Voice Conversion on VCTK
0WERGround Truth
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ground Truth2026.01 | 0 | 0 | — | — | — | — | — | 4.08 | — | — | — | — | |
| GT2026.06 | 0 | 0 | — | — | — | — | — | 4.1 | 100 | — | — | — | |
| TiCodecpoor timbre transfer=true2026.01 | 0.5 | 0.2 | — | — | — | — | 5.4 | 3.32 | 68 | 0.77 | — | — | |
| LinearVC2026.01 | 0.6 | 0.2 | — | — | — | — | 29.7 | 3.94 | 73.4 | 0.62 | — | — | |
| FreeVC2026.01 | 0.6 | 0.3 | — | — | — | — | 29 | 3.99 | 74.5 | 0.67 | — | — | |
| FACodecpoor timbre transfer=true2026.01 | 0.7 | 0.4 | — | — | — | — | 15.2 | 3.77 | 62.6 | 0.79 | — | — | |
| Kanadetoken rate=25Hz2026.01 | 0.7 | 0.3 | — | — | — | — | 30.7 | 4.16 | 77.1 | 0.71 | — | — | |
| CleanCodec@62.5Token Rate (t/s)=62.52026.06 | 0.8 | 0.3 | — | — | — | — | — | 3.93 | 70 | — | — | — | |
| CleanCodec@31.25Token Rate (t/s)=31.252026.06 | 0.9 | 0.4 | — | — | — | — | — | 3.95 | 76 | — | — | — | |
| CosyVoice 22026.01 | 1.1 | 0.5 | — | — | — | — | 31 | 4.11 | 76 | 0.64 | — | — | |
| BiCodecpoor timbre transfer=true2026.01 | 1.2 | 0.6 | — | — | — | — | 18.5 | 3.84 | 71.4 | 0.61 | — | — | |
| BiCodecToken Rate (t/s)=502026.06 | 1.3 | 0.7 | — | — | — | — | — | 3.87 | 47 | — | — | — | |
| CleanCodec@12.5Token Rate (t/s)=12.52026.06 | 1.4 | 0.7 | — | — | — | — | — | 4.09 | 81 | — | — | — | |
| Kanadetoken rate=12.5Hz2026.01 | 1.6 | 0.7 | — | — | — | — | 32 | 4.17 | 77.6 | 0.64 | — | — | |
| KanadeToken Rate (t/s)=12.52026.06 | 2.2 | 1.1 | — | — | — | — | — | 4.21 | 52 | — | — | — | |
| Seq2seq-VC2022.12 | 2.9 | 1.2 | 20.95 | 1.214 | 0.072 | 0.028 | 10.9 | — | — | — | — | — | |
| QuarkAudio2025.12 | 3.02 | — | — | — | — | — | — | — | 50 | — | 3.51 | 4.51 | |
| Metis-VC2025.12 | 4.49 | — | — | — | — | — | — | — | 50 | — | 3.48 | 4.46 | |
| SR2022.12 | 6.6 | 3.3 | 14 | 1.214 | 0.071 | 0.025 | 1.8 | — | — | — | — | — | |
| DISSCvariant=Rhythm2022.12 | 10.9 | 5.4 | 10.58 | 0.832 | 0.056 | 0.023 | 1.7 | — | — | — | — | — | |
| DISSCvariant=Both2022.12 | 13 | 6.9 | 10.53 | 0.832 | 0.056 | 0.023 | 1.7 | — | — | — | — | — | |
| DualCodecpoor timbre transfer=true, serious content degradation=true2026.01 | 21.5 | 12.9 | — | — | — | — | 6.8 | 2.51 | 52 | 0.54 | — | — | |
| PASTpoor timbre transfer=true, serious content degradation=true2026.01 | 22.9 | 15.1 | — | — | — | — | 8.2 | 1.84 | 23.3 | 0.2 | — | — | |
| AutoPST2022.12 | 40.6 | 26.7 | 21.9 | 1.379 | 0.123 | 0.037 | 24.1 | — | — | — | — | — | |
| AutoVC2022.12 | 71.3 | 47.1 | 17.68 | 1.214 | 0.072 | 0.028 | 7.5 | — | — | — | — | — | |
| STpoor timbre transfer=true, serious content degradation=true2026.01 | 74.7 | 61.7 | — | — | — | — | 10.6 | 1.54 | 35 | 0.2 | — | — | |
| Mimiserious content degradation=true2026.01 | 120.3 | 86.8 | — | — | — | — | 38.5 | 3.09 | 81.7 | 0.21 | — | — |