Research map / Point clouds / Point cloud classification
CPO: Change Robust Panorama to Point Cloud Localization
Junho Kim, Hojun Jang, Changwoon Choi, Young Min Kim
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 histogram-based candidate pose selection (ablation) | CPO · OmniScenes Extreme | accuracy: 0.20 -> 0.57 | Improved |
| add histogram-based candidate pose selection (ablation) | CPO · OmniScenes Extreme | R-error (deg): 75.60 -> 0.37 | Improved |
| add 2D score map for candidate pose selection (ablation) | CPO · OmniScenes Extreme | accuracy: 0.48 -> 0.57 | Improved |
| add 2D score map (ablation) | CPO · OmniScenes Extreme | t-error (m): 0.10 -> 0.03 | Improved |
| add 3D score map for weighted sampling loss (ablation) | CPO · OmniScenes Extreme | accuracy: 0.55 -> 0.57 | Improved |
| add 3D score map (ablation) | CPO · OmniScenes Extreme | R-error (deg): 1.56 -> 0.37 | Improved |
| use fast histogram generation versus rendering-based initialization of baselines | CPO · OmniScenes Room 3 | runtime per view (ms): 38.70 -> 0.188 | Improved |
| replace PICCOLO with CPO score-map pipeline | PICCOLO · OmniScenes Robot | accuracy: 0.06 -> 0.58 | Improved |
| give PICCOLO a gravity prior | PICCOLO · OmniScenes Robot | accuracy: 0.06 -> 0.39 | Improved |
| CPO versus structure-based method | Structure-Based · Structured3D | accuracy (0.05m, 5deg): 0.59 -> 0.56 | No clear effect |
| CPO versus structure-based method at tight threshold | Structure-Based · Structured3D | accuracy (0.01m, 1deg): 0.29 -> 0.51 | Improved |
| CPO versus PICCOLO on Stanford 2D-3D-S | PICCOLO · Stanford 2D-3D-S | accuracy: 0.53 -> 0.83 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PICCOLO — PICCOLO: Point Cloud-Centric Omnidirectional Localization (ICCV 2021) · better on 6 of 7 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
- PICCOLO: Point Cloud-Centric Omnidirectional Localization (ICCV 2021 · similarity 0.88)
- Fully Geometric Panoramic Localization (CVPR 2024 · similarity 0.86)
- Colored Point Cloud Registration Revisited (ICCV 2017 · similarity 0.86)
- LDL: Line Distance Functions for Panoramic Localization (ICCV 2023 · similarity 0.84)
- GPVK-VL: Geometry-Preserving Virtual Keyframes for Visual Localization under Large Viewpoint Changes (CVPR 2025 · similarity 0.84)
- Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World (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.
