Reinforcement and backpropagation training for an optical neural network using self-lensing effects

The optical bench training of an optical feedforward neural network, develo ped by the authors. is presented, The network uses an optical nonlinear mat erial for neuron processing and a trainable applied optical pattern as the network weights. The nonlinear material, with the applied weight pattern, m odulates the phase front of a forward propagating information beam by dynam ically altering the index of refraction profile of the material. To verify that the network can be trained in real time, six logic gates were trained using a reinforcement training paradigm. More importantly, to demonstrate o ptical backpropagation, three gates were trained via optical error backprop agation. The output error is optically backpropagated, detected with a CCD camera, and the weight pattern is updated and stored on a computer. The obt ained results lag. the ground work for the implementation of multilayer neu ral networks that are trained using optical error backpropagation and are a ble to solve more complex problems.