Research map / 3D vision / Depth and stereo
MLPHand: Real Time Multi-View 3D Hand Reconstruction via MLP Modeling
Jian Yang, Jiakun Li, Guoming Li, Huaiyu Wu, Zhen Shen, Zhaoxin Fan
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 |
|---|---|---|---|
| use MLPHand instead of POEM at ResNet34 | MLPHand · OakInk-MV | FPS: 19 -> 57 | Improved |
| replace POEM refinement with MLP Skeleton2Mesh | MLPHand · OakInk-MV skeleton-to-mesh | Multi-Adds: 6.76G -> 0.05G | Improved |
| MLPHand vs POEM accuracy on HO3D-MV | MLPHand · HO3D-MV | MPVPE (mm): 17.2 -> 18.69 | Hurt |
| add multi-view feature infuser (MGFP) to Skeleton2Mesh (ablation) | MLPHand ResNet34 · DexYCB-MV | MPVPE (mm): 6.93 -> 6.28 | Improved |
| add Global Spatial Descriptor (ablation) | Skeleton2Mesh · FreiHAND | MPVPE (MV): 6.78 -> 1.36 | Improved |
| add position encoding (ablation) | Skeleton2Mesh · FreiHAND | MPVPE (MV): 1.39 -> 1.36 | Improved |
| reduce MLP depth from 3 to 2 layers (ablation) | Skeleton2Mesh · FreiHAND | MPVPE (MV): 1.36 -> 1.71 | Hurt |
| increase MLP depth from 3 to 5 layers (ablation) | Skeleton2Mesh · FreiHAND | MPVPE (MV): 1.36 -> 1.30 | Improved |
| use per-bone parameter sharing vs separate networks (ablation) | Skeleton2Mesh · FreiHAND | MPVPE (MV): 1.23 -> 1.36; Params 7.23M -> 0.50M | Mixed |
| add two-stage training (pretrain Skeleton2Mesh first) (ablation) | MLPHand · OakInk-MV ResNet34 | MPVPE (mm): 6.64 -> 6.00 | Improved |
Later papers that compared with this method
1 later accepted paper reports MLPHand as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- JUMP-Hand: Learning Joint-wise Uncertainty to Gate Mixture of View Experts for Multi-View 3D Hand Reconstruction (CVPR 2026 · better on 3 of 3 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 0.88)
- Fine-Grained Multi-View Hand Reconstruction Using Inverse Rendering (AAAI 2024 · similarity 0.88)
- HandOS: 3D Hand Reconstruction in One Stage (CVPR 2025 · similarity 0.87)
- HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactions (CVPR 2024 · similarity 0.86)
- SRHand: Super-Resolving Hand Images and 3D Shapes via View/Pose-aware Neural Image Representations and Explicit Meshes (NeurIPS 2025 · similarity 0.86)
- Reconstructing Hands in 3D with Transformers (CVPR 2024 · similarity 0.86)
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.
