Text-to-Speech on VCTK (test)
4.4MOSReference
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ReferenceEvaluation Mode=Parallel2022.05 | 4.4 | — | — | — | |
| Reference(voc.)Evaluation Mode=Parallel, Vocoder=HiFi-GAN (V1)2022.05 | 4.37 | 4.3 | 0.96 | 0.05 | |
| GenerSpeechEvaluation Mode=Parallel, Batch size=1, Vocoder=HiFi-GAN (V1)2022.05 | 4.06 | 4.01 | 0.88 | 0.35 | |
| FG-TransformerTTSEvaluation Mode=Parallel, Batch size=1, Vocoder=HiFi-GAN (V1)2022.05 | 3.95 | 3.9 | 0.86 | 0.3 | |
| MellotronEvaluation Mode=Parallel, Batch size=1, Vocoder=HiFi-GAN (V1)2022.05 | 3.91 | 3.88 | 0.74 | 0.32 | |
| Meta-StyleSpeechEvaluation Mode=Parallel, Batch size=1, Vocoder=HiFi-GAN (V1)2022.05 | 3.9 | 3.95 | 0.83 | 0.38 | |
| StylerEvaluation Mode=Parallel, Batch size=1, Vocoder=HiFi-GAN (V1)2022.05 | 3.89 | 3.82 | 0.76 | 0.38 | |
| Expressive FS2Evaluation Mode=Parallel, Batch size=1, Vocoder=HiFi-GAN (V1)2022.05 | 3.85 | 3.87 | 0.85 | 0.41 |