Image Number 8 for United States Patent #8131542.
A system capable of separating sound source signals with high precision while improving a convergence rate and convergence precision. A process of updating a current separation matrix W.sub.k to a next separation matrix W.sub.k+1 such that a next value J(W.sub.k+1) of a cost function is closer to a minimum value J(W.sub.0) than a current value J(W.sub.k) is iteratively performed. An update amount .DELTA.W.sub.k of the separation matrix is increased as the current value J(W.sub.k) of the cost function is increased and is decreased as a current gradient .differential.J(Wk)/.differential.W of the cost function is rapid. On the basis of input signals x from a plurality of microphones M.sub.i and an optimal separation matrix W.sub.0, it is possible to separate sound source signals y(=W.sub.0x) with high precision while improving a convergence rate and convergence precision.