| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LJ Speech (test) | MOS4.54 | 36 | 4d ago | ||
| LibriTTS (test) | EVA-GAN | MOS4.9134 | 17 | 4d ago | |
| LJSpeech | MOS4.45 | 12 | 4d ago | ||
| Speech Industrial Setting | DAC | MOS Prediction4.29 | 11 | 4d ago | |
| Speech Academic Setting | BigCodec | MOS Prediction3.65 | 11 | 4d ago | |
| LibriTTS 24,000 Hz (test) | MOS4.28 | 11 | 4d ago | ||
| VCTK (OD) | PESQ4.5 | 9 | 4d ago | ||
| LibriTTS (ID) | PESQ4.5 | 9 | 4d ago | ||
| Manchu Speech Dataset (test) | MOS4.68 | 8 | 4d ago | ||
| Mozilla Common Voice (MCV) 8.0 (test) | SC-WaveRNN | English Quality Score15.155 | 7 | 4d ago | |
| Speech and 3D gesture (test) | Speech MOS4.35 | 6 | 4d ago | ||
| LJSpeech (test) | FastSpeech 2 + HiFiGAN | RTF0.011 | 6 | 4d ago | |
| EmoV-DB (test) | STOI0.93 | 3 | 4d ago | ||
| LibriSpeech 360 Clean (test) | STOI0.93 | 3 | 4d ago | ||
| VCTK v0.92 (test) | GLA-Grad++ | PESQ3.772 | 3 | 4d ago | |
| LJSpeech 26 (test) | GLA-Grad++ | PESQ3.807 | 3 | 4d ago | |
| VCTK | VITS | CMOS-N Score0.45 | 2 | 4d ago | |
| LJSpeech | NaturalSpeech | CMOS-N Score1.07 | 2 | 4d ago | |
| Internal US English female (evaluation) | Our Model | CMOS0.048 | 2 | 4d ago | |
| LibriSpeech | Vall-E | CMOS-N Score0.67 | 1 | 4d ago | |
| Manchu Laboratory Setting 20 Native Speakers | - | - | 0 | 4d ago |
