Research map / Point clouds / Point cloud classification
Deep Positional and Relational Feature Learning for Rotation-Invariant Point Cloud Analysis
Ruixuan Yu, Xin Wei, Federico Tombari, Jian Sun
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 |
|---|---|---|---|
| add geometric feature extraction (ablation) | PR-invNet · ModelNet40 | accuracy: 88.2 -> 89.2 | Improved |
| add PFE-block (ablation) | PR-invNet · ModelNet40 | accuracy: 87.8 -> 89.2 | Improved |
| add RFE-block (ablation) | PR-invNet · ModelNet40 | accuracy: 88.0 -> 89.2 | Improved |
| use pose selector instead of average-pooling over all poses (ablation) | PFE-block · ModelNet40 | accuracy: 88.1 -> 88.5 | Improved |
| use pose selector instead of max-pooling over all poses (ablation) | PFE-block · ModelNet40 | accuracy: 88.3 -> 88.5 | Improved |
| add position term to RFE filter (ablation) | RFE-block · ModelNet40 | accuracy: 88.1 -> 89.2 | Improved |
| add geometric feature term to RFE filter (ablation) | RFE-block · ModelNet40 | accuracy: 88.8 -> 89.2 | Improved |
| add point feature term to RFE filter (ablation) | RFE-block · ModelNet40 | accuracy: 89.9 -> 89.2 | Hurt |
| add geometric-feature-only term g removed from filter (ablation) | RFE-block · ModelNet40 | accuracy: 88.5 -> 89.2 | Improved |
| concatenate all PFE and RFE features instead of PFE only (ablation) | PR-invNet · ModelNet40 | accuracy: 88.5 -> 89.2 | Improved |
| concatenate all PFE and RFE features instead of RFE only (ablation) | PR-invNet · ModelNet40 | accuracy: 87.4 -> 89.2 | Improved |
| use pose selector over rot-space instead of pure PCA alignment (ablation) | PR-invNet · ModelNet40 (SO(3)/SO(3)) | accuracy: 88.3 -> 89.2 | Improved |
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
- ClusterNet: Deep Hierarchical Cluster Network With Rigorously Rotation-Invariant Representation for Point Cloud Analysis (CVPR 2019 · similarity 0.92)
- RIConv++: Effective Rotation Invariant Convolutions for 3D Point Clouds Deep Learning (IJCV 2022 · similarity 0.91)
- RISurConv: Rotation Invariant Surface Attention-Augmented Convolutions for 3D Point Cloud Classification and Segmentation (ECCV 2024 · similarity 0.90)
- SGMNet: Learning Rotation-Invariant Point Cloud Representations via Sorted Gram Matrix (ICCV 2021 · similarity 0.90)
- Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution (AAAI 2020 · similarity 0.90)
- DeepLA-Net: Very Deep Local Aggregation Networks for Point Cloud Analysis (CVPR 2025 · 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.
