In This Environment, As That Speaker: A Text-Driven Framework for Multi-Attribute Speech Conversion
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909648263577600 |
|---|---|
| author | Jin, Jiawei Yang, Zhihan Zhou, Yixuan Wu, Zhiyong |
| author_facet | Jin, Jiawei Yang, Zhihan Zhou, Yixuan Wu, Zhiyong |
| contents | We propose TES-VC (Text-driven Environment and Speaker controllable Voice Conversion), a text-driven voice conversion framework with independent control of speaker timbre and environmental acoustics. TES-VC processes simultaneous text inputs for target voice and environment, accurately generating speech matching described timbre/environment while preserving source content. Trained on synthetic data with decoupled vocal/environment features via latent diffusion modeling, our method eliminates interference between attributes. The Retrieval-Based Timbre Control (RBTC) module enables precise manipulation using abstract descriptions without paired data. Experiments confirm TES-VC effectively generates contextually appropriate speech in both timbre and environment with high content retention and superior controllability which demonstrates its potential for widespread applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_07036 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | In This Environment, As That Speaker: A Text-Driven Framework for Multi-Attribute Speech Conversion Jin, Jiawei Yang, Zhihan Zhou, Yixuan Wu, Zhiyong Sound Audio and Speech Processing We propose TES-VC (Text-driven Environment and Speaker controllable Voice Conversion), a text-driven voice conversion framework with independent control of speaker timbre and environmental acoustics. TES-VC processes simultaneous text inputs for target voice and environment, accurately generating speech matching described timbre/environment while preserving source content. Trained on synthetic data with decoupled vocal/environment features via latent diffusion modeling, our method eliminates interference between attributes. The Retrieval-Based Timbre Control (RBTC) module enables precise manipulation using abstract descriptions without paired data. Experiments confirm TES-VC effectively generates contextually appropriate speech in both timbre and environment with high content retention and superior controllability which demonstrates its potential for widespread applications. |
| title | In This Environment, As That Speaker: A Text-Driven Framework for Multi-Attribute Speech Conversion |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2506.07036 |