Blind Separation of Convolutive Mixtures of Non-stationary and Temporally Uncorrelated Sources Based on Joint Diagonalization
In this paper, we propose a new method for blindly separating convolutive mixtures of non-stationary and temporally uncorrelated sources. It estimates each source and its delayed versions up to a scale factor by Jointly Diagonalizing a set of covariance matrices in the frequency domain, contrary to most existing second-order methods which require a Block Joint Diagonalization algorithm followed by a blind deconvolution to achieve the same result. Consequently, our method is much faster than these classical methods especially for higer-order mixing filters and may lead to better performance as confirmed by our simulation results.
KeywordsCovariance Matrice Generalize Source Blind Source Separation Initial Source Blind Deconvolution
- 1.Bousbia-Salah, H., Belouchrani, A., Abed-Meraim, K.: Blind separation of convolutive mixtures using joint block diagonalization. In: ISSPA, Kuala Lumpur, Malaysia, vol. 1, pp. 13–16 (August 2001)Google Scholar
- 2.Bousbia-Salah, H., Belouchrani, A., Abed-Meraim, K.: Blind separation of non stationary sources using joint block diagonalization. In: SPWSSP, Singapore, pp. 448–451 (August 2001)Google Scholar
- 3.Févotte, C., Doncarli, C.: A unified presentation of blind source separation methods for convolutive mixtures using block-diagonalization. In: ICA 2003, Nara, Japan (April 2003)Google Scholar
- 4.Saylani, H., Hosseini, S., Deville, Y.: Blind Separation of Convolutive Mixtures of Non-stationary Sources Using Joint Block Diagonalization in the Frequency Domain. In: Vigneron, V., Zarzoso, V., Moreau, E., Gribonval, R., Vincent, E. (eds.) LVA/ICA 2010. LNCS, vol. 6365, pp. 97–105. Springer, Heidelberg (2010)CrossRefGoogle Scholar