Audio Synthesis on LJSpeech (unseen)
0.1102MAEBigVGAN
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| BigVGANZero-shot=true2022.06 | 0.1102 | 0.8554 | 4.112 | 0.7164 | 0.0957 | 96.42 | |
| BigVGAN-baseZero-shot=true2022.06 | 0.1747 | 0.9121 | 3.741 | 0.8626 | 0.1164 | 95.48 | |
| BigVGAN-baseZero-shot=true, Ablation=w/o filter2022.06 | 0.2015 | 0.9395 | 3.662 | 0.9715 | 0.1169 | 95.48 | |
| BigVGAN-baseZero-shot=true, Ablation=w/o filter & snake2022.06 | 0.2263 | 0.9946 | 3.521 | 1.132 | 0.1299 | 94.79 | |
| HiFi-GAN (V1)Zero-shot=true2022.06 | 0.3008 | 1.095 | 3.21 | 1.737 | 0.1347 | 94.56 | |
| UnivNet-c32Zero-shot=true2022.06 | 0.3356 | 1.0429 | 3.384 | 1.1356 | 0.123 | 95.03 | |
| UnivNet-c32+Zero-shot=true, Pre-training=train-clean-360 subset2022.06 | 0.3418 | 1.0613 | 3.425 | 1.1903 | 0.121 | 95.19 | |
| WaveFlow-128Zero-shot=true2022.06 | 0.3674 | 1.2402 | 3.072 | 2.7217 | 0.117 | 95.6 | |
| WaveGlow-256Zero-shot=true2022.06 | 0.4933 | 1.2893 | 3.352 | 2.9921 | 0.1182 | 95.61 | |
| SC-WaveRNNZero-shot=true2022.06 | 1.0115 | 2.6994 | 1.233 | 4.8464 | 0.3907 | 74.04 |