A generalized ICA algorithm

In this paper, we consider the source separation problem through a block al gorithm based on the maximization of contrast functions. We propose a new c ontrast with parameterized cross-cumulants. It allows us to put three class ical contrasts in a common framework. Following the same spirit of the ICA algorithm, we derive the analytical solution for the case of two sources. F inally a computer simulation is performed to illustrate the behavior of a J acobi-like algorithm for the maximization of the new contrast.