TransVIP: Speech to Speech Translation System with Voice and Isochrony Preservation
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917823657279488 |
|---|---|
| author | Le, Chenyang Qian, Yao Wang, Dongmei Zhou, Long Liu, Shujie Wang, Xiaofei Yousefi, Midia Qian, Yanmin Li, Jinyu Zhao, Sheng Zeng, Michael |
| author_facet | Le, Chenyang Qian, Yao Wang, Dongmei Zhou, Long Liu, Shujie Wang, Xiaofei Yousefi, Midia Qian, Yanmin Li, Jinyu Zhao, Sheng Zeng, Michael |
| contents | There is a rising interest and trend in research towards directly translating speech from one language to another, known as end-to-end speech-to-speech translation. However, most end-to-end models struggle to outperform cascade models, i.e., a pipeline framework by concatenating speech recognition, machine translation and text-to-speech models. The primary challenges stem from the inherent complexities involved in direct translation tasks and the scarcity of data. In this study, we introduce a novel model framework TransVIP that leverages diverse datasets in a cascade fashion yet facilitates end-to-end inference through joint probability. Furthermore, we propose two separated encoders to preserve the speaker's voice characteristics and isochrony from the source speech during the translation process, making it highly suitable for scenarios such as video dubbing. Our experiments on the French-English language pair demonstrate that our model outperforms the current state-of-the-art speech-to-speech translation model. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_17809 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | TransVIP: Speech to Speech Translation System with Voice and Isochrony Preservation Le, Chenyang Qian, Yao Wang, Dongmei Zhou, Long Liu, Shujie Wang, Xiaofei Yousefi, Midia Qian, Yanmin Li, Jinyu Zhao, Sheng Zeng, Michael Computation and Language Artificial Intelligence Sound Audio and Speech Processing There is a rising interest and trend in research towards directly translating speech from one language to another, known as end-to-end speech-to-speech translation. However, most end-to-end models struggle to outperform cascade models, i.e., a pipeline framework by concatenating speech recognition, machine translation and text-to-speech models. The primary challenges stem from the inherent complexities involved in direct translation tasks and the scarcity of data. In this study, we introduce a novel model framework TransVIP that leverages diverse datasets in a cascade fashion yet facilitates end-to-end inference through joint probability. Furthermore, we propose two separated encoders to preserve the speaker's voice characteristics and isochrony from the source speech during the translation process, making it highly suitable for scenarios such as video dubbing. Our experiments on the French-English language pair demonstrate that our model outperforms the current state-of-the-art speech-to-speech translation model. |
| title | TransVIP: Speech to Speech Translation System with Voice and Isochrony Preservation |
| topic | Computation and Language Artificial Intelligence Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2405.17809 |