Research map / Point clouds / Shape and surface completion
Geometry and Learning Co-Supported Normal Estimation for Unstructured Point Cloud
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Combine multi-scale fitting patch selection with NH-Net refinement | Point-cloud normal estimation · Fig. 6 r1 | mean angular error: 9.0° -> 8.8° | Helped |
| Combine multi-scale fitting patch selection with NH-Net refinement | Point-cloud normal estimation · Fig. 6 r2 | mean angular error: 11.1° -> 10.8° | Helped |
| Combine multi-scale fitting patch selection with NH-Net refinement | Point-cloud normal estimation · Fig. 8 r1 | mean angular error: 6.5° -> 4.5° | Helped |
| Combine multi-scale fitting patch selection with NH-Net refinement | Point-cloud normal estimation · Fig. 8 r3 | mean angular error: 5.0° -> 4.5° | Helped |
| Add NH-Net refinement to suboptimal normals (ablation) | Point-cloud normal estimation · Synthetic benchmark dataset | mean angular error: 4.81° -> 4.32° | Helped |
| Replace the proposed learning scheme with averaging filtered normals (ablation) | Point-cloud normal estimation · Synthetic benchmark dataset | mean angular error: 4.32° -> 5.01° | Hurt |
| Use only the suboptimal-normal HMP branch to produce a transformation matrix (ablation) | NH-Net · Synthetic benchmark dataset | mean angular error: 4.32° -> 4.52° | Hurt |
| Remove the gathering module while using HMPs from filtered normals (ablation) | NH-Net · Synthetic benchmark dataset | mean angular error: 4.32° -> 4.61° | Hurt |
| Remove HMPs and train using filtered normals only (ablation) | NH-Net · Synthetic benchmark dataset | mean angular error: 4.32° -> 4.48° | Hurt |
Closest papers by meaning
- CMG-Net: Robust Normal Estimation for Point Clouds via Chamfer Normal Distance and Multi-Scale Geometry (AAAI 2024 · similarity 0.90)
- High-quality Point Cloud Oriented Normal Estimation via Hybrid Angular and Euclidean Distance Encoding (CVPR 2025 · similarity 0.89)
- An Explicit Frame Construction for Normalizing 3D Point Clouds (ICML 2024 · similarity 0.88)
- Consistent Normal Orientation for 3D Point Clouds via Least Squares on Delaunay Graph (CVPR 2025 · similarity 0.87)
- PLATYPUS: Progressive Local Surface Estimator for Arbitrary-Scale Point Cloud Upsampling (AAAI 2025 · similarity 0.86)
- LiSu: A Dataset and Method for LiDAR Surface Normal Estimation (CVPR 2025 · similarity 0.86)
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.