GAPNet: Graph Attention based Point Neural Network for Exploiting Local Feature of Point Cloud

TL;DR

GAPNet introduces a graph attention mechanism within a neural network to better capture local geometric features in point clouds, significantly improving shape classification and segmentation accuracy.

Contribution

The paper presents a novel graph attention-based neural network, GAPNet, that effectively learns local features in point clouds using attention mechanisms and multi-head aggregation.

Findings

Achieves state-of-the-art results on ModelNet40 for shape classification.

Outperforms existing methods in part segmentation on ShapeNet dataset.

Demonstrates robustness and effectiveness of local feature learning in point cloud analysis.

Abstract

Exploiting fine-grained semantic features on point cloud is still challenging due to its irregular and sparse structure in a non-Euclidean space. Among existing studies, PointNet provides an efficient and promising approach to learn shape features directly on unordered 3D point cloud and has achieved competitive performance. However, local feature that is helpful towards better contextual learning is not considered. Meanwhile, attention mechanism shows efficiency in capturing node representation on graph-based data by attending over neighboring nodes. In this paper, we propose a novel neural network for point cloud, dubbed GAPNet, to learn local geometric representations by embedding graph attention mechanism within stacked Multi-Layer-Perceptron (MLP) layers. Firstly, we introduce a GAPLayer to learn attention features for each point by highlighting different attention weights on neighborhood. Secondly, in order to exploit sufficient features, a multi-head mechanism is employed to allow GAPLayer to aggregate different features from independent heads. Thirdly, we propose an attention pooling layer over neighbors to capture local signature aimed at enhancing network robustness. Finally, GAPNet applies stacked MLP layers to attention features and local signature to fully extract local geometric structures. The proposed GAPNet architecture is tested on the ModelNet40 and ShapeNet part datasets, and achieves state-of-the-art performance in both shape classification and part segmentation tasks.