NMNIST-STDP-SpiNNaker

Implementation of a Spiking Neural Network model inspired by Deihl et al. for digit recognition on the N-MNIST dataset on the Neuromophic system SpiNNaker using the software package sPyNNaker.

Dependencies

• Python3
• pyNN
• numpy
• matplotlib
• seaborn
• os
• sys
• openCV
• random
• eventvision [https://github.com/gorchard/event-Python/blob/master/eventvision.py]

To-Do

• Get Accuracy.
• Hyperparameter tuning for digit recognition of 3 digits (5, 6 and 9).
• Hyperparameter tuning for digit recognition of all digits (0 to 9).
• Testing Larger Models (with larger population sizes)