Research map / Point clouds / Point cloud classification
How Privacy-Preserving Are Line Clouds? Recovering Scene Details From 3D Lines
Kunal Chelani, Fredrik Kahl, Torsten Sattler
Check 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 |
|---|---|---|---|
| recover point locations from uniform random line clouds using line neighborhoods and closest-point estimates | 3D line clouds · 12 Scenes | point-position error: 1-3cm | Improved |
| recover point locations from uniform random line clouds using line neighborhoods and closest-point estimates | 3D line clouds · Cambridge Landmarks | point-position recovery | Mixed |
| use recovered points from line clouds as input to SfM inversion | Line2Point · Indoor and outdoor scenes | image detail recovery | Improved |
| use only 5% or less of the original lines | Line2Point with SfM inversion · 12 Scenes | image detail recovery: 5% | Hurt |
| reduce the line cloud to 1% of the original lines | Line2Point · Four indoor scenes | point-position recovery: 1% | Hurt |
| provide the true neighborhood using an oracle (ablation) | Line2Point · Four indoor scenes | point-position recovery: 1% | Hurt |
| replace 50% of true neighbors with random lines (ablation) | point estimation from line neighborhoods · Indoor scene | point-cloud recovery: 50% | No clear effect |
| replace 90% of true neighbors with random lines (ablation) | point estimation from line neighborhoods · Indoor scene | image detail recovery: 90% | Hurt |
| use a smaller true-neighbor neighborhood (ablation) | Line2Point with oracle neighborhoods · 12 Scenes | point-position error | Improved |
| use a very sparse line cloud (ablation) | Line2Point with SfM inversion · 12 Scenes | image detail recovery: 5% | Hurt |
Closest papers by meaning
- Paired-Point Lifting for Enhanced Privacy-Preserving Visual Localization (CVPR 2023 · similarity 0.90)
- Efficient Privacy-Preserving Visual Localization Using 3D Ray Clouds (CVPR 2024 · similarity 0.88)
- Gaussian Splatting Feature Fields for (Privacy-Preserving) Visual Localization (CVPR 2025 · similarity 0.86)
- GPVK-VL: Geometry-Preserving Virtual Keyframes for Visual Localization under Large Viewpoint Changes (CVPR 2025 · similarity 0.86)
- Fully Geometric Panoramic Localization (CVPR 2024 · similarity 0.85)
- Learning to Produce Semi-dense Correspondences for Visual Localization (CVPR 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.
