Abstract
Spiking Neural Networks (SNNs) are distributed systems whose computingelements, or neurons, are characterized by analog internal dynamics and bydigital and sparse inter-neuron, or synaptic, communications. The sparsity ofthe synaptic spiking inputs and the corresponding event-driven nature of neuralprocessing can be leveraged by hardware implementations to obtain significantenergy reductions as compared to conventional Artificial Neural Networks(ANNs). SNNs can be used not only as coprocessors to carry out given computingtasks, such as classification, but also as learning machines that adapt theirinternal parameters, e.g., their synaptic weights, on the basis of data and ofa learning criterion. This paper provides an overview of models, learningrules, and applications of SNNs from the viewpoint of stochastic signalprocessing.