Research map / 3D vision / Human pose estimation
Low-Dimensionality Calibration Through Local Anisotropic Scaling for Robust Hand Model Personalization
Edoardo Remelli, Anastasia Tkach, Andrea Tagliasacchi, Mark Pauly
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 |
|---|---|---|---|
| decouple pose and shape parameters (theta, beta) instead of position-based sphere parameterization (ablation) | sphere-mesh calibration of a single finger · Synthetic single-finger calibration | shape error E_beta convergence | Improved |
| add heavier damping of pose updates (omega_pose 500 or 4000 vs 0) (ablation) | decoupled calibration · Synthetic hand calibration | shape error E_beta convergence: omega_pose = 0 vs 500 vs 4000 | Improved |
| calibrate in the 5 DOF anisotropic scaling shape-space instead of the full 112 DOF space (ablation) | full hand calibration with large initial perturbation · Synthetic full-hand calibration (sigma_beta .1 to .4) | robustness of E_beta | Mixed |
| add second stage full 112 DOF refinement after the 5 DOF shape-space stage (ablation) | shape-space calibration · Synthetic full-hand calibration (sigma_beta .1 to .4) | shape error E_beta | Improved |
| use the proposed calibration instead of the template model, versus Tkach et al. calibration | hand tracking · HANDY/TEASER tracking (E3D) | tracking performance | Mixed |
Closest papers by meaning
- Fits Like a Glove: Rapid and Reliable Hand Shape Personalization (CVPR 2016 · similarity 0.89)
- Overcoming the Trade-Off Between Accuracy and Plausibility in 3D Hand Shape Reconstruction (CVPR 2023 · similarity 0.85)
- HandDGP: Camera-Space Hand Mesh Prediction with Differentiable Global Positioning (ECCV 2024 · similarity 0.85)
- Identity-Aware Hand Mesh Estimation and Personalization from RGB Images (ECCV 2022 · similarity 0.85)
- GEARS: Local Geometry-aware Hand-object Interaction Synthesis (CVPR 2024 · similarity 0.84)
- ComPose: When to Trust Hands for Object Pose Tracking (NeurIPS 2026 · similarity 0.84)
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.
