We present a blind speech watermarking algorithm that embeds the watermark data in the phase of non-voicedspeech by replacing the excitation signal of an autoregressive speech signal representation. The watermark signal is embedded in a frequency subband, which facilitates robustness against bandpass filtering channels. We derive several sets of pulse shapes that prevent intersymbol interference and that allow the passband watermark signal to be created by simple filtering. A marker-based synchronization scheme robustly detects the location of the embedded watermark data without the occurrence of insertions or deletions. In light of a potential application to analog aeronautical voice radio communication, we present experimental results for embedding a watermark in narrowband speech at a bit-rate of 450 bit/s. The recursive least-squares (RLS) equalization-based watermark detector not only compensates for the vocal tract filtering, but also recovers the watermark data in the presence of nonlinear phase and bandpass filtering, amplitude modulation, and additive white Gaussian noise (AWGN), making the watermarking scheme highly robust.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here

PROJECT TITLE :Analysing Acoustic Model Changes For Active Learning In Automatic Speech Recognition - 2017ABSTRACT:In active learning for Automatic Speech Recognition (ASR), a portion of information is automatically selected for
PROJECT TITLE :Deep Learning for Acoustic Modeling in Parametric Speech Generation: A systematic review of existing techniques and future trendsABSTRACT:Hidden Markov models (HMMs) and Gaussian mixture models (GMMs) are the 2
PROJECT TITLE :Memory-efficient buffering method and enhanced reference template for embedded automatic speech recognition systemABSTRACT:This work realises a memory-economical embedded automatic speech recognition (ASR) system
PROJECT TITLE :Toward a Universal Synthetic Speech Spoofing Detection Using Phase InformationABSTRACT:In the sector of speaker verification (SV) it is nowadays possible and comparatively easy to make a artificial voice to deceive
PROJECT TITLE :Weight-Space Viterbi Decoding Based Spectral Subtraction for Reverberant Speech RecognitionABSTRACT:One-channel blind dereverberation algorithm is proposed in this letter for distant-talking speech recognition.

Ready to Complete Your Academic MTech Project Work In Affordable Price ?

Project Enquiry