| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LRW-1000 (test) | Refined Network (SE+MixUp+Cosine LR+LS+WB) | Accuracy55.7 | 50 | 1mo ago | |
| LRW 1.0 (test) | Top-1 Accuracy94.8 | 37 | 1mo ago | ||
| LRS2 (test) | Visual Transformer Pooling (VTP) | WER22.6 | 28 | 1mo ago | |
| LRS3 1.0 (test) | AV-HuBERT + relaxed cross attention | WER25.51 | 19 | 1mo ago | |
| GRID (test) | MA-LipNet | WER1.09 | 18 | 1mo ago | |
| Tulips1 (test) | Model Params0 | 17 | 1mo ago | ||
| LRW original (test) | Ensemble | Top-1 Accuracy88.5 | 14 | 1mo ago | |
| LRW Word-level (test) | Stafylakis et al. (WB) | Accuracy88.8 | 13 | 1mo ago | |
| CMLR (test) | MA-LipNet | CER21.49 | 11 | 1mo ago | |
| LRS3 (test) | WER25.9 | 8 | 1mo ago | ||
| LRW (test) | Multi-modality Associative Bridging | Word Accuracy85.4 | 8 | 1mo ago | |
| CMLR 1.0 (test) | CSSMCM | CER3.9 | 7 | 1mo ago | |
| LRW | P3D-based visual front-end | Accuracy84.8 | 6 | 1mo ago |
