Abstract
This paper investigates the sensitivity of the joint approximate diagonalization of a set of time varying cross-spectral matrices for blind separation of convolutive mixtures of speech signals. We introduce the multitaper method of cross-spectrum estimation. Based on the work of [1] factors affecting the sensitivity of the joint approximate diagonalization problem were investigated. We studied the effect of the number of matrices in the set, and observed that there exists a link between the uniqueness of the joint diagonalizer measured by modulus of uniqueness parameter and the estimation of demixing system parameters.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Afsari, B.: Sensitivity analysis for the problem of matrix joint diagonalization. SIAM J. Matrix AA. 30, 1148–1171 (2008)
Parra, L., Spence, C.: Convolutive blind separation of non-stationary sources. IEEE Trans. Speech & Audio P. 8, 320–327 (2000)
Serviere, C., Pham, D.T.: Permutation correction in the frequency domain in blind separation of speech mixtures. EURASIP J. App. SP, 1–16 (2006)
Walden, A.T.: A unified view of multitaper multivariate spectral estimation. Biometrika 87, 767–788 (2000)
Wu, H.C., Principe, J.C.: Simultaneous diagonalization in the frequency domain (SDIF) for source separation. In: 1st Int. Workshop ICA & Signal Separation, pp. 245–250 (1999)
Matsuoka, K.: Minimal distortion principle for blind source separation. In: 41st SICE Conf., pp. 2138–2143 (2002)
Pham, D.T.: Joint approximate diagonalization of positive definite Hermitian matrices. SIAM J. Matrix AA. 22, 1136–1152 (2001)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Bulek, S., Erdol, N. (2010). Sensitivity of Joint Approximate Diagonalization in FD BSS. In: Vigneron, V., Zarzoso, V., Moreau, E., Gribonval, R., Vincent, E. (eds) Latent Variable Analysis and Signal Separation. LVA/ICA 2010. Lecture Notes in Computer Science, vol 6365. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-15995-4_47
Download citation
DOI: https://doi.org/10.1007/978-3-642-15995-4_47
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-15994-7
Online ISBN: 978-3-642-15995-4
eBook Packages: Computer ScienceComputer Science (R0)