Isolated Sign Language Recognition on WLASL 2000 (test)
64.54Per-instance Top-1 AccSigma
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Sigma2025.09 | 64.54 | — | — | — | 62.28 | — | |
| Uni-Sign2025.09 | 63.52 | — | — | — | 61.32 | — | |
| StepNetInput Modalities=RGB + Optical flow, Ensemble Strategy=1-clip2022.12 | 61.17 | — | — | 91.94 | 58.43 | 91.43 | |
| NLA-SLR2025.09 | 61.05 | — | — | — | 58.05 | — | |
| SAMInput Modalities=7 Modalities, Ensemble Strategy=5-clip2022.12 | 59.39 | — | — | 91.48 | 56.63 | 90.89 | |
| MSLU2025.09 | 56.29 | — | — | — | 53.29 | — | |
| Signbert+2022.12 | 55.59 | — | — | 89.37 | 53.33 | 88.82 | |
| Signbert (H+R)Input Modalities=Hand representation (H) + RGB (R)2022.12 | 54.69 | — | — | 87.49 | 52.08 | 86.93 | |
| BEST2022.12 | 54.59 | — | — | 88.08 | 52.12 | 87.28 | |
| SignBERT+2025.09 | 48.85 | — | — | — | 46.37 | — | |
| Signbert (H+P)Input Modalities=Hand representation (H) + Skeleton (P)2022.12 | 47.46 | — | — | 83.32 | 45.17 | 82.32 | |
| BEST2025.09 | 46.25 | — | — | — | 43.52 | — | |
| SignBERT2025.09 | 39.4 | — | — | — | 36.74 | — | |
| HMA2025.09 | 37.91 | — | — | — | 35.9 | — | |
| ST-GCN2025.09 | 34.4 | — | — | — | 32.53 | — | |
| MSLU2024.11 | — | 56.29 | 53.29 | — | — | — | |
| NLA-SLR2024.11 | — | 61.05 | 58.05 | — | — | — | |
| SHuBERT#Params=0.17M, rank=1, adaptation=LoRA2024.11 | — | 60.9 | 58.01 | — | — | — | |
| SignBERT2024.11 | — | 39.4 | 36.74 | — | — | — | |
| SignBERT+2024.11 | — | 48.85 | 46.37 | — | — | — | |
| Uni-Sign#Params=580M2024.11 | — | 63.52 | 61.32 | — | — | — |