. "Radov\u00E1, Vlasta" . "0" . . . . "[CED1F2430B37]" . . . "A method of speaker verification based on Gaussian mixture models is presented in this paper. The method works with a background model which is composed of several submodels. Several different approaches for construction of the background model from the submodels are introduced here: the log likelihood of the background model is determined either as the average of the log likelihoods of the particular submodels, or a maximum from the log likelihoods of the particular submodels is selected. A large number of experiments was performed in order to find which of the approaches gives the best result. All experiments show that procedures which use a maximum of the log likelihoods of the background submodels have better performance than the procedure which uses the average log likelihood." . "On the background model construction for speaker verification using GMM" . "P(GA102/02/0124), Z(MSM 235200004)" . . "Konstrukce modelu okol\u00ED pro verifikaci \u0159e\u010Dn\u00EDka vyu\u017E\u00EDvaj\u00EDc\u00ED GMM"@cs . "A method of speaker verification based on Gaussian mixture models is presented in this paper. The method works with a background model which is composed of several submodels. Several different approaches for construction of the background model from the submodels are introduced here: the log likelihood of the background model is determined either as the average of the log likelihoods of the particular submodels, or a maximum from the log likelihoods of the particular submodels is selected. A large number of experiments was performed in order to find which of the approaches gives the best result. All experiments show that procedures which use a maximum of the log likelihoods of the background submodels have better performance than the procedure which uses the average log likelihood."@en . "Konstrukce modelu okol\u00ED pro verifikaci \u0159e\u010Dn\u00EDka vyu\u017E\u00EDvaj\u00EDc\u00ED GMM"@cs . "425" . "577840" . "On the background model construction for speaker verification using GMM"@en . "speaker verification, score normalization, background model"@en . . "RIV/49777513:23520/04:00000034!RIV07-GA0-23520___" . "23520" . "Padrta, Ale\u0161" . . "2"^^ . "Lecture Notes in Artificial Intelligence" . . "2"^^ . . "DE - Spolkov\u00E1 republika N\u011Bmecko" . "On the background model construction for speaker verification using GMM" . . "RIV/49777513:23520/04:00000034" . . "V \u010Dl\u00E1nku je prezentov\u00E1na metoda verifikace \u0159e\u010Dn\u00EDka vyu\u017E\u00EDvaj\u00EDc\u00ED Gaussovsk\u00FDch hustotn\u00EDch sm\u011Bs\u00ED. Metoda vyu\u017E\u00EDv\u00E1 model okol\u00ED slo\u017Een\u00FD z n\u011Bkolika submodel\u016F. V \u010Dl\u00E1nku jsou uvedeny r\u016Fzn\u00E9 p\u0159\u00EDstupy pro konstrukci slo\u017Een\u00E9ho modelu: logaritmus pravd\u011Bpodobnosti slo\u017Een\u00E9ho modelu je ur\u010Den bu\u010F jako pr\u016Fm\u011Br logaritm\u016F pravd\u011Bpodobnost\u00ED d\u00EDl\u010D\u00EDch submodel\u016F, a nebo jako maximum z logaritm\u016F pravd\u011Bpodobnost\u00ED d\u00EDl\u010D\u00EDch submodel\u016F. Proveden\u00E9 experimety uk\u00E1zaly, \u017Ee prvn\u011B jmenovan\u00FD p\u0159\u00EDstup umo\u017E\u0148uje dos\u00E1hnout lep\u0161\u00EDch v\u00FDsledk\u016F."@cs . . "8"^^ . "On the background model construction for speaker verification using GMM"@en . "0302-9743" . . . . .