Video Attention Modeling on Video Attention Modeling General
111FPS (img/s)AAM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AAMBackbone=DINOv3, Input Length (Fixed/Arbitrary)=Arbi2026.06 | 111 | 21.4 | |
| TMFI-NetBackbone=Video Swin, Input Length (Fixed/Arbitrary)=Fixed2026.06 | 30 | 234 | |
| STSANetBackbone=Video Swin, Input Length (Fixed/Arbitrary)=Fixed2026.06 | 28 | 643 | |
| TASEDBackbone=3D Conv, Input Length (Fixed/Arbitrary)=Fixed2026.06 | 17 | 82 |