Semantic multimedia authentication with model vector signature

Abstract

We propose novel techniques for image/video authentication in the semantic level. These methods use statistical learning, visual object segmentation and classification schemes for semantic understanding of visual content. A robust rotation, scaling, and translation public digital signature is generated. The authentication process is executed by comparing the classification result with the information carried by the digital signature. This method leads the authentication system to learn the semantic content of multimedia data and performs authentication task in the semantic level.