creators_name: Sprekeler, Henning
creators_name: Zito, Tiziano
creators_name: Wiskott, Laurenz
creators_id: henning.sprekeler@epfl.ch
creators_id: tiziano.zito@bccn-berlin.de
creators_id: l.wiskott@biologie.hu-berlin.de
type: preprint
datestamp: 2010-10-18 11:03:37
lastmod: 2011-03-11 08:57:45
metadata_visibility: show
title: An Extension of Slow Feature Analysis for Nonlinear Blind Source Separation
subjects: comp-sci-mach-learn
full_text_status: public
keywords: slow feature analysis, nonlinear blind source separation, independent component analysis
abstract: We present and test an extension of slow feature analysis as a novel approach to nonlinear blind source separation. The algorithm relies on temporal correlations and iteratively reconstructs a set of statistically independent sources from arbitrary nonlinear instantaneous mixtures. Simulations show that it is able to invert a complicated nonlinear mixture of two audio signals with a reliability of more than $90$\%. The algorithm is based on a  mathematical analysis of slow feature analysis for the case of input data that are generated from statistically independent sources.
date: 2010-10
date_type: submitted
refereed: FALSE
citation:   Sprekeler, Dr. Henning and Zito, Tiziano and Wiskott, Dr. Laurenz  (2010) An Extension of Slow Feature Analysis for Nonlinear Blind Source Separation.  [Preprint]     
document_url: http://cogprints.org/7056/1/SprekelerZitoWiskott-Cogprints-2010.pdf