## Independent Component Analysis and Blind Signal Separation: Fifth International Conference, ICA 2004, Granada, Spain, September 22-24, 2004, ProceedingsThis book constitutes the refereed proceedings of the 5th International Conference on Independent Component Analysis and Blind Source Separation, ICA 2004, held in Granada, Spain, in September 2004. The 156 revised papers presented were carefully reviewed and selected from 203 submissions. The papers are organized in topical sections on theory and foundations, linear models, covolutive models, nonlinear models, speech processing applications, image processing applications, biomedical applications, and other applications. |

### What people are saying - Write a review

We haven't found any reviews in the usual places.

### Contents

Theory and Fundamentals | 1 |

Exploiting Spatiotemporal Information | 18 |

New EigensystemBased Method for Blind Source Separation | 33 |

Copyright | |

84 other sections not shown

### Common terms and phrases

applied approach approximation assume beamforming Berlin Heidelberg 2004 blind deconvolution blind separation Blind Signal Separation blind source separation C.G. Puntonet Cardoso coefficients column constraint convergence convolutive mixtures correlation cost function covariance defined denoising denotes density distribution eigenvalues entropy equation estimated sources extraction FastICA filter frequency domain Gaussian gradient Hyvarinen ICA algorithms ICA model IEEE IEEE Trans images Independent Component Analysis Infomax input inverse iteration joint diagonalization kernel kurtosis LNCS microphones minimization mixing matrix mutual information Neural Networks noise nonlinear number of sources obtained optimization order statistics orthogonal orthogonal matrix output parameters performance permutation Prieto Eds Proc proposed method random samples Section sensors Signal Processing simulations solution solve source signals sparse sparsity spatial spectral speech signals Springer-Verlag Berlin Heidelberg statistically independent subspace technique Theorem tion transform update values variables variance vector watermark whitening zero