Speech Synthesis on LibriTTS 24,000 Hz (test)
4.28MOSGT
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GT2024.08 | 4.28 | — | — | — | |
| PeriodWave-Turbo-LParams (M)=70.24, Steps=4, Sampler=Euler2024.08 | 4.23 | 0.52 | 24.77 | — | |
| PeriodWave-Turbo-SParams (M)=7.57, Steps=4, Sampler=Euler2024.08 | 4.22 | 2.58 | 42.26 | — | |
| BigVSANParams (M)=112.42024.08 | 4.21 | 0.25 | 61.03 | — | |
| PeriodWave-B + FreeUParams (M)=29.8, Steps=16, Sampler=Midpoint2024.08 | 4.21 | 0.21 | 4.62 | — | |
| UnivNet-c32Params (M)=14.872024.08 | 4.2 | 13.53 | 267.21 | — | |
| PeriodWave-L + FreeUParams (M)=70.24, Steps=16, Sampler=Midpoint2024.08 | 4.2 | 0.09 | 2.77 | — | |
| PeriodWave-Turbo-BParams (M)=29.8, Steps=4, Sampler=Euler2024.08 | 4.2 | 1.12 | 35.01 | — | |
| BigVGAN-v2Params (M)=112.42024.08 | 4.19 | 0.24 | 112.18 | — | |
| BigVGANParams (M)=112.42024.08 | 4.18 | 0.24 | 55.9 | — | |
| VocosParams (M)=13.532024.08 | 4.16 | 85.31 | 436.37 | — |