| | |
Summary: ICASSP 2006
JOINT DIAGONALIZATION ON THE OBLIQUE MANIFOLD FOR INDEPENDENT
COMPONENT ANALYSIS
P.-A. Absil
Department of Mathematical Engineering
Universit´e catholique de Louvain, Belgium
and Peterhouse, University of Cambridge, UK
K. A. Gallivan
School of Computational Science
Florida State University
Tallahassee, FL 32306-4120, USA
ABSTRACT
Several blind source separation algorithms obtain a separat-
ing matrix by computing the congruence transformation that
"best" diagonalizes a collection of covariance matrices. Re-
cent methods avoid a pre-whitening phase and directly at-
tempt to compute a non-orthogonal diagonalizing congruence.
However, since the magnitude of the sources is unknown,
there is a fundamental indeterminacy on the norm of the rows
of the separating matrix. We show how this indeterminacy can
|
