Performance evaluation of the symmetric tight wavelet frame in speech coding

The wavelet frame is an extension to wavelet analysis which has become an interesting subject of research in signal processing and applied mathematics. Compared to wavelet base the wavelet frame have a variety of attractive feature such as better time-frequency localization, robust and stable reconstruction, smoothness, short support and symmetry. The goal of this paper is to use the advantages of the symmetric tight wavelet frame in speech coding and to evaluate the performances. The performances of such decomposition are measured using objective criteria: the normalized root mean square error NRMSE, the peak signal to noise ratio PSNR, the segmental signal to noise ratio SNRseg, the frequency weighted segmental signal to noise ratio fwSNRseg, and the perceptual evaluation of speech quality PESQ. The experimental results show that using tight wavelet frame significantly improves the quality of the coded signal.