Action Recognition on PKU-MMD Cross-Subject
95.8AccuracyDVANet
Evaluation Results
| Method | Links | |
|---|---|---|
| DVANetModality=RGB2023.12 | 95.8 | |
| (Cheng et al. 2022) (TSM)Modality=RGB+Depth2023.12 | 93.6 | |
| (Li et al. 2018a) (IJCAI '18)Modality=Skeleton2023.12 | 92.6 | |
| (Cheng et al. 2022) (TSN)Modality=RGB+Depth2023.12 | 92.6 | |
| (Lin et al. 2020) (PAMI '22)Modality=RGB+Depth2023.12 | 91.7 | |
| (Song et al. 2017)Modality=Skeleton2023.12 | 86.9 | |
| (Elias, Sedmidubsky, and Zezula 2019) (ISM '19)Modality=Skeleton2023.12 | 86.5 | |
| (Wang et al. 2018b) (PAMI '19)Modality=RGB+Depth2023.12 | 85 | |
| MSPNet2026.05 | 74.83 | |
| InternVideo2-B/14Backbone=InternVideo2-B/142026.05 | 73.25 | |
| UniFormer V2-B/16Backbone=UniFormer V2-B/162026.05 | 72.29 | |
| VideoSwin-Swin-BBackbone=Swin-B2026.05 | 71.73 | |
| VideoMAE V2-ViT-BBackbone=ViT-B2026.05 | 71.52 | |
| VideoMAE-ViT-BBackbone=ViT-B2026.05 | 69.45 | |
| MViT V2-SBackbone=MViT V2-S2026.05 | 67.69 | |
| TimeSformer-spaceOnlyBackbone=TimeSformer2026.05 | 64.33 | |
| SlowFast-ResNet50Backbone=ResNet502026.05 | 63.84 | |
| TSM-ResNet50Backbone=ResNet502026.05 | 63.33 | |
| Decomposition and CompositionModality=J+M+B2025.12 | 62 | |
| SlowOnly-ResNet50Backbone=ResNet502026.05 | 60.4 | |
| UmURLModality=J+M+B2025.12 | 59.7 | |
| Decomposition and CompositionModality=J2025.12 | 59.2 | |
| UmURLModality=J2025.12 | 58.2 | |
| HiCoModality=J2025.12 | 56.3 | |
| CMDModality=J2025.12 | 56 | |
| ActCLRModality=J+M+B2025.12 | 55.9 | |
| TRN-ResNet50Backbone=ResNet502026.05 | 51.85 | |
| ISCModality=J2025.12 | 45.9 | |
| MS2LModality=J2025.12 | 45.8 | |
| LongT GANModality=J2025.12 | 44.8 | |
| TSN-ResNet50Backbone=ResNet502026.05 | 43.84 | |
| TSN-DenseNet161Backbone=DenseNet1612026.05 | 39.54 |