Text-to-Speech on IndicTTS Database Hindi
4.33MOSGround Truth
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Ground Truth2022.11 | 4.33 | — | — | 0.104 | |
| AI4Bharat-TTSAcoustic Model=FastPitch, Vocoder=HiFiGAN2022.11 | 4 | 8.73 | 0.21 | 0.094 | |
| Tacotron2+WaveGlowAcoustic Model=Tacotron2, Vocoder=WaveGlow2022.11 | 3.7 | 12.41 | 0.23 | 0.127 | |
| GlowTTS+HiFiGANAcoustic Model=GlowTTS, Vocoder=HiFiGAN2022.11 | 3.16 | 8.5 | 0.24 | 0.1 |