This paper addresses blind source separation problem for noisy data based on the concepts of nonlinear innovation and Gaussian moments. An objective function which incorporates Gaussian moments and the nonlinear innovation of original sources is developed. Minimizing this objective function, a noisy blind source separation algorithm is proposed when the noise covariance is known and source signals are nonstationary in the sense that the variance of each is assumed to change smoothly as a function of time. In addition, this method is further extended to the case of noise covariance unknown. Validity and performance of the described approaches are demonstrated by computer simulations. © 2008 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Zhang, H., Guo, C., Shi, Z., & Feng, E. (2008). Nonlinear innovation to noisy blind source separation based on gaussian moments. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5227 LNAI, pp. 709–716). https://doi.org/10.1007/978-3-540-85984-0_85
Mendeley helps you to discover research relevant for your work.