HAL: hal-00687408, version 1

Detailed view

Export this paper BibTeX - EndNote - TEI - RefWorks -

10th International Conference on Latent Variable Analysis and Source Separation (LVA/ICA 2012), Tel-Aviv : Israël (2012)

Separation of Sparse Signals in Overdetermined Linear-Quadratic Mixtures

Leonardo Tomazeli Duarte 1, Rafael A. Ando 1, Romis Attux 1, Yannick Deville 2, Christian Jutten (  Corresponding author ) 3

For the Université de Campinas (Brésil), UPS Toulouse collaboration(s)

(2012-03)

In this work, we deal with the problem of nonlinear blind source separation (BSS). We propose a new method for BSS in overdetermined linear-quadratic (LQ) mixtures. By exploiting the assumption that the sources are sparse in a transformed domain, we define a framework for canceling the nonlinear part of the mixing process. After that, separation can be conducted by linear BSS algorithms. Experiments with synthetic data are performed to assess the viability of our proposal.

1:	Digital Signal Processing and Communications laboratory (DSPCom)
	University of Campinas
2:	Institut de recherche en astrophysique et planétologie (IRAP)
	CNRS : UMR5277 – Université Paul Sabatier [UPS] - Toulouse III – Observatoire Midi-Pyrénées
3:	Grenoble Images Parole Signal Automatique (GIPSA-lab)
	CNRS : UMR5216 – Université Joseph Fourier - Grenoble I – Université Pierre-Mendès-France - Grenoble II – Université Stendhal - Grenoble III – Institut Polytechnique de Grenoble - Grenoble Institute of Technology

Research team: Unicamp ViBS

Subject

:

Computer Science/Signal and Image Processing Engineering Sciences/Signal and Image processing

Keyword(s): Nonlinear mixtures – sparse signals – blind source separation

hal-00687408, version 1

http://hal.archives-ouvertes.fr/hal-00687408

oai:hal.archives-ouvertes.fr:hal-00687408

From: Christian Jutten <>

Submitted on: Friday, 13 April 2012 10:07:26

Updated on: Friday, 20 April 2012 14:35:28