Backpropagation training of an optical neural network

The experimental demonstration of optical backpropagation in a feedforward optical neural network, developed by the authors, is presented. To our knowledge, this is the first report of backpropagation training in an optical system. The network uses a thermal nonlinear material as a neural processing layer and a photorefractive crystal as a phase conjugate mirror to backpropagate the optical error. The nonlinear material modulates the phase front of a forward propagating information beam by dynamically altering the index of the refraction profile of the material via a stronger weighting beam