Rotation invariant point cloud analysis: Where local geometry meets global topology. (July 2022)

• Record Type: Journal Article
• Authors: Zhao, Chen; Yang, Jiaqi; Xiong, Xin; Zhu, Angfan; Cao, Zhiguo; Li, Xin
• Is Part Of: Pattern recognition. Volume 127(2022)
• Publication Date: 2022-07
• Subjects: Point cloud analysis -- Rotation invariance -- Deep learning -- Classification -- Segmentation; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2022.108626
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We present LGR-Net which considers local geometric features and global topology-preserving features to achieve rotation invariance. The complementary relationship between shape descriptions and spatial attributes is adaptively exploited by an attention-based fusion module. LGR-Net significantly outperforms state-of-the-art methods on both synthetic and real-world datasets undergoing random 3D rotations. Abstract: Point cloud analysis is a fundamental task in 3D computer vision. Most previous works have conducted experiments on synthetic datasets with well-aligned data; while real-world point clouds are often not pre-aligned. How to achieve rotation invariance remains an open problem in point cloud analysis. To meet this challenge, we propose an approach toward achieving rotation-invariant (RI) representations by combining local geometry with global topology. In our local-global-representation (LGR)-Net, we have designed a two-branch network where one stream encodes local geometric RI features and the other encodes global topology-preserving RI features. Motivated by the observation that local geometry and global topology have different yet complementary RI responses in varying regions, two-branch RI features are fused by an innovative multi-layer perceptron (MLP) based attention module. To the best of our knowledge, this work is the first principled approach toward adaptively combining global and local information under the context of RI point cloud analysis. Extensive experiments have demonstrated that our LGR-Net achieves the state-of-the-art performance on various rotation-augmented versions of ModelNet40, ShapeNet, ScanObjectNN, and S3DIS.