Laplacian Mixture Modeling for Overcomplete Mixture Matrix Estimation in Wavelet Packet Domain by Adaptive EM-type Algorithm

Speech process has benefited a great deal from the wavelet transforms. Wavelet packets decompose signals in to broader components using linear spectral bisecting. In this paper, mixtures of speech signals are decomposed using wavelet packets, the phase difference between the two mixtures are investigated in wavelet domain. In our method Laplacian mixture model (LMM) is defined. An expectation maximization (EM) algorithm is used for training of the model and calculation of model parameters which is the mixture matrix. Therefore individual speech components of speech mixtures are separated