Visual Text-to-Speech on GRID (seen-speaker)
12.1WERProDubber
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| ProDubberSetting=Setting 1, Conditioning=video and transcript2026.06 | 12.1 | 79.7 | 3.74 | 11.74 | 6.52 | 6.24 | 2.97 | 10.21 | |
| Foley-OmniSetting=Setting 1, Conditioning=video and transcript2026.06 | 15.3 | 84.1 | 3.07 | 10.41 | 6.28 | 6.35 | 5.24 | 8.67 | |
| EmoDubberSetting=Setting 1, Conditioning=video and transcript2026.06 | 16.9 | 83.9 | 3.98 | 9.03 | 5.89 | 5.97 | 7.14 | 6.75 | |
| GTSetting=Setting 1, Conditioning=video and transcript2026.06 | 18 | 100 | 3.96 | 0 | 0 | 0 | 7.25 | 6.71 | |
| HPMDubbingSetting=Setting 1, Conditioning=video and transcript2026.06 | 44.1 | 72 | 2.13 | 12.45 | 8.11 | 8.23 | 5.87 | 9.11 |