Video Action Recognition on UCF-101 (val)
85.2Top-1 Acc (K=2)EZ-CLIP
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| EZ-CLIPBackbone=ViT-142023.12 | 85.2 | 89.7 | 92.4 | 93.3 | |
| EZ-CLIPBackbone=ViT-162023.12 | 84.7 | 88.3 | 91.3 | 92.8 | |
| ViFi-CLIP2023.12 | 80.7 | 85.1 | 90 | 92.7 | |
| EZ-CLIPBackbone=ViT-322023.12 | 79.7 | 83.7 | 87.5 | 89.8 | |
| X-CLIP2023.12 | 71.4 | 79.9 | 83.7 | 91.4 | |
| A52023.12 | 71.4 | 79.9 | 85.7 | 89.9 | |
| ActionCLIP2023.12 | 70.6 | 71.5 | 73 | 91.4 | |
| Vanilla CLIP2023.12 | 63.6 | 63.6 | 63.6 | 63.6 |