β-Divergence Two-Dimensional Nonnegative Matrix Factorization with Sparseness Constraints for Biomedical Signal Separation

Authors

  • A. M. Darsono Machine Learning & Signal Processing Research Group (MLSP), Faculty of Electronic and Computer Engineering, Universiti Teknikal Malaysia Melaka Durian Tunggal, Melaka, Malaysia.
  • C. C. Toh Machine Learning & Signal Processing Research Group (MLSP), Faculty of Electronic and Computer Engineering, Universiti Teknikal Malaysia Melaka Durian Tunggal, Melaka, Malaysia.
  • S. Saat Machine Learning & Signal Processing Research Group (MLSP), Faculty of Electronic and Computer Engineering, Universiti Teknikal Malaysia Melaka Durian Tunggal, Melaka, Malaysia.
  • N. A. Manap Machine Learning & Signal Processing Research Group (MLSP), Faculty of Electronic and Computer Engineering, Universiti Teknikal Malaysia Melaka Durian Tunggal, Melaka, Malaysia.
  • M. M. Ibrahim Machine Learning & Signal Processing Research Group (MLSP), Faculty of Electronic and Computer Engineering, Universiti Teknikal Malaysia Melaka Durian Tunggal, Melaka, Malaysia.
  • M. I. Ahmad School of Computer and Communication Engineering, Universiti Malaysia Perlis, Perlis, Malaysia.

Keywords:

Nonnegative Matrix Factorization, Sparseness Constraints, ?-Divergence, Multiplicative Rules,

Abstract

A novel of β-Divergence for nonnegative matrix factorization two-dimensional (NMF2D) with sparseness constraints is proposed in this paper. This research focuses on biomedical signal separation, which denotes a separation on the mixture of heart sound and lung sound. Initially, a mixture of heart sound and lung sound has been decomposed into an independent signal, which is an estimated heart sound signal and estimated lung sound signal. The spectrum of independent signal is modelled based on 2 dimensions, which are the temporal code and the spectral basis by using β-Divergence NMF2D algorithm with sparseness constraints. The algorithm has been updated multiplicative and iteratively via multiplicative update rules (MU rules). β-Divergence with sparseness constraints allows minimization on the vagueness of source model to be completed and oneness has been applied to it. Then, estimation of each separated audio has been analyzed via comparison with the original heart sound and lung sound signal in term of Signal-to-Distortion Ratio (SDR).

References

P. Paatero and U. Tapper. Positive matrix factorization: A nonnegative factor model with optimal utilization of error estimates of data values. Environmetrics, 5:111–126, 1994.

D. D. Lee and H. S. Seung. Learning the parts of objects by nonnegative matrix factorization. Nature, 401(6755):788–791, 1999.

D. FitzGerald, “Automatic drum transcription and source separation,” Ph.D. thesis, Dublin Institute of Technology, Dublin, Ireland, 2004.

S. Xie, Z. Yang, and Y. Fu, “Nonnegative matrix factorization applied to nonlinear speech and image cryptosystems,” IEEE Trans. on Circuits and Systems I, vol. 55, no. 8, pp. 2356-2367, Sep 2008.

I. Biciu, N. Nikolaidis, and I. Pitas, “Nonnegative matrix factorization in polynomial feature space”, IEEE Trans. On Neural Network, vol. 19, pp. 1090-1100, 2007.

C. S. Lin and E. Hasting, “Blind Source Separation of Heart and Lung Sounds Based on Nonnegative Matrix Factorization,” Intelligent Signal Processing and Communications Systems (ISPACS), 2013.

H. Pasterkamp, S. S. Kraman, and G. R. Wodicka, “Respiratory sounds: Advances beyond the stethoscope,” Amer. J. Respir. Crit. Care Med., vol. 156, pp. 974-987, 1997.

M. Morup and M. N. Schmidt, “Sparse nonnegative matrix factor 2-D deconvolution,” Techical Report, Technical University of Denmark, Copenhagen, Denmark, 2006.

D. J. Field, “Relations between the statistics of natural images and the response properties of cortical cells,” Journal of the Optical Society of America, 4:2379–2394, 1987.

M. Morup, M. N. Schmidt, “Sparse nonnegative matrix factor 2-D deconvolution, “ Technical Report, Technical University of Denmark, Copenhagen, Denmark, 2006.

S. K. Tjoa and K. J. Ray Liu, “Multiplicative Update Rules for Nonnegative Matrix Factorization with Co-Occurrence Constraints,” IEEE International Conference on Acoustics, Speech and Signal Processing, 2010.

L. Sui, X. Zhang, J. Huang, G. Zhao and Y. Yang, “Speech Enhancement Based on Sparse Nonnegative Matrix Factorization with Priors,” Systems and Informatics (ICSAI), 2012.

A.M. Darsono, N. Z. Haron, A. S. Jaafar and M. I. Ahmad, “β-Divergence Two-Dimensional Sparse Nonnegative Matrix Factorization for Audio Source Separation,” Wireless Sensor (ICWISE), 2013.

A. M. Darsono, N.Z. Haron, Shakir Saat, M.M. Ibrahim, and N.A. Manap, “Blind Audio Source Separation with Sparse Nonnegati Matrix Factorization”, Research Journal of Applied Sciences, Engineering and Tchnology, vol. 7 No 23, 2014.

A. Ozerov, and C. Févotte, “Multichannel Nonnegative Matrix Factorization in Convolutive Mixtures for Audio Source Separation,” IEEE Transactions On Audio, Speech, And Language Processing, Vol. 18, No. 3, March 2010.

S. R. Chintakindi, O. V. S. R. Varaprasad and D. V. S. S. Siva Sarma, “ Improved Hanning window based interpolated FFT for power harmonic analysis,” TENCON 2015, IEEE Region 10 Conference, 2015.

C. F´evotte, R. Gribonval and E. Vincent, “BSS EVAL Toolbox User Guide”, IRISA Technical Report 1706, Rennes, France, April 2005. http://www.irisa.fr/metiss/bss eval/.

Downloads

Published

2017-06-01

How to Cite

Darsono, A. M., Toh, C. C., Saat, S., Manap, N. A., Ibrahim, M. M., & Ahmad, M. I. (2017). β-Divergence Two-Dimensional Nonnegative Matrix Factorization with Sparseness Constraints for Biomedical Signal Separation. Journal of Telecommunication, Electronic and Computer Engineering (JTEC), 9(2-6), 7–10. Retrieved from https://jtec.utem.edu.my/jtec/article/view/2424

Most read articles by the same author(s)

1 2 > >>