Research map / 3D vision / 3D generation
Hierarchical and Holistic Open-Vocabulary Functional 3D Scene Graphs for Indoor Spaces
Xinggang Hu, Chenyangguang Zhang, Alexandros Delitzas, Xiangkui Zhang, Marc Pollefeys, Francis Engelmann, Xiangyang Ji
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 multi-cue node association (vs OpenFunGraph 3D-IoU + CLIP) (ablation) | our pipeline · FunGraph3D | overall node recall: 46.1 -> 60.2 | Improved |
| add multi-cue node association (ablation) | our pipeline · FunGraph3D | overall triplet recall: 40.9 -> 55.6 | Improved |
| add temporal graph optimization (vs Top-1 observation count) (ablation) | our pipeline · FunGraph3D | overall triplet recall: 47.1 -> 55.6 | Improved |
| add temporal graph optimization (ablation) | our pipeline · FunGraph3D | overall node recall: 55.1 -> 60.2 | Improved |
| determine hierarchy in 3D rather than 2D (ablation) | our pipeline · FunGraph3D | hierarchical triplet recall: 44.9 -> 61.9 | Improved |
| determine hierarchy in 3D rather than 2D (ablation) | our pipeline · FunGraph3D | tabletop node recall: 70.3 -> 71.4 | No clear effect |
| add hierarchy prompts to OpenFunGraph | OpenFunGraph Tabletop · FunGraph3D | overall node recall: 41.7 -> 46.3 | Improved |
| add tabletop prompts and category filter to OpenFunGraph | OpenFunGraph · FunGraph3D | tabletop triplet recall: 13.6 -> 42.1 | Improved |
| full method vs best prompt-engineered baseline OpenFunGraph Hierarchy | OpenFunGraph Hierarchy · FunGraph3D | hierarchical triplet recall (FunGraph3D): 16.1 -> 61.9 | Improved |
Closest papers by meaning
- Open-Vocabulary Functional 3D Scene Graphs for Real-World Indoor Spaces (CVPR 2025 · similarity 0.94)
- FunFact: Building Probabilistic Functional 3D Scene Graphs via Factor-Graph Reasoning (CVPR 2026 · similarity 0.92)
- SceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D Scenes (CVPR 2024 · similarity 0.90)
- TB-HSU: Hierarchical 3D Scene Understanding with Contextual Affordances (AAAI 2025 · similarity 0.88)
- InstructScene: Instruction-Driven 3D Indoor Scene Synthesis with Semantic Graph Prior (ICLR 2024 · similarity 0.88)
- Open3DSG: Open-Vocabulary 3D Scene Graphs from Point Clouds with Queryable Objects and Open-Set Relationships (CVPR 2024 · similarity 0.88)
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.
