Application of Mallat's Algorithm to the Analysis and Compression of Speech Signal
Yang Lei; Zheng Huimin; Yang Yanq (Intelligence, Language and Computer Sciences Institute, USTB, Beijing 100083, PRC)
The Mallat' s Algorithm in wavelet transforms is used to preprocess the chinese speech signal. By decomposing the speech signal, the compression ratios is improved under the certain distortion degree. The theory of wavelet transforms and the Mallat ' s Algorithm are proposed and the method is verified by the experimental results.