LinearVC: Linear transformations of self-supervised features through the lens of voice conversion
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908389587550208 |
|---|---|
| author | Kamper, Herman van Niekerk, Benjamin Zaïdi, Julian Carbonneau, Marc-André |
| author_facet | Kamper, Herman van Niekerk, Benjamin Zaïdi, Julian Carbonneau, Marc-André |
| contents | We introduce LinearVC, a simple voice conversion method that sheds light on the structure of self-supervised representations. First, we show that simple linear transformations of self-supervised features effectively convert voices. Next, we probe the geometry of the feature space by constraining the set of allowed transformations. We find that just rotating the features is sufficient for high-quality voice conversion. This suggests that content information is embedded in a low-dimensional subspace which can be linearly transformed to produce a target voice. To validate this hypothesis, we finally propose a method that explicitly factorizes content and speaker information using singular value decomposition; the resulting linear projection with a rank of just 100 gives competitive conversion results. Our work has implications for both practical voice conversion and a broader understanding of self-supervised speech representations. Samples and code: https://www.kamperh.com/linearvc/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_01510 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LinearVC: Linear transformations of self-supervised features through the lens of voice conversion Kamper, Herman van Niekerk, Benjamin Zaïdi, Julian Carbonneau, Marc-André Audio and Speech Processing Computation and Language We introduce LinearVC, a simple voice conversion method that sheds light on the structure of self-supervised representations. First, we show that simple linear transformations of self-supervised features effectively convert voices. Next, we probe the geometry of the feature space by constraining the set of allowed transformations. We find that just rotating the features is sufficient for high-quality voice conversion. This suggests that content information is embedded in a low-dimensional subspace which can be linearly transformed to produce a target voice. To validate this hypothesis, we finally propose a method that explicitly factorizes content and speaker information using singular value decomposition; the resulting linear projection with a rank of just 100 gives competitive conversion results. Our work has implications for both practical voice conversion and a broader understanding of self-supervised speech representations. Samples and code: https://www.kamperh.com/linearvc/. |
| title | LinearVC: Linear transformations of self-supervised features through the lens of voice conversion |
| topic | Audio and Speech Processing Computation and Language |
| url | https://arxiv.org/abs/2506.01510 |