Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
Springer Nature Link
Log in

Extension of Sparse, Adaptive Signal Decompositions to Semi-blind Audio Source Separation

  • Conference paper
Independent Component Analysis and Signal Separation (ICA 2009)

Part of the book series: Lecture Notes in Computer Science ((LNTCS,volume 5441))

Abstract

We apply sparse, fast and flexible adaptive lapped orthogonal transforms to underdetermined audio source separation using the time-frequency masking framework. This normally requires the sources to overlap as little as possible in the time-frequency plane.

In this work, we apply our adaptive transform schemes to the semi-blind case, in which the mixing system is already known, but the sources are unknown. By assuming that exactly two sources are active at each time-frequency index, we determine both the adaptive transforms and the estimated source coefficients using ℓ1 norm minimisation. We show average performance of 12–13 dB SDR on speech and music mixtures, and show that the adaptive transform scheme offers improvements in the order of several tenths of a dB over transforms with constant block length. Comparison with previously studied upper bounds suggests that the potential for future improvements is significant.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Similar content being viewed by others

References

  1. Bofill, P., Zibulevsky, M.: Underdetermined blind source separation using sparse representations. Signal Process. 81(11), 2353–2362 (2001)

    Article  MATH  Google Scholar 

  2. Bofill, P.: Identifying single source data for mixing matrix estimation in instantaneous blind source separation. In: Koutník, J., Kůrková, V., Neruda, R. (eds.) ICANN 2008, Part I. LNCS, vol. 5163, pp. 759–767. Springer, Heidelberg (2008)

    Chapter  Google Scholar 

  3. Huang, Y., Pollak, I., Bouman, C.A., Do, M.N.: Best basis search in lapped dictionaries. IEEE Trans. Signal Process. 54(2), 651–664 (2006)

    Article  Google Scholar 

  4. Gribonval, R.: Piecewise linear source separation. In: Proc. SPIE (Wavelets X), vol. 5207, pp. 297–310 (2003)

    Google Scholar 

  5. ISO: Information technology—Coding of audio-visual objects—Part 3: Audio (ISO/IEC 14496-3:2005). ISO, Geneva, Switzerland (2005)

    Google Scholar 

  6. Mallat, S.: A Wavelet Tour of Signal Processing, 2nd edn. Academic Press, London (1999)

    MATH  Google Scholar 

  7. Nesbit, A., Plumbley, M.D., Vincent, E.: Oracle evaluation of flexible adaptive transforms for underdetermined audio source separation. In: Proc. ICArn 2008, pp. 17–20 (2008)

    Google Scholar 

  8. Nesbit, A., Vincent, E., Plumbley, M.D.: Benchmarking flexible adaptive time-frequency transforms for underdetermined audio source separation. In: ICASSP 2009 (submitted, 2009)

    Google Scholar 

  9. Vincent, E., Gribonval, R.: Blind criterion and oracle bound for instantaneous audio source separation using adaptive time-frequency representations. In: Proc. WASPAA 2007, pp. 110–113 (2007)

    Google Scholar 

  10. Vincent, E., Gribonval, R., Plumbley, M.D.: Oracle estimators for the benchmarking of source separation algorithms. Signal Process. 87(8), 1933–1950 (2007)

    Article  MATH  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. School of Electronic Engineering and Computer Science, Queen Mary University of London, Mile End Road, London, E1 4NS, United Kingdom

    Andrew Nesbit & Mark D. Plumbley

  2. METISS Group, IRISA-INRIA, Campus de Beaulieu, 35042, Rennes Cedex, France

    Emmanuel Vincent

Authors
  1. Andrew Nesbit
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Emmanuel Vincent
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Mark D. Plumbley
    View author publications

    You can also search for this author in PubMed Google Scholar

Editor information

Editors and Affiliations

  1. Department of Computer Science and Electrical Engineering, ITE 324, University of Maryland, Baltimore County, 1000 Hilltop Circle, MD 21250, Baltimore, USA

    Tülay Adali

  2. Domaine Universitaire, GIPSA-lab, BP 46, 38402, Saint Martin d’Hères Cedex, France

    Christian Jutten

  3. Departamento de Microonda e Óptica (DMO), FEEC / Unicamp, Avenida Albert Einstein 400, 13083-852, Campinas, Sao Paulo, Brazil

    João Marcos Travassos Romano

  4. Centro Tecnológico, Curso de Engenharia Elétrica, Universidade Federal do Maranhão, Avenida dos Portugueses, s/n, Bacanga, 65080-040, São Luís, MA, Brazil

    Allan Kardec Barros

Rights and permissions

Reprints and permissions

Copyright information

© 2009 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Nesbit, A., Vincent, E., Plumbley, M.D. (2009). Extension of Sparse, Adaptive Signal Decompositions to Semi-blind Audio Source Separation. In: Adali, T., Jutten, C., Romano, J.M.T., Barros, A.K. (eds) Independent Component Analysis and Signal Separation. ICA 2009. Lecture Notes in Computer Science, vol 5441. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-00599-2_76

Download citation

  • DOI: https://doi.org/10.1007/978-3-642-00599-2_76

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-642-00598-5

  • Online ISBN: 978-3-642-00599-2

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics

Search

Navigation