Research map / 3D vision / 3D generation
Cooperative Holistic Scene Understanding: Unifying 3D Object, Layout, and Camera Pose Estimation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Train GGN and LON cooperatively rather than individually | Holistic scene model · SUN RGB-D | holistic free-space IoU: 39.6 -> 42.9 | Helped |
| Train GGN and LON cooperatively rather than individually | Holistic scene model · SUN RGB-D | 3D detection mAP: 15.24 -> 23.65 | Helped |
| Train GGN and LON cooperatively rather than individually | Holistic scene model · SUN RGB-D | camera yaw mean absolute error: 2.48 -> 2.19 | Helped |
| Add 3D box corner supervision (ablation) | Cooperative holistic scene model · SUN RGB-D holistic scene understanding | free-space IoU: 42.8 -> 43.3 | Helped |
| Add 3D box corner supervision (ablation) | Cooperative holistic scene model · SUN RGB-D 3D box estimation | 3D mIoU: 14.4 -> 17.4 | Helped |
| Add 2D projection loss (ablation) | Cooperative holistic scene model · SUN RGB-D 3D box estimation | 3D mIoU: 18.2 -> 17.4 | Hurt |
| Add 2D projection loss (ablation) | Cooperative holistic scene model · SUN RGB-D 3D box estimation | 2D mIoU: 60.7 -> 68.7 | Helped |
| Add physical constraint loss (ablation) | Cooperative holistic scene model · SUN RGB-D holistic scene understanding | free-space IoU: 41.7 -> 43.3 | Helped |
| Add 3D box, GGN, and LON losses instead of using only 2D supervision (ablation) | Cooperative holistic scene model · SUN RGB-D holistic scene understanding | free-space IoU: 35.9 -> 43.3 | Helped |
| Pre-train on synthetic data before SUN RGB-D fine-tuning (ablation) | Cooperative holistic scene model · SUN RGB-D holistic scene understanding | free-space IoU: 40.2 -> 43.3 | Helped |
| Use detected 2D boxes rather than boxes projected from ground-truth 3D boxes (ablation) | Cooperative holistic scene model · SUN RGB-D holistic scene understanding | free-space IoU: 43.0 -> 43.3 | No clear effect |
Closest papers by meaning
- Total3DUnderstanding: Joint Layout, Object Pose and Mesh Reconstruction for Indoor Scenes From a Single Image (CVPR 2020 · similarity 0.90)
- Rooms from Motion: Un-posed Indoor 3D Object Detection as Localization and Mapping (NeurIPS 2025 · similarity 0.87)
- ImageNet3D: Towards General-Purpose Object-Level 3D Understanding (NeurIPS 2024 · similarity 0.87)
- UniDet3D: Multi-dataset Indoor 3D Object Detection (AAAI 2025 · similarity 0.86)
- Cubify Anything: Scaling Indoor 3D Object Detection (CVPR 2025 · similarity 0.86)
- SceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D Scenes (CVPR 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.