Research map / Detection and segmentation / 3D detection and driving
Shape Anchor Guided Holistic Indoor Scene Understanding
Mingyue Dong, Linxi Huan, Hanjiang Xiong, Shuhan Shen, Xianwei Zheng
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 object detection shape anchors (ablation) | AncRec detector · ScanNetv2 detection | IoU@0.5 mAP: 41.12 -> 43.33 | Improved |
| add layout anchors alone (ablation) | AncRec detector · ScanNetv2 detection | IoU@0.5 mAP: 41.12 -> 40.21 | Hurt |
| add layout anchors (ablation) | AncRec detector · ScanNetv2 layout | F1 score: 60.62 -> 62.61 | Improved |
| add object detection anchors to layout estimation (ablation) | AncRec detector · ScanNetv2 layout | F1 score: 60.62 -> 60.59 | No clear effect |
| add self-attention layer to layout branch with both anchors (ablation) | AncRec · ScanNetv2 layout | F1 score: 64.00 -> 70.45 | Improved |
| add self-attention layer to layout branch with both anchors (ablation) | AncRec · ScanNetv2 detection | IoU@0.5 mAP: 41.24 -> 43.38 | Improved |
| sample instance points with AncLearn instead of segmentation (ablation) | AncRec shape predictor · ScanNetv2 reconstruction | mAP CD@0.1: 55.8 -> 56.8 | Improved |
| sample instance points with AncLearn instead of box cropping (ablation) | AncRec shape predictor · ScanNetv2 reconstruction | mAP IoU@0.5: 17.7 -> 18.9 | Improved |
| AncRec versus DIMR on IoU@0.5 | AncRec · ScanNetv2 reconstruction | mAP IoU@0.5: 12.5 -> 18.9 | Improved |
Closest papers by meaning
- Learning Shape Anchors for Holistic Indoor Scene Understanding (TPAMI 2026 · similarity 0.98)
- RfD-Net: Point Scene Understanding by Semantic Instance Reconstruction (CVPR 2021 · similarity 0.84)
- Clouds of Oriented Gradients for 3D Detection of Objects, Surfaces, and Indoor Scene Layouts (TPAMI 2019 · similarity 0.84)
- LASA: Instance Reconstruction from Real Scans using A Large-scale Aligned Shape Annotation Dataset (CVPR 2024 · similarity 0.84)
- SPGroup3D: Superpoint Grouping Network for Indoor 3D Object Detection (AAAI 2024 · similarity 0.84)
- Uncertainty Meets Diversity: A Comprehensive Active Learning Framework for Indoor 3D Object Detection (CVPR 2025 · similarity 0.83)
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.
