A blind source separation algorithm is used to estimate the mixing operator from electromagnetic emission signals through independent component analysis (ICA) technique. The mixin...
We propose a general framework for performing independent component analysis (ICA) which relies on ensemble learning and linear response theory known from statistical physics. We ...
The potential of hand shape and hand texture-based biometry is investigated and algorithms are developed. Feature extraction stage is preceded by meticulous registration of the de...
Abstract--In this paper, we study simultaneous matrix diagonalization-based techniques for the estimation of the mixing matrix in underdetermined independent component analysis (IC...
—Beyond conventional linear and kernel-based feature extraction, we present a more generalized formulation for feature extraction in this paper. Two representative algorithms usi...