Research map / 3D vision / 3D generation
Scene Generation at Absolute Scale: Utilizing Semantic and Geometric Guidance From Text for Accurate and Interpretable 3D Indoor Scene Generation
Stefan Ainetter, Thomas Deixelberger, Edoardo Dominici, Philipp Drescher, Konstantinos Vardis, Markus Steinberger
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-ControlNet (depth + semantic) conditioning (ablation) | panorama generator · Structured3D panoramas | CLIP Score: 31.17 -> 32.48 | Improved |
| add Multi-ControlNet (depth + semantic) conditioning (ablation) | panorama generator · Structured3D panoramas | Q-Align: 4.45 -> 4.54 | Improved |
| add depth ControlNet alone (ablation) | panorama generator · Structured3D panoramas | CLIP Score: 31.17 -> 31.85 | Improved |
| add semantic ControlNet alone (ablation) | panorama generator · Structured3D panoramas | CLIP Score: 31.17 -> 32.23 | Improved |
| add semantic ControlNet alone (ablation) | panorama generator · Structured3D panoramas | IS: 3.33 -> 3.41 | Improved |
| add depth and semantic ControlNets together vs semantic only (ablation) | panorama generator · Structured3D panoramas | IS: 3.41 -> 3.26 | Hurt |
| use optimized collision-aware overlapping trajectories instead of coverage-maximizing view planning (ablation) | trajectory sampling · Structured3D prompts | MEt3R: 0.029 -> 0.024 | Improved |
| add depth filter to standard view planning (ablation) | trajectory sampling · Structured3D prompts | MEt3R: 0.029 -> 0.027 | Improved |
| add depth-guided camera scale alignment (ablation) | Stable Virtual Camera trajectories · Structured3D prompts | camera pose alignment | Improved |
| replace WorldExplorer exhaustive trajectories with optimized sampling | GuidedSceneGen · Structured3D prompts | runtime per scene: 7h -> 0.75h | Improved |
| use GuidedSceneGen instead of DreamScene360 | 3D scene generation · Structured3D prompts | MEt3R: 0.026 -> 0.024 | Improved |
| use GuidedSceneGen instead of WorldExplorer | 3D scene generation · Structured3D prompts | IS: 2.27 -> 2.22 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DreamScene360 — DreamScene360: Unconstrained Text-to-3D Scene Generation with Panoramic Gaussian Splatting (ECCV 2024) · better on 4 of 4 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
- SceneCraft: Layout-Guided 3D Scene Generation (NeurIPS 2024 · similarity 0.90)
- InstructScene: Instruction-Driven 3D Indoor Scene Synthesis with Semantic Graph Prior (ICLR 2024 · similarity 0.90)
- Direct Numerical Layout Generation for 3D Indoor Scene Synthesis via Spatial Reasoning (NeurIPS 2025 · similarity 0.89)
- SceneFactor: Factored Latent 3D Diffusion for Controllable 3D Scene Generation (CVPR 2025 · similarity 0.88)
- 3D-SceneDreamer: Text-Driven 3D-Consistent Scene Generation (CVPR 2024 · similarity 0.88)
- Hierarchically-Structured Open-Vocabulary Indoor Scene Synthesis with Pre-trained Large Language Model (AAAI 2025 · 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.
