Convolutional Spiking-based GRU Cell for Spatio-temporal Data

TL;DR

This paper introduces the Convolutional Spiking GRU (CS-GRU), a novel neural network cell that combines convolutional operations with spiking neurons and GRU gating to effectively process spatio-temporal data with improved accuracy and efficiency.

Contribution

The paper proposes the CS-GRU cell, integrating convolutional operations with spiking GRUs to better capture local dependencies in spatio-temporal data, outperforming previous models.

Findings

CS-GRU achieves over 90% accuracy on sequential tasks.

It outperforms state-of-the-art GRU variants by 4.35%.

The approach is 69% more efficient than SpikGRU.

Abstract

Spike-based temporal messaging enables SNNs to efficiently process both purely temporal and spatio-temporal time-series or event-driven data. Combining SNNs with Gated Recurrent Units (GRUs), a variant of recurrent neural networks, gives rise to a robust framework for sequential data processing; however, traditional RNNs often lose local details when handling long sequences. Previous approaches, such as SpikGRU, fail to capture fine-grained local dependencies in event-based spatio-temporal data. In this paper, we introduce the Convolutional Spiking GRU (CS-GRU) cell, which leverages convolutional operations to preserve local structure and dependencies while integrating the temporal precision of spiking neurons with the efficient gating mechanisms of GRUs. This versatile architecture excels on both temporal datasets (NTIDIGITS, SHD) and spatio-temporal benchmarks (MNIST, DVSGesture, CIFAR10DVS). Our experiments show that CS-GRU outperforms state-of-the-art GRU variants by an average of 4.35%, achieving over 90% accuracy on sequential tasks and up to 99.31% on MNIST. It is worth noting that our solution achieves 69% higher efficiency compared to SpikGRU. The code is available at: https://github.com/YesmineAbdennadher/CS-GRU.