Research map / 3D vision / Face and shape
N-Penetrate: Active Learning of Neural Collision Handler for Complex 3D Mesh Deformations
Qingyang Tan, Zherong Pan, Breannan Smith, Takaaki Shiratori, Dinesh Manocha
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 active, risk-seeking data aggregation instead of random supervised sampling | Neural collision detector · SCAPE, Swing, Jump, Skirt, and Hand | accuracy: 0.9282 -> 0.9383; 0.9609 -> 0.9638; 0.9500 -> 0.9552; 0.9795 -> 0.9817; 0.9650 -> 0.9692 | Improved |
| Use active, risk-seeking data aggregation instead of random supervised sampling | Neural collision detector · SCAPE, Swing, Jump, Skirt, and Hand | false negative rate: 0.05576 -> 0.05151; 0.01766 -> 0.01485; 0.02644 -> 0.02573; 0.01956 -> 0.01808; 0.03652 -> 0.03582 | Improved |
| Use active, risk-seeking data aggregation instead of random supervised sampling | Collision handler using ALM · Unseen self-penetrating meshes | collision-handling success rate: 24.74% more collisions than Supv + bd | Improved |
| Add boundary loss for samples near the collision decision boundary (ablation) | Supervised collision detector · Collision handling | collision-handling success rate | Improved |
| Add active data aggregation that samples toward the learned decision boundary | Collision handler using ALM · Collision-handling success rate across dataset sizes | collision-handling success rate | Improved |
| Use active + bd instead of Supv + bd | Collision handler on SCAPE · SCAPE | mean relative PD reduction: 88.76% -> 94.81% | Improved |
| Use active + bd instead of Supv + bd | Collision handler on SCAPE · SCAPE | mean embedding difference: 66.11 -> 56.17 | Improved |
Closest papers by meaning
- Learning Flexible Body Collision Dynamics with Hierarchical Contact Mesh Transformer (ICLR 2024 · similarity 0.87)
- Structure from Collision (CVPR 2025 · similarity 0.84)
- Learning rigid-body simulators over implicit shapes for large-scale scenes and vision (NeurIPS 2024 · similarity 0.83)
- Implicit-ARAP: Efficient Handle-Guided Neural Field Deformation via Local Patch Meshing (NeurIPS 2025 · similarity 0.83)
- Embodied Active Defense: Leveraging Recurrent Feedback to Counter Adversarial Patches (ICLR 2024 · similarity 0.83)
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (CVPR 2025 · similarity 0.83)
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.
