-390	14	29	quality	quality	NOUN
ajst-390	14	30	of	of	ADP
ajst-390	14	31	converted	converted	ADJ
ajst-390	14	32	speech	speech	NOUN
ajst-390	14	33	.	.	PUNCT
ajst-390	15	1	this	this	PRON
ajst-390	15	2	is	be	AUX
ajst-390	15	3	the	the	DET
ajst-390	15	4	first	first	ADJ
ajst-390	15	5	time	time	NOUN
ajst-390	15	6	to	to	PART
ajst-390	15	7	apply	apply	VERB
ajst-390	15	8	the	the	DET
ajst-390	15	9	artificial	artificial	ADJ
ajst-390	15	10	neural	neural	ADJ
ajst-390	15	11	network	network	NOUN
ajst-390	15	12	model	model	NOUN
ajst-390	15	13	in	in	ADP
ajst-390	15	14	the	the	DET
ajst-390	15	15	study	study	NOUN
ajst-390	15	16	of	of	ADP
ajst-390	15	17	speech	speech	NOUN
ajst-390	15	18	style	style	NOUN
ajst-390	15	19	transfer	transfer	NOUN
ajst-390	15	20	and	and	CCONJ
ajst-390	15	21	has	have	AUX
ajst-390	15	22	made	make	VERB
ajst-390	15	23	a	a	DET
ajst-390	15	24	certain	certain	ADJ
ajst-390	15	25	breakthrough	breakthrough	NOUN
ajst-390	15	26	.	.	PUNCT
ajst-390	16	1	subsequently	subsequently	ADV
ajst-390	16	2	,	,	PUNCT
ajst-390	16	3	the	the	DET
ajst-390	16	4	research	research	NOUN
ajst-390	16	5	of	of	ADP
ajst-390	16	6	speech	speech	NOUN
ajst-390	16	7	style	style	NOUN
ajst-390	16	8	transfer	transfer	NOUN
ajst-390	16	9	based	base	VERB
ajst-390	16	10	on	on	ADP
ajst-390	16	11	neural	neural	ADJ
ajst-390	16	12	network	network	NOUN
ajst-390	16	13	becomes	become	VERB
ajst-390	16	14	the	the	DET
ajst-390	16	15	mainstream	mainstream	NOUN
ajst-390	16	16	research	research	NOUN
ajst-390	16	17	direction	direction	NOUN
ajst-390	16	18	.	.	PUNCT
ajst-390	17	1	in	in	ADP
ajst-390	17	2	1995	1995	NUM
ajst-390	17	3	,	,	PUNCT
ajst-390	17	4	narendranath	narendranath	NOUN
ajst-390	18	1	[	[	X
ajst-390	18	2	8	8	NUM
ajst-390	18	3	]	]	PUNCT
ajst-390	18	4	et	et	PROPN
ajst-390	18	5	al	al	PROPN
ajst-390	18	6	.	.	PROPN
ajst-390	18	7	took	take	VERB
ajst-390	18	8	the	the	DET
ajst-390	18	9	formant	formant	NOUN
ajst-390	18	10	parameters	parameter	NOUN
ajst-390	18	11	of	of	ADP
ajst-390	18	12	the	the	DET
ajst-390	18	13	original	original	ADJ
ajst-390	18	14	and	and	CCONJ
ajst-390	18	15	stylized	stylize	VERB
ajst-390	18	16	speech	speech	NOUN
ajst-390	18	17	signals	signal	NOUN
ajst-390	18	18	as	as	ADP
ajst-390	18	19	the	the	DET
ajst-390	18	20	inputs	input	NOUN
ajst-390	18	21	and	and	CCONJ
ajst-390	18	22	outputs	output	NOUN
ajst-390	18	23	of	of	ADP
ajst-390	18	24	the	the	DET
ajst-390	18	25	artificial	artificial	ADJ
ajst-390	18	26	neural	neural	ADJ
ajst-390	18	27	network	network	NOUN
ajst-390	18	28	(	(	PUNCT
ajst-390	18	29	ann	ann	PROPN
ajst-390	18	30	)	)	PUNCT
ajst-390	18	31	respectively	respectively	ADV
ajst-390	18	32	and	and	CCONJ
ajst-390	18	33	trained	train	VERB
ajst-390	18	34	them	they	PRON
ajst-390	18	35	by	by	ADP
ajst-390	18	36	dint	dint	NOUN
ajst-390	18	37	of	of	ADP
ajst-390	18	38	bp	bp	PROPN
ajst-390	18	39	method	method	NOUN
ajst-390	18	40	.	.	PUNCT
ajst-390	19	1	ultimately	ultimately	ADV
ajst-390	19	2	,	,	PUNCT
ajst-390	19	3	stylized	stylize	VERB
ajst-390	19	4	speech	speech	NOUN
ajst-390	19	5	signal	signal	NOUN
ajst-390	19	6	was	be	AUX
ajst-390	19	7	synthesized	synthesize	VERB
ajst-390	19	8	with	with	ADP
ajst-390	19	9	the	the	DET
ajst-390	19	10	average	average	ADJ
ajst-390	19	11	gene	gene	NOUN
ajst-390	19	12	frequency	frequency	NOUN
ajst-390	19	13	and	and	CCONJ
ajst-390	19	14	stylized	stylize	VERB
ajst-390	19	15	formant	formant	NOUN
ajst-390	19	16	parameters	parameter	NOUN
ajst-390	19	17	,	,	PUNCT
ajst-390	19	18	which	which	PRON
ajst-390	19	19	were	be	AUX
ajst-390	19	20	converted	convert	VERB
ajst-390	19	21	by	by	ADP
ajst-390	19	22	the	the	DET
ajst-390	19	23	formant	formant	NOUN
ajst-390	19	24	parameters	parameter	NOUN
ajst-390	19	25	of	of	ADP
ajst-390	19	26	the	the	DET
ajst-390	19	27	tested	test	VERB
ajst-390	19	28	speech	speech	NOUN
ajst-390	19	29	.	.	PUNCT
ajst-390	20	1	in	in	ADP
ajst-390	20	2	2002	2002	NUM
ajst-390	20	3	,	,	PUNCT
ajst-390	20	4	scholars	scholar	NOUN
ajst-390	20	5	watanabe	watanabe	PROPN
ajst-390	21	1	[	[	X
ajst-390	21	2	9	9	NUM
ajst-390	21	3	]	]	PUNCT
ajst-390	21	4	et	et	PROPN
ajst-390	21	5	al	al	PROPN
ajst-390	21	6	.	.	PROPN
ajst-390	21	7	come	come	VERB
ajst-390	21	8	up	up	ADP
ajst-390	21	9	with	with	ADP
ajst-390	21	10	an	an	DET
ajst-390	21	11	algorithm	algorithm	NOUN
ajst-390	21	12	for	for	ADP
ajst-390	21	13	lpc	lpc	PROPN
ajst-390	21	14	spectral	spectral	PROPN
ajst-390	21	15	envelope	envelope	PROPN
ajst-390	21	16	transformation	transformation	NOUN
ajst-390	21	17	using	use	VERB
ajst-390	21	18	rbf	rbf	PROPN
ajst-390	21	19	neural	neural	ADJ
ajst-390	21	20	network	network	NOUN
ajst-390	21	21	.	.	PUNCT
ajst-390	22	1	in	in	ADP
ajst-390	22	2	2007	2007	NUM
ajst-390	22	3	,	,	PUNCT
ajst-390	22	4	guldo	guldo	X
ajst-390	23	1	[	[	X
ajst-390	23	2	10	10	NUM
ajst-390	23	3	]	]	PUNCT
ajst-390	23	4	et	et	PROPN
ajst-390	23	5	al	al	PROPN
ajst-390	23	6	.	.	PROPN
ajst-390	24	1	put	put	VERB
ajst-390	24	2	forward	forward	ADV
ajst-390	24	3	a	a	DET
ajst-390	24	4	speech	speech	NOUN
ajst-390	24	5	style	style	NOUN
ajst-390	24	6	conversion	conversion	NOUN
ajst-390	24	7	algorithm	algorithm	NOUN
ajst-390	24	8	based	base	VERB
ajst-390	24	9	on	on	ADP
ajst-390	24	10	wavelet	wavelet	NOUN
ajst-390	24	11	transform	transform	NOUN
ajst-390	24	12	and	and	CCONJ
ajst-390	24	13	artificial	artificial	ADJ
ajst-390	24	14	neural	neural	ADJ
ajst-390	24	15	network	network	NOUN
ajst-390	24	16	model	model	NOUN
ajst-390	24	17	.	.	PUNCT
ajst-390	25	1	in	in	ADP
ajst-390	25	2	2014	2014	NUM
ajst-390	25	3	,	,	PUNCT
ajst-390	25	4	nirmal	nirmal	ADJ
ajst-390	25	5	[	[	X
ajst-390	25	6	11	11	NUM
ajst-390	25	7	]	]	PUNCT
ajst-390	25	8	et	et	PROPN
ajst-390	25	9	al	al	PROPN
ajst-390	25	10	.	.	PROPN
ajst-390	25	11	took	take	VERB
ajst-390	25	12	advantage	advantage	NOUN
ajst-390	25	13	of	of	ADP
ajst-390	25	14	the	the	DET
ajst-390	25	15	generalized	generalize	VERB
ajst-390	25	16	regression	regression	NOUN
ajst-390	25	17	neural	neural	ADJ
ajst-390	25	18	network	network	NOUN
ajst-390	25	19	(	(	PUNCT
ajst-390	25	20	grnn	grnn	PROPN
ajst-390	25	21	)	)	PUNCT
ajst-390	25	22	to	to	PART
ajst-390	25	23	transform	transform	VERB
ajst-390	25	24	the	the	DET
ajst-390	25	25	style	style	NOUN
ajst-390	25	26	characteristic	characteristic	ADJ
ajst-390	25	27	parameters	parameter	NOUN
ajst-390	25	28	for	for	ADP
ajst-390	25	29	speech	speech	NOUN
ajst-390	25	30	signals	signal	NOUN
ajst-390	25	31	.	.	PUNCT
ajst-390	26	1	in	in	ADP
ajst-390	26	2	2015	2015	NUM
ajst-390	26	3	,	,	PUNCT
ajst-390	26	4	ghorbandoost	ghorbandoost	ADJ
ajst-390	26	5	[	[	X
ajst-390	26	6	12	12	NUM
ajst-390	26	7	]	]	PUNCT
ajst-390	26	8	et	et	PROPN
ajst-390	26	9	al	al	PROPN
ajst-390	26	10	.	.	PROPN
ajst-390	26	11	combined	combine	VERB
ajst-390	26	12	the	the	DET
ajst-390	26	13	personality	personality	NOUN
ajst-390	26	14	characteristics	characteristic	NOUN
ajst-390	26	15	of	of	ADP
ajst-390	26	16	two	two	NUM
ajst-390	26	17	kinds	kind	NOUN
ajst-390	26	18	of	of	ADP
ajst-390	26	19	speeches	speech	NOUN
ajst-390	26	20	into	into	ADP
ajst-390	26	21	a	a	DET
ajst-390	26	22	new	new	ADJ
ajst-390	26	23	speech	speech	NOUN
ajst-390	26	24	personality	personality	NOUN
ajst-390	26	25	characteristic	characteristic	NOUN
ajst-390	26	26	,	,	PUNCT
ajst-390	26	27	and	and	CCONJ
ajst-390	26	28	realized	realize	VERB
ajst-390	26	29	the	the	DET
ajst-390	26	30	speech	speech	NOUN
ajst-390	26	31	style	style	NOUN
ajst-390	26	32	transformation	transformation	NOUN
ajst-390	26	33	through	through	ADP
ajst-390	26	34	the	the	DET
ajst-390	26	35	classical	classical	ADJ
ajst-390	26	36	gaussian	gaussian	ADJ
ajst-390	26	37	mixture	mixture	NOUN
ajst-390	26	38	model	model	NOUN
ajst-390	26	39	and	and	CCONJ
ajst-390	26	40	artificial	artificial	ADJ
ajst-390	26	41	neural	neural	ADJ
ajst-390	26	42	network	network	NOUN
ajst-390	26	43	model	model	NOUN
ajst-390	26	44	.	.	PUNCT
ajst-390	27	1	the	the	DET
ajst-390	27	2	above	above	ADJ
ajst-390	27	3	references	reference	NOUN
ajst-390	27	4	show	show	VERB
ajst-390	27	5	that	that	SCONJ
ajst-390	27	6	the	the	DET
ajst-390	27	7	performance	performance	NOUN
ajst-390	27	8	and	and	CCONJ
ajst-390	27	9	stability	stability	NOUN
ajst-390	27	10	of	of	ADP
ajst-390	27	11	generated	generate	VERB
ajst-390	27	12	speech	speech	NOUN
ajst-390	27	13	have	have	AUX
ajst-390	27	14	immensely	immensely	ADV
ajst-390	27	15	improved	improve	VERB
ajst-390	27	16	based	base	VERB
ajst-390	27	17	on	on	ADP
ajst-390	27	18	neural	neural	ADJ
ajst-390	27	19	network	network	NOUN
ajst-390	27	20	in	in	ADP
ajst-390	27	21	the	the	DET
ajst-390	27	22	research	research	NOUN
ajst-390	27	23	of	of	ADP
ajst-390	27	24	speech	speech	NOUN
ajst-390	27	25	style	style	NOUN
ajst-390	27	26	transfer	transfer	NOUN
ajst-390	27	27	.	.	PUNCT
ajst-390	28	1	however	however	ADV
ajst-390	28	2	,	,	PUNCT
ajst-390	28	3	in	in	ADP
ajst-390	28	4	the	the	DET
ajst-390	28	5	training	training	NOUN
ajst-390	28	6	stage	stage	NOUN
ajst-390	28	7	of	of	ADP
ajst-390	28	8	neural	neural	ADJ
ajst-390	28	9	network	network	NOUN
ajst-390	28	10	,	,	PUNCT
ajst-390	28	11	the	the	DET
ajst-390	28	12	difficulty	difficulty	NOUN
ajst-390	28	13	in	in	ADP
ajst-390	28	14	acquiring	acquire	VERB
ajst-390	28	15	or	or	CCONJ
ajst-390	28	16	producing	produce	VERB
ajst-390	28	17	training	training	NOUN
ajst-390	28	18	data	datum	NOUN
ajst-390	28	19	hinders	hinder	VERB
ajst-390	28	20	the	the	DET
ajst-390	28	21	study	study	NOUN
ajst-390	28	22	of	of	ADP
ajst-390	28	23	speech	speech	NOUN
ajst-390	28	24	style	style	NOUN
ajst-390	28	25	transfer	transfer	NOUN
ajst-390	28	26	.	.	PUNCT
ajst-390	29	1	on	on	ADP
ajst-390	29	2	the	the	DET
ajst-390	29	3	other	other	ADJ
ajst-390	29	4	hand	hand	NOUN
ajst-390	29	5	,	,	PUNCT
ajst-390	29	6	slow	slow	ADJ
ajst-390	29	7	training	training	NOUN
ajst-390	29	8	speed	speed	NOUN
ajst-390	29	9	can	can	AUX
ajst-390	29	10	tremendously	tremendously	ADV
ajst-390	29	11	enlarge	enlarge	VERB
ajst-390	29	12	research	research	NOUN
ajst-390	29	13	difficulty	difficulty	NOUN
ajst-390	29	14	even	even	ADV
ajst-390	29	15	with	with	ADP
ajst-390	29	16	a	a	DET
ajst-390	29	17	wealth	wealth	NOUN
ajst-390	29	18	of	of	ADP
ajst-390	29	19	data	datum	NOUN
ajst-390	29	20	.	.	PUNCT
ajst-390	30	1	therefore	therefore	ADV
ajst-390	30	2	,	,	PUNCT
ajst-390	30	3	how	how	SCONJ
ajst-390	30	4	to	to	PART
ajst-390	30	5	use	use	VERB
ajst-390	30	6	less	less	ADJ
ajst-390	30	7	training	training	NOUN
ajst-390	30	8	data	datum	NOUN
ajst-390	30	9	or	or	CCONJ
ajst-390	30	10	even	even	ADV
ajst-390	30	11	no	no	DET
ajst-390	30	12	other	other	ADJ
ajst-390	30	13	data	datum	NOUN
ajst-390	30	14	to	to	PART
ajst-390	30	15	study	study	VERB
ajst-390	30	16	the	the	DET
ajst-390	30	17	speech	speech	NOUN
ajst-390	30	18	style	style	NOUN
ajst-390	30	19	transfer	transfer	NOUN
ajst-390	30	20	model	model	NOUN
ajst-390	30	21	is	be	AUX
ajst-390	30	22	the	the	DET
ajst-390	30	23	research	research	NOUN
ajst-390	30	24	innovation	innovation	NOUN
ajst-390	30	25	of	of	ADP
ajst-390	30	26	this	this	DET
ajst-390	30	27	article	article	NOUN
ajst-390	30	28	.	.	PUNCT
ajst-390	31	1	inspired	inspire	VERB
ajst-390	31	2	by	by	ADP
ajst-390	31	3	the	the	DET
ajst-390	31	4	study	study	NOUN
ajst-390	31	5	of	of	ADP
ajst-390	31	6	image	image	NOUN
ajst-390	31	7	style	style	NOUN
ajst-390	31	8	transfer	transfer	NOUN
ajst-390	31	9	,	,	PUNCT
ajst-390	31	10	the	the	DET
ajst-390	31	11	paper	paper	NOUN
ajst-390	31	12	utilizes	utilize	VERB
ajst-390	31	13	the	the	DET
ajst-390	31	14	convolutional	convolutional	ADJ
ajst-390	31	15	neural	neural	ADJ
ajst-390	31	16	network	network	NOUN
ajst-390	31	17	to	to	PART
ajst-390	31	18	extract	extract	VERB
ajst-390	31	19	the	the	DET
ajst-390	31	20	features	feature	NOUN
ajst-390	31	21	of	of	ADP
ajst-390	31	22	the	the	DET
ajst-390	31	23	spectrograms	spectrogram	NOUN
ajst-390	31	24	for	for	ADP
ajst-390	31	25	speech	speech	NOUN
ajst-390	31	26	signals	signal	NOUN
ajst-390	31	27	,	,	PUNCT
ajst-390	31	28	so	so	SCONJ
ajst-390	31	29	as	as	SCONJ
ajst-390	31	30	to	to	PART
ajst-390	31	31	generate	generate	VERB
ajst-390	31	32	the	the	DET
ajst-390	31	33	stylized	stylize	VERB
ajst-390	31	34	spectrogram	spectrogram	NOUN
ajst-390	31	35	and	and	CCONJ
ajst-390	31	36	acquire	acquire	VERB
ajst-390	31	37	corresponding	corresponding	ADJ
ajst-390	31	38	stylized	stylize	VERB
ajst-390	31	39	speech	speech	NOUN
ajst-390	31	40	.	.	PUNCT
ajst-390	32	1	2	2	X
ajst-390	32	2	.	.	X
ajst-390	32	3	the	the	DET
ajst-390	32	4	spectrogram	spectrogram	NOUN
ajst-390	32	5	of	of	ADP
ajst-390	32	6	speech	speech	NOUN
ajst-390	32	7	signal	signal	NOUN
ajst-390	32	8	this	this	DET
ajst-390	32	9	section	section	NOUN
ajst-390	32	10	will	will	AUX
ajst-390	32	11	introduce	introduce	VERB
ajst-390	32	12	some	some	DET
ajst-390	32	13	basic	basic	ADJ
ajst-390	32	14	knowledge	knowledge	NOUN
ajst-390	32	15	of	of	ADP
ajst-390	32	16	speech	speech	NOUN
ajst-390	32	17	signal	signal	NOUN
ajst-390	32	18	and	and	CCONJ
ajst-390	32	19	focus	focus	VERB
ajst-390	32	20	on	on	ADP
ajst-390	32	21	the	the	DET
ajst-390	32	22	extraction	extraction	NOUN
ajst-390	32	23	process	process	NOUN
ajst-390	32	24	in	in	ADP
ajst-390	32	25	the	the	DET
ajst-390	32	26	2d	2d	NUM
ajst-390	32	27	spectrogram	spectrogram	NOUN
ajst-390	32	28	of	of	ADP
ajst-390	32	29	speech	speech	NOUN
ajst-390	32	30	signal	signal	NOUN
ajst-390	32	31	.	.	PUNCT
ajst-390	33	1	2.1	2.1	NUM
ajst-390	33	2	.	.	PUNCT
ajst-390	34	1	the	the	DET
ajst-390	34	2	2d	2d	PROPN
ajst-390	34	3	spectrogram	spectrogram	NOUN
ajst-390	34	4	of	of	ADP
ajst-390	34	5	speech	speech	NOUN
ajst-390	34	6	signal	signal	NOUN
ajst-390	34	7	before	before	ADP
ajst-390	34	8	the	the	DET
ajst-390	34	9	experiments	experiment	NOUN
ajst-390	34	10	of	of	ADP
ajst-390	34	11	speech	speech	NOUN
ajst-390	34	12	style	style	NOUN
ajst-390	34	13	transfer	transfer	NOUN
ajst-390	34	14	,	,	PUNCT
ajst-390	34	15	the	the	DET
ajst-390	34	16	feature	feature	NOUN
ajst-390	34	17	extraction	extraction	NOUN
ajst-390	34	18	in	in	ADP
ajst-390	34	19	the	the	DET
ajst-390	34	20	2d	2d	NUM
ajst-390	34	21	spectrogram	spectrogram	NOUN
ajst-390	34	22	of	of	ADP
ajst-390	34	23	speech	speech	NOUN
ajst-390	34	24	signal	signal	NOUN
ajst-390	34	25	is	be	AUX
ajst-390	34	26	usually	usually	ADV
ajst-390	34	27	required	require	VERB
ajst-390	34	28	.	.	PUNCT
ajst-390	35	1	that	that	PRON
ajst-390	35	2	is	be	AUX
ajst-390	35	3	to	to	PART
ajst-390	35	4	say	say	VERB
ajst-390	35	5	,	,	PUNCT
ajst-390	35	6	the	the	DET
ajst-390	35	7	discernable	discernable	ADJ
ajst-390	35	8	information	information	NOUN
ajst-390	35	9	contained	contain	VERB
ajst-390	35	10	in	in	ADP
ajst-390	35	11	speech	speech	NOUN
ajst-390	35	12	signals	signal	NOUN
ajst-390	35	13	can	can	AUX
ajst-390	35	14	be	be	AUX
ajst-390	35	15	achieved	achieve	VERB
ajst-390	35	16	through	through	ADP
ajst-390	35	17	the	the	DET
ajst-390	35	18	extracted	extract	VERB
ajst-390	35	19	feature	feature	NOUN
ajst-390	35	20	information	information	NOUN
ajst-390	35	21	.	.	PUNCT
ajst-390	36	1	moreover	moreover	ADV
ajst-390	36	2	,	,	PUNCT
ajst-390	36	3	speech	speech	NOUN
ajst-390	36	4	signals	signal	NOUN
ajst-390	36	5	are	be	AUX
ajst-390	36	6	generated	generate	VERB
ajst-390	36	7	through	through	ADP
ajst-390	36	8	the	the	DET
ajst-390	36	9	vocal	vocal	ADJ
ajst-390	36	10	tract	tract	NOUN
ajst-390	36	11	,	,	PUNCT
ajst-390	36	12	the	the	DET
ajst-390	36	13	shape	shape	NOUN
ajst-390	36	14	of	of	ADP
ajst-390	36	15	the	the	DET
ajst-390	36	16	vocal	vocal	ADJ
ajst-390	36	17	tract	tract	NOUN
ajst-390	36	18	determines	determine	VERB
ajst-390	36	19	what	what	DET
ajst-390	36	20	kind	kind	NOUN
ajst-390	36	21	of	of	ADP
ajst-390	36	22	speech	speech	NOUN
ajst-390	36	23	is	be	AUX
ajst-390	36	24	phonated	phonate	VERB
ajst-390	36	25	,	,	PUNCT
ajst-390	36	26	to	to	ADP
ajst-390	36	27	some	some	DET
ajst-390	36	28	extent	extent	NOUN
ajst-390	36	29	.	.	PUNCT
ajst-390	37	1	the	the	DET
ajst-390	37	2	shape	shape	NOUN
ajst-390	37	3	of	of	ADP
ajst-390	37	4	the	the	DET
ajst-390	37	5	vocal	vocal	ADJ
ajst-390	37	6	tract	tract	NOUN
ajst-390	37	7	can	can	AUX
ajst-390	37	8	be	be	AUX
ajst-390	37	9	shown	show	VERB
ajst-390	37	10	in	in	ADP
ajst-390	37	11	the	the	DET
ajst-390	37	12	envelope	envelope	NOUN
ajst-390	37	13	of	of	ADP
ajst-390	37	14	speech	speech	NOUN
ajst-390	37	15	short	short	ADJ
ajst-390	37	16	time	time	NOUN
ajst-390	37	17	power	power	NOUN
ajst-390	37	18	spectrum	spectrum	NOUN
ajst-390	37	19	,	,	PUNCT
ajst-390	37	20	and	and	CCONJ
ajst-390	37	21	the	the	DET
ajst-390	37	22	characteristic	characteristic	ADJ
ajst-390	37	23	information	information	NOUN
ajst-390	37	24	of	of	ADP
ajst-390	37	25	speech	speech	NOUN
ajst-390	37	26	signals	signal	NOUN
ajst-390	37	27	2d	2d	NOUN
ajst-390	37	28	spectrogram	spectrogram	NOUN
ajst-390	37	29	can	can	AUX
ajst-390	37	30	describe	describe	VERB
ajst-390	37	31	the	the	DET
ajst-390	37	32	envelope	envelope	NOUN
ajst-390	37	33	exactly	exactly	ADV
ajst-390	37	34	.	.	PUNCT
ajst-390	38	1	2	2	NUM
ajst-390	38	2	figure	figure	NOUN
ajst-390	38	3	1	1	NUM
ajst-390	38	4	.	.	PUNCT
ajst-390	39	1	the	the	DET
ajst-390	39	2	speech	speech	NOUN
ajst-390	39	3	frames	frame	VERB
ajst-390	39	4	first	first	ADV
ajst-390	39	5	,	,	PUNCT
ajst-390	39	6	the	the	DET
ajst-390	39	7	speech	speech	NOUN
ajst-390	39	8	signal	signal	NOUN
ajst-390	39	9	displayed	display	VERB
ajst-390	39	10	in	in	ADP
ajst-390	39	11	figure	figure	NOUN
ajst-390	39	12	1	1	NUM
ajst-390	39	13	is	be	AUX
ajst-390	39	14	divided	divide	VERB
ajst-390	39	15	into	into	ADP
ajst-390	39	16	14	14	NUM
ajst-390	39	17	frames	frame	NOUN
ajst-390	39	18	.	.	PUNCT
ajst-390	40	1	then	then	ADV
ajst-390	40	2	,	,	PUNCT
ajst-390	40	3	the	the	DET
ajst-390	40	4	spectrums	spectrum	NOUN
ajst-390	40	5	corresponding	correspond	VERB
ajst-390	40	6	to	to	ADP
ajst-390	40	7	those	those	DET
ajst-390	40	8	14	14	NUM
ajst-390	40	9	frames	frame	NOUN
ajst-390	40	10	of	of	ADP
ajst-390	40	11	speech	speech	NOUN
ajst-390	40	12	can	can	AUX
ajst-390	40	13	be	be	AUX
ajst-390	40	14	obtained	obtain	VERB
ajst-390	40	15	by	by	ADP
ajst-390	40	16	calculating	calculate	VERB
ajst-390	40	17	short	short	ADJ
ajst-390	40	18	fast	fast	ADJ
ajst-390	40	19	fourier	fourier	NOUN
ajst-390	40	20	transformation	transformation	NOUN
ajst-390	40	21	.	.	PUNCT
ajst-390	41	1	the	the	DET
ajst-390	41	2	spectrum	spectrum	NOUN
ajst-390	41	3	represents	represent	VERB
ajst-390	41	4	the	the	DET
ajst-390	41	5	relationship	relationship	NOUN
ajst-390	41	6	between	between	ADP
ajst-390	41	7	frequency	frequency	NOUN
ajst-390	41	8	and	and	CCONJ
ajst-390	41	9	energy	energy	NOUN
ajst-390	41	10	amplitude	amplitude	NOUN
ajst-390	41	11	,	,	PUNCT
ajst-390	41	12	and	and	CCONJ
ajst-390	41	13	the	the	DET
ajst-390	41	14	specific	specific	ADJ
ajst-390	41	15	information	information	NOUN
ajst-390	41	16	is	be	AUX
ajst-390	41	17	plotted	plot	VERB
ajst-390	41	18	on	on	ADP
ajst-390	41	19	the	the	DET
ajst-390	41	20	left	left	ADJ
ajst-390	41	21	graph	graph	NOUN
ajst-390	41	22	in	in	ADP
ajst-390	41	23	figure	figure	NOUN
ajst-390	41	24	2	2	NUM
ajst-390	41	25	.	.	PUNCT
ajst-390	41	26	figure	figure	NOUN
ajst-390	41	27	2	2	NUM
ajst-390	41	28	.	.	PUNCT
ajst-390	42	1	the	the	DET
ajst-390	42	2	spectrum	spectrum	NOUN
ajst-390	42	3	next	next	ADV
ajst-390	42	4	,	,	PUNCT
ajst-390	42	5	rotating	rotate	VERB
ajst-390	42	6	the	the	DET
ajst-390	42	7	spectrum	spectrum	NOUN
ajst-390	42	8	curve	curve	NOUN
ajst-390	42	9	by	by	ADP
ajst-390	42	10	90	90	NUM
ajst-390	42	11	degrees	degree	NOUN
ajst-390	42	12	to	to	PART
ajst-390	42	13	get	get	VERB
ajst-390	42	14	the	the	DET
ajst-390	42	15	middle	middle	ADJ
ajst-390	42	16	graph	graph	NOUN
ajst-390	42	17	in	in	ADP
ajst-390	42	18	figure	figure	NOUN
ajst-390	42	19	2	2	NUM
ajst-390	42	20	.	.	PUNCT
ajst-390	43	1	furthermore	furthermore	ADV
ajst-390	43	2	,	,	PUNCT
ajst-390	43	3	the	the	DET
ajst-390	43	4	amplitudes	amplitude	NOUN
ajst-390	43	5	in	in	ADP
ajst-390	43	6	the	the	DET
ajst-390	43	7	middle	middle	ADJ
ajst-390	43	8	graph	graph	NOUN
ajst-390	43	9	are	be	AUX
ajst-390	43	10	mapped	map	VERB
ajst-390	43	11	to	to	ADP
ajst-390	43	12	a	a	DET
ajst-390	43	13	range	range	NOUN
ajst-390	43	14	of	of	ADP
ajst-390	43	15	gray	gray	ADJ
ajst-390	43	16	level	level	NOUN
ajst-390	43	17	,	,	PUNCT
ajst-390	43	18	with	with	ADP
ajst-390	43	19	gray	gray	ADJ
ajst-390	43	20	level	level	NOUN
ajst-390	43	21	255	255	NUM
ajst-390	43	22	and	and	CCONJ
ajst-390	43	23	gray	gray	ADJ
ajst-390	43	24	level	level	NOUN
ajst-390	43	25	0	0	NUM
ajst-390	43	26	represented	represent	VERB
ajst-390	43	27	by	by	ADP
ajst-390	43	28	the	the	DET
ajst-390	43	29	black	black	ADJ
ajst-390	43	30	area	area	NOUN
ajst-390	43	31	and	and	CCONJ
ajst-390	43	32	the	the	DET
ajst-390	43	33	white	white	PROPN
ajst-390	43	34	area	area	NOUN
ajst-390	43	35	,	,	PUNCT
ajst-390	43	36	respectively	respectively	ADV
ajst-390	43	37	.	.	PUNCT
ajst-390	44	1	in	in	ADP
ajst-390	44	2	other	other	ADJ
ajst-390	44	3	words	word	NOUN
ajst-390	44	4	,	,	PUNCT
ajst-390	44	5	the	the	PRON
ajst-390	44	6	larger	large	ADJ
ajst-390	44	7	the	the	DET
ajst-390	44	8	amplitude	amplitude	NOUN
ajst-390	44	9	value	value	NOUN
ajst-390	44	10	,	,	PUNCT
ajst-390	44	11	the	the	DET
ajst-390	44	12	darker	dark	ADJ
ajst-390	44	13	the	the	DET
ajst-390	44	14	corresponding	corresponding	ADJ
ajst-390	44	15	area	area	NOUN
ajst-390	44	16	is	be	AUX
ajst-390	44	17	.	.	PUNCT
ajst-390	45	1	thus	thus	ADV
ajst-390	45	2	,	,	PUNCT
ajst-390	45	3	the	the	DET
ajst-390	45	4	right	right	ADV
ajst-390	45	5	-	-	PUNCT
ajst-390	45	6	most	most	ADJ
ajst-390	45	7	graph	graph	NOUN
ajst-390	45	8	is	be	AUX
ajst-390	45	9	acquired	acquire	VERB
ajst-390	45	10	in	in	ADP
ajst-390	45	11	figure	figure	NOUN
ajst-390	45	12	2	2	NUM
ajst-390	45	13	.	.	PUNCT
ajst-390	46	1	the	the	DET
ajst-390	46	2	aim	aim	NOUN
ajst-390	46	3	above	above	ADV
ajst-390	46	4	is	be	AUX
ajst-390	46	5	to	to	PART
ajst-390	46	6	add	add	VERB
ajst-390	46	7	the	the	DET
ajst-390	46	8	time	time	NOUN
ajst-390	46	9	dimension	dimension	NOUN
ajst-390	46	10	,	,	PUNCT
ajst-390	46	11	so	so	SCONJ
ajst-390	46	12	that	that	SCONJ
ajst-390	46	13	the	the	DET
ajst-390	46	14	spectrum	spectrum	NOUN
ajst-390	46	15	of	of	ADP
ajst-390	46	16	a	a	DET
ajst-390	46	17	speech	speech	NOUN
ajst-390	46	18	,	,	PUNCT
ajst-390	46	19	rather	rather	ADV
ajst-390	46	20	than	than	ADP
ajst-390	46	21	a	a	DET
ajst-390	46	22	frame	frame	NOUN
ajst-390	46	23	,	,	PUNCT
ajst-390	46	24	can	can	AUX
ajst-390	46	25	be	be	AUX
ajst-390	46	26	manifested	manifest	VERB
ajst-390	46	27	integrally	integrally	ADV
ajst-390	46	28	.	.	PUNCT
ajst-390	47	1	finally	finally	ADV
ajst-390	47	2	,	,	PUNCT
ajst-390	47	3	we	we	PRON
ajst-390	47	4	obtain	obtain	VERB
ajst-390	47	5	a	a	DET
ajst-390	47	6	spectrum	spectrum	NOUN
ajst-390	47	7	diagram	diagram	NOUN
ajst-390	47	8	over	over	ADP
ajst-390	47	9	time	time	NOUN
ajst-390	47	10	,	,	PUNCT
ajst-390	47	11	which	which	PRON
ajst-390	47	12	is	be	AUX
ajst-390	47	13	the	the	DET
ajst-390	47	14	2d	2d	NUM
ajst-390	47	15	spectrogram	spectrogram	NOUN
ajst-390	47	16	describing	describe	VERB
ajst-390	47	17	the	the	DET
ajst-390	47	18	speech	speech	NOUN
ajst-390	47	19	signal	signal	NOUN
ajst-390	47	20	,	,	PUNCT
ajst-390	47	21	as	as	SCONJ
ajst-390	47	22	shown	show	VERB
ajst-390	47	23	in	in	ADP
ajst-390	47	24	figure	figure	NOUN
ajst-390	47	25	3	3	NUM
ajst-390	47	26	.	.	PUNCT
ajst-390	47	27	figure	figure	NOUN
ajst-390	47	28	3	3	NUM
ajst-390	47	29	.	.	PUNCT
ajst-390	48	1	the	the	DET
ajst-390	48	2	2d	2d	PROPN
ajst-390	48	3	spectrogram	spectrogram	NOUN
ajst-390	48	4	3	3	NUM
ajst-390	48	5	3	3	NUM
ajst-390	48	6	.	.	PUNCT
ajst-390	48	7	speech	speech	NOUN
ajst-390	48	8	style	style	NOUN
ajst-390	48	9	transfer	transfer	NOUN
ajst-390	48	10	model	model	NOUN
ajst-390	48	11	based	base	VERB
ajst-390	48	12	on	on	ADP
ajst-390	48	13	neural	neural	ADJ
ajst-390	48	14	network	network	NOUN
ajst-390	48	15	this	this	DET
ajst-390	48	16	paper	paper	NOUN
ajst-390	48	17	proposes	propose	VERB
ajst-390	48	18	an	an	DET
ajst-390	48	19	innovative	innovative	ADJ
ajst-390	48	20	solution	solution	NOUN
ajst-390	48	21	the	the	DET
ajst-390	48	22	speech	speech	NOUN
ajst-390	48	23	style	style	NOUN
ajst-390	48	24	transfer	transfer	NOUN
ajst-390	48	25	model	model	NOUN
ajst-390	48	26	based	base	VERB
ajst-390	48	27	on	on	ADP
ajst-390	48	28	convolutional	convolutional	ADJ
ajst-390	48	29	neural	neural	ADJ
ajst-390	48	30	network	network	NOUN
ajst-390	48	31	.	.	PUNCT
ajst-390	49	1	3.1	3.1	NUM
ajst-390	49	2	.	.	PUNCT
ajst-390	50	1	the	the	DET
ajst-390	50	2	speech	speech	NOUN
ajst-390	50	3	style	style	NOUN
ajst-390	50	4	transfer	transfer	NOUN
ajst-390	50	5	model	model	NOUN
ajst-390	50	6	based	base	VERB
ajst-390	50	7	on	on	ADP
ajst-390	50	8	convolutional	convolutional	ADJ
ajst-390	50	9	neural	neural	ADJ
ajst-390	50	10	network	network	NOUN
ajst-390	50	11	the	the	DET
ajst-390	50	12	principle	principle	NOUN
ajst-390	50	13	of	of	ADP
ajst-390	50	14	image	image	NOUN
ajst-390	50	15	style	style	NOUN
ajst-390	50	16	transfer	transfer	NOUN
ajst-390	50	17	model	model	NOUN
ajst-390	50	18	can	can	AUX
ajst-390	50	19	be	be	AUX
ajst-390	50	20	demonstrated	demonstrate	VERB
ajst-390	50	21	through	through	ADP
ajst-390	50	22	the	the	DET
ajst-390	50	23	following	follow	VERB
ajst-390	50	24	flow	flow	NOUN
ajst-390	50	25	figure	figure	NOUN
ajst-390	50	26	4	4	NUM
ajst-390	50	27	,	,	PUNCT
ajst-390	50	28	roughly	roughly	ADV
ajst-390	50	29	.	.	PUNCT
ajst-390	51	1	figure	figure	VERB
ajst-390	51	2	4	4	NUM
ajst-390	51	3	.	.	PUNCT
ajst-390	52	1	the	the	DET
ajst-390	52	2	image	image	NOUN
ajst-390	52	3	style	style	NOUN
ajst-390	52	4	transfer	transfer	NOUN
ajst-390	52	5	model	model	NOUN
ajst-390	52	6	since	since	SCONJ
ajst-390	52	7	the	the	DET
ajst-390	52	8	convolutional	convolutional	ADJ
ajst-390	52	9	neural	neural	ADJ
ajst-390	52	10	network	network	NOUN
ajst-390	52	11	is	be	AUX
ajst-390	52	12	good	good	ADJ
ajst-390	52	13	at	at	ADP
ajst-390	52	14	dealing	deal	VERB
ajst-390	52	15	with	with	ADP
ajst-390	52	16	picture	picture	NOUN
ajst-390	52	17	type	type	NOUN
ajst-390	52	18	data	datum	NOUN
ajst-390	52	19	,	,	PUNCT
ajst-390	52	20	processing	processing	NOUN
ajst-390	52	21	speech	speech	NOUN
ajst-390	52	22	signal	signal	NOUN
ajst-390	52	23	into	into	ADP
ajst-390	52	24	the	the	DET
ajst-390	52	25	picture	picture	NOUN
ajst-390	52	26	type	type	NOUN
ajst-390	52	27	data	datum	NOUN
ajst-390	52	28	is	be	AUX
ajst-390	52	29	key	key	ADJ
ajst-390	52	30	point	point	NOUN
ajst-390	52	31	and	and	CCONJ
ajst-390	52	32	innovation	innovation	NOUN
ajst-390	52	33	point	point	NOUN
ajst-390	52	34	of	of	ADP
ajst-390	52	35	speech	speech	NOUN
ajst-390	52	36	style	style	NOUN
ajst-390	52	37	transfer	transfer	NOUN
ajst-390	52	38	.	.	PUNCT
ajst-390	53	1	according	accord	VERB
ajst-390	53	2	to	to	ADP
ajst-390	53	3	the	the	DET
ajst-390	53	4	introduction	introduction	NOUN
ajst-390	53	5	of	of	ADP
ajst-390	53	6	2d	2d	PROPN
ajst-390	53	7	spectrogram	spectrogram	NOUN
ajst-390	53	8	in	in	ADP
ajst-390	53	9	section	section	NOUN
ajst-390	53	10	2.1	2.1	NUM
ajst-390	53	11	,	,	PUNCT
ajst-390	53	12	it	it	PRON
ajst-390	53	13	can	can	AUX
ajst-390	53	14	be	be	AUX
ajst-390	53	15	known	know	VERB
ajst-390	53	16	that	that	SCONJ
ajst-390	53	17	spectrogram	spectrogram	NOUN
ajst-390	53	18	can	can	AUX
ajst-390	53	19	be	be	AUX
ajst-390	53	20	regarded	regard	VERB
ajst-390	53	21	as	as	ADP
ajst-390	53	22	2	2	NUM
ajst-390	53	23	-	-	PUNCT
ajst-390	53	24	dimensional	dimensional	ADJ
ajst-390	53	25	picture	picture	NOUN
ajst-390	53	26	type	type	NOUN
ajst-390	53	27	data	datum	NOUN
ajst-390	53	28	,	,	PUNCT
ajst-390	53	29	to	to	ADP
ajst-390	53	30	some	some	DET
ajst-390	53	31	extent	extent	NOUN
ajst-390	53	32	.	.	PUNCT
ajst-390	54	1	therefore	therefore	ADV
ajst-390	54	2	,	,	PUNCT
ajst-390	54	3	when	when	SCONJ
ajst-390	54	4	the	the	DET
ajst-390	54	5	research	research	NOUN
ajst-390	54	6	object	object	NOUN
ajst-390	54	7	is	be	AUX
ajst-390	54	8	speech	speech	NOUN
ajst-390	54	9	signal	signal	NOUN
ajst-390	54	10	,	,	PUNCT
ajst-390	54	11	content	content	NOUN
ajst-390	54	12	image	image	NOUN
ajst-390	54	13	,	,	PUNCT
ajst-390	54	14	style	style	NOUN
ajst-390	54	15	image	image	NOUN
ajst-390	54	16	and	and	CCONJ
ajst-390	54	17	generated	generate	VERB
ajst-390	54	18	image	image	NOUN
ajst-390	54	19	in	in	ADP
ajst-390	54	20	figure	figure	NOUN
ajst-390	54	21	4	4	NUM
ajst-390	54	22	are	be	AUX
ajst-390	54	23	replaced	replace	VERB
ajst-390	54	24	by	by	ADP
ajst-390	54	25	the	the	DET
ajst-390	54	26	2d	2d	PROPN
ajst-390	54	27	spectrograms	spectrogram	NOUN
ajst-390	54	28	of	of	ADP
ajst-390	54	29	content	content	NOUN
ajst-390	54	30	speech	speech	NOUN
ajst-390	54	31	,	,	PUNCT
ajst-390	54	32	style	style	NOUN
ajst-390	54	33	speech	speech	NOUN
ajst-390	54	34	and	and	CCONJ
ajst-390	54	35	generated	generate	VERB
ajst-390	54	36	speech	speech	NOUN
ajst-390	54	37	,	,	PUNCT
ajst-390	54	38	respectively	respectively	ADV
ajst-390	54	39	.	.	PUNCT
ajst-390	55	1	theoretically	theoretically	ADV
ajst-390	55	2	,	,	PUNCT
ajst-390	55	3	we	we	PRON
ajst-390	55	4	can	can	AUX
ajst-390	55	5	achieve	achieve	VERB
ajst-390	55	6	the	the	DET
ajst-390	55	7	speech	speech	NOUN
ajst-390	55	8	style	style	NOUN
ajst-390	55	9	transfer	transfer	NOUN
ajst-390	55	10	model	model	NOUN
ajst-390	55	11	.	.	PUNCT
ajst-390	56	1	the	the	DET
ajst-390	56	2	speech	speech	NOUN
ajst-390	56	3	style	style	NOUN
ajst-390	56	4	transfer	transfer	NOUN
ajst-390	56	5	model	model	NOUN
ajst-390	56	6	is	be	AUX
ajst-390	56	7	shown	show	VERB
ajst-390	56	8	in	in	ADP
ajst-390	56	9	figure	figure	NOUN
ajst-390	56	10	5	5	NUM
ajst-390	56	11	.	.	PUNCT
ajst-390	56	12	figure	figure	NOUN
ajst-390	56	13	5	5	NUM
ajst-390	56	14	.	.	PUNCT
ajst-390	57	1	the	the	DET
ajst-390	57	2	speech	speech	NOUN
ajst-390	57	3	style	style	NOUN
ajst-390	57	4	transfer	transfer	NOUN
ajst-390	57	5	model	model	NOUN
ajst-390	57	6	in	in	ADP
ajst-390	57	7	figure	figure	NOUN
ajst-390	57	8	5	5	NUM
ajst-390	57	9	,	,	PUNCT
ajst-390	57	10	the	the	DET
ajst-390	57	11	essential	essential	ADJ
ajst-390	57	12	function	function	NOUN
ajst-390	57	13	of	of	ADP
ajst-390	57	14	convolutional	convolutional	ADJ
ajst-390	57	15	neural	neural	ADJ
ajst-390	57	16	network	network	NOUN
ajst-390	57	17	is	be	AUX
ajst-390	57	18	to	to	PART
ajst-390	57	19	extract	extract	VERB
ajst-390	57	20	the	the	DET
ajst-390	57	21	feature	feature	NOUN
ajst-390	57	22	information	information	NOUN
ajst-390	57	23	of	of	ADP
ajst-390	57	24	the	the	DET
ajst-390	57	25	input	input	NOUN
ajst-390	57	26	(	(	PUNCT
ajst-390	57	27	2d	2d	NUM
ajst-390	57	28	spectrogram	spectrogram	NOUN
ajst-390	57	29	)	)	PUNCT
ajst-390	57	30	layer	layer	NOUN
ajst-390	57	31	by	by	ADP
ajst-390	57	32	layer	layer	NOUN
ajst-390	57	33	.	.	PUNCT
ajst-390	58	1	after	after	ADP
ajst-390	58	2	the	the	DET
ajst-390	58	3	layered	layered	ADJ
ajst-390	58	4	extraction	extraction	NOUN
ajst-390	58	5	of	of	ADP
ajst-390	58	6	convolutional	convolutional	ADJ
ajst-390	58	7	layer	layer	NOUN
ajst-390	58	8	,	,	PUNCT
ajst-390	58	9	pooling	pool	VERB
ajst-390	58	10	layer	layer	NOUN
ajst-390	58	11	,	,	PUNCT
ajst-390	58	12	full	full	ADJ
ajst-390	58	13	connection	connection	NOUN
ajst-390	58	14	layer	layer	NOUN
ajst-390	58	15	and	and	CCONJ
ajst-390	58	16	other	other	ADJ
ajst-390	58	17	network	network	NOUN
ajst-390	58	18	layers	layer	NOUN
ajst-390	58	19	,	,	PUNCT
ajst-390	58	20	the	the	DET
ajst-390	58	21	feature	feature	NOUN
ajst-390	58	22	information	information	NOUN
ajst-390	58	23	of	of	ADP
ajst-390	58	24	2d	2d	PROPN
ajst-390	58	25	spectrogram	spectrogram	NOUN
ajst-390	58	26	will	will	AUX
ajst-390	58	27	become	become	VERB
ajst-390	58	28	more	more	ADV
ajst-390	58	29	and	and	CCONJ
ajst-390	58	30	more	more	ADV
ajst-390	58	31	advanced	advanced	ADJ
ajst-390	58	32	and	and	CCONJ
ajst-390	58	33	abstract	abstract	ADJ
ajst-390	58	34	.	.	PUNCT
ajst-390	59	1	that	that	PRON
ajst-390	59	2	is	be	AUX
ajst-390	59	3	to	to	PART
ajst-390	59	4	say	say	VERB
ajst-390	59	5	,	,	PUNCT
ajst-390	59	6	the	the	DET
ajst-390	59	7	low	low	ADJ
ajst-390	59	8	-	-	PUNCT
ajst-390	59	9	layer	layer	NOUN
ajst-390	59	10	convolution	convolution	NOUN
ajst-390	59	11	filters	filter	NOUN
ajst-390	59	12	in	in	ADP
ajst-390	59	13	the	the	DET
ajst-390	59	14	convolutional	convolutional	ADJ
ajst-390	59	15	neural	neural	ADJ
ajst-390	59	16	network	network	NOUN
ajst-390	59	17	tend	tend	VERB
ajst-390	59	18	to	to	PART
ajst-390	59	19	extract	extract	VERB
ajst-390	59	20	content	content	NOUN
ajst-390	59	21	feature	feature	NOUN
ajst-390	59	22	information	information	NOUN
ajst-390	59	23	(	(	PUNCT
ajst-390	59	24	edge	edge	NOUN
ajst-390	59	25	,	,	PUNCT
ajst-390	59	26	texture	texture	NOUN
ajst-390	59	27	and	and	CCONJ
ajst-390	59	28	color	color	NOUN
ajst-390	59	29	etc	etc	X
ajst-390	59	30	.	.	PUNCT
ajst-390	59	31	)	)	PUNCT
ajst-390	60	1	for	for	ADP
ajst-390	60	2	2d	2d	PROPN
ajst-390	60	3	spectrogram	spectrogram	NOUN
ajst-390	60	4	,	,	PUNCT
ajst-390	60	5	the	the	DET
ajst-390	60	6	high	high	ADJ
ajst-390	60	7	-	-	PUNCT
ajst-390	60	8	level	level	NOUN
ajst-390	60	9	convolution	convolution	NOUN
ajst-390	60	10	filters	filter	NOUN
ajst-390	60	11	tend	tend	VERB
ajst-390	60	12	to	to	PART
ajst-390	60	13	extract	extract	VERB
ajst-390	60	14	style	style	NOUN
ajst-390	60	15	feature	feature	NOUN
ajst-390	60	16	information	information	NOUN
ajst-390	60	17	(	(	PUNCT
ajst-390	60	18	a	a	DET
ajst-390	60	19	rough	rough	ADJ
ajst-390	60	20	skeleton	skeleton	NOUN
ajst-390	60	21	or	or	CCONJ
ajst-390	60	22	layout	layout	NOUN
ajst-390	60	23	etc	etc	X
ajst-390	60	24	.	.	PUNCT
ajst-390	60	25	)	)	PUNCT
ajst-390	61	1	for	for	ADP
ajst-390	61	2	2d	2d	PROPN
ajst-390	61	3	spectrogram	spectrogram	NOUN
ajst-390	61	4	.	.	PUNCT
ajst-390	62	1	the	the	DET
ajst-390	62	2	2d	2d	PROPN
ajst-390	62	3	spectrograms	spectrogram	NOUN
ajst-390	62	4	of	of	ADP
ajst-390	62	5	content	content	NOUN
ajst-390	62	6	speech	speech	NOUN
ajst-390	62	7	,	,	PUNCT
ajst-390	62	8	style	style	NOUN
ajst-390	62	9	speech	speech	NOUN
ajst-390	62	10	and	and	CCONJ
ajst-390	62	11	generated	generate	VERB
ajst-390	62	12	speech	speech	NOUN
ajst-390	62	13	are	be	AUX
ajst-390	62	14	denoted	denote	VERB
ajst-390	62	15	by	by	ADP
ajst-390	62	16	vcc	vcc	PROPN
ajst-390	62	17	,	,	PUNCT
ajst-390	62	18	vs	vs	ADP
ajst-390	62	19	y	y	PROPN
ajst-390	62	20	s	s	PROPN
ajst-390	62	21	and	and	CCONJ
ajst-390	62	22	vgg	vgg	VERB
ajst-390	62	23	respectively	respectively	ADV
ajst-390	62	24	in	in	ADP
ajst-390	62	25	this	this	DET
ajst-390	62	26	paper	paper	NOUN
ajst-390	62	27	.	.	PUNCT
ajst-390	63	1	next	next	ADV
ajst-390	63	2	,	,	PUNCT
ajst-390	63	3	this	this	DET
ajst-390	63	4	paper	paper	NOUN
ajst-390	63	5	will	will	AUX
ajst-390	63	6	introduce	introduce	VERB
ajst-390	63	7	the	the	DET
ajst-390	63	8	relevant	relevant	ADJ
ajst-390	63	9	principles	principle	NOUN
ajst-390	63	10	of	of	ADP
ajst-390	63	11	the	the	DET
ajst-390	63	12	creative	creative	ADJ
ajst-390	63	13	research	research	NOUN
ajst-390	63	14	scheme	scheme	NOUN
ajst-390	63	15	the	the	DET
ajst-390	63	16	speech	speech	NOUN
ajst-390	63	17	style	style	NOUN
ajst-390	63	18	transfer	transfer	NOUN
ajst-390	63	19	model	model	NOUN
ajst-390	63	20	based	base	VERB
ajst-390	63	21	on	on	ADP
ajst-390	63	22	convolutional	convolutional	ADJ
ajst-390	63	23	neural	neural	ADJ
ajst-390	63	24	network	network	NOUN
ajst-390	63	25	.	.	PUNCT
ajst-390	64	1	figure	figure	NOUN
ajst-390	64	2	6	6	NUM
ajst-390	64	3	is	be	AUX
ajst-390	64	4	a	a	DET
ajst-390	64	5	schematic	schematic	ADJ
ajst-390	64	6	diagram	diagram	NOUN
ajst-390	64	7	of	of	ADP
ajst-390	64	8	feature	feature	NOUN
ajst-390	64	9	map	map	NOUN
ajst-390	64	10	of	of	ADP
ajst-390	64	11	the	the	DET
ajst-390	64	12	l	l	NOUN
ajst-390	64	13	th	th	X
ajst-390	64	14	layer	layer	NOUN
ajst-390	64	15	extracted	extract	VERB
ajst-390	64	16	from	from	ADP
ajst-390	64	17	the	the	DET
ajst-390	64	18	spectrogram	spectrogram	NOUN
ajst-390	64	19	by	by	ADP
ajst-390	64	20	convolution	convolution	NOUN
ajst-390	64	21	filters	filter	NOUN
ajst-390	64	22	.	.	PUNCT
ajst-390	65	1	detailedly	detailedly	ADV
ajst-390	65	2	,	,	PUNCT
ajst-390	65	3	wn	wn	INTJ
ajst-390	65	4	,	,	PUNCT
ajst-390	65	5	hn	hn	PROPN
ajst-390	65	6	and	and	CCONJ
ajst-390	65	7	cn	cn	PROPN
ajst-390	65	8	represent	represent	VERB
ajst-390	65	9	width	width	NOUN
ajst-390	65	10	,	,	PUNCT
ajst-390	65	11	height	height	NOUN
ajst-390	65	12	and	and	CCONJ
ajst-390	65	13	number	number	NOUN
ajst-390	65	14	of	of	ADP
ajst-390	65	15	channels	channel	NOUN
ajst-390	65	16	respectively	respectively	ADV
ajst-390	65	17	in	in	ADP
ajst-390	65	18	the	the	DET
ajst-390	65	19	feature	feature	NOUN
ajst-390	65	20	map	map	NOUN
ajst-390	65	21	of	of	ADP
ajst-390	65	22	the	the	DET
ajst-390	65	23	l	l	NOUN
ajst-390	65	24	th	th	X
ajst-390	65	25	layer	layer	NOUN
ajst-390	65	26	.	.	PUNCT
ajst-390	66	1	and	and	CCONJ
ajst-390	66	2	[	[	PUNCT
ajst-390	66	3	]	]	X
ajst-390	66	4	,	,	PUNCT
ajst-390	66	5	,	,	PUNCT
ajst-390	66	6	l	l	PROPN
ajst-390	66	7	i	i	PRON
ajst-390	66	8	j	j	PROPN
ajst-390	66	9	ka	ka	PROPN
ajst-390	66	10	represents	represent	VERB
ajst-390	66	11	the	the	DET
ajst-390	66	12	activation	activation	NOUN
ajst-390	66	13	value	value	NOUN
ajst-390	66	14	at	at	ADP
ajst-390	66	15	the	the	DET
ajst-390	66	16	coordinate	coordinate	NOUN
ajst-390	66	17	point	point	NOUN
ajst-390	66	18	(	(	PUNCT
ajst-390	66	19	,	,	PUNCT
ajst-390	66	20	,	,	PUNCT
ajst-390	66	21	)	)	PUNCT
ajst-390	67	1	i	i	PRON
ajst-390	67	2	j	j	PROPN
ajst-390	68	1	k	k	PROPN
ajst-390	68	2	in	in	ADP
ajst-390	68	3	the	the	DET
ajst-390	68	4	l	l	NOUN
ajst-390	68	5	th	th	X
ajst-390	68	6	layer	layer	NOUN
ajst-390	68	7	feature	feature	NOUN
ajst-390	68	8	map	map	NOUN
ajst-390	68	9	,	,	PUNCT
ajst-390	68	10	1,2	1,2	NUM
ajst-390	68	11	,	,	PUNCT
ajst-390	68	12	...	...	PUNCT
ajst-390	68	13	,	,	PUNCT
ajst-390	68	14	,	,	PUNCT
ajst-390	68	15	1,2	1,2	NUM
ajst-390	68	16	,	,	PUNCT
ajst-390	68	17	...	...	PUNCT
ajst-390	68	18	,	,	PUNCT
ajst-390	68	19	,	,	PUNCT
ajst-390	68	20	1,2,	1,2,	NUM
ajst-390	68	21	...	...	PUNCT
ajst-390	68	22	,h	,h	PUNCT
ajst-390	68	23	w	w	PROPN
ajst-390	68	24	ci	ci	PROPN
ajst-390	68	25	n	n	PRON
ajst-390	68	26	j	j	PROPN
ajst-390	68	27	n	n	CCONJ
ajst-390	68	28	k	k	PROPN
ajst-390	68	29	n	n	PROPN
ajst-390	68	30	.	.	PUNCT
ajst-390	69	1	figure	figure	VERB
ajst-390	69	2	6	6	NUM
ajst-390	69	3	.	.	PUNCT
ajst-390	70	1	the	the	DET
ajst-390	70	2	l	l	NOUN
ajst-390	70	3	th	th	X
ajst-390	70	4	layer	layer	NOUN
ajst-390	70	5	feature	feature	NOUN
ajst-390	70	6	map	map	NOUN
ajst-390	70	7	of	of	ADP
ajst-390	70	8	2d	2d	NUM
ajst-390	70	9	spectrogram	spectrogram	NOUN
ajst-390	70	10	3.1.1	3.1.1	NOUN
ajst-390	70	11	.	.	PUNCT
ajst-390	71	1	the	the	DET
ajst-390	71	2	extraction	extraction	NOUN
ajst-390	71	3	of	of	ADP
ajst-390	71	4	content	content	NOUN
ajst-390	71	5	features	feature	NOUN
ajst-390	71	6	from	from	ADP
ajst-390	71	7	the	the	DET
ajst-390	71	8	2d	2d	NUM
ajst-390	71	9	spectrogram	spectrogram	NOUN
ajst-390	71	10	of	of	ADP
ajst-390	71	11	content	content	NOUN
ajst-390	71	12	speech	speech	NOUN
ajst-390	71	13	the	the	DET
ajst-390	71	14	features	feature	NOUN
ajst-390	71	15	of	of	ADP
ajst-390	71	16	the	the	DET
ajst-390	71	17	l	l	NOUN
ajst-390	71	18	th	th	X
ajst-390	71	19	high	high	ADJ
ajst-390	71	20	layer	layer	NOUN
ajst-390	71	21	’s	’s	PART
ajst-390	71	22	feature	feature	NOUN
ajst-390	71	23	map	map	NOUN
ajst-390	71	24	extracted	extract	VERB
ajst-390	71	25	by	by	ADP
ajst-390	71	26	the	the	DET
ajst-390	71	27	convolution	convolution	NOUN
ajst-390	71	28	filters	filter	NOUN
ajst-390	71	29	are	be	AUX
ajst-390	71	30	selected	select	VERB
ajst-390	71	31	as	as	ADP
ajst-390	71	32	content	content	NOUN
ajst-390	71	33	features	feature	NOUN
ajst-390	71	34	in	in	ADP
ajst-390	71	35	the	the	DET
ajst-390	71	36	spectrogram	spectrogram	NOUN
ajst-390	71	37	of	of	ADP
ajst-390	71	38	generated	generate	VERB
ajst-390	71	39	speech	speech	NOUN
ajst-390	71	40	.	.	PUNCT
ajst-390	72	1	content	content	NOUN
ajst-390	72	2	loss	loss	NOUN
ajst-390	72	3	function	function	NOUN
ajst-390	72	4	(	(	PUNCT
ajst-390	72	5	,	,	PUNCT
ajst-390	72	6	)	)	PUNCT
ajst-390	72	7	content	content	NOUN
ajst-390	72	8	v	v	ADP
ajst-390	72	9	vj	vj	PROPN
ajst-390	72	10	c	c	PROPN
ajst-390	72	11	g	g	PROPN
ajst-390	72	12	)	)	PUNCT
ajst-390	72	13	is	be	AUX
ajst-390	72	14	defined	define	VERB
ajst-390	72	15	as	as	ADP
ajst-390	72	16	the	the	DET
ajst-390	72	17	sum	sum	NOUN
ajst-390	72	18	of	of	ADP
ajst-390	72	19	squared	square	VERB
ajst-390	72	20	errors	error	NOUN
ajst-390	72	21	of	of	ADP
ajst-390	72	22	activation	activation	NOUN
ajst-390	72	23	values	value	NOUN
ajst-390	72	24	from	from	ADP
ajst-390	72	25	the	the	DET
ajst-390	72	26	content	content	NOUN
ajst-390	72	27	speech	speech	NOUN
ajst-390	72	28	spectrogram	spectrogram	NOUN
ajst-390	72	29	vcc	vcc	NOUN
ajst-390	72	30	and	and	CCONJ
ajst-390	72	31	the	the	DET
ajst-390	72	32	generated	generate	VERB
ajst-390	72	33	speech	speech	NOUN
ajst-390	72	34	spectrogram	spectrogram	NOUN
ajst-390	72	35	vgg	vgg	NOUN
ajst-390	72	36	in	in	ADP
ajst-390	72	37	the	the	DET
ajst-390	72	38	l	l	NOUN
ajst-390	72	39	th	th	X
ajst-390	72	40	layer	layer	NOUN
ajst-390	72	41	feature	feature	NOUN
ajst-390	72	42	map	map	NOUN
ajst-390	72	43	:	:	PUNCT
ajst-390	72	44	2	2	NUM
ajst-390	72	45	[	[	PUNCT
ajst-390	72	46	]	]	X
ajst-390	72	47	(	(	PUNCT
ajst-390	72	48	)	)	PUNCT
ajst-390	73	1	[	[	PUNCT
ajst-390	73	2	]	]	X
ajst-390	73	3	(	(	PUNCT
ajst-390	73	4	)	)	PUNCT
ajst-390	73	5	[	[	PUNCT
ajst-390	73	6	]	]	X
ajst-390	73	7	(	(	PUNCT
ajst-390	73	8	)	)	PUNCT
ajst-390	73	9	[	[	PUNCT
ajst-390	73	10	]	]	X
ajst-390	73	11	(	(	PUNCT
ajst-390	73	12	)	)	PUNCT
ajst-390	73	13	2	2	NUM
ajst-390	73	14	,	,	PUNCT
ajst-390	73	15	,	,	PUNCT
ajst-390	73	16	,	,	PUNCT
ajst-390	73	17	,	,	PUNCT
ajst-390	73	18	1	1	NUM
ajst-390	73	19	1	1	NUM
ajst-390	73	20	1	1	NUM
ajst-390	73	21	1	1	NUM
ajst-390	73	22	(	(	PUNCT
ajst-390	73	23	,	,	PUNCT
ajst-390	73	24	)	)	PUNCT
ajst-390	73	25	2	2	NUM
ajst-390	73	26	1	1	NUM
ajst-390	73	27	(	(	PUNCT
ajst-390	73	28	)	)	PUNCT
ajst-390	73	29	2	2	NUM
ajst-390	73	30	v	v	NOUN
ajst-390	73	31	v	v	NOUN
ajst-390	73	32	w	w	NOUN
ajst-390	73	33	ch	ch	NOUN
ajst-390	73	34	v	v	ADP
ajst-390	73	35	v	v	NOUN
ajst-390	73	36	l	l	NOUN
ajst-390	73	37	g	g	NOUN
ajst-390	74	1	l	l	NOUN
ajst-390	74	2	c	c	NOUN
ajst-390	74	3	content	content	NOUN
ajst-390	74	4	v	v	PROPN
ajst-390	74	5	v	v	NOUN
ajst-390	74	6	n	n	CCONJ
ajst-390	74	7	nn	nn	ADP
ajst-390	74	8	l	l	NOUN
ajst-390	74	9	g	g	PROPN
ajst-390	74	10	l	l	NOUN
ajst-390	74	11	c	c	NOUN
ajst-390	75	1	i	i	PRON
ajst-390	75	2	j	j	PROPN
ajst-390	76	1	k	k	INTJ
ajst-390	76	2	i	i	PRON
ajst-390	76	3	j	j	PROPN
ajst-390	77	1	k	k	INTJ
ajst-390	77	2	i	i	PRON
ajst-390	77	3	j	j	PROPN
ajst-390	78	1	k	k	PROPN
ajst-390	78	2	j	j	PROPN
ajst-390	79	1	c	c	PROPN
ajst-390	79	2	g	g	PROPN
ajst-390	79	3	a	a	DET
ajst-390	79	4	a	a	DET
ajst-390	79	5	a	a	DET
ajst-390	79	6	a	a	DET
ajst-390	79	7	2	2	NUM
ajst-390	79	8	)	)	PUNCT
ajst-390	80	1	[	[	PUNCT
ajst-390	80	2	]	]	X
ajst-390	80	3	(	(	PUNCT
ajst-390	80	4	)	)	PUNCT
ajst-390	80	5	)	)	PUNCT
ajst-390	81	1	[	[	PUNCT
ajst-390	81	2	]	]	X
ajst-390	81	3	(	(	PUNCT
ajst-390	81	4	)	)	PUNCT
ajst-390	81	5	[	[	PUNCT
ajst-390	81	6	]	]	X
ajst-390	81	7	(	(	PUNCT
ajst-390	81	8	)	)	PUNCT
ajst-390	81	9	)	)	PUNCT
ajst-390	82	1	[	[	PUNCT
ajst-390	82	2	]	]	X
ajst-390	82	3	(	(	PUNCT
ajst-390	82	4	)	)	PUNCT
ajst-390	82	5	2	2	NUM
ajst-390	82	6	)	)	PUNCT
ajst-390	82	7	)	)	PUNCT
ajst-390	82	8	)	)	PUNCT
ajst-390	83	1	[	[	PUNCT
ajst-390	83	2	]	]	X
ajst-390	83	3	(	(	PUNCT
ajst-390	83	4	)	)	PUNCT
ajst-390	83	5	[	[	PUNCT
ajst-390	83	6	]	]	X
ajst-390	83	7	)	)	PUNCT
ajst-390	83	8	[	[	PUNCT
ajst-390	83	9	]	]	X
ajst-390	83	10	(	(	PUNCT
ajst-390	83	11	)	)	PUNCT
ajst-390	83	12	[	[	PUNCT
ajst-390	83	13	]	]	X
ajst-390	83	14	1	1	NUM
ajst-390	83	15	)	)	PUNCT
ajst-390	83	16	1	1	NUM
ajst-390	83	17	)	)	PUNCT
ajst-390	83	18	.	.	PUNCT
ajst-390	84	1	(	(	PUNCT
ajst-390	84	2	1	1	X
ajst-390	84	3	)	)	PUNCT
ajst-390	84	4	the	the	DET
ajst-390	84	5	content	content	NOUN
ajst-390	84	6	loss	loss	NOUN
ajst-390	84	7	function	function	NOUN
ajst-390	84	8	(	(	PUNCT
ajst-390	84	9	,	,	PUNCT
ajst-390	84	10	)	)	PUNCT
ajst-390	84	11	content	content	NOUN
ajst-390	84	12	v	v	ADP
ajst-390	84	13	vj	vj	PROPN
ajst-390	84	14	c	c	PROPN
ajst-390	84	15	g	g	NOUN
ajst-390	84	16	)	)	PUNCT
ajst-390	84	17	measures	measure	VERB
ajst-390	84	18	the	the	DET
ajst-390	84	19	similarity	similarity	NOUN
ajst-390	84	20	between	between	ADP
ajst-390	84	21	content	content	NOUN
ajst-390	84	22	speech	speech	NOUN
ajst-390	84	23	spectrogram	spectrogram	NOUN
ajst-390	84	24	vcc	vcc	NOUN
ajst-390	84	25	and	and	CCONJ
ajst-390	84	26	4	4	NUM
ajst-390	84	27	generated	generate	VERB
ajst-390	84	28	speech	speech	NOUN
ajst-390	84	29	spectrogram	spectrogram	NOUN
ajst-390	84	30	vgg	vgg	NOUN
ajst-390	84	31	in	in	ADP
ajst-390	84	32	content	content	NOUN
ajst-390	84	33	features	feature	NOUN
ajst-390	84	34	such	such	ADJ
ajst-390	84	35	as	as	ADP
ajst-390	84	36	skeleton	skeleton	NOUN
ajst-390	84	37	and	and	CCONJ
ajst-390	84	38	layout	layout	NOUN
ajst-390	84	39	.	.	PUNCT
ajst-390	85	1	3.1.2	3.1.2	X
ajst-390	85	2	.	.	PUNCT
ajst-390	86	1	the	the	DET
ajst-390	86	2	extraction	extraction	NOUN
ajst-390	86	3	of	of	ADP
ajst-390	86	4	style	style	NOUN
ajst-390	86	5	features	feature	NOUN
ajst-390	86	6	from	from	ADP
ajst-390	86	7	the	the	DET
ajst-390	86	8	2d	2d	NUM
ajst-390	86	9	spectrogram	spectrogram	NOUN
ajst-390	86	10	of	of	ADP
ajst-390	86	11	style	style	NOUN
ajst-390	86	12	speech	speech	NOUN
ajst-390	86	13	the	the	DET
ajst-390	86	14	features	feature	NOUN
ajst-390	86	15	of	of	ADP
ajst-390	86	16	the	the	DET
ajst-390	86	17	low	low	ADJ
ajst-390	86	18	layers	layer	NOUN
ajst-390	86	19	’	'	PUNCT
ajst-390	86	20	feature	feature	NOUN
ajst-390	86	21	maps	map	NOUN
ajst-390	86	22	extracted	extract	VERB
ajst-390	86	23	by	by	ADP
ajst-390	86	24	the	the	DET
ajst-390	86	25	convolution	convolution	NOUN
ajst-390	86	26	filters	filter	NOUN
ajst-390	86	27	are	be	AUX
ajst-390	86	28	selected	select	VERB
ajst-390	86	29	as	as	ADP
ajst-390	86	30	style	style	NOUN
ajst-390	86	31	features	feature	NOUN
ajst-390	86	32	in	in	ADP
ajst-390	86	33	the	the	DET
ajst-390	86	34	spectrogram	spectrogram	NOUN
ajst-390	86	35	of	of	ADP
ajst-390	86	36	generated	generate	VERB
ajst-390	86	37	speech	speech	NOUN
ajst-390	86	38	.	.	PUNCT
ajst-390	87	1	step	step	NOUN
ajst-390	87	2	1	1	NUM
ajst-390	87	3	the	the	DET
ajst-390	87	4	style	style	NOUN
ajst-390	87	5	matrix	matrix	NOUN
ajst-390	87	6	of	of	ADP
ajst-390	87	7	spectrogram	spectrogram	NOUN
ajst-390	87	8	,	,	PUNCT
ajst-390	87	9	also	also	ADV
ajst-390	87	10	known	know	VERB
ajst-390	87	11	as	as	ADP
ajst-390	87	12	gram	gram	NOUN
ajst-390	87	13	matrix	matrix	NOUN
ajst-390	87	14	,	,	PUNCT
ajst-390	87	15	is	be	AUX
ajst-390	87	16	used	use	VERB
ajst-390	87	17	to	to	PART
ajst-390	87	18	measure	measure	VERB
ajst-390	87	19	the	the	DET
ajst-390	87	20	associativity	associativity	NOUN
ajst-390	87	21	between	between	ADP
ajst-390	87	22	different	different	ADJ
ajst-390	87	23	sections	section	NOUN
ajst-390	87	24	(	(	PUNCT
ajst-390	87	25	1	1	NUM
ajst-390	87	26	,	,	PUNCT
ajst-390	87	27	2	2	NUM
ajst-390	87	28	...	...	PUNCT
ajst-390	87	29	)	)	PUNCT
ajst-390	87	30	in	in	ADP
ajst-390	87	31	a	a	DET
ajst-390	87	32	certain	certain	ADJ
ajst-390	87	33	feature	feature	NOUN
ajst-390	87	34	map	map	NOUN
ajst-390	87	35	.	.	PUNCT
ajst-390	88	1	define	define	VERB
ajst-390	88	2	the	the	DET
ajst-390	88	3	style	style	NOUN
ajst-390	88	4	matrix	matrix	NOUN
ajst-390	88	5	[	[	PUNCT
ajst-390	88	6	]	]	X
ajst-390	88	7	(	(	PUNCT
ajst-390	88	8	)	)	PUNCT
ajst-390	88	9	vl	vl	INTJ
ajst-390	88	10	sg	sg	PROPN
ajst-390	88	11	)	)	PUNCT
ajst-390	88	12	for	for	ADP
ajst-390	88	13	the	the	DET
ajst-390	88	14	style	style	NOUN
ajst-390	88	15	speech	speech	NOUN
ajst-390	88	16	spectrogram	spectrogram	NOUN
ajst-390	88	17	vss	vss	NOUN
ajst-390	88	18	:	:	PUNCT
ajst-390	88	19	[	[	PUNCT
ajst-390	88	20	]	]	X
ajst-390	88	21	[	[	PUNCT
ajst-390	88	22	]	]	X
ajst-390	88	23	[	[	PUNCT
ajst-390	88	24	]	]	X
ajst-390	88	25	(	(	PUNCT
ajst-390	88	26	)	)	PUNCT
ajst-390	88	27	[	[	PUNCT
ajst-390	88	28	]	]	X
ajst-390	88	29	(	(	PUNCT
ajst-390	88	30	)	)	PUNCT
ajst-390	88	31	[	[	PUNCT
ajst-390	88	32	]	]	X
ajst-390	88	33	(	(	PUNCT
ajst-390	88	34	)	)	PUNCT
ajst-390	88	35	'	'	PUNCT
ajst-390	88	36	'	'	PUNCT
ajst-390	88	37	1	1	NUM
ajst-390	88	38	1	1	NUM
ajst-390	88	39	ll	ll	AUX
ajst-390	88	40	wh	wh	VERB
ajst-390	88	41	v	v	NUM
ajst-390	88	42	v	v	PROPN
ajst-390	88	43	v	v	NOUN
ajst-390	88	44	nn	nn	PROPN
ajst-390	88	45	l	l	NOUN
ajst-390	88	46	s	s	PART
ajst-390	88	47	l	l	NOUN
ajst-390	88	48	s	s	PART
ajst-390	88	49	l	l	NOUN
ajst-390	88	50	s	s	PART
ajst-390	88	51	kk	kk	PROPN
ajst-390	88	52	ijk	ijk	PROPN
ajst-390	88	53	ijk	ijk	PROPN
ajst-390	89	1	i	i	PRON
ajst-390	89	2	j	j	PROPN
ajst-390	90	1	g	g	VERB
ajst-390	90	2	a	a	DET
ajst-390	90	3	a	a	X
ajst-390	90	4	[	[	PUNCT
ajst-390	90	5	]	]	X
ajst-390	90	6	[	[	PUNCT
ajst-390	90	7	]	]	X
ajst-390	90	8	[	[	PUNCT
ajst-390	90	9	]	]	X
ajst-390	90	10	(	(	PUNCT
ajst-390	90	11	)	)	PUNCT
ajst-390	90	12	[	[	PUNCT
ajst-390	90	13	]	]	X
ajst-390	90	14	(	(	PUNCT
ajst-390	90	15	)	)	PUNCT
ajst-390	90	16	wh	wh	PROPN
ajst-390	90	17	nn	nn	PROPN
ajst-390	90	18	)	)	PUNCT
ajst-390	91	1	[	[	PUNCT
ajst-390	91	2	]	]	X
ajst-390	91	3	(	(	PUNCT
ajst-390	91	4	)	)	PUNCT
ajst-390	91	5	[	[	PUNCT
ajst-390	91	6	]	]	X
ajst-390	91	7	(	(	PUNCT
ajst-390	91	8	)	)	PUNCT
ajst-390	91	9	[	[	PUNCT
ajst-390	91	10	]	]	X
ajst-390	91	11	(	(	PUNCT
ajst-390	91	12	)	)	PUNCT
ajst-390	91	13	[	[	PUNCT
ajst-390	91	14	]	]	X
ajst-390	91	15	(	(	PUNCT
ajst-390	91	16	)	)	PUNCT
ajst-390	91	17	[	[	PUNCT
ajst-390	91	18	]	]	X
ajst-390	91	19	(	(	PUNCT
ajst-390	91	20	)	)	PUNCT
ajst-390	91	21	;	;	PUNCT
ajst-390	91	22	(	(	PUNCT
ajst-390	91	23	2	2	X
ajst-390	91	24	)	)	PUNCT
ajst-390	91	25	define	define	VERB
ajst-390	91	26	the	the	DET
ajst-390	91	27	style	style	NOUN
ajst-390	91	28	matrix	matrix	NOUN
ajst-390	91	29	[	[	PUNCT
ajst-390	91	30	]	]	X
ajst-390	91	31	(	(	PUNCT
ajst-390	91	32	)	)	PUNCT
ajst-390	91	33	vl	vl	PROPN
ajst-390	91	34	gg	gg	PROPN
ajst-390	91	35	)	)	PUNCT
ajst-390	91	36	for	for	ADP
ajst-390	91	37	the	the	DET
ajst-390	91	38	generated	generate	VERB
ajst-390	91	39	speech	speech	NOUN
ajst-390	91	40	spectrogram	spectrogram	NOUN
ajst-390	91	41	vgg	vgg	NOUN
ajst-390	91	42	:	:	PUNCT
ajst-390	91	43	[	[	PUNCT
ajst-390	91	44	]	]	X
ajst-390	91	45	[	[	PUNCT
ajst-390	91	46	]	]	X
ajst-390	91	47	[	[	PUNCT
ajst-390	91	48	]	]	X
ajst-390	91	49	(	(	PUNCT
ajst-390	91	50	)	)	PUNCT
ajst-390	91	51	[	[	PUNCT
ajst-390	91	52	]	]	X
ajst-390	91	53	(	(	PUNCT
ajst-390	91	54	)	)	PUNCT
ajst-390	91	55	[	[	PUNCT
ajst-390	91	56	]	]	X
ajst-390	91	57	(	(	PUNCT
ajst-390	91	58	)	)	PUNCT
ajst-390	91	59	'	'	PUNCT
ajst-390	91	60	'	'	PUNCT
ajst-390	91	61	1	1	NUM
ajst-390	91	62	1	1	NUM
ajst-390	91	63	ll	ll	AUX
ajst-390	91	64	wh	wh	VERB
ajst-390	91	65	v	v	NUM
ajst-390	91	66	v	v	PROPN
ajst-390	91	67	v	v	NOUN
ajst-390	91	68	nn	nn	PROPN
ajst-390	91	69	l	l	NOUN
ajst-390	91	70	g	g	PROPN
ajst-390	91	71	l	l	NOUN
ajst-390	91	72	g	g	NOUN
ajst-390	91	73	l	l	NOUN
ajst-390	91	74	g	g	PROPN
ajst-390	91	75	kk	kk	PROPN
ajst-390	91	76	ijk	ijk	PROPN
ajst-390	91	77	ijk	ijk	PROPN
ajst-390	92	1	i	i	PRON
ajst-390	92	2	j	j	PROPN
ajst-390	93	1	g	g	VERB
ajst-390	93	2	a	a	DET
ajst-390	93	3	a	a	X
ajst-390	93	4	[	[	PUNCT
ajst-390	93	5	]	]	X
ajst-390	93	6	[	[	PUNCT
ajst-390	93	7	]	]	X
ajst-390	93	8	[	[	PUNCT
ajst-390	93	9	]	]	X
ajst-390	93	10	(	(	PUNCT
ajst-390	93	11	)	)	PUNCT
ajst-390	93	12	[	[	PUNCT
ajst-390	93	13	]	]	X
ajst-390	93	14	(	(	PUNCT
ajst-390	93	15	)	)	PUNCT
ajst-390	93	16	ll	ll	AUX
ajst-390	93	17	wh	wh	VERB
ajst-390	93	18	nn	nn	PROPN
ajst-390	93	19	)	)	PUNCT
ajst-390	94	1	[	[	PUNCT
ajst-390	94	2	]	]	X
ajst-390	94	3	(	(	PUNCT
ajst-390	94	4	)	)	PUNCT
ajst-390	94	5	[	[	PUNCT
ajst-390	94	6	]	]	X
ajst-390	94	7	(	(	PUNCT
ajst-390	94	8	)	)	PUNCT
ajst-390	94	9	[	[	PUNCT
ajst-390	94	10	]	]	X
ajst-390	94	11	(	(	PUNCT
ajst-390	94	12	)	)	PUNCT
ajst-390	94	13	[	[	PUNCT
ajst-390	94	14	]	]	X
ajst-390	94	15	(	(	PUNCT
ajst-390	94	16	[	[	PUNCT
ajst-390	94	17	]	]	X
ajst-390	94	18	(	(	PUNCT
ajst-390	94	19	)	)	PUNCT
ajst-390	94	20	]	]	PUNCT
ajst-390	94	21	)	)	PUNCT
ajst-390	94	22	.	.	PUNCT
ajst-390	95	1	(	(	PUNCT
ajst-390	95	2	3	3	X
ajst-390	95	3	)	)	PUNCT
ajst-390	95	4	figure	figure	NOUN
ajst-390	95	5	7	7	NUM
ajst-390	95	6	is	be	AUX
ajst-390	95	7	a	a	DET
ajst-390	95	8	schematic	schematic	ADJ
ajst-390	95	9	diagram	diagram	NOUN
ajst-390	95	10	of	of	ADP
ajst-390	95	11	gram	gram	NOUN
ajst-390	95	12	matrix	matrix	NOUN
ajst-390	95	13	calculation	calculation	NOUN
ajst-390	95	14	process	process	NOUN
ajst-390	95	15	including	include	VERB
ajst-390	95	16	multiplication	multiplication	NOUN
ajst-390	95	17	and	and	CCONJ
ajst-390	95	18	sum	sum	NOUN
ajst-390	95	19	operations	operation	NOUN
ajst-390	95	20	for	for	ADP
ajst-390	95	21	activation	activation	NOUN
ajst-390	95	22	values	value	NOUN
ajst-390	95	23	[	[	PUNCT
ajst-390	95	24	]	]	X
ajst-390	95	25	,	,	PUNCT
ajst-390	95	26	,	,	PUNCT
ajst-390	95	27	l	l	PROPN
ajst-390	96	1	i	i	PRON
ajst-390	96	2	j	j	PROPN
ajst-390	96	3	ka	ka	PROPN
ajst-390	96	4	and	and	CCONJ
ajst-390	96	5	[	[	PUNCT
ajst-390	96	6	]	]	X
ajst-390	96	7	,	,	PUNCT
ajst-390	96	8	,	,	PUNCT
ajst-390	96	9	'	'	PUNCT
ajst-390	96	10	l	l	NOUN
ajst-390	97	1	i	i	PRON
ajst-390	97	2	j	j	PROPN
ajst-390	97	3	ka	ka	PROPN
ajst-390	97	4	at	at	ADP
ajst-390	97	5	corresponding	correspond	VERB
ajst-390	97	6	positions	position	NOUN
ajst-390	97	7	.	.	PUNCT
ajst-390	98	1	figure	figure	VERB
ajst-390	98	2	7	7	NUM
ajst-390	98	3	.	.	PUNCT
ajst-390	99	1	the	the	DET
ajst-390	99	2	gram	gram	NOUN
ajst-390	99	3	matrix	matrix	NOUN
ajst-390	99	4	calculation	calculation	NOUN
ajst-390	99	5	process	process	NOUN
ajst-390	99	6	the	the	DET
ajst-390	99	7	gram	gram	NOUN
ajst-390	99	8	matrix	matrix	NOUN
ajst-390	99	9	of	of	ADP
ajst-390	99	10	spectrogram	spectrogram	NOUN
ajst-390	99	11	measures	measure	NOUN
ajst-390	99	12	whether	whether	SCONJ
ajst-390	99	13	or	or	CCONJ
ajst-390	99	14	not	not	PART
ajst-390	99	15	two	two	NUM
ajst-390	99	16	features	feature	NOUN
ajst-390	99	17	appear	appear	VERB
ajst-390	99	18	simultaneously	simultaneously	ADV
ajst-390	99	19	in	in	ADP
ajst-390	99	20	the	the	DET
ajst-390	99	21	spectrogram	spectrogram	NOUN
ajst-390	99	22	and	and	CCONJ
ajst-390	99	23	show	show	VERB
ajst-390	99	24	the	the	DET
ajst-390	99	25	response	response	NOUN
ajst-390	99	26	between	between	ADP
ajst-390	99	27	two	two	NUM
ajst-390	99	28	features	feature	NOUN
ajst-390	99	29	when	when	SCONJ
ajst-390	99	30	they	they	PRON
ajst-390	99	31	appear	appear	VERB
ajst-390	99	32	together	together	ADV
ajst-390	99	33	.	.	PUNCT
ajst-390	100	1	step	step	NOUN
ajst-390	100	2	2	2	NUM
ajst-390	100	3	define	define	VERB
ajst-390	100	4	style	style	NOUN
ajst-390	100	5	loss	loss	NOUN
ajst-390	100	6	function	function	NOUN
ajst-390	100	7	[	[	PUNCT
ajst-390	100	8	]	]	X
ajst-390	100	9	(	(	PUNCT
ajst-390	100	10	,	,	PUNCT
ajst-390	100	11	)	)	PUNCT
ajst-390	100	12	l	l	NOUN
ajst-390	100	13	l	l	NOUN
ajst-390	100	14	style	style	NOUN
ajst-390	100	15	v	v	ADP
ajst-390	100	16	ve	ve	PROPN
ajst-390	100	17	j	j	PROPN
ajst-390	100	18	s	s	PROPN
ajst-390	100	19	g	g	PROPN
ajst-390	100	20	)	)	PUNCT
ajst-390	100	21	for	for	ADP
ajst-390	100	22	the	the	DET
ajst-390	100	23	l	l	NOUN
ajst-390	100	24	th	th	X
ajst-390	100	25	layer	layer	NOUN
ajst-390	100	26	feature	feature	NOUN
ajst-390	100	27	map	map	NOUN
ajst-390	100	28	:	:	PUNCT
ajst-390	100	29	2	2	NUM
ajst-390	100	30	[	[	PUNCT
ajst-390	100	31	]	]	X
ajst-390	100	32	(	(	PUNCT
ajst-390	100	33	)	)	PUNCT
ajst-390	101	1	[	[	PUNCT
ajst-390	101	2	]	]	X
ajst-390	101	3	(	(	PUNCT
ajst-390	101	4	)	)	PUNCT
ajst-390	101	5	[	[	PUNCT
ajst-390	101	6	]	]	X
ajst-390	101	7	[	[	PUNCT
ajst-390	101	8	]	]	X
ajst-390	101	9	[	[	PUNCT
ajst-390	101	10	]	]	X
ajst-390	101	11	[	[	PUNCT
ajst-390	101	12	]	]	X
ajst-390	101	13	2	2	NUM
ajst-390	101	14	[	[	PUNCT
ajst-390	101	15	]	]	X
ajst-390	101	16	(	(	PUNCT
ajst-390	101	17	)	)	PUNCT
ajst-390	101	18	[	[	PUNCT
ajst-390	101	19	]	]	X
ajst-390	101	20	(	(	PUNCT
ajst-390	101	21	)	)	PUNCT
ajst-390	101	22	2	2	NUM
ajst-390	101	23	'	'	NOUN
ajst-390	101	24	'	'	PUNCT
ajst-390	101	25	[	[	PUNCT
ajst-390	101	26	]	]	X
ajst-390	101	27	[	[	PUNCT
ajst-390	101	28	]	]	X
ajst-390	101	29	[	[	PUNCT
ajst-390	101	30	]	]	X
ajst-390	101	31	2	2	NUM
ajst-390	101	32	1	1	NUM
ajst-390	101	33	'	'	NUM
ajst-390	101	34	1	1	NUM
ajst-390	101	35	1	1	NUM
ajst-390	101	36	(	(	PUNCT
ajst-390	101	37	,	,	PUNCT
ajst-390	101	38	)	)	PUNCT
ajst-390	101	39	(	(	PUNCT
ajst-390	101	40	2	2	X
ajst-390	101	41	)	)	PUNCT
ajst-390	101	42	1	1	NUM
ajst-390	101	43	(	(	PUNCT
ajst-390	101	44	)	)	PUNCT
ajst-390	101	45	(	(	PUNCT
ajst-390	101	46	2	2	X
ajst-390	101	47	)	)	PUNCT
ajst-390	101	48	v	v	ADP
ajst-390	101	49	v	v	NOUN
ajst-390	101	50	c	c	NOUN
ajst-390	101	51	c	c	NOUN
ajst-390	101	52	v	v	NOUN
ajst-390	101	53	v	v	NOUN
ajst-390	101	54	l	l	NOUN
ajst-390	101	55	s	s	PART
ajst-390	101	56	l	l	NOUN
ajst-390	101	57	gl	gl	NOUN
ajst-390	101	58	l	l	PROPN
ajst-390	101	59	style	style	NOUN
ajst-390	101	60	v	v	ADP
ajst-390	101	61	v	v	NOUN
ajst-390	101	62	l	l	NOUN
ajst-390	101	63	l	l	NOUN
ajst-390	101	64	l	l	NOUN
ajst-390	102	1	f	f	X
ajst-390	102	2	h	h	NOUN
ajst-390	102	3	w	w	PROPN
ajst-390	102	4	c	c	PROPN
ajst-390	102	5	n	n	CCONJ
ajst-390	102	6	n	n	PROPN
ajst-390	102	7	l	l	NOUN
ajst-390	102	8	s	s	PART
ajst-390	102	9	l	l	NOUN
ajst-390	102	10	g	g	NOUN
ajst-390	102	11	kk	kk	PROPN
ajst-390	102	12	kkl	kkl	PROPN
ajst-390	103	1	l	l	PROPN
ajst-390	103	2	l	l	PROPN
ajst-390	104	1	k	k	PROPN
ajst-390	104	2	kh	kh	PROPN
ajst-390	104	3	w	w	PROPN
ajst-390	104	4	c	c	PROPN
ajst-390	104	5	e	e	PROPN
ajst-390	104	6	j	j	PROPN
ajst-390	104	7	s	s	PROPN
ajst-390	104	8	g	g	PROPN
ajst-390	104	9	g	g	PROPN
ajst-390	104	10	g	g	PROPN
ajst-390	104	11	n	n	CCONJ
ajst-390	104	12	n	n	CCONJ
ajst-390	104	13	n	n	CCONJ
ajst-390	104	14	g	g	NOUN
ajst-390	104	15	g	g	PROPN
ajst-390	104	16	n	n	CCONJ
ajst-390	104	17	n	n	CCONJ
ajst-390	104	18	n	n	ADV
ajst-390	104	19	2	2	NUM
ajst-390	104	20	)	)	PUNCT
ajst-390	105	1	[	[	PUNCT
ajst-390	105	2	]	]	X
ajst-390	105	3	(	(	PUNCT
ajst-390	105	4	)	)	PUNCT
ajst-390	105	5	)	)	PUNCT
ajst-390	106	1	[	[	PUNCT
ajst-390	106	2	]	]	X
ajst-390	106	3	(	(	PUNCT
ajst-390	106	4	)	)	PUNCT
ajst-390	106	5	[	[	PUNCT
ajst-390	106	6	]	]	X
ajst-390	106	7	(	(	PUNCT
ajst-390	106	8	)	)	PUNCT
ajst-390	106	9	)	)	PUNCT
ajst-390	107	1	[	[	PUNCT
ajst-390	107	2	]	]	X
ajst-390	107	3	(	(	PUNCT
ajst-390	107	4	)	)	PUNCT
ajst-390	107	5	2	2	NUM
ajst-390	107	6	)	)	PUNCT
ajst-390	107	7	)	)	PUNCT
ajst-390	107	8	)	)	PUNCT
ajst-390	108	1	[	[	PUNCT
ajst-390	108	2	]	]	X
ajst-390	108	3	(	(	PUNCT
ajst-390	108	4	)	)	PUNCT
ajst-390	108	5	[	[	PUNCT
ajst-390	108	6	]	]	X
ajst-390	108	7	(	(	PUNCT
ajst-390	108	8	)	)	PUNCT
ajst-390	108	9	[	[	PUNCT
ajst-390	108	10	]	]	X
ajst-390	108	11	(	(	PUNCT
ajst-390	108	12	]	]	X
ajst-390	108	13	(	(	PUNCT
ajst-390	108	14	)	)	PUNCT
ajst-390	108	15	)	)	PUNCT
ajst-390	108	16	)	)	PUNCT
ajst-390	108	17	)	)	PUNCT
ajst-390	108	18	.(4	.(4	PROPN
ajst-390	108	19	)	)	PUNCT
ajst-390	108	20	step	step	NOUN
ajst-390	108	21	3	3	NUM
ajst-390	108	22	finally	finally	ADV
ajst-390	108	23	,	,	PUNCT
ajst-390	108	24	style	style	NOUN
ajst-390	108	25	loss	loss	NOUN
ajst-390	108	26	function	function	NOUN
ajst-390	108	27	(	(	PUNCT
ajst-390	108	28	,	,	PUNCT
ajst-390	108	29	)	)	PUNCT
ajst-390	108	30	style	style	NOUN
ajst-390	108	31	v	v	ADP
ajst-390	108	32	vj	vj	PROPN
ajst-390	108	33	s	s	PROPN
ajst-390	108	34	g	g	NOUN
ajst-390	108	35	)	)	PUNCT
ajst-390	108	36	of	of	ADP
ajst-390	108	37	speech	speech	NOUN
ajst-390	108	38	spectrogram	spectrogram	NOUN
ajst-390	108	39	is	be	AUX
ajst-390	108	40	defined	define	VERB
ajst-390	108	41	as	as	ADP
ajst-390	108	42	the	the	DET
ajst-390	108	43	weighted	weighted	ADJ
ajst-390	108	44	sum	sum	NOUN
ajst-390	108	45	of	of	ADP
ajst-390	108	46	multilayer	multilayer	ADJ
ajst-390	108	47	style	style	NOUN
ajst-390	108	48	loss	loss	NOUN
ajst-390	108	49	functions	function	NOUN
ajst-390	108	50	,	,	PUNCT
ajst-390	108	51	1	1	NUM
ajst-390	108	52	,	,	PUNCT
ajst-390	108	53	2,	2,	NUM
ajst-390	108	54	...	...	PUNCT
ajst-390	108	55	le	le	X
ajst-390	108	56	l	l	NOUN
ajst-390	108	57	(	(	PUNCT
ajst-390	108	58	,	,	PUNCT
ajst-390	108	59	)	)	PUNCT
ajst-390	108	60	style	style	NOUN
ajst-390	108	61	v	v	ADP
ajst-390	108	62	v	v	NOUN
ajst-390	108	63	l	l	NOUN
ajst-390	108	64	l	l	NOUN
ajst-390	108	65	l	l	X
ajst-390	108	66	j	j	PROPN
ajst-390	108	67	s	s	PROPN
ajst-390	108	68	g	g	PROPN
ajst-390	108	69	w	w	PROPN
ajst-390	108	70	e	e	NOUN
ajst-390	108	71	)	)	PUNCT
ajst-390	108	72	)	)	PUNCT
ajst-390	108	73	.	.	PUNCT
ajst-390	109	1	(	(	PUNCT
ajst-390	109	2	5	5	X
ajst-390	109	3	)	)	PUNCT
ajst-390	109	4	the	the	DET
ajst-390	109	5	style	style	NOUN
ajst-390	109	6	loss	loss	NOUN
ajst-390	109	7	function	function	NOUN
ajst-390	109	8	(	(	PUNCT
ajst-390	109	9	,	,	PUNCT
ajst-390	109	10	)	)	PUNCT
ajst-390	109	11	style	style	NOUN
ajst-390	109	12	v	v	ADP
ajst-390	109	13	vj	vj	PROPN
ajst-390	109	14	s	s	PROPN
ajst-390	109	15	g	g	NOUN
ajst-390	109	16	)	)	PUNCT
ajst-390	109	17	measures	measure	VERB
ajst-390	109	18	the	the	DET
ajst-390	109	19	similarity	similarity	NOUN
ajst-390	109	20	between	between	ADP
ajst-390	109	21	style	style	NOUN
ajst-390	109	22	speech	speech	NOUN
ajst-390	109	23	spectrogram	spectrogram	NOUN
ajst-390	109	24	vss	vss	NOUN
ajst-390	109	25	and	and	CCONJ
ajst-390	109	26	generated	generate	VERB
ajst-390	109	27	speech	speech	NOUN
ajst-390	109	28	spectrogram	spectrogram	NOUN
ajst-390	109	29	vgg	vgg	NOUN
ajst-390	109	30	in	in	ADP
ajst-390	109	31	style	style	NOUN
ajst-390	109	32	features	feature	NOUN
ajst-390	109	33	such	such	ADJ
ajst-390	109	34	as	as	ADP
ajst-390	109	35	edge	edge	NOUN
ajst-390	109	36	,	,	PUNCT
ajst-390	109	37	texture	texture	NOUN
ajst-390	109	38	and	and	CCONJ
ajst-390	109	39	color	color	NOUN
ajst-390	109	40	.	.	PUNCT
ajst-390	110	1	3.1.3	3.1.3	X
ajst-390	110	2	.	.	PUNCT
ajst-390	110	3	iterating	iterate	VERB
ajst-390	110	4	to	to	PART
ajst-390	110	5	achieve	achieve	VERB
ajst-390	110	6	generated	generate	VERB
ajst-390	110	7	speech	speech	NOUN
ajst-390	110	8	spectrogram	spectrogram	NOUN
ajst-390	110	9	,	,	PUNCT
ajst-390	110	10	and	and	CCONJ
ajst-390	110	11	converting	convert	VERB
ajst-390	110	12	it	it	PRON
ajst-390	110	13	to	to	ADP
ajst-390	110	14	audio	audio	ADJ
ajst-390	110	15	file	file	NOUN
ajst-390	110	16	in	in	ADP
ajst-390	110	17	the	the	DET
ajst-390	110	18	end	end	NOUN
ajst-390	110	19	,	,	PUNCT
ajst-390	110	20	we	we	PRON
ajst-390	110	21	define	define	VERB
ajst-390	110	22	total	total	ADJ
ajst-390	110	23	loss	loss	NOUN
ajst-390	110	24	function	function	NOUN
ajst-390	110	25	(	(	PUNCT
ajst-390	110	26	,	,	PUNCT
ajst-390	110	27	,	,	PUNCT
ajst-390	110	28	)	)	PUNCT
ajst-390	110	29	v	v	NOUN
ajst-390	110	30	v	v	NUM
ajst-390	110	31	vj	vj	INTJ
ajst-390	110	32	g	g	PROPN
ajst-390	110	33	c	c	PROPN
ajst-390	110	34	s	s	PART
ajst-390	110	35	)	)	PUNCT
ajst-390	110	36	of	of	ADP
ajst-390	110	37	the	the	DET
ajst-390	110	38	generated	generate	VERB
ajst-390	110	39	speech	speech	NOUN
ajst-390	110	40	spectrogram	spectrogram	NOUN
ajst-390	110	41	comparing	compare	VERB
ajst-390	110	42	with	with	ADP
ajst-390	110	43	the	the	DET
ajst-390	110	44	content	content	NOUN
ajst-390	110	45	speech	speech	NOUN
ajst-390	110	46	spectrogram	spectrogram	NOUN
ajst-390	110	47	and	and	CCONJ
ajst-390	110	48	style	style	NOUN
ajst-390	110	49	speech	speech	NOUN
ajst-390	110	50	spectrogram	spectrogram	NOUN
ajst-390	110	51	as	as	ADP
ajst-390	110	52	:	:	PUNCT
ajst-390	110	53	on	on	ADP
ajst-390	110	54	(	(	PUNCT
ajst-390	110	55	,	,	PUNCT
ajst-390	110	56	,	,	PUNCT
ajst-390	110	57	)	)	PUNCT
ajst-390	110	58	(	(	PUNCT
ajst-390	110	59	,	,	PUNCT
ajst-390	110	60	)	)	PUNCT
ajst-390	110	61	(	(	PUNCT
ajst-390	110	62	,	,	PUNCT
ajst-390	110	63	)	)	PUNCT
ajst-390	110	64	v	v	NOUN
ajst-390	110	65	v	v	NUM
ajst-390	110	66	v	v	NOUN
ajst-390	110	67	c	c	NOUN
ajst-390	110	68	tent	tent	NOUN
ajst-390	110	69	v	v	NOUN
ajst-390	110	70	v	v	NOUN
ajst-390	110	71	style	style	NOUN
ajst-390	110	72	v	v	ADP
ajst-390	110	73	vj	vj	INTJ
ajst-390	111	1	g	g	PROPN
ajst-390	111	2	c	c	PROPN
ajst-390	111	3	s	s	PROPN
ajst-390	111	4	j	j	PROPN
ajst-390	111	5	c	c	PROPN
ajst-390	111	6	g	g	PROPN
ajst-390	111	7	j	j	PROPN
ajst-390	111	8	s	s	PROPN
ajst-390	111	9	g	g	NOUN
ajst-390	111	10	)	)	PUNCT
ajst-390	111	11	(	(	PUNCT
ajst-390	111	12	)	)	PUNCT
ajst-390	111	13	(	(	PUNCT
ajst-390	111	14	)	)	PUNCT
ajst-390	111	15	(	(	PUNCT
ajst-390	111	16	(	(	PUNCT
ajst-390	111	17	)	)	PUNCT
ajst-390	111	18	(	(	PUNCT
ajst-390	111	19	)	)	PUNCT
ajst-390	111	20	(	(	PUNCT
ajst-390	111	21	)	)	PUNCT
ajst-390	111	22	(	(	PUNCT
ajst-390	111	23	(	(	PUNCT
ajst-390	111	24	)	)	PUNCT
ajst-390	111	25	(	(	PUNCT
ajst-390	111	26	.	.	PUNCT
ajst-390	112	1	(	(	PUNCT
ajst-390	112	2	6	6	X
ajst-390	112	3	)	)	PUNCT
ajst-390	112	4	minimizing	minimize	VERB
ajst-390	112	5	the	the	DET
ajst-390	112	6	total	total	ADJ
ajst-390	112	7	loss	loss	NOUN
ajst-390	112	8	function	function	NOUN
ajst-390	112	9	(	(	PUNCT
ajst-390	112	10	,	,	PUNCT
ajst-390	112	11	,	,	PUNCT
ajst-390	112	12	)	)	PUNCT
ajst-390	112	13	v	v	NOUN
ajst-390	112	14	v	v	NUM
ajst-390	112	15	vj	vj	INTJ
ajst-390	112	16	g	g	PROPN
ajst-390	112	17	c	c	PROPN
ajst-390	112	18	s	s	PART
ajst-390	112	19	)	)	PUNCT
ajst-390	112	20	.	.	PUNCT
ajst-390	113	1	the	the	DET
ajst-390	113	2	stylized	stylize	VERB
ajst-390	113	3	spectrogram	spectrogram	NOUN
ajst-390	113	4	(	(	PUNCT
ajst-390	113	5	):	):	PUNCT
ajst-390	113	6	v	v	NUM
ajst-390	113	7	v	v	PART
ajst-390	113	8	v	v	NOUN
ajst-390	113	9	v	v	NOUN
ajst-390	113	10	j	j	PROPN
ajst-390	113	11	gg	gg	NOUN
ajst-390	113	12	g	g	PROPN
ajst-390	113	13	g	g	PROPN
ajst-390	113	14	)	)	PUNCT
ajst-390	113	15	jg	jg	PROPN
ajst-390	113	16	g	g	PROPN
ajst-390	113	17	:	:	PUNCT
ajst-390	113	18	jjg	jjg	PROPN
ajst-390	113	19	(	(	PUNCT
ajst-390	113	20	)	)	PUNCT
ajst-390	113	21	v	v	NOUN
ajst-390	113	22	(	(	PUNCT
ajst-390	113	23	g	g	NOUN
ajst-390	113	24	)	)	PUNCT
ajst-390	113	25	of	of	ADP
ajst-390	113	26	generated	generate	VERB
ajst-390	113	27	speech	speech	NOUN
ajst-390	113	28	can	can	AUX
ajst-390	113	29	be	be	AUX
ajst-390	113	30	obtained	obtain	VERB
ajst-390	113	31	iteratively	iteratively	ADV
ajst-390	113	32	with	with	ADP
ajst-390	113	33	the	the	DET
ajst-390	113	34	aid	aid	NOUN
ajst-390	113	35	of	of	ADP
ajst-390	113	36	gradient	gradient	ADJ
ajst-390	113	37	descent	descent	NOUN
ajst-390	113	38	method	method	NOUN
ajst-390	113	39	.	.	PUNCT
ajst-390	114	1	finally	finally	ADV
ajst-390	114	2	,	,	PUNCT
ajst-390	114	3	the	the	DET
ajst-390	114	4	audio	audio	ADJ
ajst-390	114	5	file	file	NOUN
ajst-390	114	6	of	of	ADP
ajst-390	114	7	generated	generate	VERB
ajst-390	114	8	speech	speech	NOUN
ajst-390	114	9	signal	signal	NOUN
ajst-390	114	10	can	can	AUX
ajst-390	114	11	be	be	AUX
ajst-390	114	12	acquired	acquire	VERB
ajst-390	114	13	according	accord	VERB
ajst-390	114	14	to	to	ADP
ajst-390	114	15	the	the	DET
ajst-390	114	16	stylized	stylize	VERB
ajst-390	114	17	spectrogram	spectrogram	NOUN
ajst-390	114	18	.	.	PUNCT
ajst-390	115	1	figure	figure	NOUN
ajst-390	115	2	8	8	NUM
ajst-390	115	3	is	be	AUX
ajst-390	115	4	an	an	DET
ajst-390	115	5	iteration	iteration	NOUN
ajst-390	115	6	acquisition	acquisition	NOUN
ajst-390	115	7	diagram	diagram	NOUN
ajst-390	115	8	for	for	ADP
ajst-390	115	9	the	the	DET
ajst-390	115	10	audio	audio	ADJ
ajst-390	115	11	file	file	NOUN
ajst-390	115	12	of	of	ADP
ajst-390	115	13	generated	generate	VERB
ajst-390	115	14	speech	speech	NOUN
ajst-390	115	15	.	.	PUNCT
ajst-390	116	1	figure	figure	NOUN
ajst-390	116	2	8	8	NUM
ajst-390	116	3	.	.	PUNCT
ajst-390	117	1	the	the	DET
ajst-390	117	2	iteration	iteration	NOUN
ajst-390	117	3	acquisition	acquisition	NOUN
ajst-390	117	4	diagram	diagram	NOUN
ajst-390	117	5	for	for	ADP
ajst-390	117	6	the	the	DET
ajst-390	117	7	audio	audio	ADJ
ajst-390	117	8	file	file	NOUN
ajst-390	117	9	of	of	ADP
ajst-390	117	10	generated	generate	VERB
ajst-390	117	11	speech	speech	NOUN
ajst-390	117	12	5	5	NUM
ajst-390	117	13	references	reference	NOUN
ajst-390	117	14	[	[	X
ajst-390	117	15	1	1	NUM
ajst-390	117	16	]	]	PUNCT
ajst-390	117	17	childers	childer	NOUN
ajst-390	117	18	d	d	X
ajst-390	117	19	g	g	PROPN
ajst-390	117	20	,	,	PUNCT
ajst-390	117	21	wu	wu	PROPN
ajst-390	117	22	k	k	PROPN
ajst-390	117	23	,	,	PUNCT
ajst-390	117	24	hicks	hicks	PROPN
ajst-390	117	25	d	d	PROPN
ajst-390	117	26	m	m	PROPN
ajst-390	117	27	,	,	PUNCT
ajst-390	117	28	et	et	PROPN
ajst-390	117	29	al	al	PROPN
ajst-390	117	30	.	.	PUNCT
ajst-390	118	1	voice	voice	PROPN
ajst-390	118	2	conversion[j	conversion[j	PROPN
ajst-390	118	3	]	]	PUNCT
ajst-390	118	4	.	.	PUNCT
ajst-390	119	1	speech	speech	NOUN
ajst-390	119	2	communication	communication	NOUN
ajst-390	119	3	,	,	PUNCT
ajst-390	119	4	1989	1989	NUM
ajst-390	119	5	,	,	PUNCT
ajst-390	119	6	8(2):147	8(2):147	NUM
ajst-390	119	7	-	-	SYM
ajst-390	119	8	158	158	NUM
ajst-390	119	9	.	.	PUNCT
ajst-390	120	1	[	[	X
ajst-390	120	2	2	2	X
ajst-390	120	3	]	]	PUNCT
ajst-390	120	4	byron	byron	PROPN
ajst-390	120	5	d	d	PROPN
ajst-390	120	6	k	k	PROPN
ajst-390	120	7	,	,	PUNCT
ajst-390	120	8	pikovsky	pikovsky	ADJ
ajst-390	120	9	a	a	X
ajst-390	120	10	,	,	PUNCT
ajst-390	120	11	woods	woods	PROPN
ajst-390	120	12	e.	e.	PROPN
ajst-390	120	13	text	text	PROPN
ajst-390	120	14	-	-	PUNCT
ajst-390	120	15	to	to	ADP
ajst-390	120	16	-	-	PUNCT
ajst-390	120	17	speech	speech	NOUN
ajst-390	120	18	for	for	ADP
ajst-390	120	19	digital	digital	ADJ
ajst-390	120	20	literature	literature	NOUN
ajst-390	120	21	,	,	PUNCT
ajst-390	120	22	us9183831[p	us9183831[p	PROPN
ajst-390	120	23	]	]	X
ajst-390	120	24	.	.	NOUN
ajst-390	120	25	2015	2015	NUM
ajst-390	120	26	.	.	PUNCT
ajst-390	121	1	[	[	X
ajst-390	121	2	3	3	NUM
ajst-390	121	3	]	]	X
ajst-390	121	4	schwardt	schwardt	PROPN
ajst-390	121	5	l.	l.	PROPN
ajst-390	121	6	c.	c.	PROPN
ajst-390	121	7	,	,	PUNCT
ajst-390	121	8	du	du	PROPN
ajst-390	121	9	preez	preez	PROPN
ajst-390	121	10	j.	j.	PROPN
ajst-390	121	11	a.	a.	PROPN
ajst-390	121	12	,	,	PUNCT
ajst-390	121	13	voice	voice	NOUN
ajst-390	121	14	conversion	conversion	NOUN
ajst-390	121	15	based	base	VERB
ajst-390	121	16	on	on	ADP
ajst-390	121	17	static	static	ADJ
ajst-390	121	18	speaker	speaker	NOUN
ajst-390	121	19	characteristics	characteristic	NOUN
ajst-390	121	20	ieee	ieee	PROPN
ajst-390	121	21	comsig-98	comsig-98	PROPN
ajst-390	121	22	,	,	PUNCT
ajst-390	121	23	cape	cape	NOUN
ajst-390	121	24	town	town	NOUN
ajst-390	121	25	,	,	PUNCT
ajst-390	121	26	september	september	PROPN
ajst-390	121	27	1998	1998	NUM
ajst-390	121	28	,	,	PUNCT
ajst-390	121	29	57~62	57~62	NUM
ajst-390	121	30	.	.	PUNCT
ajst-390	122	1	[	[	X
ajst-390	122	2	4	4	X
ajst-390	122	3	]	]	PUNCT
ajst-390	122	4	qi	qi	PROPN
ajst-390	122	5	yingyong	yingyong	NOUN
ajst-390	122	6	,	,	PUNCT
ajst-390	122	7	weinbery	weinbery	PROPN
ajst-390	122	8	b.	b.	PROPN
ajst-390	122	9	bi	bi	PROPN
ajst-390	122	10	ning	ning	PROPN
ajst-390	122	11	.	.	PUNCT
ajst-390	122	12	enhancement	enhancement	NOUN
ajst-390	122	13	of	of	ADP
ajst-390	122	14	female	female	ADJ
ajst-390	122	15	esophageal	esophageal	ADJ
ajst-390	122	16	and	and	CCONJ
ajst-390	122	17	tracheoesophageal	tracheoesophageal	NOUN
ajst-390	122	18	speech	speech	NOUN
ajst-390	122	19	,	,	PUNCT
ajst-390	122	20	j.	j.	PROPN
ajst-390	122	21	acoust	acoust	PROPN
ajst-390	122	22	.	.	PUNCT
ajst-390	123	1	soc	soc	PROPN
ajst-390	123	2	.	.	PUNCT
ajst-390	124	1	am	be	AUX
ajst-390	124	2	.	.	PUNCT
ajst-390	124	3	,	,	PUNCT
ajst-390	124	4	nov	nov	PROPN
ajst-390	124	5	.	.	PROPN
ajst-390	124	6	1998	1998	NUM
ajst-390	124	7	,	,	PUNCT
ajst-390	124	8	(	(	PUNCT
ajst-390	124	9	5	5	NUM
ajst-390	124	10	):	):	PUNCT
ajst-390	124	11	2461~2465	2461~2465	NUM
ajst-390	124	12	.	.	PUNCT
ajst-390	125	1	[	[	X
ajst-390	125	2	5	5	NUM
ajst-390	125	3	]	]	X
ajst-390	125	4	sundermann	sundermann	PROPN
ajst-390	125	5	d.	d.	PROPN
ajst-390	125	6	,	,	PUNCT
ajst-390	125	7	ney	ney	PROPN
ajst-390	125	8	h.	h.	PROPN
ajst-390	125	9	,	,	PUNCT
ajst-390	125	10	hoge	hoge	PROPN
ajst-390	125	11	h.	h.	PROPN
ajst-390	125	12	,	,	PUNCT
ajst-390	125	13	vtln	vtln	NOUN
ajst-390	125	14	-	-	PUNCT
ajst-390	125	15	based	base	VERB
ajst-390	125	16	crosslanguage	crosslanguage	NOUN
ajst-390	125	17	voice	voice	NOUN
ajst-390	125	18	conversion	conversion	NOUN
ajst-390	125	19	.	.	PUNCT
ajst-390	126	1	in	in	ADP
ajst-390	126	2	ieee	ieee	NOUN
ajst-390	126	3	automatic	automatic	ADJ
ajst-390	126	4	speech	speech	NOUN
ajst-390	126	5	recognition	recognition	NOUN
ajst-390	126	6	and	and	CCONJ
ajst-390	126	7	understanding	understanding	NOUN
ajst-390	126	8	workshop	workshop	NOUN
ajst-390	126	9	,	,	PUNCT
ajst-390	126	10	2003	2003	NUM
ajst-390	126	11	,	,	PUNCT
ajst-390	126	12	676~681	676~681	NUM
ajst-390	126	13	.	.	PUNCT
ajst-390	127	1	[	[	X
ajst-390	127	2	6	6	NUM
ajst-390	127	3	]	]	PUNCT
ajst-390	127	4	m.	m.	NOUN
ajst-390	127	5	abe	abe	PROPN
ajst-390	127	6	,	,	PUNCT
ajst-390	127	7	s.	s.	PROPN
ajst-390	127	8	nakamura	nakamura	PROPN
ajst-390	127	9	,	,	PUNCT
ajst-390	127	10	k.	k.	PROPN
ajst-390	127	11	shikano	shikano	PROPN
ajst-390	127	12	,	,	PUNCT
ajst-390	127	13	and	and	CCONJ
ajst-390	127	14	h.	h.	PROPN
ajst-390	127	15	kuwabara	kuwabara	PROPN
ajst-390	127	16	,	,	PUNCT
ajst-390	127	17	“	"	PUNCT
ajst-390	127	18	voice	voice	NOUN
ajst-390	127	19	conversion	conversion	NOUN
ajst-390	127	20	through	through	ADP
ajst-390	127	21	vector	vector	NOUN
ajst-390	127	22	quantization	quantization	NOUN
ajst-390	127	23	,	,	PUNCT
ajst-390	127	24	”	"	PUNCT
ajst-390	127	25	icassp	icassp	PROPN
ajst-390	127	26	,	,	PUNCT
ajst-390	127	27	1988:655658	1988:655658	NUM
ajst-390	127	28	.	.	PUNCT
ajst-390	128	1	[	[	X
ajst-390	128	2	7	7	X
ajst-390	128	3	]	]	X
ajst-390	128	4	m.	m.	NOUN
ajst-390	128	5	savic	savic	PROPN
ajst-390	128	6	,	,	PUNCT
ajst-390	128	7	and	and	CCONJ
ajst-390	128	8	i.	i.	PROPN
ajst-390	128	9	nam	nam	PROPN
ajst-390	128	10	,	,	PUNCT
ajst-390	128	11	“	"	PUNCT
ajst-390	128	12	voice	voice	NOUN
ajst-390	128	13	personality	personality	NOUN
ajst-390	128	14	transformation	transformation	NOUN
ajst-390	128	15	,	,	PUNCT
ajst-390	128	16	”	"	PUNCT
ajst-390	128	17	digital	digital	ADJ
ajst-390	128	18	signal	signal	NOUN
ajst-390	128	19	process	process	NOUN
ajst-390	128	20	,	,	PUNCT
ajst-390	128	21	no.1	no.1	NUM
ajst-390	128	22	,	,	PUNCT
ajst-390	128	23	pp.107	pp.107	PROPN
ajst-390	128	24	-	-	SYM
ajst-390	128	25	110	110	NUM
ajst-390	128	26	,	,	PUNCT
ajst-390	128	27	1991	1991	NUM
ajst-390	128	28	.	.	PUNCT
ajst-390	129	1	[	[	X
ajst-390	129	2	8	8	NUM
ajst-390	129	3	]	]	X
ajst-390	129	4	m.	m.	NOUN
ajst-390	129	5	narendranath	narendranath	NOUN
ajst-390	129	6	,	,	PUNCT
ajst-390	129	7	h.	h.	PROPN
ajst-390	129	8	a.	a.	PROPN
ajst-390	129	9	murthy	murthy	PROPN
ajst-390	129	10	,	,	PUNCT
ajst-390	129	11	s.	s.	PROPN
ajst-390	129	12	rajendran	rajendran	PROPN
ajst-390	129	13	,	,	PUNCT
ajst-390	129	14	and	and	CCONJ
ajst-390	129	15	b.	b.	PROPN
ajst-390	129	16	yegnanarayana	yegnanarayana	PROPN
ajst-390	129	17	,	,	PUNCT
ajst-390	129	18	“	"	PUNCT
ajst-390	129	19	transformation	transformation	NOUN
ajst-390	129	20	of	of	ADP
ajst-390	129	21	formants	formant	NOUN
ajst-390	129	22	for	for	ADP
ajst-390	129	23	voice	voice	NOUN
ajst-390	129	24	conversion	conversion	NOUN
ajst-390	129	25	using	use	VERB
ajst-390	129	26	artificial	artificial	ADJ
ajst-390	129	27	neural	neural	ADJ
ajst-390	129	28	networks	network	NOUN
ajst-390	129	29	,	,	PUNCT
ajst-390	129	30	”	"	PUNCT
ajst-390	129	31	speech	speech	NOUN
ajst-390	129	32	communication	communication	NOUN
ajst-390	129	33	,	,	PUNCT
ajst-390	129	34	vol.16	vol.16	PROPN
ajst-390	129	35	,	,	PUNCT
ajst-390	129	36	no.2	no.2	PROPN
ajst-390	129	37	,	,	PUNCT
ajst-390	129	38	pp.207	pp.207	NOUN
ajst-390	129	39	-	-	SYM
ajst-390	129	40	216	216	NUM
ajst-390	129	41	,	,	PUNCT
ajst-390	129	42	1995	1995	NUM
ajst-390	129	43	.	.	PUNCT
ajst-390	130	1	[	[	X
ajst-390	130	2	9	9	NUM
ajst-390	130	3	]	]	PUNCT
ajst-390	130	4	t.	t.	PROPN
ajst-390	130	5	watanabe	watanabe	PROPN
ajst-390	130	6	,	,	PUNCT
ajst-390	130	7	t.	t.	PROPN
ajst-390	130	8	murakami	murakami	NOUN
ajst-390	130	9	,	,	PUNCT
ajst-390	130	10	m.	m.	NOUN
ajst-390	130	11	namba	namba	PROPN
ajst-390	130	12	,	,	PUNCT
ajst-390	130	13	t.	t.	PROPN
ajst-390	130	14	hoya	hoya	PROPN
ajst-390	130	15	,	,	PUNCT
ajst-390	130	16	and	and	CCONJ
ajst-390	130	17	y.	y.	PROPN
ajst-390	130	18	ishida	ishida	PROPN
ajst-390	130	19	,	,	PUNCT
ajst-390	130	20	“	"	PUNCT
ajst-390	130	21	transformation	transformation	NOUN
ajst-390	130	22	of	of	ADP
ajst-390	130	23	spectral	spectral	ADJ
ajst-390	130	24	envelope	envelope	NOUN
ajst-390	130	25	for	for	ADP
ajst-390	130	26	voice	voice	NOUN
ajst-390	130	27	conversion	conversion	NOUN
ajst-390	130	28	based	base	VERB
ajst-390	130	29	on	on	ADP
ajst-390	130	30	radial	radial	ADJ
ajst-390	130	31	basis	basis	NOUN
ajst-390	130	32	function	function	NOUN
ajst-390	130	33	networks	network	NOUN
ajst-390	130	34	,	,	PUNCT
ajst-390	130	35	”	"	PUNCT
ajst-390	130	36	international	international	ADJ
ajst-390	130	37	conference	conference	NOUN
ajst-390	130	38	on	on	ADP
ajst-390	130	39	spoken	spoken	ADJ
ajst-390	130	40	language	language	NOUN
ajst-390	130	41	processing	processing	NOUN
ajst-390	130	42	,	,	PUNCT
ajst-390	130	43	pp.789	pp.789	NOUN
ajst-390	130	44	-	-	SYM
ajst-390	130	45	793	793	NUM
ajst-390	130	46	,	,	PUNCT
ajst-390	130	47	2002	2002	NUM
ajst-390	130	48	.	.	PUNCT
ajst-390	131	1	[	[	X
ajst-390	131	2	10	10	NUM
ajst-390	131	3	]	]	X
ajst-390	131	4	r.	r.	PROPN
ajst-390	131	5	c.	c.	PROPN
ajst-390	131	6	guido	guido	PROPN
ajst-390	131	7	,	,	PUNCT
ajst-390	131	8	l.	l.	PROPN
ajst-390	131	9	sasso	sasso	PROPN
ajst-390	131	10	.	.	PUNCT
ajst-390	132	1	vieira	vieira	PROPN
ajst-390	132	2	,	,	PUNCT
ajst-390	132	3	s.	s.	PROPN
ajst-390	132	4	barbon	barbon	PROPN
ajst-390	132	5	junior	junior	PROPN
ajst-390	132	6	,	,	PUNCT
ajst-390	132	7	f.	f.	PROPN
ajst-390	132	8	l.	l.	PROPN
ajst-390	132	9	sanchez	sanchez	PROPN
ajst-390	132	10	,	,	PUNCT
ajst-390	132	11	c.	c.	PROPN
ajst-390	132	12	dias	dias	PROPN
ajst-390	132	13	maciel	maciel	PROPN
ajst-390	132	14	,	,	PUNCT
ajst-390	132	15	e.	e.	PROPN
ajst-390	132	16	silva	silva	PROPN
ajst-390	132	17	fonseca	fonseca	PROPN
ajst-390	132	18	,	,	PUNCT
ajst-390	132	19	and	and	CCONJ
ajst-390	132	20	j.	j.	PROPN
ajst-390	132	21	carlos	carlos	PROPN
ajst-390	132	22	pereira	pereira	PROPN
ajst-390	132	23	,	,	PUNCT
ajst-390	132	24	“	"	PUNCT
ajst-390	132	25	a	a	DET
ajst-390	132	26	neural	neural	ADJ
ajst-390	132	27	-	-	PUNCT
ajst-390	132	28	wavelet	wavelet	NOUN
ajst-390	132	29	architecture	architecture	NOUN
ajst-390	132	30	for	for	ADP
ajst-390	132	31	voice	voice	NOUN
ajst-390	132	32	conversion	conversion	NOUN
ajst-390	132	33	,	,	PUNCT
ajst-390	132	34	”	"	PUNCT
ajst-390	132	35	neurocomputing	neurocomputing	NOUN
ajst-390	132	36	,	,	PUNCT
ajst-390	132	37	vol.71	vol.71	NOUN
ajst-390	132	38	,	,	PUNCT
ajst-390	132	39	no.1	no.1	NOUN
ajst-390	132	40	-	-	SYM
ajst-390	132	41	3	3	NUM
ajst-390	132	42	,	,	PUNCT
ajst-390	132	43	pp.174	pp.174	NOUN
ajst-390	132	44	-	-	PUNCT
ajst-390	132	45	180	180	NUM
ajst-390	132	46	,	,	PUNCT
ajst-390	132	47	dec.2007	dec.2007	ADJ
ajst-390	132	48	.	.	PUNCT
ajst-390	133	1	[	[	X
ajst-390	133	2	11	11	NUM
ajst-390	133	3	]	]	X
ajst-390	133	4	nirmal	nirmal	PROPN
ajst-390	133	5	j	j	PROPN
ajst-390	133	6	,	,	PUNCT
ajst-390	133	7	zaveri	zaveri	PROPN
ajst-390	133	8	m	m	PROPN
ajst-390	133	9	,	,	PUNCT
ajst-390	133	10	patnaik	patnaik	PROPN
ajst-390	133	11	s	s	PROPN
ajst-390	133	12	,	,	PUNCT
ajst-390	133	13	et	et	PROPN
ajst-390	133	14	al	al	PROPN
ajst-390	133	15	.	.	PUNCT
ajst-390	133	16	voice	voice	NOUN
ajst-390	133	17	conversion	conversion	NOUN
ajst-390	133	18	using	use	VERB
ajst-390	133	19	general	general	ADJ
ajst-390	133	20	regression	regression	NOUN
ajst-390	133	21	neural	neural	ADJ
ajst-390	133	22	network[j	network[j	PROPN
ajst-390	133	23	]	]	PUNCT
ajst-390	133	24	.	.	PUNCT
ajst-390	134	1	applied	apply	VERB
ajst-390	134	2	soft	soft	ADJ
ajst-390	134	3	computing	computing	NOUN
ajst-390	134	4	,	,	PUNCT
ajst-390	134	5	2014	2014	NUM
ajst-390	134	6	,	,	PUNCT
ajst-390	134	7	24(24):1–12	24(24):1–12	NUM
ajst-390	134	8	.	.	PUNCT
ajst-390	135	1	[	[	X
ajst-390	135	2	12	12	NUM
ajst-390	135	3	]	]	X
ajst-390	135	4	ghorbandoost	ghorbandoost	PROPN
ajst-390	135	5	m	m	PROPN
ajst-390	135	6	,	,	PUNCT
ajst-390	135	7	sayadiyan	sayadiyan	NOUN
ajst-390	135	8	a	a	X
ajst-390	135	9	,	,	PUNCT
ajst-390	135	10	ahangar	ahangar	NOUN
ajst-390	135	11	m	m	PROPN
ajst-390	135	12	,	,	PUNCT
ajst-390	135	13	et	et	PROPN
ajst-390	135	14	al	al	PROPN
ajst-390	135	15	.	.	PUNCT
ajst-390	135	16	voice	voice	PROPN
ajst-390	135	17	conversion	conversion	NOUN
ajst-390	135	18	based	base	VERB
ajst-390	135	19	on	on	ADP
ajst-390	135	20	feature	feature	NOUN
ajst-390	135	21	combination	combination	NOUN
ajst-390	135	22	with	with	ADP
ajst-390	135	23	limited	limited	ADJ
ajst-390	135	24	training	training	NOUN
ajst-390	135	25	data[j	data[j	NOUN
ajst-390	135	26	]	]	X
ajst-390	135	27	.	.	PUNCT
ajst-390	136	1	speech	speech	NOUN
ajst-390	136	2	communication	communication	NOUN
ajst-390	136	3	,	,	PUNCT
ajst-390	136	4	2015	2015	NUM
ajst-390	136	5	,	,	PUNCT
ajst-390	136	6	67(67):113	67(67):113	NUM
ajst-390	136	7	-	-	SYM
ajst-390	136	8	128	128	NUM
ajst-390	136	9	.	.	PUNCT