| Paper | Code | Mean Opinion Score | ModelName | ReleaseDate |
|---|---|---|---|---|
| BDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis | ✓ Link | 4.48 | BDDM vocoder | 2022-03-25 |
| DiffWave: A Versatile Diffusion Model for Audio Synthesis | ✓ Link | 4.44 | DiffWave LARGE | 2020-09-21 |
| Neural HMMs are all you need (for high-quality attention-free TTS) | ✓ Link | 3.24 | Neural HMM | 2021-08-30 |
| Neural HMMs are all you need (for high-quality attention-free TTS) | ✓ Link | 2.68 | Neural HMM Ablation with 1 state per phone | 2021-08-30 |