Research map / 3D vision / Human pose estimation
Identity-Aware Hand Mesh Estimation and Personalization from RGB Images
Deying Kong, Linguang Zhang, Liangjian Chen, Haoyu Ma, Xiangyi Yan, Shanlin Sun, Xingwei Liu, Kun Han, Xiaohui Xie
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 ground-truth identity (MANO shape) input | Baseline hand mesh estimator · DexYCB (no inference optimization) | MPJPE: 21.58 -> 18.83 | Improved |
| add ground-truth identity input | Baseline hand mesh estimator · HUMBI (no inference optimization) | MPJPE: 12.13 -> 11.41 | Improved |
| add ground-truth identity input with inference optimization | Baseline hand mesh estimator · DexYCB (with optimization) | MPJPE: 18.03 -> 16.60 | Improved |
| add ground-truth identity input with inference optimization | Baseline hand mesh estimator · HUMBI (with optimization) | MPJPE: 10.75 -> 10.17 | Improved |
| use calibrated shape instead of GT shape | Identity-aware model · DexYCB (with optimization) | MPJPE: 16.60 -> 16.81 | No clear effect |
| use calibrated shape instead of GT shape | Identity-aware model · HUMBI (with optimization) | MPJPE: 10.17 -> 10.31 | No clear effect |
| add optimization augmented inference | Baseline hand mesh estimator · DexYCB | MPJPE: 21.58 -> 18.03 | Improved |
| add optimization augmented inference | Ours, Calibrated · HUMBI | MPJPE: 11.51 -> 10.31 | Improved |
| add optimization module to Boukhayma et al. | Boukhayma et al. · DexYCB | MPJPE: 27.94 -> 21.20 | Improved |
| add confidence-based attention in calibration (ablation) | Personalization pipeline · HUMBI | MSE_mano: 0.084 -> 0.07 | Improved |
| add confidence-based attention in calibration (ablation) | Personalization pipeline · HUMBI | L-error: 1.93 -> 1.71 | Improved |
| add MANO parameter regressor initialization before optimization (ablation) | Optimization augmented inference · DexYCB and HUMBI | MPJPE: >50 mm without regressor | Improved |
Closest papers by meaning
- Authentic Hand Avatar from a Phone Scan via Universal Hand Model (CVPR 2024 · similarity 0.88)
- Analyzing the Synthetic-to-Real Domain Gap in 3D Hand Pose Estimation (CVPR 2025 · similarity 0.87)
- Monocular 3D Hand Mesh Recovery via Dual Noise Estimation (AAAI 2024 · similarity 0.87)
- SiMA-Hand: Boosting 3D Hand-Mesh Reconstruction by Single-to-Multi-View Adaptation (AAAI 2024 · similarity 0.87)
- HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactions (CVPR 2024 · similarity 0.87)
- SRHand: Super-Resolving Hand Images and 3D Shapes via View/Pose-aware Neural Image Representations and Explicit Meshes (NeurIPS 2025 · 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.
