Action Recognition on Epic-Kitchens (Pairwise PXX Comparison)
69.06Average Comparison ScoreMM+Mixup
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| MM+MixupBackbone=ViT-Base, Framework=D2M2026.05 | 69.06 | — | — | — | — | — | — | 65.06 | 72.93 | 69.2 | |
| MM+CLBackbone=ViT-Base, Framework=D2M2026.05 | 69.06 | — | — | — | — | — | — | 64.6 | 73.6 | 68.99 | |
| MM+MIROBackbone=ViT-Base, Framework=D2M2026.05 | 67.83 | — | — | — | — | — | — | 63.68 | 72.67 | 67.15 | |
| MM+MixupBackbone=CNN-Based, Framework=D2M2026.05 | 66.98 | — | — | — | — | — | — | 63.45 | 71.47 | 66.02 | |
| MM+MMDBackbone=ViT-Base, Framework=D2M2026.05 | 66.92 | — | — | — | — | — | — | 63.22 | 71.33 | 66.22 | |
| MM+CLBackbone=CNN-Based, Framework=D2M2026.05 | 66.56 | — | — | — | — | — | — | 61.38 | 71.47 | 66.84 | |
| MM+MMDBackbone=CNN-Based, Framework=D2M2026.05 | 66.5 | — | — | — | — | — | — | 62.3 | 71.07 | 66.12 | |
| MM+ERMBackbone=ViT-Base, Framework=D2M2026.05 | 66.39 | — | — | — | — | — | — | 61.15 | 71.6 | 66.43 | |
| MM+ERMBackbone=CNN-Based, Framework=D2M2026.05 | 66.33 | — | — | — | — | — | — | 61.84 | 70.53 | 66.63 | |
| MM+MIROBackbone=CNN-Based, Framework=D2M2026.05 | 66.3 | — | — | — | — | — | — | 63.45 | 71.07 | 64.37 | |
| MM+MixupBackbone=ViT-Base, Framework=M2D2026.05 | 65.88 | — | — | — | — | — | — | 62.3 | 71.47 | 63.86 | |
| MM+MMDBackbone=ViT-Base, Framework=M2D2026.05 | 65.67 | — | — | — | — | — | — | 63.45 | 68.67 | 64.89 | |
| MM+MixupBackbone=CNN-Based, Framework=M2D2026.05 | 65.51 | — | — | — | — | — | — | 60 | 70.9 | 66.02 | |
| MM+MIROBackbone=ViT-Base, Framework=M2D2026.05 | 65.18 | — | — | — | — | — | — | 61.84 | 71.07 | 62.63 | |
| MM+ERMBackbone=ViT-Base, Framework=M2D2026.05 | 65.17 | — | — | — | — | — | — | 62.3 | 70.27 | 62.94 | |
| MM+ERMBackbone=CNN-Based, Framework=M2D2026.05 | 65.04 | — | — | — | — | — | — | 62.3 | 68.13 | 64.68 | |
| MM+MMDBackbone=CNN-Based, Framework=M2D2026.05 | 65.01 | — | — | — | — | — | — | 61.84 | 67.73 | 65.91 | |
| SimMMDGBackbone=CNN-Based2026.05 | 64.8 | — | — | — | — | — | — | 63.21 | 67.73 | 63.45 | |
| MDJATBackbone=CNN-Based2026.05 | 64.75 | — | — | — | — | — | — | 62.99 | 70.8 | 60.47 | |
| MM+CLBackbone=ViT-Base, Framework=M2D2026.05 | 64.74 | — | — | — | — | — | — | 62.99 | 68.8 | 62.42 | |
| MM+MIROBackbone=CNN-Based, Framework=M2D2026.05 | 64.2 | — | — | — | — | — | — | 60.92 | 66.27 | 65.4 | |
| CMRFBackbone=CNN-Based2026.05 | 64.12 | — | — | — | — | — | — | 55.4 | 69.2 | 67.76 | |
| MM+CLBackbone=CNN-Based, Framework=M2D2026.05 | 63.56 | — | — | — | — | — | — | 62.3 | 66.67 | 61.7 | |
| MDJATBackbone=ViT-Base2026.05 | 63.47 | — | — | — | — | — | — | 58.39 | 67.86 | 64.17 | |
| MAD-DGBackbone=CNN-Based2026.05 | 62.73 | — | — | — | — | — | — | 60.7 | 65.06 | 62.42 | |
| MAD-DGBackbone=ViT-Base2026.05 | 62.27 | — | — | — | — | — | — | 60.05 | 64.69 | 62.09 | |
| DeepAllBackbone=CNN-Based2026.05 | 62.04 | — | — | — | — | — | — | 55.7 | 66 | 62.42 | |
| CMRFBackbone=ViT-Base2026.05 | 62.04 | — | — | — | — | — | — | 57.7 | 66 | 62.42 | |
| Supervised-target (Tsup)Backbone=I3D2026.05 | 61.5 | 64 | 63.7 | 57 | 63.7 | 57 | 64 | — | — | — | |
| DeepAllBackbone=ViT-Base2026.05 | 61.2 | — | — | — | — | — | — | 57.47 | 64.93 | 61.19 | |
| SimMMDGBackbone=ViT-Base2026.05 | 61.19 | — | — | — | — | — | — | 55.86 | 66.53 | 61.19 | |
| TranSVAEBackbone=I3D2026.05 | 52.6 | 50.5 | 50.3 | 50.3 | 58.6 | 48 | 58 | — | — | — | |
| MetaTransBackbone=I3D2026.05 | 51 | 48 | 50.4 | 47.4 | 56.6 | 48.5 | 55.1 | — | — | — | |
| DFRABackbone=I3D2026.05 | 46.6 | 46 | 43.8 | 48 | 48.7 | 43.7 | 49.2 | — | — | — | |
| MCTBackbone=I3D2026.05 | 46.1 | 43.2 | 51.4 | 48.3 | 40.8 | 48.3 | 44.6 | — | — | — | |
| HCTBackbone=I3D2026.05 | 45.3 | 45 | 43.3 | 43.2 | 47.6 | 46.2 | 46.4 | — | — | — | |
| MixDANNBackbone=I3D2026.05 | 44.7 | 48 | 43 | 43.4 | 43.9 | 43 | 46.9 | — | — | — | |
| C-RNABackbone=I3D2026.05 | 44.7 | 43.1 | 48.9 | 43.1 | 41.7 | 49.6 | 41.9 | — | — | — | |
| STCDABackbone=I3D2026.05 | 44.6 | 47.7 | 41.2 | 44.4 | 47.6 | 41.1 | 45.5 | — | — | — | |
| MM-SADABackbone=I3D2026.05 | 43.9 | 45 | 39.7 | 41.7 | 46.1 | 42.1 | 46.1 | — | — | — | |
| CleanAdaptBackbone=I3D2026.05 | 43.9 | 44.5 | 40.9 | 44.6 | 45.7 | 40.7 | 47.1 | — | — | — | |
| CoMixBackbone=I3D2026.05 | 43.2 | 42.9 | 40.9 | 38.6 | 45.2 | 42.3 | 49.2 | — | — | — | |
| STDNBackbone=I3D2026.05 | 41.6 | 40.5 | 38.6 | 38.5 | 44 | 40.4 | 47.2 | — | — | — | |
| TA3NBackbone=I3D2026.05 | 39.9 | 34.2 | 37.4 | 40.9 | 42.8 | 39.9 | 44.2 | — | — | — | |
| DANNBackbone=I3D2026.05 | 39.2 | 37.7 | 36.6 | 38.3 | 41.9 | 38.8 | 42.1 | — | — | — | |
| ADDABackbone=I3D2026.05 | 37.4 | 35.4 | 34.9 | 36.3 | 40.8 | 36.1 | 41.4 | — | — | — | |
| Source-only (Sonly)Backbone=I3D2026.05 | 35.3 | 32.8 | 34.1 | 35.4 | 39.1 | 34.6 | 35.8 | — | — | — |