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.
CITATION STYLE
Yue, T. W., & Chiang, S. (2001). The general neural-network paradigm for visual cryptography. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2084 LNCS, pp. 196–206). Springer Verlag. https://doi.org/10.1007/3-540-45720-8_23
Mendeley helps you to discover research relevant for your work.