Research map / Point clouds / Registration and matching
Unsupervised RGB-D Point Cloud Registration for Scenes with Low Overlap and Photometric Inconsistency
Yejun Shou, Haocheng Wang, Lingfeng Shen, Qian Zheng, Gang Pan, Yanlong Cao
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 GRPA Gaussian rendering module (ablation) | OG-UPCR baseline · ScanNet1500 | rotation accuracy <5: 56.3 -> 65.1 | Improved |
| add GRPA Gaussian rendering module (ablation) | OG-UPCR baseline · ScanNet | rotation accuracy <5: 96.0 -> 97.4 | Improved |
| add full OCID module (RO and GL) (ablation) | OG-UPCR baseline · ScanNet1500 | rotation accuracy <5: 56.3 -> 78.5 | Improved |
| add full OCID module (RO and GL) (ablation) | OG-UPCR baseline · ScanNet1500 | translation accuracy <5 cm: 36.5 -> 51.1 | Improved |
| add full OCID module (RO and GL) (ablation) | OG-UPCR baseline · ScanNet1500 | chamfer accuracy <1 mm: 48.5 -> 68.9 | Improved |
| add random overlapping region prediction (RO) only (ablation) | OG-UPCR baseline · ScanNet1500 | rotation accuracy <5: 56.3 -> 74.7 | Improved |
| add global-to-local inlier detection (GL) only (ablation) | OG-UPCR baseline · ScanNet1500 | rotation accuracy <5: 56.3 -> 73.9 | Improved |
| add RO or GL only (ablation) | OG-UPCR baseline · ScanNet | rotation accuracy <5: 96.0 -> 97.2 | Improved |
| add GRPA on top of OCID (ablation) | OG-UPCR · ScanNet1500 | rotation accuracy <5: 78.5 -> 81.3 | Improved |
| add GRPA on top of OCID (ablation) | OG-UPCR · ScanNet | translation accuracy <5 cm: 85.7 -> 88.8 | Improved |
| swap training set from 3DMatch to ScanNet | OG-UPCR · ScanNet1500 | rotation accuracy <5: 76.9 -> 81.3 | Improved |
Closest papers by meaning
- Reliable Inlier Evaluation for Unsupervised Point Cloud Registration (AAAI 2022 · similarity 0.90)
- Zero-shot RGB-D Point Cloud Registration with Pre-trained Large Vision Model (CVPR 2025 · similarity 0.90)
- Colored Point Cloud Registration Revisited (ICCV 2017 · similarity 0.90)
- Improving RGB-D Point Cloud Registration by Learning Multi-Scale Local Linear Transformation (ECCV 2022 · similarity 0.90)
- A robust inlier identification algorithm for point cloud registration via $\mathbf{\ell_0}$-minimization (NeurIPS 2024 · similarity 0.89)
- Inlier Confidence Calibration for Point Cloud Registration (CVPR 2024 · similarity 0.88)
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.
