Research map / 3D vision / Human pose estimation
HandVoxNet: Deep Voxel-Based Network for 3D Hand Shape and Pose Estimation From a Single Depth Map
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| add V2V-SynNet and S2V-SynNet weak-supervision synthesizers (ablation) | HandVoxNet · SynHand5M test | 3D vertex error (mm): 2.92 -> 2.67 | Helped |
| add 3D joint heatmaps to V2S-Net input (ablation) | V2S-Net with voxelized depth input · SynHand5M | 3D surface vertex error (mm): 8.78 -> 3.36 | Helped |
| add voxelized depth input to V2S-Net (ablation) | V2S-Net with heatmaps · SynHand5M | 3D surface vertex error (mm): 3.54 -> 3.36 | Helped |
| add 3D joint heatmaps to V2V-ShapeNet input (ablation) | V2V-ShapeNet with voxelized depth input · SynHand5M | 3D voxelized shape error: 0.016 -> 0.005 | Helped |
| add voxelized depth input to V2V-ShapeNet (ablation) | V2V-ShapeNet with heatmaps · SynHand5M | 3D voxelized shape error: 0.007 -> 0.005 | Helped |
| register the estimated surface to the voxelized shape with DispVoxNet (ablation) | HandVoxNet · SynHand5M | 3D surface vertex error (mm): 3.36mm to 2.67mm | Helped |
| apply 3D augmentation to voxelized depth maps (ablation) | V2V-PoseNet · NYU | 3D joint error (mm): 9.22 -> 8.72 | Helped |
| apply 3D augmentation to voxelized depth maps (ablation) | V2V-PoseNet · BigHand2.2M | 3D joint error (mm): 9.95 -> 9.27 | Helped |
| train pose estimation jointly with the shape-regression pipeline | HandVoxNet on SynHand5M · SynHand5M test | 3D joint error (mm) | Helped |
Closest papers by meaning
- V2V-PoseNet: Voxel-to-Voxel Prediction Network for Accurate 3D Hand and Human Pose Estimation From a Single Depth Map (CVPR 2018 · similarity 0.94)
- 3D Hand Shape and Pose Estimation From a Single RGB Image (CVPR 2019 · similarity 0.92)
- 3D Convolutional Neural Networks for Efficient and Robust Hand Pose Estimation From Single Depth Images (CVPR 2017 · similarity 0.91)
- 3D Hand Shape and Pose From Images in the Wild (CVPR 2019 · similarity 0.91)
- Reconstructing Hands in 3D with Transformers (CVPR 2024 · similarity 0.89)
- Fine-Grained Multi-View Hand Reconstruction Using Inverse Rendering (AAAI 2024 · similarity 0.87)
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.