Action Recognition on HMDB51 (1-shot and 5-shot Accuracy)
96.85-shot AccuracyManta
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MantaPre-Backbone=ImageNet-ViT2024.12 | 96.8 | 88.9 | |
| MantaPre-Backbone=ImageNet-VM2024.12 | 96.6 | 89.1 | |
| MantaPre-Backbone=ImageNet-RN502024.12 | 96.4 | 86.8 | |
| STARPre-training=CLIP-ViT-B2026.05 | 89.2 | 86 | |
| TSAMPre-training=CLIP-ViT-B2026.05 | 88.9 | 84.5 | |
| AMFARPre-Backbone=ImageNet-RN502024.12 | 87.8 | 73.9 | |
| CLIP-FSARPre-training=CLIP-ViT-B2026.05 | 87.7 | 77.1 | |
| STARPre-training=CLIP-RN502026.05 | 85 | 74.9 | |
| MoLoPre-Backbone=ImageNet-VM2024.12 | 84.5 | 71.1 | |
| MoLoPre-Backbone=ImageNet-ViT2024.12 | 84.4 | 71.3 | |
| TSAMPre-training=CLIP-RN502026.05 | 82.8 | 72.9 | |
| HyRSMPre-Backbone=ImageNet-VM2024.12 | 82.6 | 70.2 | |
| MVP-ShotPre-training=CLIP-RN502026.05 | 82.5 | 72.5 | |
| HyRSMPre-Backbone=ImageNet-ViT2024.12 | 82.2 | 69.6 | |
| SA-CTPre-Backbone=ImageNet-ViT2024.12 | 81.6 | — | |
| SA-CT(ViT)Pre-training=INet-ViT-B2026.05 | 81.6 | — | |
| STRMPre-Backbone=ImageNet-ViT2024.12 | 81.3 | — | |
| CLIP-FSARPre-training=CLIP-RN502026.05 | 80.7 | 69.4 | |
| GCSMPre-Backbone=ImageNet-RN502024.12 | 79.3 | 61.3 | |
| MGCSMPre-training=INet-RN502026.05 | 79.3 | 61.3 | |
| TRXPre-Backbone=ImageNet-ViT2024.12 | 78.8 | 66.9 | |
| TRXPre-Backbone=ImageNet-VM2024.12 | 78.5 | 66.7 | |
| SA-CTPre-training=INet-RN502026.05 | 78.3 | 60.4 | |
| SloshNetPre-Backbone=ImageNet-RN502024.12 | 77.5 | — | |
| Lite-MKDPre-Backbone=ImageNet-ViT2024.12 | 77.4 | 71.1 | |
| MoLoPre-training=INet-RN502026.05 | 77.4 | 60.8 | |
| MoLoPre-Backbone=ImageNet-RN502024.12 | 77.3 | 67.1 | |
| Lite-MKDPre-Backbone=ImageNet-VM2024.12 | 77.2 | 71.5 | |
| SA-CTPre-Backbone=ImageNet-RN502024.12 | 76.9 | 61.2 | |
| GgHMPre-Backbone=ImageNet-RN502024.12 | 76.9 | 61.2 | |
| STRMPre-Backbone=ImageNet-RN502024.12 | 76.3 | — | |
| HyRSMPre-Backbone=ImageNet-RN502024.12 | 76.2 | 60.1 | |
| SPRNPre-training=INet-RN502026.05 | 76.2 | 61.6 | |
| HyRSMPre-training=INet-RN502026.05 | 76 | 60.3 | |
| TRXPre-Backbone=ImageNet-RN502024.12 | 75.8 | 63.5 | |
| TRXPre-training=INet-RN502026.05 | 75.6 | — | |
| AmeFu-NetPre-Backbone=ImageNet-RN502024.12 | 75.5 | 60.2 | |
| AmeFuNetPre-training=INet-RN502026.05 | 75.5 | 60.2 | |
| Lite-MKDPre-Backbone=ImageNet-RN502024.12 | 74.7 | 66.9 | |
| MTFANPre-Backbone=ImageNet-RN502024.12 | 74.6 | 59 |