Inlier-based ICA with an application to superimposed images

Frank C. Meinecke, Stefan Harmeling, Klaus Muller

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

This paper proposes a new independent component analysis (ICA) method which is able to unmix overcomplete mixtures of sparce or structured signals like speech, music or images. Furthermore, the method is designed to be robust against outliers, which is a favorable feature for ICA algorithms since most of them are extremely sensitive to outliers. Our approach is based on a simple outlier index. However, instead of robustifying an existing algorithm by some outlier rejection technique we show how this index can be used directly to solve the ICA problem for super-Gaussian sources. The resulting inlier-based ICA (IBICA) is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals).

Original languageEnglish
Pages (from-to)48-55
Number of pages8
JournalInternational Journal of Imaging Systems and Technology
Volume15
Issue number1
DOIs
Publication statusPublished - 2005 Oct 5
Externally publishedYes

Fingerprint

Independent component analysis
music
rejection

Keywords

  • Blind Source Separation (BSS)
  • Independent Component Analysis (ICA)
  • Outlier robustness
  • Overcomplete ICA

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Atomic and Molecular Physics, and Optics
  • Computer Vision and Pattern Recognition

Cite this

Inlier-based ICA with an application to superimposed images. / Meinecke, Frank C.; Harmeling, Stefan; Muller, Klaus.

In: International Journal of Imaging Systems and Technology, Vol. 15, No. 1, 05.10.2005, p. 48-55.

Research output: Contribution to journalArticle

Meinecke, Frank C. ; Harmeling, Stefan ; Muller, Klaus. / Inlier-based ICA with an application to superimposed images. In: International Journal of Imaging Systems and Technology. 2005 ; Vol. 15, No. 1. pp. 48-55.
@article{44f432a396cd4239b8429da5e5921c92,
title = "Inlier-based ICA with an application to superimposed images",
abstract = "This paper proposes a new independent component analysis (ICA) method which is able to unmix overcomplete mixtures of sparce or structured signals like speech, music or images. Furthermore, the method is designed to be robust against outliers, which is a favorable feature for ICA algorithms since most of them are extremely sensitive to outliers. Our approach is based on a simple outlier index. However, instead of robustifying an existing algorithm by some outlier rejection technique we show how this index can be used directly to solve the ICA problem for super-Gaussian sources. The resulting inlier-based ICA (IBICA) is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals).",
keywords = "Blind Source Separation (BSS), Independent Component Analysis (ICA), Outlier robustness, Overcomplete ICA",
author = "Meinecke, {Frank C.} and Stefan Harmeling and Klaus Muller",
year = "2005",
month = "10",
day = "5",
doi = "10.1002/ima.20037",
language = "English",
volume = "15",
pages = "48--55",
journal = "International Journal of Imaging Systems and Technology",
issn = "0899-9457",
publisher = "John Wiley and Sons Inc.",
number = "1",

}

TY - JOUR

T1 - Inlier-based ICA with an application to superimposed images

AU - Meinecke, Frank C.

AU - Harmeling, Stefan

AU - Muller, Klaus

PY - 2005/10/5

Y1 - 2005/10/5

N2 - This paper proposes a new independent component analysis (ICA) method which is able to unmix overcomplete mixtures of sparce or structured signals like speech, music or images. Furthermore, the method is designed to be robust against outliers, which is a favorable feature for ICA algorithms since most of them are extremely sensitive to outliers. Our approach is based on a simple outlier index. However, instead of robustifying an existing algorithm by some outlier rejection technique we show how this index can be used directly to solve the ICA problem for super-Gaussian sources. The resulting inlier-based ICA (IBICA) is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals).

AB - This paper proposes a new independent component analysis (ICA) method which is able to unmix overcomplete mixtures of sparce or structured signals like speech, music or images. Furthermore, the method is designed to be robust against outliers, which is a favorable feature for ICA algorithms since most of them are extremely sensitive to outliers. Our approach is based on a simple outlier index. However, instead of robustifying an existing algorithm by some outlier rejection technique we show how this index can be used directly to solve the ICA problem for super-Gaussian sources. The resulting inlier-based ICA (IBICA) is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals).

KW - Blind Source Separation (BSS)

KW - Independent Component Analysis (ICA)

KW - Outlier robustness

KW - Overcomplete ICA

UR - http://www.scopus.com/inward/record.url?scp=25444462012&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=25444462012&partnerID=8YFLogxK

U2 - 10.1002/ima.20037

DO - 10.1002/ima.20037

M3 - Article

AN - SCOPUS:25444462012

VL - 15

SP - 48

EP - 55

JO - International Journal of Imaging Systems and Technology

JF - International Journal of Imaging Systems and Technology

SN - 0899-9457

IS - 1

ER -