Text-to-Speech on EN (N-MOS, S-MOS)
4.02N-MOSGround Truth
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Ground Truthzero-shot=true2026.05 | 4.02 | 4.53 | |
| SemaVoicezero-shot=true2026.05 | 3.98 | 3.89 | |
| CosyVoice 2zero-shot=true2026.05 | 3.96 | 3.78 | |
| SemaVoice-Emiliazero-shot=true2026.05 | 3.86 | 3.69 | |
| IndexTTS 2zero-shot=true2026.05 | 3.75 | 3.93 |