Audio-Visual Speech Recognition on LRS3 (clean)
1.38WERVisG AV-HuBERT Large
Evaluation Results
| Method | Links | |
|---|---|---|
| VisG AV-HuBERT LargeParams=484M2026.04 | 1.38 | |
| AV-HuBERT LargeParams=477M2026.04 | 1.4 | |
| VisG AV-HuBERT BaseParams=162M2026.04 | 3.78 | |
| AV-HuBERT BaseParams=160M2026.04 | 4.1 |