Research map / 3D vision / Human pose estimation
MaskHand: Generative Masked Modeling for Robust Hand Mesh Reconstruction in the Wild
Muhammad Usama Saleem, Ekkasit Pinyoanuntapong, Mayur Jagdishbhai Patel, Hongfei Xue, Ahmed Helmy, Srijan Das, Pu Wang
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 graph-based anatomical pose refinement (GAPR) (ablation) | MaskHand · HO3Dv3 (zero-shot) ablation | PA-MPJPE (mm): 7.3 -> 7.0 | Improved |
| add upsampler (ablation) | MaskHand · HO3Dv3 (zero-shot) ablation | PA-MPJPE (mm): 7.2 -> 7.0 | Improved |
| add 2D pose context (ablation) | MaskHand · HO3Dv3 (zero-shot) ablation | PA-MPJPE (mm): 7.1 -> 7.0 | Improved |
| increase iterative confidence-guided sampling from 1 to 5 iterations (ablation) | MaskHand · HO3Dv3 (zero-shot) | PA-MPJPE (mm): 7.2 -> 7.0 | Improved |
| increase sampling iterations from 1 to 5 (ablation) | MaskHand · FreiHAND | PA-MPJPE (mm): 5.6 -> 5.5 | Improved |
| train with mask ratio U(0,0.7) instead of U(0,0.3) (ablation) | MaskHand · HO3Dv3 (zero-shot) masking ratio | PA-MPJPE (mm): 7.2 -> 7.0 | Improved |
| train with mask ratio U(0,0.7) instead of U(0,1.0) (ablation) | MaskHand · HO3Dv3 (zero-shot) masking ratio | PA-MPVPE (mm): 7.3 -> 7.0 | Improved |
| use all loss terms (L_mask, L_MANO, L_3D, L_2D, beta) with differential sampling (ablation) | MaskHand · HO3Dv2 | PA-MPJPE (mm): 7.6 -> 7.0 | Improved |
| use all loss terms instead of L_3D and beta only (ablation) | MaskHand · FreiHAND | PA-MPJPE (mm): 6.8 -> 5.7 | Improved |
| increase VQ-MANO pose tokens from 16 to 64 (ablation) | MaskHand · VQ-MANO FreiHAND reconstruction | PA-MPJPE (mm): 1.01 -> 0.47 | Improved |
| increase VQ-MANO pose tokens from 64 to 128 (ablation) | MaskHand · VQ-MANO FreiHAND reconstruction | PA-MPJPE (mm): 0.47 -> 0.41 | Improved |
| increase codebook from 1024x256 to 2048x256 (ablation) | MaskHand · VQ-MANO FreiHAND reconstruction | PA-MPJPE (mm): 0.66 -> 0.47 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- H2ONet — H2ONet: Hand-Occlusion-and-Orientation-Aware Network for Real-Time 3D Hand Mesh Reconstruction (CVPR 2023) · better on 1 of 1 shared result
- Deformer — DeFormer: Integrating Transformers with Deformable Models for 3D Shape Abstraction from a Single Image (ICCV 2023) · better on 1 of 1 shared result
- HandOccNet — HandOccNet: Occlusion-Robust 3D Hand Mesh Estimation Network (CVPR 2022) · better on 1 of 1 shared result
- MobRecon — MobRecon: Mobile-Friendly Hand Mesh Reconstruction From Monocular Image (CVPR 2022) · better on 1 of 1 shared result
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
- HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactions (CVPR 2024 · similarity 0.91)
- A Simple Baseline for Efficient Hand Mesh Reconstruction (CVPR 2024 · similarity 0.90)
- Monocular 3D Hand Mesh Recovery via Dual Noise Estimation (AAAI 2024 · similarity 0.90)
- Fine-Grained Multi-View Hand Reconstruction Using Inverse Rendering (AAAI 2024 · similarity 0.89)
- HHMR: Holistic Hand Mesh Recovery by Enhancing the Multimodal Controllability of Graph Diffusion Models (CVPR 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 0.89)
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.
