STDP Learning of Image Patches with Convolutional Spiking Neural Networks

TL;DR

This paper introduces convolutional spiking neural networks trained with STDP for unsupervised image feature detection, analyzing their efficiency and comparing performance on the MNIST dataset.

Contribution

It presents a novel convolutional SNN architecture with competitive learning, exploring shared and independent feature representations, and evaluates its efficiency and effectiveness.

Findings

Convolutional SNNs can learn image features unsupervised.

Shared and independent feature learning modes have different advantages.

The proposed networks show promising performance on MNIST.

Abstract

Spiking neural networks are motivated from principles of neural systems and may possess unexplored advantages in the context of machine learning. A class of \textit{convolutional spiking neural networks} is introduced, trained to detect image features with an unsupervised, competitive learning mechanism. Image features can be shared within subpopulations of neurons, or each may evolve independently to capture different features in different regions of input space. We analyze the time and memory requirements of learning with and operating such networks. The MNIST dataset is used as an experimental testbed, and comparisons are made between the performance and convergence speed of a baseline spiking neural network.