Research map / 3D vision / Human pose estimation
HuCollisionField: Resolving Self-Collisions via Neural Fields for Human Prediction
Zhengyuan Li, Zeyun Deng, Yifan Shen, Liang-Yan Gui, Miaolan Xie, Joseph Campbell, Xifeng Gao, Kui Wu, Zherong Pan, Aniket Bera
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 PoseShield instead of COAP | collision resolution · HwC | success rate: 0.446 -> 0.958 | Improved |
| add weighted distance (WD) in pose space (ablation) | PoseShield · HwC | mean vertex distance: 0.067 -> 0.059 | Improved |
| add weighted distance (WD) in pose space (ablation) | PoseShield · HwC | success rate: 0.960 -> 0.958 | Mixed |
| add weighted distance (WD) in pose space (ablation) | PoseShield · PROX | success rate: 0.850 -> 0.800 | Hurt |
| use LTD instead of Lgrad (ablation) | PoseShield · HwC | success rate: 0.862 -> 0.958 | Improved |
| use LTD instead of Lgrad (ablation) | PoseShield · PROX | success rate: 0.520 -> 0.800 | Improved |
| add Lgrad on top of LTD (ablation) | PoseShield · HwC | success rate: 0.958 -> 0.870 | Hurt |
| add Eikonal gradient terms (Lgrad or LTD) (ablation) | PoseShield · HwC | success rate: 0.068 -> 0.958 | Improved |
| add Eikonal gradient terms (Lgrad or LTD) (ablation) | PoseShield · PROX | success rate: 0.050 -> 0.800 | Improved |
| increase constraint margin Cl (ablation) | PoseShield · HwC | SCC vs MVD | Mixed |
| use neural field instead of classifier probability constraint | latent-space optimization · HwC | success rate: 0.056 -> 0.958 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- VolumetricSMPL — VolumetricSMPL: A Neural Volumetric Body Model for Efficient Interactions, Contacts, and Collisions (ICCV 2025) · better on 5 of 5 shared results
- COAP — COAP: Compositional Articulated Occupancy of People (CVPR 2022) · better on 5 of 5 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
- LCollision: Fast Generation of Collision-Free Human Poses using Learned Non-Penetration Constraints (AAAI 2021 · similarity 0.89)
- CLOAF: CoLlisiOn-Aware Human Flow (CVPR 2024 · similarity 0.85)
- Optimal-state Dynamics Estimation for Physics-based Human Motion Capture from Videos (NeurIPS 2024 · similarity 0.85)
- CoDA: Coordinated Diffusion Noise Optimization for Whole-Body Manipulation of Articulated Objects (NeurIPS 2025 · similarity 0.84)
- NIFTY: Neural Object Interaction Fields for Guided Human Motion Synthesis (CVPR 2024 · similarity 0.84)
- InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactions (CVPR 2025 · similarity 0.84)
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.
