Action Recognition on Kinetics (1-shot to 5-shot evaluation)
97.1Accuracy (5-shot)TSAM
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| TSAMBackbone=CLIP-ViT-B/162025.12 | 97.1 | 96.2 | — | — | — | — | |
| TS-FSARBackbone=CLIP-ViT-B/162025.12 | 96.6 | 96.3 | — | — | — | — | |
| MantaPre-Backbone=ImageNet-ViT2024.12 | 96.3 | 84.2 | — | — | — | — | |
| CLIP-M2DFPre-training=CLIP-ViT-B2025.12 | 96.2 | 90.9 | — | 95.1 | — | — | |
| MantaPre-Backbone=ImageNet-VM2024.12 | 96.2 | 84.4 | — | — | — | — | |
| CLIP-MEIPre-training=CLIP-ViT-B2025.12 | 96 | 87.7 | — | 95.3 | — | — | |
| MA-FSARBackbone=CLIP-ViT-B/162025.12 | 96 | 95.7 | — | — | — | — | |
| MoLoBackbone=ViT-B2024.07 | 95.8 | 78.9 | — | — | — | — | |
| MoLoPre-Backbone=ImageNet-ViT2024.12 | 95.8 | 78.9 | — | — | — | — | |
| MoLoPre-Backbone=ImageNet-VM2024.12 | 95.6 | 79.4 | — | — | — | — | |
| CLIP-CPM2CPre-training=CLIP-ViT-B2025.12 | 95.5 | 91 | — | 95.1 | — | — | |
| D2ST-AdapterBackbone=CLIP-ViT-B/162025.12 | 95.5 | 89.3 | — | — | — | — | |
| SOAP-NetBackbone=ViT-B2024.07 | 95.5 | 89.9 | — | — | — | — | |
| CLIP-FSARBackbone=CLIP-ViT-B/162025.12 | 95.4 | 94.8 | — | — | — | — | |
| MVP-shotBackbone=CLIP-ViT-B/162025.12 | 95.1 | 91 | — | — | — | — | |
| CLIP-FSARPre-training=CLIP-ViT-B2025.12 | 95 | 89.7 | — | 94.2 | — | — | |
| CLIP-FSARBackbone=ViT-B2024.07 | 95 | 89.7 | — | — | — | — | |
| CLIP-SPMPre-training=CLIP-ViT-B2025.12 | 94.3 | 92.8 | — | 94.1 | — | — | |
| MantaPre-Backbone=ImageNet-RN502024.12 | 94.2 | 82.4 | — | — | — | — | |
| SOAP-NetBackbone=ResNet-502024.07 | 93.8 | 81.1 | — | — | — | — | |
| CLIP-MEIPre-training=CLIP-RN502025.12 | 93.6 | 90.1 | — | 92.9 | — | — | |
| CLIP-M2DFPre-training=CLIP-RN502025.12 | 93.5 | 90.1 | — | 92 | — | — | |
| CLIP-SPMPre-training=CLIP-RN502025.12 | 93.5 | 90.9 | — | 92.9 | — | — | |
| EMP-NetBackbone=CLIP-ViT-B/162025.12 | 93.5 | 89.1 | — | — | — | — | |
| CLIP-CPM2CPre-training=CLIP-RN502025.12 | 93.1 | 88.6 | — | 91.5 | — | — | |
| AMFARBackbone=ResNet-502024.07 | 92.6 | 80.1 | — | — | — | — | |
| AMFARPre-Backbone=ImageNet-RN502024.12 | 92.6 | 80.1 | — | — | — | — | |
| HyRSMPre-Backbone=ImageNet-VM2024.12 | 92.4 | 76.6 | — | — | — | — | |
| HyRSMBackbone=ViT-B2024.07 | 92.3 | 76.8 | — | — | — | — | |
| HyRSMPre-Backbone=ImageNet-ViT2024.12 | 92.3 | 76.8 | — | — | — | — | |
| CLIP-FSARPre-training=CLIP-RN502025.12 | 91.9 | 87.6 | — | 90.7 | — | — | |
| STENPre-training=CLIP-ViT-B2025.12 | 91.4 | 83 | — | — | — | — | |
| STRMBackbone=ViT-B2024.07 | 91.2 | — | — | — | — | — | |
| SA-CTBackbone=ViT-B2024.07 | 91.2 | — | — | — | — | — | |
| STRMPre-Backbone=ImageNet-ViT2024.12 | 91.2 | — | — | — | — | — | |
| SA-CTPre-Backbone=ImageNet-ViT2024.12 | 91.2 | — | — | — | — | — | |
| Lite-MKDPre-Backbone=ImageNet-VM2024.12 | 90.8 | 78.5 | — | — | — | — | |
| DPNPre-training=INet-RN502025.12 | 90.7 | 76.3 | — | — | — | — | |
| Lite-MKDPre-Backbone=ImageNet-ViT2024.12 | 90.6 | 78.8 | — | — | — | — | |
| TEAMPre-training=INet-RN502025.12 | 88.2 | 75.1 | — | — | — | — | |
| TEAMBackbone=IN-RN502025.12 | 88.2 | 75.1 | — | — | — | — | |
| GCSMBackbone=ResNet-502024.07 | 88.2 | 74.2 | — | — | — | — | |
| GCSMPre-Backbone=ImageNet-RN502024.12 | 88.2 | 74.2 | — | — | — | — | |
| Lite-MKDBackbone=ResNet-502024.07 | 87.5 | 75 | — | — | — | — | |
| TRXBackbone=ViT-B2024.07 | 87.5 | 76.3 | — | — | — | — | |
| TRXPre-Backbone=ImageNet-ViT2024.12 | 87.5 | 76.3 | — | — | — | — | |
| Lite-MKDPre-Backbone=ImageNet-RN502024.12 | 87.5 | 75 | — | — | — | — | |
| MTFANPre-training=INet-RN502025.12 | 87.4 | 74.6 | — | — | — | — | |
| NguyenBackbone=IN-RN502025.12 | 87.4 | 74.3 | — | — | — | — | |
| GgHMBackbone=IN-RN502025.12 | 87.4 | 74.9 | — | — | — | — | |
| GgHMBackbone=ResNet-502024.07 | 87.4 | 74.9 | — | — | — | — | |
| MTFANBackbone=ResNet-502024.07 | 87.4 | 74.6 | — | — | — | — | |
| GgHMPre-Backbone=ImageNet-RN502024.12 | 87.4 | 74.9 | — | — | — | — | |
| MTFANPre-Backbone=ImageNet-RN502024.12 | 87.4 | 74.6 | — | — | — | — | |
| TRXPre-Backbone=ImageNet-VM2024.12 | 87.2 | 76.2 | — | — | — | — | |
| SA-CTBackbone=ResNet-502024.07 | 87.1 | 71.9 | — | — | — | — | |
| SA-CTPre-Backbone=ImageNet-RN502024.12 | 87.1 | 71.9 | — | — | — | — | |
| SloshNetBackbone=IN-RN502025.12 | 87 | 70.4 | — | — | — | — | |
| SloshNetBackbone=ResNet-502024.07 | 87 | — | — | — | — | — | |
| SloshNetPre-Backbone=ImageNet-RN502024.12 | 87 | — | — | — | — | — | |
| AmeFu-NetBackbone=ResNet-502024.07 | 86.8 | 74.1 | — | — | — | — | |
| AmeFu-NetPre-Backbone=ImageNet-RN502024.12 | 86.8 | 74.1 | — | — | — | — | |
| STRMPre-training=INet-RN502025.12 | 86.7 | 65.3 | — | — | — | — | |
| STRMBackbone=IN-RN502025.12 | 86.7 | 65.1 | — | — | — | — | |
| STRMBackbone=ResNet-502024.07 | 86.7 | — | — | — | — | — | |
| STRMPre-Backbone=ImageNet-RN502024.12 | 86.7 | — | — | — | — | — | |
| LSTCBackbone=ResNet-502024.07 | 86.5 | 73.4 | — | — | — | — | |
| HyRSMBackbone=ResNet-502024.07 | 86.2 | 73.5 | — | — | — | — | |
| HyRSMPre-Backbone=ImageNet-RN502024.12 | 86.2 | 73.5 | — | — | — | — | |
| HyRSMPre-training=INet-RN502025.12 | 86.1 | 73.7 | — | 83.5 | — | — | |
| HyRSMBackbone=IN-RN502025.12 | 86.1 | 73.7 | — | — | — | — | |
| TRXPre-training=INet-RN502025.12 | 85.9 | 63.6 | — | — | — | — | |
| TRXBackbone=IN-RN502025.12 | 85.9 | 65.1 | — | — | — | — | |
| TRXBackbone=ResNet-502024.07 | 85.9 | 74.9 | — | — | — | — | |
| TRXPre-Backbone=ImageNet-RN502024.12 | 85.9 | 74.9 | — | — | — | — | |
| TA2NPre-training=INet-RN502025.12 | 85.8 | 72.8 | — | — | — | — | |
| OTAMBackbone=IN-RN502025.12 | 85.8 | 73 | — | — | — | — | |
| HCLBackbone=IN-RN502025.12 | 85.8 | 73.7 | — | — | — | — | |
| OTAMBackbone=ResNet-502024.07 | 85.8 | 73 | — | — | — | — | |
| TA2NBackbone=ResNet-502024.07 | 85.8 | 72.8 | — | — | — | — | |
| HCLBackbone=ResNet-502024.07 | 85.8 | 73.7 | — | — | — | — | |
| MoLoBackbone=ResNet-502024.07 | 85.7 | 75.2 | — | — | — | — | |
| MoLoPre-Backbone=ImageNet-RN502024.12 | 85.7 | 74.2 | — | — | — | — | |
| MoLoPre-training=INet-RN502025.12 | 85.6 | 74 | — | 83.7 | — | — | |
| OTAMPre-training=INet-RN502025.12 | 85.5 | 73 | — | — | — | — | |
| STENPre-training=CLIP-RN502025.12 | 85.4 | 67 | — | — | — | — | |
| ITANetBackbone=ResNet-502024.07 | 84.3 | 73.6 | — | — | — | — | |
| ARNShots=5-shot2020.01 | 82.4 | — | — | — | — | — | |
| HCRPre-training=INet-RN502025.12 | 81.4 | 70.2 | — | — | — | — | |
| TARNFeatures=C3D2019.07 | 80.66 | 66.55 | 74.56 | 77.33 | 78.89 | — | |
| CMNFeatures=ResNet-502019.07 | 78.9 | 60.5 | 70 | 75.6 | 77.3 | — | |
| CMNShots=5-shot2020.01 | 78.9 | — | — | — | — | — | |
| CMNBackbone=ResNet-502024.07 | 78.9 | 60.5 | — | — | — | — | |
| TARN-fFeatures=ResNet-502019.07 | 78.52 | 64.83 | 72.94 | 76.22 | 78.02 | — | |
| TARN-singleFeatures=C3D2019.07 | 76.88 | 62.84 | 69.8 | 73.96 | 74.88 | — | |
| TARN-f-singleFeatures=ResNet-502019.07 | 70.64 | 57.92 | 63.96 | 65.9 | 68.52 | — | |
| ARNShots=1-shot2020.01 | 63.7 | — | — | — | — | — | |
| CMNShots=1-shot2020.01 | 60.5 | — | — | — | — | — | |
| ARNshot=5-shot, way=5-way, Backbone=Conv-3D2021.01 | — | — | — | — | — | 82.4 | |
| ARNBackbone=ResNet-502021.12 | — | — | — | — | — | 82.4 |