Speech Synthesis On Ljspeech
Metrics
Mean Opinion Score
Results
Performance results of various models on this benchmark
Model Name | Mean Opinion Score | Paper Title | Repository |
|---|---|---|---|
| Neural HMM Ablation with 1 state per phone | 2.68 | Neural HMMs are all you need (for high-quality attention-free TTS) | |
| Neural HMM | 3.24 | Neural HMMs are all you need (for high-quality attention-free TTS) | |
| DiffWave LARGE | 4.44 | DiffWave: A Versatile Diffusion Model for Audio Synthesis | |
| BDDM vocoder | 4.48 | BDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis |
0 of 4 row(s) selected.