The Community for Technology Leaders
Green Image
Issue No. 05 - May (2010 vol. 32)
ISSN: 0162-8828
pp: 875-888
Fa-Yu Wang , National Tsing Hua University, Hsinchu
Chong-Yung Chi , National Tsing Hua University, Hsinchu
Tsung-Han Chan , National Tsing Hua University, Hsinchu
Yue Wang , Virginia Polytechnic Institute and State University, Arlington
Although significant efforts have been made in developing nonnegative blind source separation techniques, accurate separation of positive yet dependent sources remains a challenging task. In this paper, a joint correlation function of multiple signals is proposed to reveal and confirm that the observations after nonnegative mixing would have higher joint correlation than the original unknown sources. Accordingly, a new nonnegative least-correlated component analysis (n{\rm LCA}) method is proposed to design the unmixing matrix by minimizing the joint correlation function among the estimated nonnegative sources. In addition to a closed-form solution for unmixing two mixtures of two sources, the general algorithm of n{\rm LCA} for the multisource case is developed based on an iterative volume maximization (IVM) principle and linear programming. The source identifiability and required conditions are discussed and proven. The proposed n{\rm LCA} algorithm, denoted by n{\rm LCA\hbox{-}IVM}, is evaluated with both simulation data and real biomedical data to demonstrate its superior performance over several existing benchmark methods.
Nonnegative blind source separation, nonnegative least-correlated component analysis, dependent sources, joint correlation function of multiple signals, iterative volume maximization.

C. Chi, T. Chan, Y. Wang and F. Wang, "Nonnegative Least-Correlated Component Analysis for Separation of Dependent Sources by Volume Maximization," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 32, no. , pp. 875-888, 2009.
92 ms
(Ver 3.3 (11022016))