Research map / Detection and segmentation / CNN features
Rescan: Inductive Instance Segmentation for Indoor RGBD Scans
Maciej Halber, Yifei Shi, Kai Xu, Thomas Funkhouser
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 Coverage Term to the objective (ablation) | Rescan · Rescan benchmark | Instance Transfer (mean IoU): 0.048 -> 0.650 | Improved |
| add Coverage Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Label (mean class IoU): 0.061 -> 0.859 | Improved |
| add Coverage Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Instance (mAP@0.5): 0.058 -> 0.837 | Improved |
| add Hysteresis Term to the objective (ablation) | Rescan · Rescan benchmark | Instance Transfer (mean IoU): 0.226 -> 0.650 | Improved |
| add Hysteresis Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Label (mean class IoU): 0.870 -> 0.859 | Hurt |
| add Hysteresis Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Instance (mAP@0.5): 0.818 -> 0.837 | Improved |
| add Intersection Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Instance (mAP@0.5): 0.781 -> 0.837 | Improved |
| add Intersection Term to the objective (ablation) | Rescan · Rescan benchmark | Instance Transfer (mean IoU): 0.584 -> 0.650 | Improved |
| add Intersection Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Label (mean class IoU): 0.859 -> 0.859 | No clear effect |
| add Geometry Term to the objective (ablation) | Rescan · Rescan benchmark | Instance Transfer (mean IoU): 0.617 -> 0.650 | Improved |
| add Geometry Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Instance (mAP@0.5): 0.825 -> 0.837 | Improved |
| add Geometry Term to the objective (ablation) | Rescan · Rescan benchmark | Semantic Label (mean class IoU): 0.853 -> 0.859 | Improved |
Later papers that compared with this method
1 later accepted paper reports RESCAN as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Beyond Monocular Deraining: Parallel Stereo Deraining Network Via Semantic Prior (IJCV 2022 · better on 1 of 1 shared result)
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
- ReScene4D: Temporally Consistent Semantic Instance Segmentation of Evolving Indoor 3D Scenes (CVPR 2026 · similarity 0.89)
- 3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans (CVPR 2019 · similarity 0.88)
- RevealNet: Seeing Behind Objects in RGB-D Scans (CVPR 2020 · similarity 0.87)
- MultiScan: Scalable RGBD scanning for 3D environments with articulated objects (NeurIPS 2022 · similarity 0.87)
- UnScene3D: Unsupervised 3D Instance Segmentation for Indoor Scenes (CVPR 2024 · similarity 0.86)
- Geometric Exploitation for Indoor Panoramic Semantic Segmentation (NeurIPS 2024 · similarity 0.85)
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.
