Skip to main content
top

Bibliography

Conference Paper (international conference)

On Computation of Approximate Joint Block-Diagonalization Using Ordinary AJD

Tichavský Petr, Yeredor A., Koldovský Zbyněk

: Latent Variable Analysis and Signal Separation, p. 163-171 , Eds: Theis Fabian

: Latent Variable Analysis and Signal Separation,10th International Conference, LVA/ICA 2012, (Tel Aviv, IL, 12.03.2012-15.03.2012)

: 1M0572, GA MŠk, GA102/09/1278, GA ČR

: joint block diagonalization, independent subspace analysis

: 10.1007/978-3-642-28551-6_21

: http://library.utia.cas.cz/separaty/2012/SI/tichavsky-on computation of approximate joint block-diagonalization using ordinary ajd.pdf

(eng): Approximate joint block diagonalization (AJBD) of a set of matrices has applications in blind source separation, e.g., when the signal mixtures contain mutually independent subspaces of dimension higher than one. The main message of this paper is that certain ordinary approximate joint diagonalization (AJD) methods can also be used successfully for AJBD, but not all are suitable equally well. In particular, we prove that when the set is exactly jointly block-diagonalizable, perfect block-diagonalization is attainable by the recently proposed AJD algorithm “U-WEDGE" (uniformly weighted exhaustive diagonalization with Gaussian iteration) - but this basic consistency property is not shared by some other popular AJD algorithms. In addition, we show using simulation, that in the more general noisy case, the subspace identification accuracy of U-WEDGE compares favorably to competitors.

: BB