Orthogonal and Non-Orthogonal Joint Blind Source Separation in the Least-Squares Sense
Résumé
We present two new algorithms for the orthogonal and nonorthogonal joint blind source separation (JBSS), a flexible framework that extends the well-known blind source separation to the case when multiple datasets are decomposed simultaneously. The algorithms minimize the total sum of squares of the off-diagonal terms by means of very simple gradient ascent iterations.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...