Advertisement

Performance Evaluation of Directionally Constrained Filterbank ICA on Blind Source Separation of Noisy Observations

  • Chandra Shekhar Dhir
  • Hyung-Min Park
  • Soo-Young Lee
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4232)

Abstract

Separation performance of directionally constrained filter- bank ICA is evaluated in presence of noise with different spectral properties. Stationarity of mixing channels is exploited to introduce directional constraint on the adaptive subband separation networks of filterbank- based blind source separation approach. Directional constraints on demixing network improves separation of source signals from noisy convolved mixtures, when significant spectral overlap exists between the noise and the convolved mixtures. Observations corrupted with low frequency noises exhibit slight improvement in the separation performance as there is less spectral overlap. Initialization and constraining of subband demixing network in accordance to the spatial location of source signals results in faster convergence and effective permutation correction, irrespective, of the nature of additive noise.

Keywords

Independent Component Analysis Speech Signal Independent Component Analysis Blind Source Separation Separation Performance 
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.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Hyvarinen, A., Karhunen, J., Oja, E.: Independent Component Analysis. John Wiley & Sons, Inc., Chichester (2001)CrossRefGoogle Scholar
  2. 2.
    Lee, T.W.: Independent component analysis - Theory and applications. Kluwer Academic Publisher, Boston (1998)MATHGoogle Scholar
  3. 3.
    Comon, P.: Independent component analysis, a new concept? Signal Processing 36, 287–314 (1994)MATHCrossRefGoogle Scholar
  4. 4.
    Haykins, S. (ed.): Unsupervised Adaptive Filtering. John Wiley & Sons, Inc., Chichester (2000)Google Scholar
  5. 5.
    Park, H.-M., Dhir, C.S., Oh, S.-H., Lee, S.-Y.: A filter bank approach to independent component analysis for convolved mixtures. Neurocomputing (in press)Google Scholar
  6. 6.
    Park, H.-M., Dhir, C.S., Oh, D.-K., Lee, S.-Y.: Filterbank-based BSS with estimated sound direction. In: Proc. IEEE Int. Symp. on Circuits and Systems, pp. 5874–5877 (2005)Google Scholar
  7. 7.
    Dhir, C.S., Park, H.-M., Oh, D.-K., Lee, S.-Y.: Blind source separation of noisy observations using directionally constrained Filterbank ICA. In: Int. conf. on Neural Information Processing, pp. 544–549 (2005)Google Scholar
  8. 8.
    Dhir, C.S.: Performance improvement of filterbank ICA with sound localization. M.S thesis, KAIST, Daejeon (2006)Google Scholar
  9. 9.
    Torkkola, K.: Blind separation of convolved sources based on information maximization. In: Proc. of IEEE Int. workshop on NNSP, pp. 423–432 (1996)Google Scholar
  10. 10.
    Weib, S.: On Adaptive Filtering on Oversampled Subbands. Ph.D. thesis, Signal Processing Division, Univ. Strathclyde, Glasgow (May 1998)Google Scholar
  11. 11.
    Schobben, D., Torkkola, K., Smaragdis, P.: Evaluation of blind signal separation methods. In: Proc. Int. Conf. on ICA and BSS, pp. 261–266 (1999)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Chandra Shekhar Dhir
    • 1
    • 3
  • Hyung-Min Park
    • 4
  • Soo-Young Lee
    • 1
    • 2
    • 3
  1. 1.Department of Biosystems 
  2. 2.Department of Electrical Engineering and Computer Science 
  3. 3.Brain Science Research CenterKorea Advanced Institute of Science and TechnologyDaejeonKorea
  4. 4.Language Technologies Institute and Department of Electrical and Computer Engg.Carnegie Mellon UniversityPittsburghU.S.A

Personalised recommendations