An Information-Maximization Approach to Blind Separation and Blind Deconvolution

Neural Computation - Tập 7 Số 6 - Trang 1129-1159 - 1995
Anthony J. Bell1, Terrence J. Sejnowski1
1Howard Hughes Medical Institute, Computational Neurobiology Laboratory, The Salk Institute, 10010 N. Torrey Pines Road, La Jolla, CA 92037 USA and Department of Biology, University of California, San Diego, La Jolla, CA 92093 USA

Tóm tắt

We derive a new self-organizing learning algorithm that maximizes the information transferred in a network of nonlinear units. The algorithm does not assume any knowledge of the input distributions, and is defined here for the zero-noise limit. Under these conditions, information maximization has extra properties not found in the linear case (Linsker 1989). The nonlinearities in the transfer function are able to pick up higher-order moments of the input distributions and perform something akin to true redundancy reduction between units in the output representation. This enables the network to separate statistically independent components in the inputs: a higher-order generalization of principal components analysis. We apply the network to the source separation (or cocktail party) problem, successfully separating unknown mixtures of up to 10 speakers. We also show that a variant on the network architecture is able to perform blind deconvolution (cancellation of unknown echoes and reverberation in a speech signal). Finally, we derive dependencies of information transfer on time delays. We suggest that information maximization provides a unifying framework for problems in "blind" signal processing.

Từ khóa


Tài liệu tham khảo

10.1088/0954-898X/3/2/009

10.1162/neco.1993.5.1.45

10.1162/neco.1989.1.3.295

10.1162/neco.1992.4.2.141

10.1038/355161a0

10.1016/S0893-6080(05)80090-5

10.1109/82.365346

10.1016/0165-1684(94)90029-9

10.1016/0165-1684(91)90080-3

10.1162/neco.1994.6.4.559

10.1073/pnas.88.15.6462

10.1016/0165-1684(91)90079-X

10.1016/0893-6080(94)90060-4

10.1515/znc-1981-9-1040

Li S., 1994, I E E E J. Oceanic Eng., 20, 73

10.1162/neco.1992.4.5.691

10.1103/PhysRevLett.72.3634

10.1088/0954-898X/5/4/008

10.1088/0954-898X/6/1/004

10.1162/neco.1992.4.6.863

10.1103/PhysRevA.46.2131

10.1016/0165-1684(91)90081-S

10.1007/978-1-4613-1639-8_3

10.1109/78.301850

Thông tin
Thông tin xuất bản

Neural Computation

Tập 7 Số 6

1129-1159

Thông tin tác giả