Fast fixed-point independent vector analysis algorithms for convolutive blind source separation

摘要

A new type of independent component analysis (ICA) model showed excellence in tackling the blind source separation problem in the frequency domain. The new model, called independent vector analysis, is an extension of ICA for (independent) multivariate sources where the sources are mixed component-wise. In this work we examine available contrasts for the new formulation that can solve the frequency-domain blind source separation problem. Also, we introduce a quadratic Taylor polynomial in the notations of complex variables which is very useful in directly applying Newton's method to a contrast function of complex-valued variables. The use of the form makes the derivation of a Newton update rule simple and clear. Fast fixed-point blind source separation algorithms are derived and the performance is shown by experimental results.