Underdetermined blind source separation of convolutive mixtures by hierarchical clustering and L1-norm minimization

Winter S, Kellermann W, Sawada H, Makino S (2007)


Publication Language: English

Publication Type: Book chapter / Article in edited volumes

Publication year: 2007

Publisher: Springer

Edited Volumes: Blind Speech Separation

Series: Signals and Communication Technology

City/Town: Dordrecht

Pages Range: 271-304

ISBN: 978-1-4020-6479-1

DOI: 10.1007/978-1-4020-6479-1_10

Abstract

In this chapter we present a complete solution for underdetermined blind source separation (BSS) of convolutive speech mixtures based on two stages. In the first stage, the mixing system is estimated, for which we employ hierarchical clustering. Based on the estimated mixing system, the source signals are estimated in the second stage. The solution for the second stage utilizes the common assumption of independent and identically distributed sources. Modeling the sources by a Laplacian distribution leads to ℓ1-norm minimization.©Springer

Authors with CRIS profile

How to cite

APA:

Winter, S., Kellermann, W., Sawada, H., & Makino, S. (2007). Underdetermined blind source separation of convolutive mixtures by hierarchical clustering and L1-norm minimization. In S. Makino, Te-Won Lee, H. Sawada (Eds.), Blind Speech Separation. (pp. 271-304). Dordrecht: Springer.

MLA:

Winter, Stefan, et al. "Underdetermined blind source separation of convolutive mixtures by hierarchical clustering and L1-norm minimization." Blind Speech Separation. Ed. S. Makino, Te-Won Lee, H. Sawada, Dordrecht: Springer, 2007. 271-304.

BibTeX: Download