Action Recognition on UCF101 (Few-Shot Accuracy)
97Top-1 Accuracy (1-shot)CLIP-M2DF
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CLIP-M2DFPre-training=CLIP-ViT-B2025.12 | 97 | 98.1 | 99.3 | |
| CLIP-FSARPre-training=CLIP-ViT-B2025.12 | 96.6 | 98.4 | 99 | |
| CLIP-SPMPre-training=CLIP-ViT-B2025.12 | 96.2 | 98.2 | 98.7 | |
| CLIP-CPM2CPre-training=CLIP-ViT-B2025.12 | 95 | 98.1 | 98.6 | |
| CLIP-MEIPre-training=CLIP-ViT-B2025.12 | 94.9 | 98.2 | 98.9 | |
| CLIP-M2DFPre-training=CLIP-RN502025.12 | 94.3 | 98 | 98.8 | |
| CLIP-SPMPre-training=CLIP-RN502025.12 | 93.8 | 97.2 | 97.7 | |
| CLIP-MEIPre-training=CLIP-RN502025.12 | 93.7 | 97 | 97.5 | |
| STENPre-training=CLIP-ViT-B2025.12 | 91.5 | — | 96.8 | |
| CLIP-FSARPre-training=CLIP-RN502025.12 | 91.3 | 95.1 | 97 | |
| CLIP-CPM2CPre-training=CLIP-RN502025.12 | 91.1 | 96.2 | 97.4 | |
| DPNPre-training=INet-RN502025.12 | 87.7 | — | 98 | |
| TEAMPre-training=INet-RN502025.12 | 87.2 | — | 96.2 | |
| MoLoPre-training=INet-RN502025.12 | 86 | 93.5 | 95.5 | |
| MTFANPre-training=INet-RN502025.12 | 84.8 | — | 95.1 | |
| HyRSMPre-training=INet-RN502025.12 | 83.9 | 93 | 94.7 | |
| STENPre-training=CLIP-RN502025.12 | 83.5 | — | 96.8 | |
| TA2NPre-training=INet-RN502025.12 | 81.9 | — | 95.1 | |
| OTAMPre-training=INet-RN502025.12 | 79.9 | — | 88.9 | |
| STRMPre-training=INet-RN502025.12 | 79.2 | — | 96.9 | |
| TRXPre-training=INet-RN502025.12 | 77.3 | — | 96.1 | |
| HCRPre-training=INet-RN502025.12 | 71.8 | — | 87.3 |