Figure 8From: Simplified spiking neural network architecture and STDP learning algorithm applied to image classification Network structure used in the simulation. Input space of 10 × 10 is converted into a spike train by a matrix of 10 × 10 input neurons with 5 × 5 receptive field. The generated spike train is fed to the hidden layer of 9 simplified LIF neurons with training. Lateral inhibition connections are shown in red. Not all connections between the input space and encoding layer are shown.Back to article page