U2-VC: one-shot voice conversion using two-level nested U-structure
Abstract Voice conversion is to transform a source speaker to the target one, while keeping the linguistic content unchanged. Recently, one-shot voice conversion gradually becomes a hot topic for its potentially wide range of applications, where it has the capability to convert the voice from any so...
Guardado en:
Autores principales: | , , , , |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
SpringerOpen
2021
|
Materias: | |
Acceso en línea: | https://doaj.org/article/9774f3e85b684d4c8afe408040baa9c9 |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:9774f3e85b684d4c8afe408040baa9c9 |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:9774f3e85b684d4c8afe408040baa9c92021-11-28T12:27:53ZU2-VC: one-shot voice conversion using two-level nested U-structure10.1186/s13636-021-00226-31687-4722https://doaj.org/article/9774f3e85b684d4c8afe408040baa9c92021-11-01T00:00:00Zhttps://doi.org/10.1186/s13636-021-00226-3https://doaj.org/toc/1687-4722Abstract Voice conversion is to transform a source speaker to the target one, while keeping the linguistic content unchanged. Recently, one-shot voice conversion gradually becomes a hot topic for its potentially wide range of applications, where it has the capability to convert the voice from any source speaker to any other target speaker even when both the source speaker and the target speaker are unseen during training. Although a great progress has been made in one-shot voice conversion, the naturalness of the converted speech remains a challenging problem. To further improve the naturalness of the converted speech, this paper proposes a two-level nested U-structure (U2-Net) voice conversion algorithm called U2-VC. The U2-Net can extract both local feature and multi-scale feature of log-mel spectrogram, which can help to learn the time-frequency structures of the source speech and the target speech. Moreover, we adopt sandwich adaptive instance normalization (SaAdaIN) in decoder for speaker identity transformation to retain more content information of the source speech while maintaining the speaker similarity between the converted speech and the target speech. Experiments on VCTK dataset show that U2-VC outperforms many SOTA approaches including AGAIN-VC and AdaIN-VC in terms of both objective and subjective measurements.Fangkun LiuHui WangRenhua PengChengshi ZhengXiaodong LiSpringerOpenarticleVoice conversionU2-NetSandwich adaptive instance normalizationAcoustics. SoundQC221-246Electronic computers. Computer scienceQA75.5-76.95ENEURASIP Journal on Audio, Speech, and Music Processing, Vol 2021, Iss 1, Pp 1-15 (2021) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
Voice conversion U2-Net Sandwich adaptive instance normalization Acoustics. Sound QC221-246 Electronic computers. Computer science QA75.5-76.95 |
spellingShingle |
Voice conversion U2-Net Sandwich adaptive instance normalization Acoustics. Sound QC221-246 Electronic computers. Computer science QA75.5-76.95 Fangkun Liu Hui Wang Renhua Peng Chengshi Zheng Xiaodong Li U2-VC: one-shot voice conversion using two-level nested U-structure |
description |
Abstract Voice conversion is to transform a source speaker to the target one, while keeping the linguistic content unchanged. Recently, one-shot voice conversion gradually becomes a hot topic for its potentially wide range of applications, where it has the capability to convert the voice from any source speaker to any other target speaker even when both the source speaker and the target speaker are unseen during training. Although a great progress has been made in one-shot voice conversion, the naturalness of the converted speech remains a challenging problem. To further improve the naturalness of the converted speech, this paper proposes a two-level nested U-structure (U2-Net) voice conversion algorithm called U2-VC. The U2-Net can extract both local feature and multi-scale feature of log-mel spectrogram, which can help to learn the time-frequency structures of the source speech and the target speech. Moreover, we adopt sandwich adaptive instance normalization (SaAdaIN) in decoder for speaker identity transformation to retain more content information of the source speech while maintaining the speaker similarity between the converted speech and the target speech. Experiments on VCTK dataset show that U2-VC outperforms many SOTA approaches including AGAIN-VC and AdaIN-VC in terms of both objective and subjective measurements. |
format |
article |
author |
Fangkun Liu Hui Wang Renhua Peng Chengshi Zheng Xiaodong Li |
author_facet |
Fangkun Liu Hui Wang Renhua Peng Chengshi Zheng Xiaodong Li |
author_sort |
Fangkun Liu |
title |
U2-VC: one-shot voice conversion using two-level nested U-structure |
title_short |
U2-VC: one-shot voice conversion using two-level nested U-structure |
title_full |
U2-VC: one-shot voice conversion using two-level nested U-structure |
title_fullStr |
U2-VC: one-shot voice conversion using two-level nested U-structure |
title_full_unstemmed |
U2-VC: one-shot voice conversion using two-level nested U-structure |
title_sort |
u2-vc: one-shot voice conversion using two-level nested u-structure |
publisher |
SpringerOpen |
publishDate |
2021 |
url |
https://doaj.org/article/9774f3e85b684d4c8afe408040baa9c9 |
work_keys_str_mv |
AT fangkunliu u2vconeshotvoiceconversionusingtwolevelnestedustructure AT huiwang u2vconeshotvoiceconversionusingtwolevelnestedustructure AT renhuapeng u2vconeshotvoiceconversionusingtwolevelnestedustructure AT chengshizheng u2vconeshotvoiceconversionusingtwolevelnestedustructure AT xiaodongli u2vconeshotvoiceconversionusingtwolevelnestedustructure |
_version_ |
1718407971261644800 |