An optical neural network based on distributed holographic gratings for ATR

We describe a laser diode-based optoelectronic implementation of artificial neural networks which utilizes real-time holography in photorefractive crystals. The use of a laser diode light source reduces the system size and power requirements. The holographic material is rhodium-doped BaTiO/sub 3/ which has enhanced sensitivity at the laser-diode wavelength of 830 nm. A balanced coherent-detection method is used to represent bipolar optical neurons and weights. In addition, by distributing each neuron weight among a set of spatially and angularly distributed gratings using beam fanning, Bragg degeneracy and its associated inter-neuron optical crosstalk is virtually eliminated. The structure of the neural network is programmable and we have implemented a variety of neural networks including backpropagation and Kohonen-style self-organizing maps with up to 10,000 neurons and performance of up to 10/sup 8/ weights processed per second during learning and readout. We also discuss the weight decay in photorefractive materials, specifically its relative effect in the neural network and data storage domains. Applications to ATR are discussed.