Research map / 3D vision / Face and shape
Efficient 3D Room Shape Recovery From a Single Panorama
Hao Yang, Hui Zhang
Check 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 the proposed reconstruction instead of GC | Surface label classification · Indoor panoramas, all pixels | classification error (%): 44.5 -> 26.9 | Improved |
| use the proposed reconstruction instead of GC | Surface label classification · Indoor panoramas, GCclutter < 0.7 | classification error (%): 44.4 -> 26.8 | Improved |
| use the proposed reconstruction instead of GC | Surface label classification · Indoor panoramas, GCclutter < 0.3 | classification error (%): 48.7 -> 26.7 | Improved |
| remove geometric context prior (ablation) | Proposed reconstruction · Surface label classification, all pixels | classification error (%): 26.9 -> 31.8 | Hurt |
| remove geometric context prior (ablation) | Proposed reconstruction · Surface label classification, GCclutter < 0.7 | classification error (%): 26.8 -> 31.7 | Hurt |
| remove geometric context prior (ablation) | Proposed reconstruction · Surface label classification, GCclutter < 0.3 | classification error (%): 26.7 -> 28.6 | Hurt |
| use the proposed reconstruction instead of aligned cuboids | Room depth estimation · Indoor panoramas | cosine distance (×10⁻²): 5.23 -> 4.27 | Improved |
| use the proposed reconstruction instead of aligned cuboids | Room depth estimation · Indoor panoramas | L2 distance (×10⁻²): 28.48 -> 27.02 | Improved |
| assume no occlusions (ablation) | Proposed reconstruction · Room depth estimation | cosine distance (×10⁻²): 4.27 -> 4.53 | Hurt |
| assume no occlusions (ablation) | Proposed reconstruction · Room depth estimation | L2 distance (×10⁻²): 27.02 -> 27.85 | Hurt |
| use ground-truth occlusions (ablation) | Proposed reconstruction · Room depth estimation | cosine distance (×10⁻²): 4.27 -> 4.23 | No clear effect |
| use ground-truth occlusions (ablation) | Proposed reconstruction · Room depth estimation | L2 distance (×10⁻²): 27.02 -> 27.09 | No clear effect |
Closest papers by meaning
- Automatic 3D Indoor Scene Modeling From Single Panorama (CVPR 2018 · similarity 0.92)
- Total3DUnderstanding: Joint Layout, Object Pose and Mesh Reconstruction for Indoor Scenes From a Single Image (CVPR 2020 · similarity 0.87)
- Shading Meets Motion: Self-supervised Indoor 3D Reconstruction Via Simultaneous Shape-from-Shading and Structure-from-Motion (CVPR 2025 · similarity 0.85)
- Behind the Veil: Enhanced Indoor 3D Scene Reconstruction with Occluded Surfaces Completion (CVPR 2024 · similarity 0.85)
- MvFFN: Multi-view Floor-Plan Feed-Forward Network for Unposed Wide-Baseline Panorama Layout Reconstruction (NeurIPS 2026 · similarity 0.84)
- Unified Panoramic Geometry Estimation via Multi-View Foundation Models (NeurIPS 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.
