Action Recognition on FineGYM
96AccuracyACLNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ACLNet2026.01 | 96 | — | |
| ProtoGCNModality=Skeleton2024.11 | 95.9 | — | |
| PoseC3DModality=R+P†, Pre-trained=true2022.12 | 95.6 | — | |
| PoseConv3D (J+L)Input Modality=joint/limb, Skeleton Source=high-quality 2D skeletons (Sec 3.1)2021.04 | 94.3 | — | |
| PoseConv3DModality=Skeleton+Limb2024.11 | 94.3 | — | |
| PoseConv3DPublication=CVPR 20222026.01 | 94.3 | — | |
| PYSKLModality=Skeleton2024.11 | 94.1 | — | |
| PoseConv3D (J)Input Modality=joint, Skeleton Source=high-quality 2D skeletons (Sec 3.1)2021.04 | 93.2 | — | |
| VA-ARPublication=AAAI 20252026.01 | 92.8 | — | |
| MS-G3D++Skeleton Source=high-quality 2D skeletons (Sec 3.1)2021.04 | 92.6 | — | |
| SkeletonMAEModality=Skeleton2024.11 | 91.8 | — | |
| Ske2GridPublication=ICML 20232026.01 | 91.8 | — | |
| SkeletonMAEPublication=ICCV 20232026.01 | 91.8 | — | |
| TQNModality=R, Pre-trained=true2022.12 | 90.6 | — | |
| 3D deformable transformerModality=R+P†, Pre-trained=false2022.12 | 90.3 | — | |
| LT-S3DModality=RGB2024.11 | 88.9 | — | |
| LT-S3DPublication=ECCV 20182026.01 | 88.9 | — | |
| TSMModality=RGB+Flow2024.11 | 81.2 | — | |
| TSNModality=RGB+Flow2024.11 | 79.8 | — | |
| Heterogeneous Skeleton-Based Action Representation LearningModality=Skeleton2025.06 | 75.3 | — | |
| HyperbolicModality=RGB2025.06 | 73.4 | — | |
| MoLoModality=RGB+Point2025.06 | 73.3 | — | |
| BEARModality=RGB2025.06 | 69.6 | — | |
| EuclideanModality=RGB2025.06 | 68.2 | — | |
| I3DModality=RGB2024.11 | 64.4 | — | |
| SVTModality=RGB2025.06 | 62.3 | — | |
| CARLModality=RGB2025.06 | 41.8 | — | |
| ST-GCNSkeleton Source=N/A (Reported by [49])2021.04 | 25.2 | — | |
| ST-GCNModality=P†, Pre-trained=false2022.12 | 25.2 | — | |
| EVERESTBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 23.3 | |
| MGMBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 25.8 | |
| MGMAEBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 26.1 | |
| MMEBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 29 | |
| MVDBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 22.7 | |
| SIGMABackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 30.1 | |
| SMILEBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 30.2 | |
| SMILE w/o motionBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 27.9 | |
| VideoMAEBackbone=ViT-B, Pre-training Dataset=K400, Evaluation Protocol=Linear Probing2025.04 | — | 23.9 |