Whisper-to-Normal speech conversion on WTIMIT English (test)
3.75UTMOSWhispEar-Scaled
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| WhispEar-Scaled2026.03 | 3.75 | 0.577 | 22.44 | 3.76 | 4.38 | 0.513 | — | — | — | |
| WESPER2026.03 | 3.63 | 0.064 | 42.01 | 3.91 | 4.37 | 0.373 | — | — | — | |
| WESPERPre-trained=true2025.11 | 3.496 | — | — | 3.202 | — | — | 30.724 | 0.531 | 0.543 | |
| WhispEar2026.03 | 3.45 | 0.554 | 30.74 | 3.7 | 4.15 | 0.485 | — | — | — | |
| Seed-VC2025.11 | 3.321 | — | — | 3.108 | — | — | 16.709 | 0.839 | 0.926 | |
| WhisperVC2025.11 | 3.276 | — | — | 2.894 | — | — | 11.389 | 0.594 | 0.724 | |
| FreeVC2025.11 | 2.643 | — | — | 2.657 | — | — | 26.534 | 0.702 | 0.885 | |
| CosyVoice22026.03 | 2.51 | 0.409 | 36.69 | 2.54 | 3.59 | 0.052 | — | — | — | |
| DistillW2N2026.03 | 1.97 | 0.189 | 50.99 | 3.61 | 2.25 | 0.113 | — | — | — | |
| MaskCycleGAN2026.03 | 1.97 | 0.237 | 48.53 | 3.24 | 3.32 | 0.023 | — | — | — | |
| DistillW2NPre-trained=true2025.11 | 1.894 | — | — | 3.012 | — | — | 36.028 | 0.642 | 0.806 | |
| Whispered Speech2026.03 | 1.46 | 0.333 | 12.36 | 1.81 | 1.66 | 0.029 | — | — | — |