Research map / 3D vision / 3D generation
InterDiff: Generating 3D Human-Object Interactions with Physics-Informed Diffusion
Sirui Xu, Zhengyuan Li, Yu-Xiong Wang, Liang-Yan Gui
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 interaction correction step to diffusion (ablation) | InterDiff · BEHAVE | Penetration: 228 -> 164 | Improved |
| add interaction correction step (ablation) | InterDiff · BEHAVE | Rot. Err.: 256 -> 226 | Improved |
| add interaction correction step (ablation) | InterDiff · BEHAVE | MPJPE-H: 140 -> 140 | No clear effect |
| add interaction correction on unseen objects (ablation) | InterDiff · Human-Object Interaction dataset | MPJPE-O: 117 -> 84 | Improved |
| add interaction correction on long-term forecasting (ablation) | InterDiff · BEHAVE 100 frames, 1 sample | Pene.: 236 -> 88 | Improved |
| remove coordinate transformation to contact reference system (w/o relative) (ablation) | interaction correction · BEHAVE | physical fidelity user study: full wins 67.8% | Hurt |
| remove contact and penetration losses (w/o contact) (ablation) | interaction predictor · BEHAVE | realism (qualitative) | Hurt |
| apply correction at every step without scheduler (w/o schedule) (ablation) | interaction correction · BEHAVE | contact floating (qualitative) | Hurt |
| increase number of DCT bases in interaction predictor (ablation) | interaction predictor · BEHAVE long-term | translation vs rotation error | Mixed |
| encode history as condition instead of infilling ground-truth past | interaction diffusion · BEHAVE | performance | Improved |
Closest papers by meaning
- CG-HOI: Contact-Guided 3D Human-Object Interaction Generation (CVPR 2024 · similarity 0.92)
- InterAct: Advancing Large-Scale Versatile 3D Human-Object Interaction Generation (CVPR 2025 · similarity 0.92)
- HOI-Dyn: Learning Interaction Dynamics for Human-Object Motion Diffusion (NeurIPS 2025 · similarity 0.92)
- InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction (NeurIPS 2024 · similarity 0.90)
- Guiding Human-Object Interactions with Rich Geometry and Relations (CVPR 2025 · similarity 0.89)
- Template Free Reconstruction of Human-object Interaction with Procedural Interaction Generation (CVPR 2024 · similarity 0.89)
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.
