Detection of Limb automatisms on Seizure Video Dataset
67.8AccuracyCNN
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CNNModel Architecture=CNN, Signal Enhancement (Pose)=false2026.05 | 67.8 | 36.7 | 31.5 | 31.6 | |
| Pose+QwenModel Architecture=Qwen, Signal Enhancement (Pose)=true2026.05 | 39.5 | 25.4 | 75 | 38 | |
| QwenModel Architecture=Qwen, Signal Enhancement (Pose)=false2026.05 | 35.6 | 22.4 | 71.4 | 34.1 | |
| InternModel Architecture=Intern, Signal Enhancement (Pose)=false2026.05 | 32.2 | 23 | 81 | 35.8 | |
| Pose+InternModel Architecture=Intern, Signal Enhancement (Pose)=true2026.05 | 31 | 25.6 | 100 | 40.8 | |
| ViViTModel Architecture=ViViT, Signal Enhancement (Pose)=false2026.05 | 30 | 18.3 | 54.6 | 27.4 |