Continuous Sign Language Recognition on CSL-Daily (dev)
25.4Word Error Rate (WER)TwoStream-SLR (Ours)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| TwoStream-SLR (Ours)Modality=RGB + Pose2022.11 | 25.4 | — | — | |
| TwoStream-SLRModality=RGB+Keypoints2023.09 | 25.4 | — | — | |
| TwoStream-SLRExtra Cues (Keypoints)=true2023.05 | 25.4 | — | — | |
| TS-SLRModality=Pose + RGB2025.01 | 25.4 | — | — | |
| SlowFastSignModality=RGB Only2023.09 | 25.5 | — | — | |
| SlowFastSignVenue=ICASSP 20242025.12 | 25.5 | — | — | |
| DCA2023.05 | 25.6 | 9.1 | 2.8 | |
| C2ST2023.05 | 25.9 | 9.3 | 2.7 | |
| C2STModality=Pose2025.01 | 25.9 | — | — | |
| CLIP-SLAVenue=CVPR 20252025.12 | 26 | — | — | |
| Sigma2025.09 | 26.09 | — | — | |
| Uni-SignModality=Pose + RGB2025.01 | 26.7 | — | — | |
| Uni-Sign2025.09 | 26.7 | — | — | |
| SlowFastSign (Reproduced)Venue=ICASSP 2024, reproducibility=Reproduced2025.12 | 26.9 | — | — | |
| USTMsmallModality=RGB2025.12 | 27.7 | — | — | |
| USTMbaseModality=RGB2025.12 | 27.7 | — | — | |
| USTMsmallBackbone=Swin-small2025.12 | 27.7 | — | — | |
| USTMbaseBackbone=Swin-base2025.12 | 27.7 | — | — | |
| USTMsmallVariant=small2025.12 | 27.7 | — | — | |
| USTMbaseVariant=base2025.12 | 27.7 | — | — | |
| USTMtinyModality=RGB2025.12 | 27.8 | — | — | |
| USTMtinyBackbone=Swin-tiny2025.12 | 27.8 | — | — | |
| USTMtinyVariant=tiny2025.12 | 27.8 | — | — | |
| CoSign2023.05 | 28.1 | — | — | |
| CoSignModality=Pose2025.01 | 28.1 | — | — | |
| CoSign2025.09 | 28.1 | — | — | |
| Uni-SignModality=Pose2025.01 | 28.2 | — | — | |
| MSKAVenue=PR 20252025.12 | 28.2 | — | — | |
| Swin-MSTPVenue=Neurocomp. 2025, Backbone=Swin-small2025.12 | 28.3 | — | — | |
| STNetVenue=IEEE Acc. 20252025.12 | 28.3 | — | — | |
| ACMCVenue=IMAVIS 20252025.12 | 28.3 | — | — | |
| TB-Net2023.05 | 28.4 | — | — | |
| MSLUModality=Pose2025.01 | 28.6 | — | — | |
| CorrNet+Venue=arXiv 20242025.12 | 28.6 | — | — | |
| MP-SLRVenue=TPAMI 20252025.12 | 28.6 | — | — | |
| MSLU2025.09 | 28.6 | — | — | |
| TCNet2024.03 | 29.7 | — | — | |
| SGN2023.05 | 30.4 | 10.8 | 3.1 | |
| CorrNet2023.03 | 30.6 | — | — | |
| CorrnetModality=RGB Only2023.09 | 30.6 | — | — | |
| CorrNet2024.03 | 30.6 | — | — | |
| CorrNet2023.05 | 30.6 | — | — | |
| CorrNetModality=Pose2025.01 | 30.6 | — | — | |
| CorrNetVenue=CVPR 20232025.12 | 30.6 | — | — | |
| CorrNet2025.09 | 30.6 | — | — | |
| SEN2022.11 | 31.1 | — | — | |
| AdaBrowsealpha=0.04, GFLOPS=287, Throughput (videos/s)=12.312023.08 | 31.1 | — | — | |
| AdaBrowse+alpha=0.05, GFLOPS=241, Throughput (videos/s)=15.122023.08 | 31.1 | — | — | |
| SEN2024.03 | 31.1 | — | — | |
| SEN2023.05 | 31.1 | — | — | |
| SENModality=Pose2025.01 | 31.1 | — | — | |
| SEN2025.09 | 31.1 | — | — | |
| AdaBrowse+GFLOPS per video=1712023.08 | 31.2 | — | — | |
| AdaBrowse+alpha=0.1, GFLOPS=171, Throughput (videos/s)=17.582023.08 | 31.2 | — | — | |
| AdaBrowseModality=Pose2025.01 | 31.2 | — | — | |
| AdaBrowseVenue=ACM MM 20232025.12 | 31.2 | — | — | |
| AdaBrowse2025.09 | 31.2 | — | — | |
| AdaBrowseGFLOPS per video=2542023.08 | 31.3 | — | — | |
| AdaBrowsealpha=0.15, GFLOPS=254, Throughput (videos/s)=14.262023.08 | 31.3 | — | — | |
| CTCA2023.05 | 31.3 | 9.2 | 2.5 | |
| CTCAVenue=CVPR 20232025.12 | 31.3 | — | — | |
| BaselineGFLOPS=364, Throughput (videos/s)=12.222023.08 | 31.5 | — | — | |
| C2SLREnd-to-end=true, Modalities (Training)=R+P, Modalities (Inference)=R2022.12 | 31.9 | — | — | |
| TIN-Iterative2022.11 | 32.8 | — | — | |
| Baseline2022.11 | 32.8 | — | — | |
| DNF (RGB)Modality=RGB2022.11 | 32.8 | — | — | |
| TIN-Iterative2023.03 | 32.8 | — | — | |
| TIN-IterativeGFLOPS per video=6322023.08 | 32.8 | — | — | |
| DNF (RGB)Modality=RGB Only2023.09 | 32.8 | — | — | |
| DNFEnd-to-end=false, Modalities (Training)=R+F, Modalities (Inference)=R+F2022.12 | 32.8 | — | — | |
| TIN2024.03 | 32.8 | — | — | |
| DNFModality=RGB+Op.Flow2025.12 | 32.8 | — | — | |
| Joint-SLRT2022.11 | 33.1 | — | — | |
| Joint-SLRTModality=RGB2022.11 | 33.1 | — | — | |
| Joint-SLRT2023.03 | 33.1 | — | — | |
| Joint-SLRTGFLOPS per video=3162023.08 | 33.1 | — | — | |
| Joint-SLRT+Modality=RGB Only2023.09 | 33.1 | — | — | |
| SLTEnd-to-end=true, Modalities (Training)=R, Modalities (Inference)=R2022.12 | 33.1 | — | — | |
| Joint-SLRT2024.03 | 33.1 | — | — | |
| TwoStream-SLRModality=RGB+Pose2025.12 | 33.1 | — | — | |
| FCN2022.11 | 33.2 | — | — | |
| FCNModality=RGB2022.11 | 33.2 | — | — | |
| SignBTModality=RGB2022.11 | 33.2 | — | — | |
| FCN2023.03 | 33.2 | — | — | |
| FCNGFLOPS per video=2862023.08 | 33.2 | — | — | |
| FCN+Modality=RGB Only2023.09 | 33.2 | — | — | |
| FCNEnd-to-end=true, Modalities (Training)=R, Modalities (Inference)=R2022.12 | 33.2 | — | — | |
| FCN2024.03 | 33.2 | — | — | |
| SignBTModality=Pose2025.01 | 33.2 | — | — | |
| SignBT2025.09 | 33.2 | — | — | |
| BN-TIN2022.11 | 33.6 | — | — | |
| BN-TIN2023.03 | 33.6 | — | — | |
| BN-TINGFLOPS per video=3162023.08 | 33.6 | — | — | |
| SignBTModality=RGB Only2023.09 | 33.6 | — | — | |
| TIN + TransformerEnd-to-end=true, Modalities (Training)=R, Modalities (Inference)=R2022.12 | 33.6 | — | — | |
| BN-TIN2024.03 | 33.6 | — | — | |
| BN-TIN+Transf2023.05 | 33.6 | 13.9 | 3.4 | |
| SignBERTModality=RGB+Pose2025.12 | 33.6 | — | — | |
| LS-HAN2022.11 | 39 | — | — | |
| LS-HANModality=RGB2022.11 | 39 | — | — |