Samadzadeh Ali ; Far Fatemeh Sadat Tabatabaei ; Javadi Ali ; Nickabadi Ahmad ; Chehreghani Morteza Haghir

Abstract

Spiking neural networks (SNNs) can be used in low-power and embedded systems(such as emerging neuromorphic chips) due to their event-based nature. Also,they have the advantage of low computation cost in contrast to conventionalartificial neural networks (ANNs), while preserving ANN's properties. However,temporal coding in layers of convolutional spiking neural networks and othertypes of SNNs has yet to be studied. In this paper, we provide insight intospatio-temporal feature extraction of convolutional SNNs in experimentsdesigned to exploit this property. The shallow convolutional SNN outperformsstate-of-the-art spatio-temporal feature extractor methods such as C3D,ConvLstm, and similar networks. Furthermore, we present a new deep spikingarchitecture to tackle real-world problems (in particular classification tasks)which achieved superior performance compared to other SNN methods on NMNIST(99.6%), DVS-CIFAR10 (69.2%) and DVS-Gesture (96.7%) and ANN methods on UCF-101(42.1%) and HMDB-51 (21.5%) datasets. It is also worth noting that the trainingprocess is implemented based on variation of spatio-temporal backpropagationexplained in the paper.