Abstract
In this article, we present a hybrid optoelectronic system for pattern recognition based on a structured two-layer holographic neural network. An auxiliary subsection with local connections is added to the first layer to improve the convergence of the iterative process. We propose a new class of high-performance optical processors: linear intensity, holographic correlators that utilize global connections in both layers of the neural network.
Get full access to this article
View all access options for this article.