Global Attention-Guided Dual-Domain Point Cloud Feature Learning for Classification and Segmentation

TL;DR

This paper introduces a novel point cloud analysis network that employs global attention-guided dual-domain feature learning to improve input embedding and aggregation efficiency, enhancing classification and segmentation tasks.

Contribution

It proposes the CPT module with global attention for better input embedding and the DKFF method for dual-domain feature aggregation, addressing key limitations in existing point cloud models.

Findings

Achieves superior performance on classification and segmentation tasks

Demonstrates the effectiveness of CPT and DKFF modules

Outperforms existing methods in multiple point cloud benchmarks

Abstract

Previous studies have demonstrated the effectiveness of point-based neural models on the point cloud analysis task. However, there remains a crucial issue on producing the efficient input embedding for raw point coordinates. Moreover, another issue lies in the limited efficiency of neighboring aggregations, which is a critical component in the network stem. In this paper, we propose a Global Attention-guided Dual-domain Feature Learning network (GAD) to address the above-mentioned issues. We first devise the Contextual Position-enhanced Transformer (CPT) module, which is armed with an improved global attention mechanism, to produce a global-aware input embedding that serves as the guidance to subsequent aggregations. Then, the Dual-domain K-nearest neighbor Feature Fusion (DKFF) is cascaded to conduct effective feature aggregation through novel dual-domain feature learning which appreciates both local geometric relations and long-distance semantic connections. Extensive experiments on multiple point cloud analysis tasks (e.g., classification, part segmentation, and scene semantic segmentation) demonstrate the superior performance of the proposed method and the efficacy of the devised modules.