Research map / 3D vision / Human pose estimation
Point-to-Point Regression PointNet for 3D Hand Pose Estimation
Liuhao Ge, Zhou Ren, Junsong Yuan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| stack two PointNet modules instead of one (ablation) | P2P Reg. PointNet · NYU | mean error distance: 10.3mm -> 9.5mm | Improved |
| output unit vector fields instead of offset fields (ablation) | two-stacked P2P Reg. · NYU | mean error distance: 10.8mm -> 9.5mm | Improved |
| use point-to-point regression instead of direct regression | two-stacked PointNet · NYU | mean error distance: 10.8mm -> 9.5mm | Improved |
| add direct-regression replacement for divergent joints (ablation) | two-stacked P2P Reg. · NYU | mean error distance: 9.5mm -> 9.2mm | Improved |
| add PCA constraint on the pose (ablation) | P2P Reg.+Direct Reg. · NYU | mean error distance: 9.2mm -> 9.1mm | Improved |
| increase candidate points M from 25 to 55 (weighted average) (ablation) | P2P Reg. inference · NYU | mean error distance: 9.46mm -> 9.71mm | Hurt |
| use weighted instead of unweighted average at M=55 (ablation) | P2P Reg. inference · NYU | mean error distance: 9.84mm -> 9.71mm | Improved |
| use weighted instead of unweighted average at M=5 (ablation) | P2P Reg. inference · NYU | mean error distance: 9.50mm -> 9.50mm | No clear effect |
| use P2P Reg. instead of DenseReg | hand pose estimation · NYU | mean error distance: 10.2mm -> 9.1mm | Improved |
| use P2P Reg. instead of DenseReg | hand pose estimation · ICVL | mean error distance: 7.3mm -> 6.3mm | Improved |
| use P2P Reg. instead of DenseReg | hand pose estimation · MSRA | mean error distance: 7.2mm -> 7.7mm | Hurt |
Closest papers by meaning
- Hand PointNet: 3D Hand Pose Estimation Using Point Sets (CVPR 2018 · similarity 0.96)
- Two Heads Are Better than One: Image-Point Cloud Network for Depth-Based 3D Hand Pose Estimation (AAAI 2023 · similarity 0.94)
- 3D Convolutional Neural Networks for Efficient and Robust Hand Pose Estimation From Single Depth Images (CVPR 2017 · similarity 0.92)
- Real-Time 3D Hand Pose Estimation with 3D Convolutional Neural Networks (TPAMI 2018 · similarity 0.92)
- SO-HandNet: Self-Organizing Network for 3D Hand Pose Estimation With Semi-Supervised Learning (ICCV 2019 · similarity 0.92)
- Generalized Feedback Loop for Joint Hand-Object Pose Estimation (TPAMI 2019 · similarity 0.91)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
