Research map / 3D vision / Depth and stereo
3D Room Layout Estimation from a Cubemap of Panorama Image via Deep Manhattan Hough Transform
Yining Zhao, Chao Wen, Zhou Xue, Yue Gao
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 horizontal line detection (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 43.58 -> 85.48 | Improved |
| add vertical line detection (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 83.28 -> 85.48 | Improved |
| add detection of lines passing the center (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 84.15 -> 85.48 | Improved |
| use DMHT decoder instead of HorizonNet-like LSTM (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 81.44 -> 85.48 | Improved |
| use cubemap instead of equirectangular features (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 83.99 -> 85.48 | Improved |
| add 100 post-processing optimization steps (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 83.65 -> 85.48 | Improved |
| add 100 post-processing optimization steps (ablation) | DMH-Net · Stanford 2D-3D (cuboid) | 3DIoU: 81.66 -> 84.93 | Improved |
| use full voting bins instead of 1/4 (ablation) | DMH-Net · Stanford 2D-3D (cuboid) | 3DIoU: 78.81 -> 84.93 | Improved |
| use full voting bins instead of 1/2 (ablation) | DMH-Net · PanoContext (cuboid) | 3DIoU: 84.73 -> 85.48 | Improved |
| swap backbone to ResNet-50 from DRN-38 | DMH-Net · PanoContext (cuboid) | 3DIoU: 85.48 -> 85.90 | Improved |
| use DMH-Net instead of HorizonNet with ResNet-18 | ResNet-18 · PanoContext (cuboid) (ResNet-18) | 3DIoU: 79.02 -> 82.96 | Improved |
| use DMH-Net instead of AtlantaNet on non-cuboid layouts | Matterport3D · Matterport3D (non-cuboid, overall) | 3DIoU: 81.59 -> 78.97 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- HorizonNet — HorizonNet: Learning Room Layout With 1D Representation and Pano Stretch Data Augmentation (CVPR 2019) · better on 6 of 7 shared results
- LayoutNet — LayoutNet: Reconstructing the 3D Room Layout From a Single RGB Image (CVPR 2018) · better on 3 of 3 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
- AtlantaNet: Inferring the 3D Indoor Layout from a Single 360(∘) Image beyond the Manhattan World Assumption (ECCV 2020 · similarity 0.90)
- GeoLayout: Geometry Driven Room Layout Estimation Based on Depth Maps of Planes (ECCV 2020 · similarity 0.90)
- Manhattan Room Layout Reconstruction from a Single $$360^{\circ }$$ Image: A Comparative Study of State-of-the-Art Methods (IJCV 2021 · similarity 0.90)
- PolyLayout: Multi-room Manhattan Layout Estimation (ECCV 2026 · similarity 0.89)
- Joint 3D Layout and Depth Prediction from a Single Indoor Panorama Image (ECCV 2020 · similarity 0.89)
- HouseLayout3D: A Benchmark and Training-free Baseline for 3D Layout Estimation in the Wild (NeurIPS 2025 · 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.
