Action Recognition on EGTEA
29.21Noun Top-1 AccX-MIC
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| X-MICBackbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 29.21 | 31.41 | 31.21 | 30.12 | |
| Co-CoOpBackbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 27.9 | 27.74 | 28.84 | 24.1 | |
| CoOpBackbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 23.9 | 26.45 | 26.29 | 24.35 | |
| A5Backbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 23.7 | 28.03 | 27.76 | 27.14 | |
| CLIP-Adapter *Backbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 22.4 | 26.57 | 25.34 | 26.93 | |
| CLIP-AdapterBackbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 21.41 | 26.51 | 24.98 | 24.52 | |
| ZS CLIPBackbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 19.7 | 18.71 | 9.07 | 3.51 | |
| Vita-CLIPBackbone=CLIP ViT-B/16, Evaluation Protocol=Zero-shot, Training Dataset=Ego4D2024.03 | 17.24 | 27.63 | 22.76 | 24.89 |