Blind Separation of Nonstationary Sources by Spectral Decorrelation

  • Shahram Hosseini
  • Yannick Deville
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3195)


This paper demonstrates and exploits some interesting frequency-domain properties of nonstationary signals. Considering these properties, two new methods for blind separation of linear instantaneous mixtures of mutually uncorrelated, nonstationary sources are proposed. These methods are based on spectral decorrelation of the sources. The second method is particularly important because it allows the existing time-domain algorithms developed for stationary, temporally correlated sources to be applied to nonstationary, temporally uncorrelated sources just by mapping the mixtures in the frequency domain. Moreover, it sets no constraint on the variance profile, unlike previously reported methods.


Spectral Function Independent Component Analysis Source Separation Blind Source Separation Nonstationary Signal 
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.
    Cardoso, J.-F.: The three easy routes to independent component analysis: contrast and geometry. In: Proc. ICA 2001, San Diego, pp. 1–6 (2001)Google Scholar
  2. 2.
    Matsuoka, K., Ohya, M., Kawamoto, M.: A neural net for blind separation of nonstationary signals. Neural Networks 8(3), 411–419 (1995)CrossRefGoogle Scholar
  3. 3.
    Souloumiac, A.: Blind source detection and separation using second-order nonstationarity. In: Proc. ICASSP, pp. 1912–1915 (1995)Google Scholar
  4. 4.
    Hyvarinen, A.: Blind source separation by nonstationarity of variance: a cumulant based approach. IEEE Trans. on Neural Networks 12(6), 1471–1474 (2001)CrossRefGoogle Scholar
  5. 5.
    Belouchrani, A., Amin, M.G.: Blind source separation based on time-frequency signal representation. IEEE Trans. on Signal Processing 46(11) (November 1998)Google Scholar
  6. 6.
    Abrard, F., Deville, Y., White, P.: From blind source separation to blind source cancellation in the underdetermined case: a new approach based on time-frequency analysis. In: Proc. ICA 2001, San Diego, USA, pp. 734–739 (2001)Google Scholar
  7. 7.
    Deville, Y.: Temporal and time-frequency correlation-based blind source separation methods. In: Proc. ICA 2003, Nara, Japan, pp. 1059–1064 (2003)Google Scholar
  8. 8.
    Albouy, B., Deville, Y.: A time-frequency blind source separation method based on segmented coherence function. In: Proc. IWANN 2003, Mao, Menorca, Spain, vol. 2, pp. 289–296 (2003)Google Scholar
  9. 9.
    Pham, D.-T., Cardoso, J.-F.: Blind separation of independent mixtures of nonstationary sources. IEEE Trans. on Signal Processing 49(9) (2001)Google Scholar
  10. 10.
    Papoulis, A., Pillai, S.U.: Probability, random variables and stochastic processes, 4th edn. McGraw-Hill, New York (2002)Google Scholar
  11. 11.
    Tong, L., Soon, V.: Indeterminacy and identifiability of blind identification. IEEE Trans. Circuits Syst. 38, 499–509 (1991)zbMATHCrossRefGoogle Scholar
  12. 12.
    Molgedey, L., Schuster, H.G.: Separation of a mixture of independent signals using time delayed correlation. Physical Review Letters 72, 3634–3636 (1994)CrossRefGoogle Scholar
  13. 13.
    Belouchrani, A., Abed Meraim, K., Cardoso, J.-F., Moulines, E.: A blind source separation technique based on second order statistics. IEEE Trans. on Signal Processing 45, 434–444 (1997)CrossRefGoogle Scholar
  14. 14.
    Ziehe, A., Muller, K.R.: TDSEP - an efficient algorithm for blind separation using time structure. In: Proceedings of Int. Conf. on Artificial Neural Networks, Skovde, Sweden, pp. 675–680 (1998)Google Scholar
  15. 15.
    Degerine, S., Malki, R.: Second order blind separation of sources based on canonical partial innovations. IEEE Trans. on Signal Processing 48, 629–641 (2000)CrossRefGoogle Scholar
  16. 16.
    Hosseini, S., Jutten, C., Pham, D.-T.: Markovian source separation. IEEE Trans. on Signal Processing 51, 3009–3019 (2003)CrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Shahram Hosseini
    • 1
  • Yannick Deville
    • 1
  1. 1.Laboratoire d’Acoustique, Métrologie, InstrumentationUniversité Paul SabatierToulouse CedexFrance

Personalised recommendations