Research map / Point clouds / Registration and matching
LSG-CPD: Coherent Point Drift With Local Surface Geometry for Point Cloud Registration
Weixiao Liu, Hongtao Wu, Gregory S. Chirikjian
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 confidence filtering without truncation (ablation) | LSG-CPD · Stanford Lounge | angle error (deg): 0.51 -> 0.42 | Improved |
| add confidence filtering with truncation of low-confidence points (ablation) | LSG-CPD · Stanford Lounge | frame rate (FPS): 5 FPS -> 19 FPS | Improved |
| use too large a truncation threshold (ablation) | LSG-CPD with CF · Stanford Lounge | angle error | Hurt |
| use LSG-CPD instead of CPD | registration · Bunny, no outliers | runtime (ms): 450 -> 89 | Improved |
| use LSG-CPD instead of CPD | registration · Bunny, outlier ratio 1 | runtime (ms): 9940 -> 346 | Improved |
| use anisotropic local-geometry covariance instead of isotropic | LSG-CPD vs CPD · KITTI odometry seq 07 | relative translation error (m): 0.190 -> 0.022 | Improved |
| use point-to-plane FilterReg instead of point-to-point (ablation) | FilterReg · Bunny noise and outliers | robustness | Hurt |
| increase outlier ratio (ablation) | CPD and FilterReg · Bunny with outliers | registration error | Hurt |
| increase noise level (ablation) | LSG-CPD · Bunny with noise | penalization coefficient alpha | Mixed |
| use LSG-CPD instead of multi-view EMPMR | registration · Happy Stand | registration error (mm): 1.07 -> 1.08 | No clear effect |
| use LSG-CPD instead of FilterReg | pairwise registration · Dragon Stand | registration error (mm): 1.13 -> 0.97 | Improved |
Later papers that compared with this method
1 later accepted paper reports LSG-CPD as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- GraphCPD: Coherent Point Drift for Point Cloud Registration via Graph Signal Processing (ECCV 2026 · better on 2 of 3 shared results)
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
- GraphCPD: Coherent Point Drift for Point Cloud Registration via Graph Signal Processing (ECCV 2026 · similarity 0.91)
- Generalized-CVO: Fast and Correspondence-Free Local Point Cloud Registration with Second Order Riemannian Optimization (CVPR 2026 · similarity 0.89)
- HGMR: Hierarchical Gaussian Mixtures for Adaptive 3D Registration (ECCV 2018 · similarity 0.88)
- Unsupervised Deep Probabilistic Approach for Partial Point Cloud Registration (CVPR 2023 · similarity 0.88)
- GRICP: Granular-Ball Iterative Closest Point with Multikernel Correntropy for Point Cloud Fine Registration (AAAI 2025 · similarity 0.88)
- A robust inlier identification algorithm for point cloud registration via $\mathbf{\ell_0}$-minimization (NeurIPS 2024 · similarity 0.87)
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.
