Research map / 3D vision / 3D generation
Stability-Driven Contact Reconstruction From Monocular Color Images
Zimeng Zhao, Binghui Zuo, Wei Xie, Yangang Wang
Check 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 stability-driven optimization to the regressed state | Hand-object reconstruction · FPHB | Max Pene. (mm)↓: 15.07 -> 11.43 | Improved |
| Add stability-driven optimization to the regressed state | Hand-object reconstruction · HO3Dv3 | Max Pene. (mm)↓: 16.85 -> 11.36 | Improved |
| Add stability-driven optimization to the regressed state | Hand-object reconstruction · DexYCB | Max Pene. (mm)↓: 7.32 -> 6.72 | Improved |
| Add physics-based stability optimization | Hand-object contact recovery · ContactPose | SC.↓: 1.03 -> 0.27 | Improved |
| Add physics-based stability optimization | Hand-object contact recovery · GRAB_rh50 | SC.↓: 1.28 -> 0.44 | Improved |
| Add collision loss L_in during training (ablation) | Hand-object reconstruction pipeline · ContactPose | SC.↓: 0.86 -> 0.31 | Improved |
| Add semi-supervised pre-training (ablation) | Hand-object reconstruction pipeline · ContactPose | SC.↓: 0.51 -> 0.31 | Improved |
| Add state-consistency term C_S (ablation) | Stability-cost optimization · ContactPose | SC.↓: 4.62 -> 0.31 | Improved |
| Add contact-count term L_cnt (ablation) | Stability-cost optimization · ContactPose | SC.↓: 0.58 -> 0.31 | Improved |
| Add resultant-force term L_frc (ablation) | Stability-cost optimization · ContactPose | SC.↓: 0.73 -> 0.31 | Improved |
| Add torque term L_tau (ablation) | Stability-cost optimization · ContactPose | SC.↓: 0.64 -> 0.31 | Improved |
| Optimize with the full stability cost rather than omit it (ablation) | Hand-object contact recovery · ContactPose | SC.↓: 1.44 -> 0.31 | Improved |
Closest papers by meaning
- Learning Explicit Contact for Implicit Reconstruction of Hand-Held Objects from Monocular Images (AAAI 2024 · similarity 0.89)
- ContactPose: A Dataset of Grasps with Object Contact and Hand Pose (ECCV 2020 · similarity 0.89)
- HOLD: Category-agnostic 3D Reconstruction of Interacting Hands and Objects from Video (CVPR 2024 · similarity 0.88)
- Learning Dense Hand Contact Estimation from Imbalanced Data (NeurIPS 2025 · similarity 0.87)
- How Do I Do That? Synthesizing 3D Hand Motion and Contacts for Everyday Interactions (CVPR 2025 · similarity 0.87)
- Hand-held Object Reconstruction from RGB Video with Dynamic Interaction (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.
