Research map / 3D vision / Human pose estimation
Weakly-Supervised 3D Hand Reconstruction with Knowledge Prior and Uncertainty Guidance
Yufei Zhang, Jeffrey Kephart, Qiang Ji
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 |
|---|---|---|---|
| add hand biomechanics loss (ablation) | weakly-supervised 3D hand reconstruction · FreiHAND | E_J (mm): 22.4 -> 10.9 | Improved |
| add functional anatomy loss (ablation) | weakly-supervised 3D hand reconstruction · FreiHAND | E_J (mm): 10.9 -> 9.6 | Improved |
| add physics non-penetration loss (ablation) | weakly-supervised 3D hand reconstruction · FreiHAND | penetration rate (%): 11.4 -> 1.9 | Improved |
| add non-penetration loss (ablation) | weakly-supervised 3D hand reconstruction · FreiHAND | E_J (mm): 9.6 -> 9.4 | Improved |
| train with NLL uncertainty loss instead of MSE (ablation) | weakly-supervised 3D hand reconstruction · FreiHAND | E_J (mm): 9.4 -> 8.5 | Improved |
| use generic hand knowledge prior instead of data-driven FreiHAND-3D prior (ablation) | DexYCB training with 2D labels · DexYCB | E_J (mm): 8.3 -> 6.7 | Improved |
| add 10% 3D annotations to the weakly-supervised method (ablation) | weakly-supervised hand reconstruction · FreiHAND | E_J (mm): 8.52 -> 8.26 | Improved |
| weakly-supervised method without NLL versus second-best prior method | weakly-supervised 3D hand reconstruction · FreiHAND | E_J (mm): 10.7 -> 9.4 | Improved |
Closest papers by meaning
- Weakly Supervised 3D Hand Pose Estimation via Biomechanical Constraints (ECCV 2020 · similarity 0.92)
- Model-Based 3D Hand Reconstruction via Self-Supervised Learning (CVPR 2021 · similarity 0.90)
- Consistent 3D Hand Reconstruction in Video via Self-Supervised Learning (TPAMI 2023 · similarity 0.90)
- HOLD: Category-agnostic 3D Reconstruction of Interacting Hands and Objects from Video (CVPR 2024 · similarity 0.89)
- HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactions (CVPR 2024 · similarity 0.88)
- Reconstructing Hands in 3D with Transformers (CVPR 2024 · similarity 0.88)
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.
