Speech Watermarking Based on Significant Events of Speech

Abstract

With the increased use of the Internet in everyday life, data transmission over the Internet needs to be secure, so it necessitates the adoption of watermarking. Water- marking is the best solution for securing digital content and ensuring the integrity of the original data. It also certifies digital content and provides the best form of concealing information for permission. Speech watermarking is a functioning research area for the most recent decades. Digital speech watermarking is a hearty method to cover up and consequently secure information from any purposeful or accidental control over the transmission. Speech can be considered as a sequence of events. The event is defined as the sequence of any significant changes during the production of speech. The main issue related to speech watermarking is to maintain the high security to protect our original copy of watermarked speech signals. For security issues, we have explored two speech events vowel onset and vowel offset points for segmenting speech into vowel and non-vowel segments. The temper detection and authentication issues have been addressed using epoch speech events based on LPA, logistic mapping, and longest common subsequence (LCS) methods. Epoch detection provides 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 tempered locations. The event-based speech watermarking technique provides us with a robust and secure result because it produces variable-length frame segments. newlineThe major contributions of the thesis are: A method based on Fast Fourier Transform and Arnold transform has been proposed for speech watermarking. The LP analysis-based method has been proposed for robust speech watermark- ing. newline Vowel onset and offset points-based digital speech watermarking technique has bee