Visual Text-to-Speech on GRID
10.9WERStyleDubber
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| StyleDubber2025.09 | 10.9 | 51.4 | 3.74 | 12.85 | 7.81 | 7.91 | 6.33 | 8.77 | |
| GT2025.09 | 12.9 | 100 | 4.04 | 0 | 0 | 0 | 5.49 | 8.56 | |
| GT-vocoder2025.09 | 13.4 | 64.1 | 3.37 | 5.02 | 3.88 | 3.89 | 6.83 | 8.16 | |
| VSSFlow-SModel Size=Small2025.09 | 15.4 | 52.1 | 3.29 | 8.78 | 5.77 | 5.78 | 6.84 | 8.14 | |
| VSSFlow-MModel Size=Medium2025.09 | 15.8 | 52.3 | 3.25 | 8.45 | 5.61 | 5.61 | 6.8 | 8.19 | |
| EmoDubber2025.09 | 15.9 | 50.5 | 3.98 | 15.52 | 5.89 | 9.83 | 3.48 | 10.35 | |
| VSSFlow-LModel Size=Large2025.09 | 16.2 | 51.4 | 3.3 | 8.62 | 5.73 | 5.73 | 6.81 | 8.18 | |
| E2-TTS2025.09 | 20.2 | 44.2 | 3.62 | 15.64 | 5.62 | 6.98 | 2.36 | 12.08 | |
| HPMDubbing2025.09 | 27.6 | 31.3 | 2.11 | 12.31 | 8.05 | 8.23 | 6.02 | 8.85 | |
| DSU2025.09 | 34.3 | 5.9 | 3.55 | 13.82 | 10.55 | 10.57 | 5.63 | 8.73 |