Automatic Speech Recognition on LRS3 Samples with > 100 Frames (test)
2.6WER (10 dB)USR 2.0
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| USR 2.0Beam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 2.6 | 4.3 | 10.4 | 26 | 10.8 | |
| USRBeam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 2.7 | 4.6 | 11.6 | 29.1 | 12 | |
| USR 2.0Beam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 3.8 | 10.6 | 42.8 | 98.3 | 38.9 | |
| AV-HuBERTBeam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 4.4 | 8 | 18 | 48.7 | 19.8 | |
| USRBeam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 4.9 | 12.6 | 48.1 | 111.7 | 44.3 | |
| BRAVEnBeam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 6.4 | 13.8 | 49.2 | 108.4 | 44.5 | |
| AV-HuBERTBeam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 6.5 | 16.2 | 49.3 | 97.2 | 42.3 | |
| BRAVEnBeam size=30, Zero-shot evaluation=true, Noise augmentation=None, Labelled training data=30 hours2026.02 | 6.5 | 12.7 | 30.9 | 55.5 | 26.4 |