This paper aims to sho" that the signal compression algorithms using the implementation of a wavelet filter bank based on the "traditional lifting scheme" could be improved through the concept of "adaptive lifting". In the context described next, the term of "adaptive" means selecting not only the best wavelet basis, but also the most appropriate Euclidean lifting scheme that minimizes the signal entropy or the Golomb-Rice code length. Three type of signals have been considered in the experimental tests: artificially generated AR signals, real speech and standard MPEG audio signals. All tests have shown that the adaptive lifting performs better than the traditional lifting.
signal compression, wavelets, Shannon entropy, Golomb-Rice code length.
All articles published by Studies in Informatics and Control (SIC) are published in accordance with the terms and conditions of the Creative Commons Attribution License (CC BY-NC-SA 4.0).