Research map / Point clouds / Point cloud classification
SGMNet: Learning Rotation-Invariant Point Cloud Representations via Sorted Gram Matrix
Jianyun Xu, Xin Tang, Yushi Zhu, Jie Sun, Shiliang Pu
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 |
|---|---|---|---|
| replace distance-only features with SGM (ablation) | SGMNet · ModelNet40 | overall accuracy (%): 80.6 -> 90.0 | Improved |
| replace angle-only features with SGM (ablation) | SGMNet · ModelNet40 | overall accuracy (%): 85.2 -> 90.0 | Improved |
| replace simplified distance+angle PPFs with dense SGM (ablation) | SGMNet · ModelNet40 | overall accuracy (%): 86.6 -> 90.0 | Improved |
| combine distance and angle rather than distance alone (ablation) | simplified PPF · ModelNet40 | overall accuracy (%): 80.6 -> 86.6 | Improved |
| shrink SGMNet channels from big to mini (ablation) | SGMNet · ModelNet40 z/SO(3) | overall accuracy (%): 90.0 -> 88.8 | Mixed |
| use SGM descriptor in place of xyz coordinates in PointNet++ | PointNet++ · ModelNet40 z/SO(3) | overall accuracy (%): 18.4 -> 90.0 | Improved |
| add SGM module to RSCNN | RSCNN · ModelNet40 | overall accuracy (%): 88.5 in all three settings | Improved |
| add SGM module to DGCNN | DGCNN · ModelNet40 | overall accuracy (%): 92.2 -> 88.9 (z/z), 20.6 -> 88.9 (z/SO(3)) | Mixed |
| use SGMNet instead of PointNet++ on part segmentation under arbitrary rotation | PointNet++ · ShapeNet part z/SO(3) | mIoU (%): 45.9 -> 79.3 | Improved |
| use SGMNet instead of PointNet++ in aligned setting | PointNet++ · ShapeNet part z/z | mIoU (%): 80.6 -> 79.3 | Hurt |
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.
- ClusterNet — ClusterNet: Deep Hierarchical Cluster Network With Rigorously Rotation-Invariant Representation for Point Cloud Analysis (CVPR 2019) · better on 3 of 3 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
- Deep Positional and Relational Feature Learning for Rotation-Invariant Point Cloud Analysis (ECCV 2020 · similarity 0.90)
- Rethinking Rotation Invariance with Point Cloud Registration (AAAI 2023 · similarity 0.89)
- Equivariant Point Network for 3D Point Cloud Analysis (CVPR 2021 · similarity 0.89)
- PARE-Net: Position-Aware Rotation-Equivariant Networks for Robust Point Cloud Registration (ECCV 2024 · similarity 0.89)
- RIConv++: Effective Rotation Invariant Convolutions for 3D Point Clouds Deep Learning (IJCV 2022 · similarity 0.89)
- A Linear Time and Space Local Point Cloud Geometry Encoder via Vectorized Kernel Mixture (VecKM) (ICML 2024 · similarity 0.88)
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.
