dc.contributor.author
Tapia, Ernesto
dc.contributor.author
Rojas, Raúl
dc.date.accessioned
2018-06-08T07:32:05Z
dc.date.available
2009-10-13T13:12:03.011Z
dc.identifier.uri
https://refubium.fu-berlin.de/handle/fub188/18168
dc.identifier.uri
http://dx.doi.org/10.17169/refubium-21877
dc.description.abstract
A neural architecture based on linear predictability is used to separate
linear mixtures of signals. The architecture is divided in two parameterers
groups, one modeling the linear mixture of signals and the other computing the
linear predictions of the reconstructed signals. The network weights
correspond to the mixing matrices and coefficients of the linear predictions,
while the values computed by the network units correspond to the predicted and
reconstructed signal values. A quadratic error is iteratively minimized to
approximate the mixing matrix and to maximize the linear predictability.
Experiments with toy and acoustic signals show the feasibility of the
architecture.
en
dc.relation.ispartofseries
urn:nbn:de:kobv:188-fudocsseries000000000021-2
dc.rights.uri
http://www.fu-berlin.de/sites/refubium/rechtliches/Nutzungsbedingungen
dc.subject.ddc
000 Informatik, Informationswissenschaft, allgemeine Werke::000 Informatik, Wissen, Systeme::003 Systeme
dc.title
A neural architecture for blind source separation
refubium.affiliation
Mathematik und Informatik
de
refubium.affiliation.other
Institut für Informatik
refubium.mycore.fudocsId
FUDOCS_document_000000003868
refubium.resourceType.isindependentpub
no
refubium.series.name
Freie Universität Berlin, Fachbereich Mathematik und Informatik
refubium.series.reportNumber
06-4
refubium.mycore.derivateId
FUDOCS_derivate_000000000731
dcterms.accessRights.openaire
open access