Convolutional Spiking-based GRU Cell for Spatio-temporal Data
Yesmine Abdennadher, Eleonora Cicciarella, Michele Rossi
TL;DR
The paper addresses the challenge of learning from spatio-temporal event data while preserving local structure and achieving energy-efficient computation. It introduces the Convolutional Spiking GRU (CS-GRU), a cell that combines convolutional processing with spiking neuron dynamics and GRU-style gating, enhanced by four design modifications and an arctan surrogate gradient. Empirical results show CS-GRU outperforms SpikGRU on temporal benchmarks NTIDIGITS and SHD and on spatio-temporal datasets MNIST, DVSGesture, and CIFAR10-DVS, with accuracies reaching up to $99.31\%$ and substantial energy efficiency gains (≈$69\%$ fewer spikes) relative to SpikGRU. The work suggests CS-GRU as a versatile building block for efficient neuromorphic architectures and future integration into larger networks such as ensembles, ResNets, and VGG-style models, enabling more capable and energy-aware event-based learning.
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.