Research map / 3D vision / 3D generation
AnyHome: Open-Vocabulary Large-Scale Indoor Scene Generation with First-Person View Exploration
Rao Fu, Zehao Wen, Zichen Liu, Srinath Sridhar
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 |
|---|---|---|---|
| use LLM-generated constrained graph layout instead of few-shot LayoutGPT placement (ablation) | layout generation · text-to-scene comparison | OOB rate: 69.4 -> 34.2 | Improved |
| add egocentric depth-conditioned inpainting for texture (ablation) | CG+Retrieval · text-to-scene comparison | CLIP-sim: 11.5 -> 27.9 | Improved |
| add SDS-based layout refinement (ablation) | CG+Inpainting · text-to-scene comparison | OOB rate: 34.2 -> 23.7 | Improved |
| add SDS-based layout refinement (ablation) | CG+Inpainting · text-to-scene comparison | CLIP-style-sim: 17.5 -> 17.5 | No clear effect |
| add SDS-based layout refinement (ablation) | CG+Inpainting · text-to-scene comparison | Caption-sim: 15.9 -> 16.2 | Improved |
| use MVDiffusion images with back-projection instead of single-view Stable Diffusion depth inpainting | texture inpainting · qualitative gothic living room | view consistency | Improved |
| random viewpoint perturbation of camera at evaluation | AnyHome · text-to-scene comparison | variance of scores: 16.2±0.1, 29.3±0.6, 17.5±0.5 | No clear effect |
Later papers that compared with this method
2 later accepted papers report AnyHome as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Edit-As-Act: Goal-Regressive Planning for Open-Vocabulary 3D Indoor Scene Editing (CVPR 2026 · better on 5 of 5 shared results)
- Global-Local Tree Search in VLMs for 3D Indoor Scene Generation (CVPR 2025 · better on 10 of 10 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
- Hierarchically-Structured Open-Vocabulary Indoor Scene Synthesis with Pre-trained Large Language Model (AAAI 2025 · similarity 0.87)
- SceneCraft: Layout-Guided 3D Scene Generation (NeurIPS 2024 · similarity 0.87)
- Direct Numerical Layout Generation for 3D Indoor Scene Synthesis via Spatial Reasoning (NeurIPS 2025 · similarity 0.86)
- InstructScene: Instruction-Driven 3D Indoor Scene Synthesis with Semantic Graph Prior (ICLR 2024 · similarity 0.86)
- SceneWeaver: All-in-One 3D Scene Synthesis with an Extensible and Self-Reflective Agent (NeurIPS 2025 · similarity 0.85)
- Tokenization Allows Multimodal Large Language Models to Understand, Generate and Edit Architectural Floor Plans (CVPR 2026 · similarity 0.84)
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.
