Action Recognition on MSRAction3D
97.56AccuracyGATS
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GATSInput=point2026.03 | 97.56 | — | — | — | |
| GATSInput=point, Frames=242026.03 | 97.56 | — | — | — | |
| PST-Transformer MotionPDE2026.06 | 96.63 | — | — | — | |
| KAN-HyperpointNetReference=ICASSP 252026.06 | 95.59 | — | — | — | |
| PST-Transformer + DiMPVenue=This Paper2026.05 | 95.51 | — | — | — | |
| ST-NBMIM2018.05 | 95.3 | — | — | — | |
| PSTNet++ MotionPDE2026.06 | 95.22 | — | — | — | |
| ARRN-LSTM2018.05 | 95 | — | — | — | |
| PST-Transformer + M2PSCVenue=ECCV242026.05 | 94.84 | — | — | — | |
| ST-LSTM2018.05 | 94.8 | — | — | — | |
| Graph-Based2018.05 | 94.8 | — | — | — | |
| ST-NBNN2018.05 | 94.8 | — | — | — | |
| PvNeXtInput=point, Frames=242026.03 | 94.77 | — | — | — | |
| PPTr+C2PInput=point, Frames=242026.03 | 94.76 | — | — | — | |
| GATSInput=point, Frames=202026.03 | 94.54 | — | — | — | |
| HBRNN2018.05 | 94.5 | — | — | — | |
| PSTNet MotionPDE2026.06 | 94.28 | — | — | — | |
| MaST-PreInput=point, Frames=242026.03 | 94.08 | — | — | — | |
| PST-Transformer + MaST-PreVenue=ICCV232026.05 | 94.08 | — | — | — | |
| SRENet#Frame=242026.06 | 94.07 | — | — | — | |
| P4Transformer + DiMPVenue=This Paper2026.05 | 93.97 | — | — | — | |
| X4D#Frame=242026.06 | 93.9 | — | — | — | |
| Leaf#Frame=242026.06 | 93.84 | — | — | — | |
| PST-TransformerEvaluation Protocol=Supervised Learning, #Frames=242023.05 | 93.73 | — | — | — | |
| PST-Transformer# frames=242023.03 | 93.73 | — | — | — | |
| PST-TransformerInput=point, Frames=242026.03 | 93.73 | — | — | — | |
| PST-TransformerVenue=TPAMI222026.05 | 93.73 | — | — | — | |
| PST-TransformerReference=IEEE TPAMI 232026.06 | 93.73 | — | — | — | |
| PST-Trans.#Frame=242026.06 | 93.73 | — | — | — | |
| MAMBA4DInput=point, Frames=242026.03 | 93.38 | — | — | — | |
| Mamba4DReference=CVPR 252026.06 | 93.38 | — | — | — | |
| KinetModalities=Point Cloud, # of Frames=242022.03 | 93.27 | — | — | — | |
| KinetEvaluation Protocol=Supervised Learning, #Frames=242023.05 | 93.27 | — | — | — | |
| PSTNet + PointCMPEvaluation Protocol=End-to-end Fine-tuning, #Frames=242023.05 | 93.27 | — | — | — | |
| Kinet# frames=242023.03 | 93.27 | — | — | — | |
| PSTNet + PointCMPVenue=CVPR232026.05 | 93.27 | — | — | — | |
| KinetReference=CVPR 222026.06 | 93.27 | — | — | — | |
| Kinet#Frame=242026.06 | 93.27 | — | — | — | |
| PPTr MotionPDE2026.06 | 93.11 | — | — | — | |
| P4Transformer + M2PSCVenue=ECCV242026.05 | 93.03 | — | — | — | |
| PSTNet + PointCMPEvaluation Protocol=Linear Probing, #Frames=242023.05 | 92.93 | — | — | — | |
| PSTNet++Evaluation Protocol=Supervised Learning, #Frames=242023.05 | 92.68 | — | — | — | |
| PSTNet++Input=point, Frames=242026.03 | 92.68 | — | — | — | |
| PointCPSCInput=point, Frames=242026.03 | 92.68 | — | — | — | |
| PointCPSCReference=ICCV 232026.06 | 92.68 | — | — | — | |
| PSTNet++Reference=IEEE TPAMI 222026.06 | 92.68 | — | — | — | |
| PSTNet++#Frame=242026.06 | 92.68 | — | — | — | |
| MAMBA4D#Frame=242026.06 | 92.68 | — | — | — | |
| SequentialPointNetInput=point, Frames=242026.03 | 92.64 | — | — | — | |
| X4D#Frame=162026.06 | 92.56 | — | — | — | |
| Lie Group2018.05 | 92.5 | — | — | — | |
| PPTrEvaluation Protocol=Supervised Learning, #Frames=242023.05 | 92.33 | — | — | — | |
| PPTrReference=ECCV 222026.06 | 92.33 | — | — | — | |
| PSTNet + PointCMPEvaluation Protocol=End-to-end Fine-tuning, #Frames=162023.05 | 92.26 | — | — | — | |
| 3DinAction# frames=242023.03 | 92.23 | — | — | — | |
| 3DInActionInput=skeleton, Frames=242026.03 | 92.23 | — | — | — | |
| 3DInActionReference=CVPR 232026.06 | 92.23 | — | — | — | |
| 3DInAction#Frame=242026.06 | 92.23 | — | — | — | |
| PST-TransformerEvaluation Protocol=Supervised Learning, #Frames=162023.05 | 91.98 | — | — | — | |
| PST-Transformer# frames=162023.03 | 91.98 | — | — | — | |
| PST-Trans.#Frame=162026.06 | 91.98 | — | — | — | |
| SRENet#Frame=162026.06 | 91.98 | — | — | — | |
| SequentialPointNetReference=IEEE TII 222026.06 | 91.94 | — | — | — | |
| KinetModalities=Point Cloud, # of Frames=162022.03 | 91.92 | — | — | — | |
| KinetEvaluation Protocol=Supervised Learning, #Frames=162023.05 | 91.92 | — | — | — | |
| PSTNet + PointCMPEvaluation Protocol=Linear Probing, #Frames=162023.05 | 91.92 | — | — | — | |
| Kinet# frames=162023.03 | 91.92 | — | — | — | |
| Kinet#Frame=162026.06 | 91.92 | — | — | — | |
| PointMapNetReference=Symmetry 232026.06 | 91.91 | — | — | — | |
| PSTNet + PointCMPEvaluation Protocol=End-to-end Fine-tuning, #Frames=122023.05 | 91.58 | — | — | — | |
| PSTNet + PointCMPInput=Point, Number of Frames=162023.05 | 91.58 | — | — | — | |
| Leaf#Frame=162026.06 | 91.5 | — | — | — | |
| PSTNetModalities=Point Cloud, # of Frames=242022.03 | 91.2 | — | — | — | |
| PSTNetInput=point, #Frames=242022.05 | 91.2 | — | — | — | |
| PSTNetEvaluation Protocol=Supervised Learning, #Frames=242023.05 | 91.2 | — | — | — | |
| PSTNet# frames=242023.03 | 91.2 | — | — | — | |
| PSTNetInput=point, Frames=242026.03 | 91.2 | — | — | — | |
| PSTNetVenue=ICLR212026.05 | 91.2 | — | — | — | |
| PSTNetReference=ICLR 212026.06 | 91.2 | — | — | — | |
| PPTr#Frame=242026.06 | 91.2 | — | — | — | |
| PSTNet#Frame=242026.06 | 91.2 | — | — | — | |
| P4TransformerModalities=Point Cloud, # of Frames=242022.03 | 90.94 | — | — | — | |
| P4TransformerEvaluation Protocol=Supervised Learning, #Frames=242023.05 | 90.94 | — | — | — | |
| P4Transformer# frames=242023.03 | 90.94 | — | — | — | |
| P4TransformerInput=point, Frames=242026.03 | 90.94 | — | — | — | |
| P4TransformerVenue=CVPR212026.05 | 90.94 | — | — | — | |
| P4TransformerReference=CVPR 212026.06 | 90.94 | — | — | — | |
| P4Trans.#Frame=242026.06 | 90.94 | — | — | — | |
| 3DinAction# frames=162023.03 | 90.57 | — | — | — | |
| 3DInAction#Frame=162026.06 | 90.57 | — | — | — | |
| PPTrEvaluation Protocol=Supervised Learning, #Frames=162023.05 | 90.31 | — | — | — | |
| PPTr#Frame=162026.06 | 90.31 | — | — | — | |
| PSTNet++Evaluation Protocol=Supervised Learning, #Frames=162023.05 | 90.24 | — | — | — | |
| PSTNet + PointCMPEvaluation Protocol=Linear Probing, #Frames=122023.05 | 90.24 | — | — | — | |
| GATSInput=point, Frames=122026.03 | 90.24 | — | — | — | |
| PSTNet++#Frame=162026.06 | 90.24 | — | — | — | |
| PSTNetModalities=Point Cloud, # of Frames=162022.03 | 89.9 | — | — | — | |
| PSTNetInput=point, #Frames=162022.05 | 89.9 | — | — | — | |
| PSTNetEvaluation Protocol=Supervised Learning, #Frames=162023.05 | 89.9 | — | — | — | |
| PSTNet# frames=162023.03 | 89.9 | — | — | — |