Epoch event based speech watermarking for tamper detection

Abstract

There are big challenges to handling content authentication and tamper detection in speech watermarking techniques. These challenges can be addressed using the detection of epoch locations, logistic mapping, verification of the tamper locations, and the Min algorithm. Epoch events provide the variable length epoch intervals and it is used to find the variable lengths of frame segments in this work. The logistic mapping provides strong security for original signals. They can provide a robust and secure method that addresses the issues of copyright, illegal intentional, or unintentional modifications, and tampered locations. In this paper, we have implemented an epoch event-based speech watermarking technique for tamper detection using the logistic map method. We observe that the simulation results demonstrated that the PRE value is approximately more than
82% and the PESQ value lies between excellent and good. It means that there are no significant differences between original and watermarked signals and we have obtained a highly imperceptible, robust, and secure system. Moreover, the overall simulation results of the proposed method provide a better solution as well as more security than existing methods.