End-to-End Speech Synthesis on End-to-End Speech Synthesis Tacotron2 pipeline
4.23MOSGround Truth
Evaluation Results
| Method | Links | |
|---|---|---|
| Ground Truth2020.10 | 4.23 | |
| HiFi-GAN V1Fine-tuning=true2020.10 | 4.18 | |
| HiFi-GAN V2Fine-tuning=true2020.10 | 4.12 | |
| HiFi-GAN V3Fine-tuning=true2020.10 | 4.02 | |
| HiFi-GAN V1Fine-tuning=false2020.10 | 3.91 | |
| HiFi-GAN V3Fine-tuning=false2020.10 | 3.89 | |
| HiFi-GAN V2Fine-tuning=false2020.10 | 3.88 | |
| WaveGlowFine-tuning=false2020.10 | 3.69 | |
| WaveGlowFine-tuning=true2020.10 | 3.66 |