Text-to-Speech on LibriTTS zero-shot
4.3026UTMOSPeriodWave-Turbo-L
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| PeriodWave-Turbo-LSampling Steps=4, ODE Solver=Euler2024.08 | 4.3026 | 4.32 | 0.5 | 1.49 | 0.579 | 0.835 | — | — | |
| PeriodWave-Turbo-BSampling Steps=4, ODE Solver=Euler2024.08 | 4.2932 | 4.33 | 0.52 | 1.52 | 0.576 | 0.835 | — | — | |
| PeriodWave-Turbo-SSampling Steps=4, ODE Solver=Euler2024.08 | 4.2716 | 4.34 | 0.64 | 1.64 | 0.575 | 0.835 | — | — | |
| PeriodWave-L + FreeUSampling Steps=16, ODE Solver=Midpoint, FreeU=true2024.08 | 4.2643 | 4.29 | 0.57 | 1.6 | 0.56 | 0.833 | — | — | |
| PeriodWavesampling steps=16, zero-shot=true, FreeU=true2024.08 | 4.2621 | — | — | — | — | — | — | — | |
| PeriodWave-B + FreeUSampling Steps=16, ODE Solver=Midpoint, FreeU=true2024.08 | 4.2621 | 4.31 | 0.59 | 1.65 | 0.556 | 0.835 | — | — | |
| PeriodWave-LSampling Steps=16, ODE Solver=Midpoint2024.08 | 4.2464 | 4.27 | 0.59 | 1.65 | 0.557 | 0.821 | — | — | |
| PeriodWavesampling steps=16, zero-shot=true2024.08 | 4.2209 | — | — | — | — | — | — | — | |
| PeriodWave-BSampling Steps=16, ODE Solver=Midpoint2024.08 | 4.2209 | 4.3 | 0.59 | 1.65 | 0.553 | 0.837 | — | — | |
| BigVGANzero-shot=true, training_steps=5M2024.08 | 4.0424 | — | — | — | — | — | — | — | |
| BigVGAN2024.08 | 4.0424 | 4.28 | 0.59 | 1.6 | 0.576 | 0.836 | — | — | |
| BigVGAN-v22024.08 | 3.9894 | 4.24 | 0.54 | 1.53 | 0.579 | 0.835 | — | — | |
| BigVSANzero-shot=true, training_steps=10M2024.08 | 3.9732 | — | — | — | — | — | — | — | |
| BigVSAN2024.08 | 3.9732 | 4.26 | 0.58 | 1.63 | 0.573 | 0.835 | — | — | |
| CosyVoice 2Source ZS-TTS Model=CV2, Training Data=Real 10% + Synth 90%, DC=✓, OS=✓2026.03 | — | — | — | — | — | — | — | 61.8 | |
| Fish-SpeechSource ZS-TTS Model=FS, Training Data=Real 10% + Synth 90%, DC=-, OS=-2026.03 | — | — | — | — | — | — | 3.86 | — | |
| Fish-SpeechSource ZS-TTS Model=FS, Training Data=Real 10% + Synth 90%, DC=✓, OS=✓2026.03 | — | — | — | — | — | — | 3.92 | 70.8 | |
| Ground Truth (Recordings)Training Data=-, DC=-, OS=-2026.03 | — | — | — | — | — | — | 4.31 | — | |
| Real-OnlyTraining Data=Real 10%, DC=-, OS=-2026.03 | — | — | — | — | — | — | 3.58 | — | |
| Real-OnlyTraining Data=Real 100%, DC=-, OS=-2026.03 | — | — | — | — | — | — | 3.67 | — |