Research map / Point clouds / Point cloud classification
RIConv++: Effective Rotation Invariant Convolutions for 3D Point Clouds Deep Learning
Zhiyuan Zhang, Binh‐Son Hua, Sai-Kit Yeung
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 neighbor-neighbor relation features to IRIF (ablation) | RIConv++ · ModelNet40 | accuracy: 88.4 -> 91.2 | Improved |
| add distance features alongside angle features (ablation) | RIConv++ · ModelNet40 | accuracy: 90.5 -> 91.2 | Improved |
| add angle features alongside distance features (ablation) | RIConv++ · ModelNet40 | accuracy: 83.3 -> 91.2 | Improved |
| increase kernel size from 1 to 3 (ablation) | RIConv++ · ModelNet40 | accuracy: 91.2 -> 91.2 | No clear effect |
| increase kernel size from 1 to 7 (ablation) | RIConv++ · ModelNet40 | accuracy: 91.2 -> 91.0 | Hurt |
| replace xyz input with IRIF features | PointNet++ · ModelNet40 (z/SO3) | accuracy: 28.6 -> 89.0 | Improved |
| replace xyz input with IRIF features | DGCNN · ModelNet40 (z/SO3) | accuracy: 20.6 -> 90.2 | Improved |
| replace xyz input with IRIF features | DGCNN · ModelNet40 (z/z) | accuracy: 92.2 -> 90.2 | Hurt |
| use IRIF with LRA instead of RIConv features | RIConv · ModelNet40 (3-layer) | accuracy: 86.5 -> 89.8 | Improved |
| use IRIF instead of LRF-transform features of GCAConv | GCAConv · ModelNet40 (3-layer) | accuracy: 89.0 -> 89.8 | Improved |
| use normal vectors as reference axis instead of LRA | RIConv++ · ModelNet40 | accuracy: 91.2 -> 91.3 | No clear effect |
| add background points to objects (OBJ_ONLY to OBJ_BG) | RIConv++ · ScanObjectNN (z/z) | accuracy: 86.2 -> 85.6 | No clear effect |
Later papers that compared with this method
2 later accepted papers report RIConv++ as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Learning Coordinate-based Convolutional Kernels for Continuous SE(3) Equivariant and Efficient Point Cloud Analysis (CVPR 2026 · better on 1 of 1 shared result)
- RISurConv: Rotation Invariant Surface Attention-Augmented Convolutions for 3D Point Cloud Classification and Segmentation (ECCV 2024 · better on 4 of 4 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
- RISurConv: Rotation Invariant Surface Attention-Augmented Convolutions for 3D Point Cloud Classification and Segmentation (ECCV 2024 · similarity 0.93)
- PointCNN++: Performant Convolution on Native Points (CVPR 2026 · similarity 0.92)
- PointConv: Deep Convolutional Networks on 3D Point Clouds (CVPR 2019 · similarity 0.91)
- Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution (AAAI 2020 · similarity 0.91)
- Deep Positional and Relational Feature Learning for Rotation-Invariant Point Cloud Analysis (ECCV 2020 · similarity 0.91)
- ShellNet: Efficient Point Cloud Convolutional Neural Networks Using Concentric Shells Statistics (ICCV 2019 · similarity 0.90)
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.
