Research map / 3D vision / Face and shape
PhysSkin: Real-Time and Generalizable Physics-Based Animation via Self-Supervised Neural Skinning
Yuanhang Lei, Tao Cheng, Xingxuan Li, Boming Zhao, Siyuan Huang, Ruizhen Hu, Peter Yichen Chen, Hujun Bao, Zhaopeng Cui
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 on-the-fly skinning-field normalization (ablation) | PhysSkin · RigNet dataset | Orthogonality Ωorth x 10^-2: 6.5533 -> 0.0033 | Improved |
| add on-the-fly skinning-field normalization (ablation) | PhysSkin · RigNet dataset | Spectrum entropy Hspec: 0.8113 -> 0.9999 | Improved |
| add ConFIG conflict-aware gradient correction (ablation) | PhysSkin · RigNet dataset | Orthogonality Ωorth x 10^-2: 8.9247 -> 0.0033 | Improved |
| add ConFIG conflict-aware gradient correction (ablation) | PhysSkin · RigNet dataset | Log-condition number κlog: 11.8595 -> 1.0453 | Improved |
| add orthogonality loss Lorth (ablation) | PhysSkin · RigNet dataset | Orthogonality Ωorth x 10^-2: 100.0 -> 0.0033 | Improved |
| add ONI orthogonalization layer (ablation) | PhysSkin · RigNet dataset | Orthogonality Ωorth x 10^-2: 0.0081 -> 0.0033 | Improved |
| add spatial smoothness loss Lsmooth (ablation) | PhysSkin · RigNet dataset | Orthogonality Ωorth x 10^-2: 0.0050 -> 0.0033 | Improved |
| add potential energy loss Lpot (ablation) | PhysSkin · RigNet dataset | physical interpretability (qualitative) | Improved |
| generalizable feed-forward skinning versus per-object Simplicits | PhysSkin · ShapeNet | Orthogonality Ωorth x 10^-2: 0.2621 -> 0.0098 | Improved |
| self-supervised skinning versus supervised rigging methods | PhysSkin · RigNet dataset (unseen objects) | Log-condition number κlog: 2.6093 -> 1.0453 | Improved |
| subspace skinning dynamics instead of full-space FEM | PhysSkin · Bag (121K vertices) | Step cost (ms): 3012.47 -> 13.39 | Improved |
| subspace skinning dynamics instead of explicit MPM | PhysSkin · Airplane (10K vertices) | Step cost (ms): 141.83 -> 12.26 | 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.
- RigNet — RigNet: Repetitive Image Guided Network for Depth Completion (ECCV 2022) · better on 4 of 4 shared results
- Puppeteer — Puppeteer: Rig and Animate Your 3D Models (NeurIPS 2025) · 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
- PhysHO: Physics-Based Dynamic 3D Gaussian Human and Object from Monocular Video (CVPR 2026 · similarity 0.88)
- PhysCtrl: Generative Physics for Controllable and Physics-Grounded Video Generation (NeurIPS 2025 · similarity 0.86)
- PhysAnimator: Physics-Guided Generative Cartoon Animation (CVPR 2025 · similarity 0.86)
- PhysX-3D: Physical-Grounded 3D Asset Generation (NeurIPS 2025 · similarity 0.84)
- PhyRecon: Physically Plausible Neural Scene Reconstruction (NeurIPS 2024 · similarity 0.84)
- PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics (CVPR 2024 · 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.
