Title :
Optical neural network using Kerr-type nonlinear materials
Author :
Skinner, Steven R. ; Steck, James E. ; Behrman, Elizabeth C.
Author_Institution :
Wichita State Univ., KS, USA
Abstract :
An innovative optical implementation of a feedforward artificial neural network is presented using Kerr-type nonlinear optical materials. Many Kerr-type materials have ultrafast response times and allow both weighted connections and nonlinear neuron processing to be implemented using only thin layers of Kerr-type material separated by free space. Nonlinear neuron processing results from the nonlinear Kerr-type effect (self-focusing or self-defocusing) and interference of the optical signals within the device. In addition, the Kerr effect is used to create optical connections by applying patterns of irradiance to thin layers of the nonlinear media. This is a result of a variation of the refractive index profile of the nonlinear media in response to the applied irradiance
Keywords :
feedforward neural nets; optical Kerr effect; optical materials; optical neural nets; Kerr-type nonlinear materials; feedforward artificial neural network; nonlinear neuron processing; optical neural network; patterns of irradiance; refractive index profile; self-defocusing; self-focusing; weighted connections; Artificial neural networks; Neural networks; Neurons; Nonlinear optics; Optical computing; Optical fiber networks; Optical materials; Optical refraction; Optical variables control; Ultrafast optics;
Conference_Titel :
Microelectronics for Neural Networks and Fuzzy Systems, 1994., Proceedings of the Fourth International Conference on
Conference_Location :
Turin
Print_ISBN :
0-8186-6710-9
DOI :
10.1109/ICMNN.1994.593141