On-line blind separation of non-stationary signals
DOI:
https://doi.org/10.2298/YJOR0501079TKeywords:
blind source separation, decorrelaton, neural networks, extended Kalman filterAbstract
This paper addresses the problem of blind separation of non-stationary signals. We introduce an on-line separating algorithm for estimation of independent source signals using the assumption of non-stationary of sources. As a separating model, we apply a self-organizing neural network with lateral connections, and define a contrast function based on correlation of the network outputs. A separating algorithm for adaptation of the network weights is derived using the state-space model of the network dynamics, and the extended Kalman filter. Simulation results obtained in blind separation of artificial and real-world signals from their artificial mixtures have shown that separating algorithm based on the extended Kalman filter outperforms stochastic gradient based algorithm both in convergence speed and estimation accuracy.References
Amari, S.-I., Cichocki, A., and Yang, H., “Blind signal separation and extraction: Neural and information–theoretic approach”, in: Haykin, S. (ed.), Unsupervised Adaptive Filtering, Vol. I, John Willey & Sons, Ltd, 2000.
Bell, A.J., and Sejnowski, T.J., “An information maximization approach to blind separation and blind deconvolution”, Neural Computation, 7 (6) (1995) 1129-1159.
Cao, X.-R., and Liu, R.-W., “General approach to blind source separation”, IEEE Transactions on Signal Processing, 44 (3) (1996) 562-571.
Cardoso, J.-F., “High-order contrasts for independent component analysis”, Neural Computation, Massachusetts Institute of Technology, 11 (1999) 157-192.
Cichocki, A., and Amari, S.-I., Adaptive Blind Signal and Image Processing: Learning Algorithms and Applications, John Willey & Sons, Ltd, 2002.
Choi, S., Cichocki, A., and Amari, S.-I., “Equivariant nonstationary source separation”, Neural Networks, 15 (2003) 121-131.
Comon, P., “Independent component analysis, a new concept”, Signal Processing, 36 (1994) 287-314.
De Freitas, J.F.G., Niranjan, M., and Gee, A.H., “Hierarchical Bayesian-Kalman models for regularization and ARD in sequential learning”, Technical Report CUED/F-INFENG/TR 307, Cambridge University, 1997.
Haykin, S., (ed.), Kalman Filtering and Neural Networks, John Willey & Sons, Ltd, 2001.
Mansour, A., Kardec Barros, A., and Ohnishi, N., “Blind separation of sources: Methods, assumptions and applications”, IEICE Trans. Fund., E83-A (8) (2000).
Mansour, A., “The blind separation of nonstationary signals by only using the second order statistics”, Proc. of the ISSPA’99, Australia, 1999, 235-238.
Matsuoka, K., Ohya, M., and Kawamoto, M., “A neural net for blind separation of nonstationary signals”, Neural Networks, 8 (3) (1995) 411-419.
Todorović, B., Stanković, M., and Todorović-Zarkula, S., “Structurally adaptive RBF network in non-stationary time series prediction”, Proc. IEEE AS-SPCC, Lake Louise, Alberta, Canada, Oct. 1-4, 2000, 224-229.
Todorovic-Zarkula, S., Todorovic, B., and Moraga, C., “Blind separation of non-stationary signals using extended Kalman filter”, Proc. of 1st BCI, Greece, 2003, 382-392.
Todorovic-Zarkula, S., Todorovic, B., and Stankovic, M., “On-line blind separation of sounds using neural network and extended Kalman filter”, Proc. of the SYMOPIS’2003, 2003, 49-52.
Downloads
Published
Issue
Section
License
Copyright (c) 2005 YUJOR
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
