Recently, several techniques have been presented for blind source separation using linear or non-linear mixture models. The problem is to recover the original source signals without knowing apriori information about the mixture model. Accordingly, several statistic and information theory-based objective functions are used in literature to estimate the original signals without providing mixture ...

PURPOSE To investigate the relationship between outflow facility and separation between the inner wall of the aqueous plexus and the juxtacanalicular connective tissue (JCT) during washout in the bovine eye. METHODS Facility was recorded during 3 hours of anterior chamber perfusion at 15 mm Hg in eight pairs of bovine eyes. One eye of each pair was then lowered to 0 mm Hg for 1 hour, whereas ...

Noisy ICA model offers a viable solution for blind speech signal separation under real life scenarios, but it also creates new problem which need to be tackled, namely: estimating of the de-mixing matrix with non-invertible model. In this paper, a new algorithm based on short-time ICA is proposed for accurate estimation of the de-mixing matrix in the presence of noise. Without any assumption or...

Presented here is a generalization of the relative Newton method, recently proposed for quasimaximum likelihood blind source separation. Special structure of the Hessian matrix allows performing block-coordinate Newton descent, which significantly reduces the algorithm computational complexity and boosts its performance. Simulations based on artificial and real data showed that the separation q...