Research map / Point clouds / Shape and surface completion
HGNet: Learning Hierarchical Geometry From Points, Edges, and Surfaces
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 |
|---|---|---|---|
| Integrate hierarchical point-, edge-, and surface-level features (ablation) | HGNet · ScanObjectNN | OA: 87.7 -> 88.9 | Helped |
| Add edge-level features to point-level features (ablation) | HGNet · ScanObjectNN | OA: 87.7 -> 88.5 | Helped |
| Add surface-level features to point-level features (ablation) | HGNet · ScanObjectNN | OA: 87.7 -> 88.6 | Helped |
| Use surface-level rather than point-level features alone (ablation) | HGNet · ScanObjectNN | OA: 87.7 -> 87.8 | Helped |
| Replace max pooling with Transformer self-attention aggregation (ablation) | HGNet with point-, edge-, and surface-level features · ScanObjectNN | OA: 88.9 -> 89.2 | Helped |
| Replace max pooling with Transformer self-attention aggregation (ablation) | HGNet with point-, edge-, and surface-level features · ScanObjectNN | mAcc: 87.3 -> 87.5 | Helped |
| Use hierarchical point-, edge-, surface-, and super-surface geometry | HGNet · ScanObjectNN PB_T50_RS | OA: 87.7 -> 89.2 | Helped |
| Use hierarchical geometry features | HGNet · PartNet test | part-category mIoU: 53.8 -> 55.6 | Helped |
| Use hierarchical geometry features | HGNet · S3DIS Area-5 | mIoU: 69.0 -> 70.8 | Helped |
| Use hierarchical geometry features | HGNet · ModelNet40 test | OA: 94.4 -> 94.5 | Helped |
Closest papers by meaning
- A Hierarchical Graph Network for 3D Object Detection on Point Clouds (CVPR 2020 · similarity 0.88)
- A Linear Time and Space Local Point Cloud Geometry Encoder via Vectorized Kernel Mixture (VecKM) (ICML 2024 · similarity 0.86)
- LinNet: Linear Network for Efficient Point Cloud Representation Learning (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)
- Improved MLP Point Cloud Processing with High-Dimensional Positional Encoding (AAAI 2024 · similarity 0.86)
- Orthogonal Dictionary Guided Shape Completion Network for Point Cloud (AAAI 2024 · 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.