Title:	On speaker adaptive training of artificial neural networks
Other Titles:	Speaker adaptive training pro ANN
Authors:	Trmal, Jan Zelinka, Jan Müller, Luděk
Citation:	TRMAL, Jan; ZELINKA, Jan; MÜLLER, Luděk. On speaker adaptive training of artificial neural networks. In: Proceedings of ICSPL 2010: 11th Annual Conference of the International Speech Communication Association 2010, 26-30 September 2010, Makuhari, Chiba, Japan. [Baixas]: ISCA, 2010, p. [1-4]. ISSN 1990-9772.
Issue Date:	2010
Publisher:	ISCA
Document type:	článek article
URI:	http://www.kky.zcu.cz/cs/publications/TrmalJan_2010_OnSpeakerAdaptive http://hdl.handle.net/11025/16965
ISSN:	1990-9772
Keywords:	speaker adaptive training;TRAPS;VTLN;neuronové sítě;rozpoznávání
Keywords in different language:	speaker adaptive training;TRAPS;VTLN;neural networks;recognition
Abstract in different language:	In the paper we present two techniques improving the recognition accuracy of multilayer perceptron neural networks (MLP ANN) by means of adopting Speaker Adaptive Training. The use of the MLP ANN, usually in combination with the TRAPS parametrization, includes applications in speech recognition tasks, discriminative features production for GMM-HMM and other. In the first SAT experiments, we used the VTLN as a speaker normalization technique. Moreover, we developed a novel speaker normalization technique called Minimum Error Linear Transform (MELT) that resembles the cMLLR/fMLLR method \cite{gales96variance} with respect to the possible application either on the model or features. We tested these two methods extensively on telephone speech corpus SpeechDat-East. The results obtained in these experiments suggest that incorporation of SAT into MLP ANN training process is beneficial and depending on the setup leads to significant decrease of phoneme error rate (3% -- 8% absolute, 12% -- 25% relative).
Rights:	© Jan Trmal - Jan Zelinka - Luděk Müller
Appears in Collections:	Články / Articles (KKY)

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.