Performance of Graph Neural Networks for Point Cloud Applications

TL;DR

This paper analyzes the performance of EdgeConv-based Dynamic Graph Neural Networks for point cloud classification, highlighting the computational costs of dynamic graph generation and proposing a quasi-DGNN to reduce latency while maintaining accuracy.

Contribution

It provides a detailed profiling of DGNNs' inference performance and introduces a quasi-DGNN approach to optimize latency without sacrificing accuracy.

Findings

Dynamic graph generation via kNN dominates inference latency (up to 95% on GPU).

Halting dynamic graph updates at a certain network depth reduces latency significantly.

Proposed quasi-DGNN maintains accuracy while improving inference speed.

Abstract

Graph Neural Networks (GNNs) have gained significant momentum recently due to their capability to learn on unstructured graph data. Dynamic GNNs (DGNNs) are the current state-of-the-art for point cloud applications; such applications (viz. autonomous driving) require real-time processing at the edge with tight latency and memory constraints. Conducting performance analysis on such DGNNs, thus, becomes a crucial task to evaluate network suitability. This paper presents a profiling analysis of EdgeConv-based DGNNs applied to point cloud inputs. We assess their inference performance in terms of end-to-end latency and memory consumption on state-of-the-art CPU and GPU platforms. The EdgeConv layer has two stages: (1) dynamic graph generation using k-Nearest Neighbors (kNN) and, (2) node feature updation. The addition of dynamic graph generation via kNN in each (EdgeConv) layer enhances network performance compared to networks that work with the same static graph in each layer; such performance enhancement comes, however, at the added computational cost associated with the dynamic graph generation stage (via kNN algorithm). Understanding its costs is essential for identifying the performance bottleneck and exploring potential avenues for hardware acceleration. To this end, this paper aims to shed light on the performance characteristics of EdgeConv-based DGNNs for point cloud inputs. Our performance analysis on a state-of-the-art EdgeConv network for classification shows that the dynamic graph construction via kNN takes up upwards of 95% of network latency on the GPU and almost 90% on the CPU. Moreover, we propose a quasi-Dynamic Graph Neural Network (qDGNN) that halts dynamic graph updates after a specific depth within the network to significantly reduce the latency on both CPU and GPU whilst matching the original networks inference accuracy.