Analysis of ICA techniques in terms of Failure percentage and Average CPU Time for Real World BSS Task

Analysis of ICA techniques in terms of Failure percentage and Average CPU Time for Real World BSS Task

  IJETT-book-cover  International Journal of Engineering Trends and Technology (IJETT)          
  
© 2015 by IJETT Journal
Volume-30 Number-6
Year of Publication : 2015
Authors : Naveen Dubey, Rajesh Mehra
DOI :  10.14445/22315381/IJETT-V30P252

Citation 

Naveen Dubey, Rajesh Mehra"Analysis of ICA techniques in terms of Failure percentage and Average CPU Time for Real World BSS Task", International Journal of Engineering Trends and Technology (IJETT), V30(6),276-280 December 2015. ISSN:2231-5381. www.ijettjournal.org. published by seventh sense research group

Abstract
Taking assumption the hidden sources are statistically independent, Independent component analysis technique separates these sources from a linear mixture of audio signals, communication signals generated by equally spaced independent audio sources. Since, in Audio applications source exhibit non - dependence. Mutual information minimization corresponds to minimization of entropy, that ensures quality of separation and exploits non-Gaussianity, noncircularity and sample dependence simultaneously. In this paper these properties are exploited with the help of Cramer-Rao lower bound, modified convex divergence based ICA, Fast ICA and JADE. The performance of these techniques are examined with the help of a number of example and a comparative analysis presented in term of failure percentage and average CPU time taken for execution.

 References

[1] Vincent Emmanuel, Bertin Nancy, G.Remi, Bimbot Frederic, ?From Blind to Guided Audio Source Separation, IEEE Signal Proc. Magazine, pp. 107-115, May 2014
[2] J. F. Cardoso and A Souloumiac, ?Blind beamforming for non-Gaussian signals, Proc. Inst. Electri. Eng. F-Radar Signal processing, vol. 140, No. 6, pp. 362-370, Nov 1998.
[3] Jen-Tzung Chien, Hsin-Lung Hsieh, ?Convex Divergence ICA for Blind Source Separation, IEEE Transc. On Audio, Speech and Language Prosc., Vol. 20, No. 01, pp. 302-312.
[4] E. Bingham and A. Hyvarinen, ?A fast fixed point algorithm for independent component analysis of complex valued signals Int, J. Neural syst. Vol. 10, No. 01, pp. 1-8, Feb 2000.
[5] Ritesh K Yadav, R. Mehra, N. Dubey, ?Blind Audio Source Separation using Weight Initialized Independent Component Analysis, in proc. Of IEEE conference on Next Generation Computing, UPES, Dehradun, India, Sept. 2015.
[6] T. Adali, T.Kim, and V. Calhoun, ?Independent component analysis by complex nonlinearities, In proc. Of Int. conf. Acoust. Signal processing, 2004, Vol. 5, pp. 525-528.
[7] H. Li and T. Adali, ?A class of complex ICA algorithm based on the kurtosis cost function, IEEE Transc. Neural Netw. Vol. 19, No. 3, pp. 408-420, Mar. 2008
[8] X.-L. Li, T. Adali, ?Complex Independent Component Analysis vy entropy bound minimization IEEE Transc. Circuit Syst. , Vol. 57, No. 7, pp. 1417-1430, Jul 2010.
[9] B. Loesch and B. Yang, ?Cramer-Rao Bound for circular complex independent component analysis, IEEE Transc. Signal Proc., Vol. 61, No. 2, pp. 365-379, Jan 2013.
[10] X-L Li, T. Adali, ?Blind separation of noncircular correlated source using Gaussian entropy rate, IEEE Transc. Signal Proc., Vol. 59, No. 6, pp. 2969-2975, Jun 2011.
[11] X-L Li and X D Zang, ?Non orthogonal joint diagonalization free of degenerate solution, IEEE Transc. Signal Processing, Vol. 55, No. 10, pp. 1803-1814, May 2007.
[12] M. Anderson, X-L Li, P Rodriguez, T. Adali, ? An effective coupling method for matrix optimization and its application to ICA problem, in porc. Of IEEE int. conf. Acous. Sig, Processing. March. 2012.
[13] T. M. Cover and J. A Thomas, ?Elements of information theory, publisher John Wiley and sons, Hoboken pp; 640, ISBN 0471748811, 1991.
[14] C. E. Shannon, ?A mathematical theory of communication, Bell Syst.Tech. J., vol. 27, pp. 379–423, 1948.
[15] D. Xu, J. C. Principe, J. Fisher, III, and H.-C. Wu, ?A novel measurefor independent component analysis (ICA), in Proc. Int. Conf. Acoust.,Speech, Signal Process., 1998, vol. 2, pp. 1161–1164.
[16] S. Amari, Differential-Geometrical Methods in Statistics. New York: Springer-Verlag, 1985.
[17] J. Zhang, ?Divergence function, duality, and convex analysis, Neural Comput., vol. 16, pp. 159–195, 2004.
[18] Csiszár and P. C. Shields, ?Information theory and statistics: A tutorial, Foundations Trends Commun. Inf. Theory, vol. 1, no. 4, pp.417–528, 2004.
[19] J. Lin, ?Divergence measures based on the Shannon entropy, IEEETrans. Inf. Theory, vol. 37, no. 1, pp. 145– 151, 1991.
[20] Naveen Dubey, Rajesh Mehra, ?Unsupervised Learning Based Modified C-ICA for Audio Source Separation in Blind Scenario, Accepted in Int. J. Information Technology and Computer Sc.
[21] N. Dubey, R. Mehra, ?Blind Audio Source Separation (BASS): An Unsupervised Approach, Edit-2015, Int. J. of Electrical and Electronics Eng. Vol.2 Spl issue. 01 , pp. 29- 33, 2015
[22] Naveen Dubey, Rajesh Mehra, ?Blind Audio Source Separation in Time Domain Using ICA Decomposition, IJCA, Vol 132, No. 6, pp.48-56, 2015
[23] Shaminder Kaur , Rajesh Mehra, ?FPGA Implimentation of OFDM Transceiver using FFT Algorithms, IJEST, Vol.04, No. 04, pp. 1532-1537, April 2012

Keywords
Independent Component Analysis, Blind Source Separation, Convex Divergence, Independence, non-Gaussianity.