Research map / 3D vision / Human pose estimation
Towards Accurate Alignment in Real-Time 3D Hand-Mesh Reconstruction
Xiao Tang, Tianyu Wang, Chi-Wing Fu
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 joint stage (ablation) | mesh+refine pipeline · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 92.42 -> 92.95 | Improved |
| add refine stage (ablation) | joint+mesh pipeline · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 92.05 -> 92.95 | Improved |
| add refine stage (ablation) | joint+mesh pipeline · FreiHAND train-split validation (ablation, hand-level alignment) | HD: 5.93 -> 4.70 | Improved |
| add local projection unit (ablation) | refine stage · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 92.55 -> 92.95 | Improved |
| add global broadcast unit (ablation) | refine stage · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 92.63 -> 92.95 | Improved |
| regress per-vertex offsets instead of direct vertex coordinates (ablation) | refine stage Graph-CNN · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 88.01 -> 92.95 | Improved |
| regress per-vertex offsets instead of direct vertex coordinates (ablation) | refine stage Graph-CNN · FreiHAND train-split validation (ablation, hand-level alignment) | Mesh AUC: 0.782 -> 0.866 | Improved |
| add segmentation branch and silhouette loss (ablation) | full method · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 92.50 -> 92.95 | Improved |
| add render loss (ablation) | full method · FreiHAND train-split validation (ablation, hand-level alignment) | mIoU: 92.80 -> 92.95 | Improved |
| add render loss (ablation) | full method · FreiHAND train-split validation (ablation, hand-level alignment) | HD: 4.97 -> 4.70 | Improved |
| add ObMan synthetic data to training | full method · FreiHAND test | ME (10mm): 0.71 -> 0.67 | Improved |
| replace I2L-MeshNet with decoupled three-stage method | 3D hand mesh reconstruction · FreiHAND finger-level alignment | mIoU: 71.07 -> 77.33 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- I2L-MeshNet — I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human Pose and Mesh Estimation from a Single RGB Image (ECCV 2020) · better on 8 of 8 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- SiMA-Hand: Boosting 3D Hand-Mesh Reconstruction by Single-to-Multi-View Adaptation (AAAI 2024 · similarity 0.89)
- SRHand: Super-Resolving Hand Images and 3D Shapes via View/Pose-aware Neural Image Representations and Explicit Meshes (NeurIPS 2025 · similarity 0.88)
- Monocular 3D Hand Mesh Recovery via Dual Noise Estimation (AAAI 2024 · similarity 0.87)
- Fine-Grained Multi-View Hand Reconstruction Using Inverse Rendering (AAAI 2024 · similarity 0.87)
- Learning Explicit Contact for Implicit Reconstruction of Hand-Held Objects from Monocular Images (AAAI 2024 · similarity 0.87)
- HandOS: 3D Hand Reconstruction in One Stage (CVPR 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.
