Optical Neural Networks

In recent years, there have been extensive research efforts on photorefractive phenomena, materials, devices and applications. Much of the progress has been achieved and some potential applications are currently under extensive investigations [1-3]. Holographic optical information storage and optical information processing are two important areas that are considered to be of great potential for practical applications [4-6]. The merits of holographic storage come from the capacity of information storage in the material volumes as well as the parallel access characteristics of the optical systems, whereas the potential for optical information processing is obtained from the dynamic characteristics of the photorefractive materials. The dynamic behavior of the photorefractive materials has led to many interesting applications such as optical wave mixing, optical phase conjugation, and optical logic operations. In this chapter, we present the application of photorefractive dynamic memory for optical neural networks. In particular, we focus on the optical implementation of neural networks, discussing the characteristics of the photorefractive neural networks, and describing its application for real-time image recognition.