An Adaptive Multiscale Framework for Compressed Sensing of Speech Signal

In this paper, a matrix form of Sym wavelet synthesis is deduced, keeping the length of the coefficient no more than the length of original speech signals, and then we propose an Adaptive Multiscale Compressed Sensing (AMCS) method, which design the sensing matrix and the num of level of wavelet decomposition adaptively, according to the sparsity of each level wavelet coefficients of the speech signals. We compare AMCS with Multiscale Compressed Sensing (MCS) by applying both methods to speech compression and reconstruction, and the reconstructed speech signal is evaluated by the objective and subjective evaluation. The experimental results show that the reconstruction performance of speech signal based on AMCS is superior to MCS.