Action Prediction on Epic-Kitchens systematic shift
59IID AccuracyCDE (ViT-CLIP)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CDE (ViT-CLIP)Encoder Backbone=ViT-CLIP2025.08 | 59 | 34 | 25 | |
| CDE (ViT-DINO)Encoder Backbone=ViT-DINO2025.08 | 54 | 33 | 21 | |
| Group-tokenFeature Aggregation Strategy=Group-token2025.08 | 52 | 27 | 25 | |
| Group-denseFeature Aggregation Strategy=Group-dense2025.08 | 50 | 26 | 24 | |
| CDE (ViT-MAE)Encoder Backbone=ViT-MAE2025.08 | 50 | 30 | 20 | |
| Group-avgFeature Aggregation Strategy=Group-avg2025.08 | 47 | 24 | 23 | |
| CLIPEncoder Backbone=CLIP2025.08 | 45 | 24 | 21 | |
| ResNetEncoder Backbone=ResNet2025.08 | 42 | 17 | 25 |