This HTML5 document contains 51 embedded RDF statements represented using HTML+Microdata notation.

The embedded RDF content will be recognized by any processor of HTML5 Microdata.

Namespace Prefixes

PrefixIRI
n24http://linked.opendata.cz/ontology/domain/vavai/riv/typAkce/
dctermshttp://purl.org/dc/terms/
n23http://purl.org/net/nknouf/ns/bibtex#
n20http://localhost/temp/predkladatel/
n10http://linked.opendata.cz/resource/domain/vavai/projekt/
n8http://linked.opendata.cz/resource/domain/vavai/riv/tvurce/
n15http://linked.opendata.cz/resource/domain/vavai/subjekt/
n14http://linked.opendata.cz/ontology/domain/vavai/
n7https://schema.org/
shttp://schema.org/
skoshttp://www.w3.org/2004/02/skos/core#
rdfshttp://www.w3.org/2000/01/rdf-schema#
n3http://linked.opendata.cz/ontology/domain/vavai/riv/
n22http://bibframe.org/vocab/
n2http://linked.opendata.cz/resource/domain/vavai/vysledek/
rdfhttp://www.w3.org/1999/02/22-rdf-syntax-ns#
n12http://linked.opendata.cz/resource/domain/vavai/vysledek/RIV%2F49777513%3A23520%2F13%3A43919379%21RIV14-TA0-23520___/
n5http://linked.opendata.cz/ontology/domain/vavai/riv/klicoveSlovo/
n19http://linked.opendata.cz/ontology/domain/vavai/riv/duvernostUdaju/
xsdhhttp://www.w3.org/2001/XMLSchema#
n13http://linked.opendata.cz/ontology/domain/vavai/riv/jazykVysledku/
n4http://linked.opendata.cz/ontology/domain/vavai/riv/aktivita/
n18http://linked.opendata.cz/ontology/domain/vavai/riv/obor/
n11http://linked.opendata.cz/ontology/domain/vavai/riv/druhVysledku/
n16http://reference.data.gov.uk/id/gregorian-year/

Statements

Subject Item
n2:RIV%2F49777513%3A23520%2F13%3A43919379%21RIV14-TA0-23520___
rdf:type
skos:Concept n14:Vysledek
rdfs:seeAlso
http://link.springer.com/chapter/10.1007%2F978-3-319-01931-4_13
dcterms:description
An estimation of parameters of a multivariate Gaussian Mixture Model is usually based on a criterion (e.g. Maximum Likelihood) that is focused mostly on training data. Therefore, testing data, which were not seen during the training procedure, may cause problems. Moreover, numerical instabilities can occur (e.g. for low-occupied Gaussians especially when working with full-covariance matrices in high-dimensional spaces). Another question concerns the number of Gaussians to be trained for a specific data set. The approach proposed in this paper can handle all these issues. It is based on an assumption that the training and testing data were generated from the same source distribution. The key part of the approach is to use a criterion based on the source distribution rather than using the training data itself. It is shown how to modify an estimation procedure in order to fit the source distribution better (despite the fact that it is unknown), and subsequently new estimation algorithm for diagonal- as well as full-covariance matrices is derived and tested. An estimation of parameters of a multivariate Gaussian Mixture Model is usually based on a criterion (e.g. Maximum Likelihood) that is focused mostly on training data. Therefore, testing data, which were not seen during the training procedure, may cause problems. Moreover, numerical instabilities can occur (e.g. for low-occupied Gaussians especially when working with full-covariance matrices in high-dimensional spaces). Another question concerns the number of Gaussians to be trained for a specific data set. The approach proposed in this paper can handle all these issues. It is based on an assumption that the training and testing data were generated from the same source distribution. The key part of the approach is to use a criterion based on the source distribution rather than using the training data itself. It is shown how to modify an estimation procedure in order to fit the source distribution better (despite the fact that it is unknown), and subsequently new estimation algorithm for diagonal- as well as full-covariance matrices is derived and tested.
dcterms:title
Covariance Matrix Enhancement Approach to Train Robust Gaussian Mixture Models of Speech Data Covariance Matrix Enhancement Approach to Train Robust Gaussian Mixture Models of Speech Data
skos:prefLabel
Covariance Matrix Enhancement Approach to Train Robust Gaussian Mixture Models of Speech Data Covariance Matrix Enhancement Approach to Train Robust Gaussian Mixture Models of Speech Data
skos:notation
RIV/49777513:23520/13:43919379!RIV14-TA0-23520___
n14:predkladatel
n15:orjk%3A23520
n3:aktivita
n4:P
n3:aktivity
P(TA01011264)
n3:dodaniDat
n16:2014
n3:domaciTvurceVysledku
n8:6579760 n8:8612889 n8:2963671 n8:4396855
n3:druhVysledku
n11:D
n3:duvernostUdaju
n19:S
n3:entitaPredkladatele
n12:predkladatel
n3:idSjednocenehoVysledku
67224
n3:idVysledku
RIV/49777513:23520/13:43919379
n3:jazykVysledku
n13:eng
n3:klicovaSlova
Automatic Speech Recognition; Regularization; Full Covariance Matrix; Full Covariance; Gaussian Mixture Models
n3:klicoveSlovo
n5:Full%20Covariance n5:Full%20Covariance%20Matrix n5:Regularization n5:Automatic%20Speech%20Recognition n5:Gaussian%20Mixture%20Models
n3:kontrolniKodProRIV
[0989B2E77021]
n3:mistoKonaniAkce
Pilzen, Czech Republic
n3:mistoVydani
Cham
n3:nazevZdroje
Speech and Computer
n3:obor
n18:JD
n3:pocetDomacichTvurcuVysledku
4
n3:pocetTvurcuVysledku
4
n3:projekt
n10:TA01011264
n3:rokUplatneniVysledku
n16:2013
n3:tvurceVysledku
Psutka, Josef Psutka jr., Josef Vaněk, Jan Machlica, Lukáš
n3:typAkce
n24:WRD
n3:zahajeniAkce
2013-09-01+02:00
s:issn
0302-9743
s:numberOfPages
8
n22:doi
10.1007/978-3-319-01931-4_13
n23:hasPublisher
Springer-Verlag
n7:isbn
978-3-319-01930-7
n20:organizacniJednotka
23520