/spiking-neural-networks-mnist-classification

Pure python implementation of unsupervised MNIST classification using Spiking Neural Networks (using STDP)

Primary LanguagePython

This repository is created to implement classification of MNIST dataset using SNN

Following papers have been used as a reference:-

1.STDP-based Unsupervised Feature Learning using Convolution-over-time in Spiking Neural Networks for Energy-Efficient Neuromorphic Computing by Gopalakrishnan Srinivasan, Priyadarshini Panda and Kaushik Roy

2.Unsupervised learning of digit recognition using spike-timing-dependent plasticity by Peter U. Diehl and Matthew Cook Institute

Observations so far:-

  1. Basic Network
Number of input neurons : 784 (image 28x28 -> 784 x 1 vector)
Number of hidden neurons : 0
Number of output neurons : 800 neurons trained on 80 samples of each digit
Classification accuracy on MNIST Test set = 71.49%