Research map / 3D vision / 3D generation
Controllable Human-Object Interaction Synthesis
Jiaman Li, Alexander Clegg, Roozbeh Mottaghi, Jiajun Wu, Xavier Puig, C. Karen 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 |
|---|---|---|---|
| add object geometry loss during training (ablation) | CHOIS · FullBodyManipulation test | T_xy (cm): 8.44 -> 2.69 | Improved |
| add object geometry loss during training (ablation) | CHOIS · FullBodyManipulation test | T_e (cm): 14.16 -> 6.61 | Improved |
| add contact and floor guidance at inference (ablation) | CHOIS · FullBodyManipulation test | contact recall: 0.49 -> 0.64 | Improved |
| add guidance at inference (ablation) | CHOIS · FullBodyManipulation test | H_feet (cm): 6.64 -> 4.20 | Improved |
| add guidance at inference (ablation) | CHOIS · FullBodyManipulation test | FID: 3.58 -> 0.69 | Improved |
| add guidance at inference (ablation) | CHOIS · FullBodyManipulation test | T_xy (cm): 2.69 -> 2.87 | Hurt |
| add hand-object contact guidance (ablation) | CHOIS · FullBodyManipulation test | contact percentage: 0.41 -> 0.54 | Improved |
| add feet-floor contact guidance (ablation) | CHOIS · FullBodyManipulation test | H_feet (cm): 6.65 -> 4.20 | Improved |
| add guidance at inference on unseen objects (ablation) | CHOIS · 3D-FUTURE (unseen objects) | FID: 4.78 -> 1.60 | Improved |
| add guidance in long-term synthesis (ablation) | CHOIS · Long-term synthesis (FullBodyManipulation) | contact percentage: 0.49 -> 0.63 | Mixed |
Closest papers by meaning
- CG-HOI: Contact-Guided 3D Human-Object Interaction Generation (CVPR 2024 · similarity 0.91)
- NIFTY: Neural Object Interaction Fields for Guided Human Motion Synthesis (CVPR 2024 · similarity 0.90)
- InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction (NeurIPS 2024 · similarity 0.90)
- InteractAnything: Zero-shot Human Object Interaction Synthesis via LLM Feedback and Object Affordance Parsing (CVPR 2025 · similarity 0.90)
- InterControl: Zero-shot Human Interaction Generation by Controlling Every Joint (NeurIPS 2024 · similarity 0.90)
- InterAct: Advancing Large-Scale Versatile 3D Human-Object Interaction Generation (CVPR 2025 · 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.
