A speech coding algorithm based on wavelet transform
暂无分享,去创建一个
Publisher Summary Wavelet transform has begun to be used in audio compression and speech parameter abstraction; it has not been used in speech compression. In this chapter, a speech coding algorithm is given based on wavelet transform. It has the features of high compression gain, good quality, wide applicability, and no artifacts. It is effective not only to speech signals but also to music signals, so it helps to implement low bit rate audio compression. The shortcoming of the algorithm is that because of the high compression gain, too much of the high frequency part of the signal is lost. Therefore, the reconstructed speeches sound low and deep. Because this algorithm belongs to waveform coding methods of transform domain, the use of perceptual effect is necessary, which will greatly improve the quality of the reconstructed signals.
[1] S. Mallat. Multiresolution approximations and wavelet orthonormal bases of L^2(R) , 1989 .
[2] I. Daubechies. Orthonormal bases of compactly supported wavelets , 1988 .
[3] Deepen Sinha,et al. Low bit rate transparent audio compression using adapted wavelets , 1993, IEEE Trans. Signal Process..
[4] Deepen Sinha,et al. On the optimal choice of a wavelet for signal representation , 1992, IEEE Trans. Inf. Theory.