The optimal wavelet for speech compression

Abstract

The main idea of speech compression is to reduce the bite rate of the speech for communication or storage without significant loss of quality. There are mainly three functional categories of speech processing methods in use. The new technique called wavelet transform is being used for speech signal analysis and synthesis. The major issues regarding the design of real time wavelet based speech coder are choosing optimal wavelets for the compression, and selecting suitable frame size etc. The performance of the different wavelets families on speech compression is evaluated and compared based on different parameters. Male and female speech is used for the comparison and analysis. © 2011 Springer-Verlag.