Blind Source Separation via Unsupervised Learning

  • B. Freisleben
  • C. Hagen
  • M. Borschbach


In this paper, a two-layer neural network is presented that organizes itself to perform blind source separation, i.e. it extracts the unknown independent source signals out of their linear mixtures. The convergence behaviour of the network is analyzed, and experimental results of separating historical speeches of four different speakers are presented.


Learning Rule Independent Component Analysis Convergence Behaviour Blind Source Separation Linear Mixture 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    P. Comon. Independent Component Analysis — A New Concept. Signal Processing, 36:287–314, 1994.MATHCrossRefGoogle Scholar
  2. [2]
    P. Comon, C. Jutten, J. Herault. Blind Separation of Sources, Part II: Problem Statements. Signal Processing, 24:11–20, 1991.MATHCrossRefGoogle Scholar
  3. [3]
    N. Delfosse, P. Loubaton. Adaptive Blind Separation of Independent Sources: A Deflation Approach. Signal Processing, 45:59–83, 1995.MATHCrossRefGoogle Scholar
  4. [4]
    J. Karhunen. Neural Approaches to Independent Component Analysis and Source Separation. Proc. of the 4th European Symp. on Artificial Neural Networks, Bruges, Belgium, 1996.Google Scholar
  5. [5]
    J. Karhunen, E. Oja, L. Wang, R. Vigario, J. Joutsensalo. A Class of Neural Networks for Independent Component Analysis. Report A 28, Helsinki University, 1995.Google Scholar

Copyright information

© Springer-Verlag Wien 1998

Authors and Affiliations

  • B. Freisleben
    • 1
  • C. Hagen
    • 2
  • M. Borschbach
    • 1
  1. 1.Department of Electrical Engineering and Computer Science (FB12)University of SiegenSiegenGermany
  2. 2.Department of Computer Science (FB20)University of DarmstadtDarmstadtGermany

Personalised recommendations