Blind deconvolution of timely correlated sources by gradient descent search

• Autorzy: Kasprzak W.
• Języki publikacji: EN

Abstrakty

EN

In multichannel blind decon volution (MB D) the goal is to calculate possibly scaled and delayed estimates of source signals from their convoluted mixtures, using approximate knowledge of the source characteristics oniy. Nearly all of the solutions to MBD proposed so far require from source signals to be pairwise statistically independent and to be timely not correlated. In practice, this can only be satisfied by specific synthetic signals. In this paper we describe how to modify gradient-based iterative algorithms in order to perform the MBD task on timely correlated sources. Implementation issues are discussed and specific tests on synthetic and real 2-D images are documented.

Słowa kluczowe

EN

blind source deconvolution; gradient descent search; speech and image processing; unsupervised learning

PL

przetwarzanie mowy; przetwarzanie obrazu

• Wydawca: Instytut Telekomunikacji i Informatyki Uniwersytetu Technologiczno-Przyrodniczego w Bydgoszczy
• Czasopismo: Image Processing & Communications
• Rocznik: 2003
• Tom: Vol. 9, no 1-2
• Strony: 33--52
• Opis fizyczny: Bibliogr. 15 poz., fot., rys.

Twórcy

autor

Kasprzak W.

• Warsaw University of Technology Institute of Control and Computation Engineering ul. Nowowiejska 15-19, PL - 00-665 Warsaw POLAND,

Bibliografia

• [1] Amari S.. Douglas S.C.. Cichocki A., Yang H.Y.: Novel on-line adaptive learning algorithms for blind deconvolution using the natural gradient approach. IEEE Signal Proc. Workshop on Signal Processing Advances in Wireless Communications. April 1997. Paris, 107 112.
• [2] Cardoso J.F., Laheld B.: Equivariant adaptive source separation, IEEE Trans, on Signal Processing, vol. 44(1996), December 1996, 3017-3030.
• [3] Cichocki A., Amari S., Adaptive Blind Signal and Image Processing , John Wiley. Chichester. UK. 2002.
• [4] Cichocki A., Unbehauen R., Rummert E.: Robust learning algorithm for blind separation of signals, Electronics Letters, vol. 30(1994), August 1994. 1386-1387.
• [5] Duda R.O., Hart P.E. : Pattern classification and scene analysis, John Wiley & Sons, New York. 1973.
• [6] Hua Y.: Fast maximum-likelihood for blind identification of multiple FIR channels, IEEE Trans. Signal Processing, vol. 44 (1996), 661-672.
• [7] Hyvarinen A., Oja E.: Simple neuron models for independent component analysis. Int. J. Neural Systems, vol. 7, 1996. 671687.
• [8] Karhunen J., Oja E., Wang L., Vigario R., Joutsensalo J.: A class of neural networks for independent component analysis, IEEE Trans, on Neural Networks, vol. 8(1997). May 1997, 486-504.
• [9] Kasprzak W.: ICA algorithms for image restoration. Chapter 3 of the monograph work: Kasprzak W.: Adaptive computation methods in digital image sequence analysis. Prace Naukowe - Elektronika, Nr. 127 (2000), Warsaw University of Techology Press, Warszawa, 170 pages.
• [10] Kasprzak W., Cichocki A.: Hidden Image Separation From Incomplete Image Mixtures by Independent Component Analysis, 13th Int. Conf. on Pattern Recognition, ICPR’96, Proceedings. IEEE Computer Society Press, Los Alamitos CA, 1996, vol.II, 394-398.
• [11] Kasprzak W., Cichocki A., Amari S.: Blind Source Separation with Convolutive Noise Cancellation, Neural Computing and Applications, Springer Ltd., London, vol. 6 (1997), 127-141.
• [12] Niemann H.: Pattern Analysis and Understanding, Springer. Berlin, 1990.
• [13] Pajunen P.: An Algorithm for Binary Blind Source Separation. Helsinki University of Technology, Lab. of Computer and Information Science, Report A36, July 1996.
• [14] Sabaa L: Multichannel deconvolution and separation of statistically independent signals for unknown dynamic systems. Ph.D. dissertation. Warsaw University of Technology, Dep. of Electrical Engineering, Warsaw, 1998.
• [15] Widrow B., Stearns S.D.: Adaptive signal processing, Prentice Hall. Englewood Cliffs. NJ. 1985.