Blind identification of underdetermined mixtures by simultaneous matrix diagonalization

Lieven De Lathauwer, Jos├ęphine Castaing

Abstract

In this paper, we study simultaneous matrix diagonalization-based techniques for the estimation of the mixing matrix in underdetermined independent component analysis (ICA). This includes a generalization to underdetermined mixtures of the well-known SOBI algorithm. The problem is reformulated in terms of the parallel factor decomposition (PARAFAC) of a higher-order tensor. We present conditions under which the mixing matrix is unique and discuss several algorithms for its computation.

Code description

This package provides an implementation of the SOBIUM algorithm discussed in the paper on blind identification by simultaneous diagonalization. Both algorithm 1 and 2 from the paper are available.

Reference

L. De Lathauwer, J. Castaing, "Blind identification of underdetermined mixtures by simultaneous matrix diagonalization," IEEE Transactions on Signal Processing, Vol. 56, No. 3, pp. 1096-1105, Mar. 2008.

Download code



This repository can be cited as:
S. Hendrikx, M. Boussé, N. Vervliet, M. Vandecappelle, R. Kenis, and L. De Lathauwer, Tensorlab⁺, Available online, Version of Dec 2022 downloaded from https://www.tensorlabplus.net.