论文信息 - A NMF algorithm for blind separation of uncorrelated signals

A NMF algorithm for blind separation of uncorrelated signals

Most of the proposed algorithms for blind sources separation are not able to extract the source signals when the unknown sources are not mutually statistically independent. In this paper, the blind separation problem for uncorrelated signals is explored. A novel algorithm is proposed based on the nonnegative matrix factorization methods with the least correlated component constraints. The algorithm relaxes the source independence assumption and has low-complexity algebraic computations, and thus is computationally efficient. Simulation results show that the proposed algorithm can provide an efficient separation performance for the uncorrelated source signals.

Ye Zhang | Yong Fang

[1] P. Paatero,et al. Positive matrix factorization: A non-negative factor model with optimal utilization of error estimates of data values† , 1994 .

[2] H. Sebastian Seung,et al. Learning the parts of objects by non-negative matrix factorization , 1999, Nature.

[3] H. Sebastian Seung,et al. Algorithms for Non-negative Matrix Factorization , 2000, NIPS.

[4] Andrzej Cichocki,et al. Adaptive blind signal and image processing , 2002 .

[5] Mark D. Plumbley. Algorithms for nonnegative independent component analysis , 2003, IEEE Trans. Neural Networks.

[6] E. Oja,et al. Independent Component Analysis , 2001 .

[7] Erkki Oja,et al. Blind Separation of Positive Sources by Globally Convergent Gradient Search , 2004, Neural Computation.

[8] Patrik O. Hoyer,et al. Non-negative Matrix Factorization with Sparseness Constraints , 2004, J. Mach. Learn. Res..

[9] Dietrich Lehmann,et al. Nonsmooth nonnegative matrix factorization (nsNMF) , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10] Andrzej Cichocki,et al. New Algorithms for Non-Negative Matrix Factorization in Applications to Blind Source Separation , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.