Research map / 3D vision / 3D generation
General 3D Room Layout from a Single View by Render-and-Compare
Sinisa Stekovic, Shreyas Hampali, Mahdi Rad, Sayan Deb Sarkar, Friedrich Fraundorfer, Vincent Lepetit
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 general polygon layout optimization instead of cuboid assumption (Hirzer et al.) | Ours (RGB) · ScanNet-Layout | IoU (%): 48.6 ± 22.2 -> 63.5 ± 25.2 | Improved |
| add depth map input (RGBD instead of RGB) | Ours · ScanNet-Layout | IoU (%): 63.5 ± 25.2 -> 75.9 ± 23.4 | Improved |
| add depth map input (RGBD instead of RGB) | Ours · ScanNet-Layout | PE (%): 16.3 ± 14.7 -> 9.9 ± 12.9 | Improved |
| add depth map input (RGBD instead of RGB) | Ours · ScanNet-Layout | EE: 22.3 ± 14.9 -> 11.9 ± 13.2 | Improved |
| add depth map input (RGBD instead of RGB) | Ours · ScanNet-Layout | RMSE: 0.5 ± 0.5 -> 0.2 ± 0.4 | Improved |
| add depth map input (RGBD instead of RGB) | Ours · NYUv2 303 | PE (%): 13.0 -> 8.9 | Improved |
| use general method instead of cuboid-specific Zhang et al. (RGBD) | Ours (RGBD) · NYUv2 303 (RGBD) | PE (%): 8.04 -> 8.9 | Hurt |
| use general method instead of cuboid-specific Hirzer et al. (RGB) | Ours (RGB) · NYUv2 303 (RGB) | PE (%): 8.49 -> 13.0 | Hurt |
| use general method instead of RoomNet (RGB) | Ours (RGB) · NYUv2 303 (RGB) | PE (%): 12.31 -> 13.0 | No clear effect |
| use general method instead of Howard-Jenkins et al. (RGB) | Ours (RGB) · NYUv2 303 (RGB) | PE (%): 12.19 -> 13.0 | No clear effect |
Closest papers by meaning
- LayoutNet: Reconstructing the 3D Room Layout From a Single RGB Image (CVPR 2018 · similarity 0.92)
- PolyLayout: Multi-room Manhattan Layout Estimation (ECCV 2026 · similarity 0.90)
- HouseLayout3D: A Benchmark and Training-free Baseline for 3D Layout Estimation in the Wild (NeurIPS 2025 · similarity 0.90)
- 3D-FRONT: 3D Furnished Rooms With layOuts and semaNTics (ICCV 2021 · similarity 0.90)
- Unposed Sparse Views Room Layout Reconstruction in the Age of Pretrain Model (ICLR 2025 · similarity 0.87)
- Seg2Reg: Differentiable 2D Segmentation to 1D Regression Rendering for 360 Room Layout Reconstruction (CVPR 2024 · similarity 0.87)
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.
