Action Recognition on EPIC-Kitchens v1 (test s2 (unseen))
29.6Actions Top-1 AccStudent
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| StudentBackbone=Swin-T, Input Modality=RGB, lambda=1, gamma=302025.12 | 29.6 | 54.1 | — | 43.7 | — | — | — | — | — | — | — | — | |
| OmnivoreBackbone=Swin-T, Input Modality=RGB2025.12 | 27.9 | 54.5 | — | 38.5 | — | — | — | — | — | — | — | — | |
| Leaderboard2019.05 | 27.4 | 60 | 82 | 38.1 | 63.8 | 45.2 | — | — | — | — | — | — | |
| G-BlendModality=Audio-Visual2019.05 | 26.6 | 58.3 | 81.3 | 36.7 | 60.3 | 43.6 | — | — | — | — | — | — | |
| R(2+1)Ddepth=152, pre-training=IG-Kinetics, short edge scaling=true2019.05 | 25.6 | 57.3 | 81.1 | 35.7 | 58.7 | 42.7 | — | — | — | — | — | — | |
| R(2+1)Ddepth=152, pre-training=IG-Kinetics2019.05 | 25.4 | 55.3 | 80.3 | 34.7 | 58.2 | 40.7 | — | — | — | — | — | — | |
| Baidu-UTS2019.05 | 25.1 | 59.7 | 82.7 | 34.1 | 62.4 | 46 | — | — | — | — | — | — | |
| Visual: ip-CSN-152Modality=Visual2019.05 | 25.1 | 56.2 | 80.9 | 35.8 | 59.6 | 41.2 | — | — | — | — | — | — | |
| Ours SAPPre-training=Kinetics2020.02 | 23.9 | 53.2 | 78.2 | 33 | 58 | 40.5 | — | — | — | — | — | — | |
| R(2+1)Ddepth=34, pre-training=IG-Kinetics2019.05 | 23.7 | 55.5 | 80.9 | 33.6 | 56.7 | 39.1 | — | — | — | — | — | — | |
| R(2+1)Ddepth=34, pre-training=IG-Kinetics, short edge scaling=true2019.05 | 23.6 | 56 | 80.6 | 32.4 | 55.6 | 39.5 | — | — | — | — | — | — | |
| LFB Maxoperator=Max2018.12 | 21.2 | 50.9 | 77.6 | 31.5 | 57.8 | 39.4 | — | — | — | — | — | — | |
| LFB MaxPre-training=Kinetics+ImageNet2020.02 | 21.2 | 50.9 | 77.6 | 31.5 | 57.8 | 39.4 | — | — | — | — | — | — | |
| Leaderboard2019.05 | 21 | 54.5 | 81.2 | 30.4 | 55.7 | 39.4 | — | — | — | — | — | — | |
| TBN EnsembleEnsemble=5 models2019.05 | 21 | 54.5 | 81.2 | 30.4 | 55.7 | 39.4 | — | — | — | — | — | — | |
| TBN Single2019.05 | 19.1 | 52.7 | 79.9 | 27.9 | 53.8 | 36.5 | — | — | — | — | — | — | |
| R(2+1)Ddepth=34, pre-training=Sports-1M2019.05 | 18.3 | 47.2 | 77.4 | 28.7 | 50 | 31.6 | — | — | — | — | — | — | |
| R(2+1)Ddepth=34, pre-training=Kinetics2019.05 | 16.8 | 48.4 | 77.2 | 26.6 | 50.4 | 31.2 | — | — | — | — | — | — | |
| R(2+1)D 34Pre-training=Kinetics2020.02 | 16.8 | 48.4 | 77.2 | 26.6 | 50.4 | 31.2 | — | — | — | — | — | — | |
| LSTAModality=two stream2018.11 | 16.63 | 47.32 | 77.02 | 22.16 | 43.15 | 30.93 | 31.57 | 17.91 | 8.97 | 26.17 | 17.8 | 11.92 | |
| LSTAPre-training=ImageNet2020.02 | 15.9 | — | — | — | — | — | — | — | — | — | — | — | |
| LSTAModality=RGB2018.11 | 15.88 | 45.51 | 75.25 | 23.46 | 43.16 | 30.01 | 26.19 | 17.58 | 8.44 | 20.8 | 19.67 | 11.29 | |
| TSN FusionPre-training=ImageNet2020.02 | 14.8 | 46.1 | 76.7 | 24.3 | 49.3 | 29.8 | — | — | — | — | — | — | |
| TSN FlowPre-training=ImageNet2020.02 | 13.5 | 47.4 | 77 | 21.2 | 42.5 | 27.5 | — | — | — | — | — | — | |
| TSN RGBPre-training=ImageNet2020.02 | 11.3 | 36.5 | 74.4 | 22.6 | 46.9 | 26.3 | — | — | — | — | — | — | |
| TSN Fusionmodality=Fusion2018.12 | 10.9 | 39.4 | 74.3 | 22.7 | 45.7 | 25.3 | — | — | — | — | — | — | |
| TSNModality=two stream2018.11 | 10.89 | 39.4 | 74.29 | 22.7 | 45.72 | 25.26 | 22.54 | 15.33 | 6.21 | 13.06 | 17.52 | 6.49 | |
| 2SCNNModality=RGB2018.11 | 10.11 | 34.89 | 74.56 | 21.82 | 45.34 | 25.33 | 19.48 | 14.67 | 5.32 | 11.22 | 17.24 | 6.34 | |
| TSNModality=RGB2018.11 | 10.11 | 34.89 | 74.56 | 21.82 | 45.34 | 25.33 | 19.48 | 14.67 | 5.32 | 11.22 | 17.24 | 6.34 | |
| TSN RGBmodality=RGB2018.12 | 10.1 | 34.9 | 74.6 | 21.8 | 45.3 | 25.3 | — | — | — | — | — | — | |
| 2SCNNModality=two stream2018.11 | 7.31 | 36.16 | 71.97 | 18.03 | 38.41 | 19.49 | 18.11 | 15.31 | 3.19 | 10.52 | 12.55 | 3 | |
| TSN Flowmodality=Flow2018.12 | 6.7 | 40.1 | 73.4 | 14.5 | 33.8 | 18.6 | — | — | — | — | — | — |