New paper on extraction of a single source from multichannel data

[Michael Zibulevsky]

Announcing a paper ...

Title:   Extraction of a single source from  multichannel data
         using sparse decomposition

Authors: M. Zibulevsky and Y.Y. Zeevi 

ABSTRACT:

It was discovered recently that use of sparse decompositions in signal
dictionaries improves dramatically quality of blind source
separation. In this work we exploit sparse decomposition of a single
source in order to extract it from the multidimensional sensor data, when
in addition a rough template of the source is known. This leads to a
convex  optimization problem, which is solved by a Newton-type
method. Complete and overcomplete  dictionaries are considered.
Simulations with synthetic evoked responses mixed into natural 122-channel
MEG data  show significant improvement in  accuracy of signal restoration.

URL of gzipped ps file:

http://ie.technion.ac.il/~mcib/extr_onesrc2.ps.gz

Contact:  mzib at ee.technion.ac.il