NEURAL NETWORKS: CONTINUOUS APPROACH AND APPLICATIONS TO IMAGE PROCESSING
Georges‐Henri Cottet · Journal of Biological Systems · 1995
Our goal here is to describe elementary tools allowing to translate collective behavior of large neural networks in a continous, with respect to time and space, framework. Our starting point will be the Mac-Culloch and Pitts formal neuron, which operates in the following way: if (xi) a set of neurons connected through a synaptic matrix J(xi, xj), at every time step t, the states V t i of the neurons will be updated according to the rule