Memory-Efficient Graph Convolutional Networks for Object Classification and Detection with Event Cameras

TL;DR

This paper introduces a memory-efficient graph convolutional network approach for object classification and detection using event camera data, achieving significant reductions in model size and data representation while maintaining high accuracy and processing speed.

Contribution

It provides a comparative analysis of graph convolution operations focusing on both memory and computational efficiency, leading to a novel low-memory GCN model for event-based vision tasks.

Findings

450-fold reduction in model parameters

4.5-fold reduction in data size

Achieved 52.3% classification accuracy and 53.7% mAP@0.5 for detection

Abstract

Recent advances in event camera research emphasize processing data in its original sparse form, which allows the use of its unique features such as high temporal resolution, high dynamic range, low latency, and resistance to image blur. One promising approach for analyzing event data is through graph convolutional networks (GCNs). However, current research in this domain primarily focuses on optimizing computational costs, neglecting the associated memory costs. In this paper, we consider both factors together in order to achieve satisfying results and relatively low model complexity. For this purpose, we performed a comparative analysis of different graph convolution operations, considering factors such as execution time, the number of trainable model parameters, data format requirements, and training outcomes. Our results show a 450-fold reduction in the number of parameters for the feature extraction module and a 4.5-fold reduction in the size of the data representation while maintaining a classification accuracy of 52.3%, which is 6.3% higher compared to the operation used in state-of-the-art approaches. To further evaluate performance, we implemented the object detection architecture and evaluated its performance on the N-Caltech101 dataset. The results showed an accuracy of 53.7 % mAP@0.5 and reached an execution rate of 82 graphs per second.