Graph Structure from Point Clouds: Geometric Attention is All You Need

TL;DR

This paper introduces GravNetNorm, an attention-based method for learning graph structures from point clouds, improving top jet tagging accuracy while reducing computational costs.

Contribution

It presents a novel attention mechanism that learns graph topology in a geometric space, addressing the Topology Problem in point cloud analysis.

Findings

Competitive top jet tagging accuracy

Fewer computational resources used

Effective learned graph structure

Abstract

The use of graph neural networks has produced significant advances in point cloud problems, such as those found in high energy physics. The question of how to produce a graph structure in these problems is usually treated as a matter of heuristics, employing fully connected graphs or K-nearest neighbors. In this work, we elevate this question to utmost importance as the Topology Problem. We propose an attention mechanism that allows a graph to be constructed in a learned space that handles geometrically the flow of relevance, providing one solution to the Topology Problem. We test this architecture, called GravNetNorm, on the task of top jet tagging, and show that it is competitive in tagging accuracy, and uses far fewer computational resources than all other comparable models.