Video Action Recognition on HMDB51 5-way 5-shot
88.7AccuracyDIST
Evaluation Results
| Method | Links | |
|---|---|---|
| DISTPre-training=CLIP-ViT-B2026.02 | 88.7 | |
| CLIP-FSARPre-training=CLIP-ViT-B2026.02 | 87.7 | |
| TEAMPre-training=INet-ViT-B2026.02 | 85.5 | |
| STRMBackbone=ViT2021.12 | 81.3 | |
| CLIP-FSARPre-training=CLIP-RN502026.02 | 80.3 | |
| MVP-shotPre-training=CLIP-ViT-B2026.02 | 80.3 | |
| TRXBackbone=ViT2021.12 | 79.7 | |
| CapFSARPre-training=BLIP-ViT-B2026.02 | 78.6 | |
| SloshNetPre-training=INet-RN502026.02 | 77.5 | |
| MoLoPre-training=INet-RN502026.02 | 77.4 | |
| STRMBackbone=ResNet-502021.12 | 77.3 | |
| STRMPre-training=INet-RN502026.02 | 77.3 | |
| CLIP-FreezePre-training=CLIP-ViT-B2026.02 | 77 | |
| GgHMPre-training=INet-RN502026.02 | 76.9 | |
| HCLPre-training=INet-RN502026.02 | 76.3 | |
| HyRSMPre-training=INet-RN502026.02 | 76 | |
| TRXshot=5-shot, way=5-way, Backbone=ResNet-50, cardinalities=Ω={2,3}2021.01 | 75.6 | |
| TRXBackbone=ResNet-502021.12 | 75.6 | |
| TRXPre-training=INet-RN502026.02 | 75.6 | |
| AmeFu-NetPre-training=INet-RN502026.02 | 75.5 | |
| Lite-MKDPre-training=INet-RN502026.02 | 74.7 | |
| MTFANPre-training=INet-RN502026.02 | 74.6 | |
| TA2NPre-training=INet-RN502026.02 | 73.9 | |
| ProtoNetPre-training=INet-RN502026.02 | 68.4 | |
| HF-ARBackbone=ResNet-502021.12 | 62.2 | |
| ARNshot=5-shot, way=5-way, Backbone=Conv-3D2021.01 | 60.6 | |
| ARNBackbone=ResNet-502021.12 | 60.6 | |
| ARNPre-training=C3D2026.02 | 60.6 |