Voice Conversion on LibriTTS test-clean test-other and VCTK (merged)
3.668WERSource Speech
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Source SpeechMode=Original2026.06 | 3.668 | 0.09 | 1 | — | — | — | 3.897 | 4.253 | |
| ProsoCodecSetting=Zero-shot2026.06 | 4.451 | 0.565 | 0.167 | 4.167 | 0.428 | 3.852 | 3.853 | 4 | |
| VevoSetting=Zero-shot2026.06 | 4.826 | 0.478 | 0.243 | 4.13 | 0.464 | 3.506 | 3.8 | 4.056 | |
| Seed-VCSetting=Zero-shot2026.06 | 5.078 | 0.531 | 0.239 | 4.21 | 0.473 | 3.463 | 3.712 | 3.611 | |
| FACodecSetting=Zero-shot2026.06 | 5.454 | 0.354 | 0.347 | 3.247 | 0.455 | 3.364 | 3.544 | 2.302 | |
| HierSpeech++Setting=Zero-shot2026.06 | 7.568 | 0.381 | 0.312 | 3.494 | 0.461 | 3.389 | 3.869 | 2.679 | |
| UniAudioSetting=Zero-shot2026.06 | 9.076 | 0.276 | 0.178 | 2.809 | 0.47 | 2.475 | 3.631 | 3.006 | |
| DDDM-VCSetting=Zero-shot2026.06 | 9.186 | 0.297 | 0.296 | 3.309 | 0.472 | 3.012 | 3.624 | 2.006 |