Research map / 3D vision / 3D generation
HSImul3R: Physics-in-the-Loop Reconstruction of Simulation-Ready Human–Scene Interactions
Yukang Cao, Haozhe Xie, Fangzhou Hong, Long Zhuo, Zhaoxi Chen, Liang Pan, Ziwei 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 |
|---|---|---|---|
| replace HSfM with full HSImul3R pipeline | HSfM · HSIBench Easy | Stability-HSI: 10.52 -> 53.68 | Improved |
| replace HSfM with full HSImul3R pipeline | HSfM · HSIBench | SP-3D penetration: 69.51 -> 22.9 | Improved |
| add scene-targeted distance minimization (Eq. scene loss) (ablation) | HSImul3R · HSIBench Easy | Stability-HSI: 39.56 -> 53.68 | Improved |
| use object-surface distance instead of center-point distance (ablation) | HSImul3R · HSIBench Easy | Stability-HSI: 42.57 -> 53.68 | Improved |
| add DSRO fine-tuning of image-to-3D model (ablation) | HSImul3R · HSIBench Easy | Stability-HSI: 29.56 -> 53.68 | Improved |
| add DSRO fine-tuning of image-to-3D model (ablation) | HSImul3R · HSIBench Hard | Stability-HSI: 5.17 -> 13.92 | Improved |
| add DSRO fine-tuning of image-to-3D model (ablation) | HSImul3R · HSIBench | W-MPJPE: 4.57 -> 4.09 | Improved |
| use DSRO instead of DSO for fine-tuning MIDI | MIDI · HSIBench Easy | Stability-HSI: 38.75 -> 53.68 | Improved |
| add DSRO fine-tuning to MIDI | MIDI · HSIBench | Stability-Gravity: 79.19 -> 91.50 | Improved |
| use DSRO instead of DSO | MIDI · HSIBench | Chamfer Distance: 0.191 -> 0.173 | Improved |
| add 3D structural prior alignment, scene RL and contact optimization to HSfM+MIDI (ablation) | V1 · HSIBench Easy | Stability-HSI: 13.96 -> 29.56 | Improved |
| increase input views from 4 to 16 | HSImul3R · HSIBench Easy | Stability-HSI: 53.68 -> 55.16 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MIDI — MIDI: Multi-Instance Diffusion for Single Image to 3D Scene Generation (CVPR 2025) · better on 2 of 2 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
- Real2Sim in HOI: Toward Physically Plausible HOI Reconstruction from Monocular Videos (NeurIPS 2026 · similarity 0.89)
- Simulation-Ready Compositional 3D Scene Reconstruction from a Single Image (NeurIPS 2026 · similarity 0.88)
- PhyMotion: Structured 3D Motion Reward for Physics-Grounded Human Video Generation (NeurIPS 2026 · similarity 0.87)
- PRISM: Programming Interactive Scenes from Monocular Images for Embodied Simulation (NeurIPS 2026 · similarity 0.87)
- InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactions (CVPR 2025 · similarity 0.87)
- Scaling Up Dynamic Human-Scene Interaction Modeling (CVPR 2024 · 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.
