Text-to-Speech on IndicTTS Database Odia
4.77MOSGround Truth
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Ground Truth2022.11 | 4.77 | — | — | — | |
| AI4Bharat-TTSAcoustic Model=FastPitch, Vocoder=HiFiGAN2022.11 | 4.19 | 8.9 | 0.22 | — | |
| GlowTTS+HiFiGANAcoustic Model=GlowTTS, Vocoder=HiFiGAN2022.11 | 3.56 | 11.07 | 0.24 | — | |
| Tacotron2+WaveGlowAcoustic Model=Tacotron2, Vocoder=WaveGlow2022.11 | 3.04 | 15.72 | 0.33 | — |