The Discrete Wavelet Transform is the most powerful and new signal compression technique which uses multi-resolution analysis for analyzing speech signal. This article proposes the frequency loss less comparison of the modern wavelets based on speech compression technique. Haar wavelet (Haar), Biorthogonal wavelet (Bior) and Discrete approximation of Meyer wavelet (Dmey) are investigated in this experiment with eighty speech signal from male and female. The experiments indicated that the Bior provides the best performance in time and frequency domain. Also it can also gain the compression rate as 2.67 to the original speech signal.
Published in:
Computer, Consumer and Control (IS3C), 2012 International Symposium on
Date of Conference: 4-6 June 2012