Research map / 3D vision / 3D generation
InterFusion: Text-Driven Generation of 3D Human-Object Interaction
Sisi Dai, Wenhao Li, Haowen Sun, Haibin Huang, Chongyang Ma, Hui Huang, Kai Xu, Ruizhen Hu
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-level SDS (whole HOI scene guidance) (ablation) | InterFusion · 61 prompts | CLIP score: 0.3164 -> 0.3308 | Improved |
| add interaction-level SDS (ablation) | InterFusion · 61 prompts | GPT-4V select (%): 1.64 -> 77.05 | Improved |
| add object-level SDS (ablation) | InterFusion · 61 prompts | CLIP score: 0.3293 -> 0.3308 | No clear effect |
| add object-level SDS (ablation) | InterFusion · 61 prompts | GPT-4V select (%): 4.92 -> 77.05 | Improved |
| add object geometric constraint (keep object out of anchor pose) (ablation) | InterFusion · 61 prompts | CLIP score: 0.3171 -> 0.3308 | Improved |
| add object geometric constraint (ablation) | InterFusion · 61 prompts | GPT-4V select (%): 16.39 -> 77.05 | Improved |
| use human pose prior instead of object prior (ShapeNet object) (ablation) | InterFusion · 61 prompts | GPT-4V select (%): 13.11 -> 65.57 | Improved |
| use InterFusion instead of direct text-to-3D methods | Text-to-3D (DreamFusion, Magic3D, TextMesh) · 61 prompts | CLIP score: 0.3179 (Magic3D) -> 0.3308 | Improved |
| use synthesized-image pseudo poses instead of AMASS mocap codebook (AvatarCLIP) (ablation) | Anchor pose generation · Qualitative | pose fit to interaction | Improved |
Later papers that compared with this method
1 later accepted paper reports InterFusion as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- AnchorHOI: Zero-shot Generation of 4D Human-Object Interaction via Anchor-based Prior Distillation (AAAI 2026 · 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
- InteractAnything: Zero-shot Human Object Interaction Synthesis via LLM Feedback and Object Affordance Parsing (CVPR 2025 · similarity 0.92)
- InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction (NeurIPS 2024 · similarity 0.91)
- InterAct: Advancing Large-Scale Versatile 3D Human-Object Interaction Generation (CVPR 2025 · similarity 0.91)
- CG-HOI: Contact-Guided 3D Human-Object Interaction Generation (CVPR 2024 · similarity 0.90)
- Semantic-Aware Human Object Interaction Image Generation (ICML 2024 · similarity 0.90)
- GenZI: Zero-Shot 3D Human-Scene 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.
