PRA-Net: Point Relation-Aware Network for 3D Point Cloud Analysis

TL;DR

PRA-Net is a novel framework that enhances 3D point cloud analysis by jointly learning intra-region structures and inter-region relations, improving feature representation for various tasks.

Contribution

It introduces a unified approach combining intra-region structure learning and inter-region relation learning for better point cloud features.

Findings

Effective across shape classification, keypoint estimation, and part segmentation.

Outperforms existing methods on several 3D benchmarks.

Demonstrates strong generalization ability.

Abstract

Learning intra-region contexts and inter-region relations are two effective strategies to strengthen feature representations for point cloud analysis. However, unifying the two strategies for point cloud representation is not fully emphasized in existing methods. To this end, we propose a novel framework named Point Relation-Aware Network (PRA-Net), which is composed of an Intra-region Structure Learning (ISL) module and an Inter-region Relation Learning (IRL) module. The ISL module can dynamically integrate the local structural information into the point features, while the IRL module captures inter-region relations adaptively and efficiently via a differentiable region partition scheme and a representative point-based strategy. Extensive experiments on several 3D benchmarks covering shape classification, keypoint estimation, and part segmentation have verified the effectiveness and the generalization ability of PRA-Net. Code will be available at https://github.com/XiwuChen/PRA-Net .