Research map / 3D vision / 3D generation
Peek-a-Boo: Occlusion Reasoning in Indoor Scenes With Plane Representations
Ziyu Jiang, Buyu Liu, Samuel Schulter, Zhangyang Wang, Manmohan Chandraker
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 |
|---|---|---|---|
| Add separate object and layout branches (ablation) | PlaneRCNN-OR · ScanNet | APH0.4: 0.068 -> 0.092 | Improved |
| Add separate object and layout branches (ablation) | PlaneRCNN-OR · ScanNet | AP0.4: 0.319 -> 0.315 | Hurt |
| Add semantic merging to DualRPN (ablation) | DualRPN · ScanNet | APH0.4: 0.092 -> 0.097 | Improved |
| Add semantic merging to DualRPN (ablation) | DualRPN · ScanNet | AP0.4: 0.315 -> 0.331 | Improved |
| Add plane-warping loss to DualRPN with semantic merging (ablation) | DualRPN + Semantic Merging · ScanNet | APH0.4: 0.097 -> 0.097 | No clear effect |
| Add plane-warping loss to DualRPN with semantic merging (ablation) | DualRPN + Semantic Merging · ScanNet | APH0.9: 0.114 -> 0.118 | Improved |
| Add plane-warping loss to DualRPN with semantic merging (ablation) | DualRPN + Semantic Merging · ScanNet | AP0.9: 0.395 -> 0.405 | Improved |
| Add occlusion-aware complete-mask reasoning (ablation) | PlaneRCNN · ScanNet | Acc1: 0.78 -> 0.79 | Improved |
| Add separate object and layout branches (ablation) | PlaneRCNN-OR · ScanNet | RMSE: 0.34 -> 0.32 | Improved |
| Train and evaluate using refined masks with hole filling (ablation) | PlaneRCNN-OR · ScanNet | APH0.4: 0.007 -> 0.068 | Improved |
Closest papers by meaning
- Behind the Veil: Enhanced Indoor 3D Scene Reconstruction with Occluded Surfaces Completion (CVPR 2024 · similarity 0.86)
- Slice3D: Multi-Slice Occlusion-Revealing Single View 3D Reconstruction (CVPR 2024 · similarity 0.86)
- AirPlanes: Accurate Plane Estimation via 3D-Consistent Embeddings (CVPR 2024 · similarity 0.86)
- 3D Occupancy Prediction with Low-Resolution Queries via Prototype-aware View Transformation (CVPR 2025 · similarity 0.86)
- Video Spatial Reasoning with Object-Centric 3D Rollout (AAAI 2026 · similarity 0.86)
- RadOcc: Learning Cross-Modality Occupancy Knowledge through Rendering Assisted Distillation (AAAI 2024 · similarity 0.86)
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.
