Emotional Text-to-Speech on IEMOCAP (Mid-mismatch)
0.913E-SIMCoCoEmo
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| CoCoEmoModel=CosyVoice2, Steering strength α=62026.02 | 0.913 | 0.576 | 0.884 | 6.61 | |
| CoCoEmoModel=CosyVoice2, Steering strength α=32026.02 | 0.901 | 0.446 | 0.888 | 6.8 | |
| InstructionModel=CosyVoice22026.02 | 0.896 | 0.566 | 0.877 | 2.26 | |
| CoCoEmoModel=IndexTTS2, Steering strength α=32026.02 | 0.888 | 0.456 | 0.882 | 5.42 | |
| CoCoEmoModel=IndexTTS2, Steering strength α=62026.02 | 0.887 | 0.626 | 0.879 | 7.95 | |
| No-steerModel=CosyVoice22026.02 | 0.882 | 0.328 | 0.887 | 7.06 | |
| Emo vectorModel=IndexTTS22026.02 | 0.88 | 0.632 | 0.855 | 5.39 | |
| No-steerModel=IndexTTS22026.02 | 0.875 | 0.326 | 0.881 | 5.31 |