The semipublic encryption for visual cryptography using Q’tron neural networks

The paper proposes the semipublic encrypting scheme for visual cryptography using the Q’tron neural-network (Q’tron NN) model. This encrypting scheme hides only the true secret from the public. That is, the pictorial meaning appearing in a public share describes the public information of a document while leaving its confidential part undisclosed. A piece of confidential information is retrievable if and only if a right user share is available. The principle to construct the Q’tron NN to fulfill the aforementioned scheme will be thoroughly discussed. An application that uses the scheme for key distribution in a public area will be demonstrated.