Rotation invariance and equivariance in 3D deep learning: a survey

J Fei, Z Deng - Artificial Intelligence Review, 2024‏ - Springer
Deep neural networks (DNNs) in 3D scenes show a strong capability of extracting high-level
semantic features and significantly promote research in the 3D field. 3D shapes and scenes …

Robust 3D shape classification via non-local graph attention network

S Qin, Z Li, L Liu - … of the IEEE/CVF Conference on …, 2023‏ - openaccess.thecvf.com
We introduce a non-local graph attention network (NLGAT), which generates a novel global
descriptor through two sub-networks for robust 3D shape classification. In the first sub …

Incorporating rotation invariance with non-invariant networks for point clouds

J Fei, Z Deng - 2024 international conference on 3D vision …, 2024‏ - ieeexplore.ieee.org
Rotation invariance is a fundamental requirement of point cloud processing when input point
clouds are not aligned. Many non-invariant networks performing well on aligned point …

Self-supervised rotation-equivariant spherical vector network for learning canonical 3D point cloud orientation

H Chen, J Zhao, K Chen, Y Chen - Engineering Applications of Artificial …, 2024‏ - Elsevier
The perception of orientation in augmented reality, robot gras**, and 3D scene
understanding is commonly addressed through the utilization of hand-crafted geometric …

Tsi-Gcn: Translation and Scaling Invariant Gcn for 3d Point Cloud Analysis

Z Du, J Liang, K Yao, F Cao - Available at SSRN 4949329, 2024‏ - papers.ssrn.com
Point cloud is a crucial data format for 3D vision, but its irregularity makes it challenging to
comprehend the associated geometric information. Although some previous research has …

[PDF][PDF] Supplementary Materials for Robust 3D Shape Classification via Non-local Graph Attention Network

S Qin, Z Li, L Liu‏ - openaccess.thecvf.com
Theorem 1. For any two points xi and xj on the point cloud model, their neighborhood
matrices are **s, Xjs, and their Gram matrices are G (**s), G (Xjs), respectively. If G (**s) and …