Text-to-Speech on Biaobei (test)
0DTWGT
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GT2023.05 | 0 | 4.59 | 0 | — | — | |
| GT(voc.)2023.05 | 0 | 4.43 | 0 | — | — | |
| VB + CLAPSpeechnumber of parameters=41.54M2023.05 | 13.48 | 4.22 | 37.07 | — | — | |
| VB + BERTnumber of parameters=132.69M2023.05 | 13.63 | 3.91 | 38.41 | — | — | |
| VB + A³Tnumber of parameters=59.73M2023.05 | 13.97 | 4.04 | 39.15 | — | — | |
| VBnumber of parameters=23.02M2023.05 | 14.22 | 3.75 | 40.31 | — | — | |
| PB + CLAPSpeechnumber of parameters=30.51M2023.05 | 16.04 | 3.62 | 27.6 | — | — | |
| PB + BERTnumber of parameters=109.48M2023.05 | 16.79 | 3.43 | 28.06 | — | — | |
| PB + A³Tnumber of parameters=48.25M2023.05 | 17.18 | 3.51 | 28.44 | — | — | |
| PBnumber of parameters=11.99M2023.05 | 18.01 | 3.37 | 28.79 | — | — | |
| Dict-TTSVocoder=HiFi-GAN (V3)2022.06 | 52.4 | — | — | — | — | |
| Phoneme (pypinyin)Vocoder=HiFi-GAN (V3)2022.06 | 52.6 | — | — | — | — | |
| CharacterVocoder=HiFi-GAN (V3)2022.06 | 53.1 | — | — | — | — | |
| Phoneme (G2PM)Vocoder=HiFi-GAN (V3)2022.06 | 53.1 | — | — | — | — | |
| NLRVocoder=HiFi-GAN (V3)2022.06 | 53.3 | — | — | — | — | |
| BERT EmbeddingVocoder=HiFi-GAN (V3)2022.06 | 55 | — | — | — | — |