The general neural-network paradigm for visual cryptography

Abstract

This paper proposes the general paradigm to build Q'tron neural networks (NNs) for visual cryptography. Given a visual encryption scheme, usually described using an access structure, it was formulated as a optimization problem of integer programming by which the a Q'tron NN with the so-called integer-programming-type energy function is, then, built to fulfill that scheme. Remarkably, this type of energy function has the so-called known-energy property, which allows us to inject bounded noises persistently into Q'trons in the NN to escape local minima. The so-built Q'tron NN, as a result, will settle down onto a solution state if and only if the instance of the given encryption scheme is realizable. © Springer-Verlag Berlin Heidelberg 2001.