Sign Language Recognition on WLASL 1.0 (test)
86.5Top-1 AccuracySiformer
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SiformerInput Modality=Skeleton-based2025.03 | 86.5 | — | — | — | — | |
| SignBERT+Input Modality=RGB-based2025.03 | 84.11 | — | — | — | — | |
| SignBERT+Input Modality=Skeleton-based2025.03 | 79.84 | — | — | — | — | |
| TCKInput Modality=RGB-based2025.03 | 77.52 | — | — | — | — | |
| Fusion-3Input Modality=RGB-based2025.03 | 75.67 | — | — | — | — | |
| I3DInput Modality=RGB-based2025.03 | 65.89 | — | — | — | — | |
| SPOTERInput Modality=Skeleton-based, Source=Original authors2025.03 | 63.18 | — | — | — | — | |
| SPOTERInput Modality=Skeleton-based, Source=Reproduction2025.03 | 58.52 | — | — | — | — | |
| Pose-TGCNInput Modality=Skeleton-based2025.03 | 55.43 | — | — | — | — | |
| ST-GCNInput Modality=Skeleton-based2025.03 | 50.78 | — | — | — | — | |
| BESTModalities=RGB2023.03 | — | 81.01 | 94.19 | 81.63 | 94.67 | |
| Fusion-3Modalities=RGB2023.03 | — | 75.67 | 86 | — | — | |
| I3DModalities=RGB2023.03 | — | 65.89 | 84.11 | — | — | |
| NLA-SLRInference crops=1-crop2023.03 | — | 91.47 | 96.9 | 92.17 | 97.17 | |
| NLA-SLRInference crops=3-crop2023.03 | — | 92.64 | 96.9 | 93.08 | 97.17 | |
| PSLRModalities=Skeleton2023.03 | — | 60.15 | 83.98 | — | — | |
| SignBERT+Extra data=true2023.03 | — | 82.56 | 94.96 | 83.3 | 95 | |
| ST-GCNModalities=Keypoints2023.03 | — | 50.78 | 79.07 | 51.62 | 79.47 | |
| TCK+Modalities=Skeleton2023.03 | — | 77.52 | 91.08 | 77.55 | 91.42 |