Research map / 3D vision / 3D generation
ManiDext: Hand-Object Manipulation Synthesis via Continuous Correspondence Embeddings and Residual-Guided Diffusion
Jiajun Zhang, Yuxiang Zhang, Liang An, Mengcheng Li, Hongwen Zhang, Zonghai Hu, Yebin Liu
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 residual-guided diffusion (ablation) | C&E-Map* · ARCTIC | Pen: 5.38 -> 4.65 | Improved |
| add residual-guided diffusion (ablation) | C&E-Map* · ARCTIC | V-MPVPE: 3.27 -> 2.98 | Improved |
| add residual-guided diffusion (ablation) | C&E-Map* · GRAB | V-Contact Ratio: 0.951 -> 0.972 | Improved |
| add residual-guided diffusion (ablation) | C&E-Map* · GRAB | P-Score: 3.60 -> 4.02 | Improved |
| add contact map in a two-stage framework (ablation) | MDM† · ARCTIC | V-Contact Ratio: 0.826 -> 0.937 | Improved |
| add contact map in a two-stage framework (ablation) | MDM† · ARCTIC | Pen: 5.74 -> 6.54 | Hurt |
| add continuous correspondence embedding map (ablation) | C-Map · GRAB | Pen: 5.31 -> 4.53 | Improved |
| add correspondence losses (L_att, L_consist) (ablation) | C&E-Map · ARCTIC | V-MPVPE: 3.39 -> 3.27 | Improved |
| use object canonical coordinates instead of world coordinates (ablation) | ManiDext · ARCTIC | Pen: 6.12 -> 4.65 | Improved |
| use object canonical coordinates instead of world coordinates (ablation) | ManiDext · ARCTIC | V-MPVPE: 4.21 -> 2.98 | Improved |
| use continuous correspondence embedding instead of discrete part labels (ablation) | ManiDext · ARCTIC | V-Contact Ratio: 0.941 -> 0.965 | Improved |
| use continuous correspondence embedding instead of canonical vertex coordinates (ablation) | ManiDext · ARCTIC | V-MPVPE: 3.51 -> 2.98 | Improved |
Closest papers by meaning
- Contact Map Transfer with Conditional Diffusion Model for Generalizable Dexterous Grasp Generation (NeurIPS 2025 · similarity 0.91)
- DiffGrasp: Whole-Body Grasping Synthesis Guided by Object Motion Using a Diffusion Model (AAAI 2025 · similarity 0.90)
- LatentHOI: On the Generalizable Hand Object Motion Generation with Latent Hand Diffusion. (CVPR 2025 · similarity 0.90)
- CoDA: Coordinated Diffusion Noise Optimization for Whole-Body Manipulation of Articulated Objects (NeurIPS 2025 · similarity 0.90)
- HandDiffuse: Generative Controllers for Two-Hand Interactions via Diffusion Models (AAAI 2025 · similarity 0.89)
- Single-view Image to Novel-view Generation for Hand-Object Interactions (AAAI 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.
