Research map / 3D vision / 3D generation
Scenethesis: A Language and Vision Agentic Framework for 3D Scene Generation
Lu Ling, Chen-Hsuan Lin, Tsung-Yi Lin, Yifan Ding, Yu Zeng, Yichen Sheng, Yunhao Ge, Ming-Yu Liu, Aniket Bera, Max Li
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 pose alignment (ablation) | Scenethesis raw layout · Physics-aware optimization ablation | pose alignment: 0.536 -> 0.732 | Improved |
| add pose alignment (ablation) | Scenethesis raw layout · Physics-aware optimization ablation | collision rate: 22.7% -> 10.6% | Improved |
| add collision constraint (ablation) | Scenethesis · Physics-aware optimization ablation | collision rate: 10.6% -> 3.6% | Improved |
| add collision constraint (ablation) | Scenethesis · Physics-aware optimization ablation | instability rate: 74.2% -> 69.8% | Improved |
| add stability constraint (ablation) | Scenethesis · Physics-aware optimization ablation | instability rate: 69.8% -> 3.2% | Improved |
| add stability constraint (ablation) | Scenethesis · Physics-aware optimization ablation | collision rate: 3.6% -> 0.8% | Improved |
| add stability constraint (ablation) | Scenethesis · Physics-aware optimization ablation | pose alignment: 0.755 -> 0.836 | Improved |
| use Scenethesis vision-guided layout instead of Holodeck LLM-only layout | Scenethesis · Indoor scenes, text-image alignment | BLIP score: 46.25 -> 77.17 | Improved |
| use Scenethesis instead of Holodeck | Scenethesis · Indoor scenes, physical plausibility | Col-S: 21% -> 6% | Improved |
| use Scenethesis instead of Holodeck for walkability | Scenethesis · Indoor scenes, interactivity | Walk: 0.96 -> 0.96 | No clear effect |
| use Scenethesis instead of Holodeck for reachability | Scenethesis · Indoor scenes, interactivity | Reach: 0.90 -> 0.94 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- Holodeck — Holodeck: Language Guided Generation of 3D Embodied AI Environments (CVPR 2024) · better on 8 of 9 shared results
- DiffuScene — DiffuScene: Denoising Diffusion Models for Generative Indoor Scene Synthesis (CVPR 2024) · better on 4 of 4 shared results
- SceneTeller — SceneTeller: Language-to-3D Scene Generation (ECCV 2024) · better on 4 of 4 shared results
- PhyScene — PhyScene: Physically Interactable 3D Scene Synthesis for Embodied AI (CVPR 2024) · better on 1 of 1 shared result
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
- SceneWeaver: All-in-One 3D Scene Synthesis with an Extensible and Self-Reflective Agent (NeurIPS 2025 · similarity 0.91)
- Direct Numerical Layout Generation for 3D Indoor Scene Synthesis via Spatial Reasoning (NeurIPS 2025 · similarity 0.90)
- InstructScene: Instruction-Driven 3D Indoor Scene Synthesis with Semantic Graph Prior (ICLR 2024 · similarity 0.90)
- SceneCraft: An LLM Agent for Synthesizing 3D Scenes as Blender Code (ICML 2024 · similarity 0.89)
- LayoutVLM: Differentiable Optimization of 3D Layout via Vision-Language Models (CVPR 2025 · similarity 0.88)
- 3D Space as a Scratchpad for Editable Text-to-Image Generation (CVPR 2026 · 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.
