Proceedings ArticleDOI
Multichannel blind deconvolution and equalization using the natural gradient
Shun-ichi Amari,Scott C. Douglas,Andrzej Cichocki,H.H. Yang +3 more
- pp 101-104
TLDR
It is proved that the doubly-infinite multichannel equalizer based on the maximum entropy cost function with natural gradient possesses the so-called "equivariance property" such that its asymptotic performance depends on the normalized stochastic distribution of the source signals and not on the characteristics of the unknown channel.Abstract:
Multichannel deconvolution and equalization is an important task for numerous applications in communications, signal processing, and control. We extend the efficient natural gradient search method of Amari, Cichocki and Yang (see Advances in Neural Information Processing Systems, p.752-63, 1995) to derive a set of on-line algorithms for combined multichannel blind source separation and time-domain deconvolution/equalization of additive, convolved signal mixtures. We prove that the doubly-infinite multichannel equalizer based on the maximum entropy cost function with natural gradient possesses the so-called "equivariance property" such that its asymptotic performance depends on the normalized stochastic distribution of the source signals and not on the characteristics of the unknown channel. Simulations indicate the ability of the algorithm to perform efficient simultaneous multichannel signal deconvolution and source separation.read more
Citations
More filters
Journal ArticleDOI
Natural gradient works efficiently in learning
TL;DR: In this paper, the authors used information geometry to calculate the natural gradients in the parameter space of perceptrons, the space of matrices (for blind source separation), and the spaces of linear dynamical systems for blind source deconvolution, and proved that Fisher efficient online learning has asymptotically the same performance as the optimal batch estimation of parameters.
Independent Component Analysis.
TL;DR: The standardization of the IC model is talked about, and on the basis of n independent copies of x, the aim is to find an estimate of an unmixing matrix Γ such that Γx has independent components.
Book
Adaptive blind signal and image processing
Andrzej Cichocki,Shun-ichi Amari +1 more
TL;DR: Find the secret to improve the quality of life by reading this adaptive blind signal and image processing and make the words as your good value to your life.
Journal ArticleDOI
Convolutive blind separation of non-stationary sources
Lucas C. Parra,Clay D. Spence +1 more
TL;DR: This work tackles the problem of source separation by explicitly exploiting the nonstationarity of the acoustic sources, and finds an FIR backward model, which generates well separated model sources.
Journal ArticleDOI
A robust and precise method for solving the permutation problem of frequency-domain blind source separation
TL;DR: By utilizing the harmonics of signals, the new method is robust even for low frequencies where DOA estimation is inaccurate, and provides an almost perfect solution to the permutation problem for a case where two sources were mixed in a room whose reverberation time was 300 ms.
References
More filters
Journal ArticleDOI
An information-maximization approach to blind separation and blind deconvolution
TL;DR: It is suggested that information maximization provides a unifying framework for problems in "blind" signal processing and dependencies of information transfer on time delays are derived.
Book
Theory and Practice of Recursive Identification
Lennart Ljung,Torsten Söderström +1 more
TL;DR: Methods of recursive identification deal with the problem of building mathematical models of signals and systems on-line, at the same time as data is being collected.
Journal ArticleDOI
Natural gradient works efficiently in learning
TL;DR: In this paper, the authors used information geometry to calculate the natural gradients in the parameter space of perceptrons, the space of matrices (for blind source separation), and the spaces of linear dynamical systems for blind source deconvolution, and proved that Fisher efficient online learning has asymptotically the same performance as the optimal batch estimation of parameters.
Proceedings Article
A New Learning Algorithm for Blind Signal Separation
TL;DR: A new on-line learning algorithm which minimizes a statistical dependency among outputs is derived for blind separation of mixed signals and has an equivariant property and is easily implemented on a neural network like model.
Journal ArticleDOI
Equivariant adaptive source separation
TL;DR: A class of adaptive algorithms for source separation that implements an adaptive version of equivariant estimation and is henceforth called EASI, which yields algorithms with a simple structure for both real and complex mixtures.