3D Hand Pose Estimation on NYU (test)
7.06Mean Error (mm)HandMCM
Evaluation Results
| Method | Links | |
|---|---|---|
| HandMCMInput=Depth2026.02 | 7.06 | |
| HandDAGTInput=Depth2026.02 | 7.12 | |
| IPNetInput=Depth2026.02 | 7.17 | |
| HandR2N2Input=Point2026.02 | 7.27 | |
| SSRNInput=Depth2026.02 | 7.37 | |
| HandDiffInput=Depth2026.02 | 7.38 | |
| PHGInput=Depth2026.02 | 7.39 | |
| HandGCNFormerInput=Point2026.02 | 7.43 | |
| OursLabel Usage=100%, Augmented Set=No2023.03 | 8.01 | |
| Ourstraining_annotations=100%2023.03 | 8.01 | |
| OursLabel Usage=75%, Augmented Set=No2023.03 | 8.06 | |
| OursLabel Usage=50%, Augmented Set=No2023.03 | 8.11 | |
| OursLabel Usage=25%, Augmented Set=No2023.03 | 8.14 | |
| Ourstraining_annotations=25%2023.03 | 8.14 | |
| OursNumber of Labeled Examples=10,0002023.03 | 8.16 | |
| JGR-P2O2023.03 | 8.29 | |
| Fang et al.FPS(s)=111.202023.04 | 8.29 | |
| JGR-P2OInput=Depth2026.02 | 8.29 | |
| HambaInput=Depth2026.02 | 8.38 | |
| V2V-PoseNet2017.11 | 8.42 | |
| HandFoldingNet2023.03 | 8.42 | |
| V2VInput=Voxel2026.02 | 8.42 | |
| A2J-TransformerFPS(s)=24.812023.04 | 8.43 | |
| V2V-PoseNet2023.03 | 8.58 | |
| HandFoldingInput=Point2026.02 | 8.58 | |
| OursNumber of Labeled Examples=1,0002023.03 | 8.6 | |
| A2JFPS=105.062019.08 | 8.61 | |
| A2J2023.03 | 8.61 | |
| Xiong et al.FPS(s)=105.062023.04 | 8.61 | |
| OursLabel Usage=1%, Augmented Set=No2023.03 | 8.71 | |
| P2PFPS=41.82019.08 | 9.045 | |
| Point-to-Point2023.03 | 9.1 | |
| Point-to-PointInput=Point2026.02 | 9.1 | |
| V2VFPS=352019.08 | 9.22 | |
| Moon et al.FPS(s)=352023.04 | 9.22 | |
| NARHT2023.03 | 9.8 | |
| MURAUERNumber of Labeled Examples=10,0002023.03 | 9.9 | |
| CrossInfoNet2023.03 | 10.08 | |
| CrossInfoNetInput=Depth2026.02 | 10.08 | |
| DenseRegFPS=27.82019.08 | 10.2 | |
| DenseReg2023.03 | 10.2 | |
| DenseRegInput=Depth2026.02 | 10.2 | |
| HandPointNetFPS=482019.08 | 10.5 | |
| HandPointNet2023.03 | 10.54 | |
| HandPointNetInput=Point2026.02 | 10.54 | |
| SHPR-NetInput=Point2026.02 | 10.78 | |
| Generalized-Feedback2023.03 | 10.89 | |
| MURAUERNumber of Labeled Examples=1,0002023.03 | 10.9 | |
| SO-HandNetLabel Usage=100%, Augmented Set=No2023.03 | 11.2 | |
| Pose-REN2017.11 | 11.81 | |
| Pose-REN2019.08 | 11.81 | |
| Pose-REN2023.03 | 11.81 | |
| Pose-RenInput=Depth2026.02 | 11.81 | |
| OursNumber of Labeled Examples=1002023.03 | 12.11 | |
| MURAUERNumber of Labeled Examples=1002023.03 | 12.2 | |
| DeepPrior++2017.11 | 12.24 | |
| DeepPrior++FPS=302019.08 | 12.24 | |
| DeepPrior++2023.03 | 12.24 | |
| DeepPrior++Input=Depth2026.02 | 12.24 | |
| REN-9x6x62017.11 | 12.69 | |
| REN-9x6x62019.08 | 12.69 | |
| REN-9x6x62023.03 | 12.69 | |
| SO-HandNetLabel Usage=75%, Augmented Set=No2023.03 | 12.8 | |
| REN-4x6x62017.11 | 13.39 | |
| REN-4x6x62019.08 | 13.39 | |
| REN-4x6x62023.03 | 13.39 | |
| 3DCNN2017.11 | 14.1 | |
| Baek et al.Label Usage=100%, Augmented Set=Yes, 10 times2023.03 | 14.1 | |
| SO-HandNetLabel Usage=50%, Augmented Set=No2023.03 | 14.1 | |
| 3DCNN2023.03 | 14.1 | |
| Lie-X2017.11 | 14.51 | |
| Lie-X2019.08 | 14.51 | |
| Baek et al.(w/o refine)Label Usage=100%, Augmented Set=Yes, 10 times2023.03 | 14.9 | |
| SO-HandNetLabel Usage=25%, Augmented Set=No2023.03 | 14.9 | |
| LSPSLabel Usage=100%, Augmented Set=No2023.03 | 15.4 | |
| LSPSLabel Usage=50%, Augmented Set=No2023.03 | 15.45 | |
| LSPSLabel Usage=75%, Augmented Set=No2023.03 | 15.45 | |
| Crossing NetLabel Usage=100%, Augmented Set=No2023.03 | 15.5 | |
| Global-to-Local2017.11 | 15.6 | |
| Global-to-LocalFPS=502019.08 | 15.6 | |
| LSPSLabel Usage=25%, Augmented Set=No2023.03 | 15.7 | |
| Crossing NetLabel Usage=75%, Augmented Set=No2023.03 | 15.9 | |
| Feedback2017.11 | 15.97 | |
| Crossing NetLabel Usage=50%, Augmented Set=No2023.03 | 16 | |
| Crossing NetLabel Usage=25%, Augmented Set=No2023.03 | 16.1 | |
| Baek et al.(w/o aug.; refine)Label Usage=100%, Augmented Set=No2023.03 | 16.4 | |
| MURAUERNumber of Labeled Examples=102023.03 | 16.4 | |
| JTSC2017.11 | 16.8 | |
| JTSC2019.08 | 16.8 | |
| DeepModel2017.11 | 17.04 | |
| DeepModel2019.08 | 17.04 | |
| DeepModel2023.03 | 17.04 | |
| Baek et al.(baseline)Label Usage=100%, Augmented Set=No2023.03 | 17.3 | |
| Hand3D2017.11 | 17.6 | |
| Hand3DFPS=302019.08 | 17.6 | |
| DeepPrior2017.11 | 19.73 | |
| DeepPrior2023.03 | 19.73 | |
| DISCO2017.11 | 20.7 | |
| DISCO2019.08 | 20.7 | |
| OursNumber of Labeled Examples=102023.03 | 25.82 |