Research map / Point clouds / Point cloud classification
NerVE: Neural Volumetric Edges for Parametric Curve Extraction From Point Cloud
Read the paperCheck 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 |
|---|---|---|---|
| Predict structured volumetric edges with NerVE rather than point-wise edge estimates | Edge estimation on ABC test set · ABC test set | CD: 0.0226 -> 0.0012 (VCM); 0.0037 -> 0.0012 (EC-NET); 0.0074 -> 0.0012 (PIE-NET) | Helped |
| Use NerVE for parametric curve extraction instead of keypoint-based baselines | Parametric curve extraction on DEF-Sim · DEF-Sim | CD: 0.97 -> 0.008 (PIE-NET); 0.04 -> 0.008 (DEF) | Helped |
| Increase grid resolution from 32³ to 64³ (ablation) | NerVE edge and PWL prediction · ABC | CD: 0.0012 -> 0.0009 | Helped |
| Increase grid resolution from 32³ to 64³ (ablation) | NerVE occupancy prediction · ABC | R_o: 0.965 -> 0.958 | Hurt |
| Increase grid resolution from 64³ to 128³ (ablation) | NerVE edge and PWL prediction · ABC | CD: 0.0009 -> 0.0008 | Helped |
| Use midpoint position instead of DC QEF for each cube (ablation) | Ground-truth NerVE curve restoration · Ground-truth NerVE | CD: 0.002799 -> 0.000136 | Helped |
| Use k=8 instead of k=4 in the point encoder (ablation) | NerVE at 32³ · ABC edge estimation | CD: 0.0017 -> 0.0012 | Helped |
| Use k=16 instead of k=8 in the point encoder (ablation) | NerVE at 32³ · ABC edge estimation | CD: 0.0012 -> 0.0012 | No clear effect |
| Remove PointNet++ features (ablation) | NerVE · ABC edge estimation | CD: 0.0012 -> 0.0025 | Hurt |
| Remove 3DCNN features (ablation) | NerVE · ABC edge estimation | CD: 0.0012 -> 0.0020 | Hurt |
| Remove topology post-processing (ablation) | NerVE parametric curve extraction · Parametric curves | CD: 0.008 -> 0.067 | Hurt |
| Change random train/test split (ablation) | NerVE at 32³ · ABC | CD: 0.0012; 0.0013; 0.0016 | No clear effect |
Closest papers by meaning
- Improving Neural Network Surface Processing with Principal Curvatures (NeurIPS 2024 · similarity 0.86)
- STAR-Edge: Structure-aware Local Spherical Curve Representation for Thin-walled Edge Extraction from Unstructured Point Clouds (CVPR 2025 · similarity 0.86)
- 3D Neural Edge Reconstruction (CVPR 2024 · similarity 0.86)
- Flatten Anything: Unsupervised Neural Surface Parameterization (NeurIPS 2024 · similarity 0.85)
- Small Steps and Level Sets: Fitting Neural Surface Models with Point Guidance (CVPR 2024 · similarity 0.85)
- EdgeMovingNet: Edge-preserving Point Cloud Reconstruction via Joint Geometry Features (CVPR 2025 · 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.