A New Method for Discriminative Model Combination in Speech Recognition

Author

Wu Yahui ; Liu Gang ; Guo Jun

Author_Institution

Lab. of Pattern Recognition & Intell. Syst., Beijing Univ. of Posts & Telecommun., Beijing

Volume

1

fYear

2008

fDate

13-17 Dec. 2008

Firstpage

200

Lastpage

203

Abstract

A new method based on discriminative model combination for acoustic model training is proposed. The MPE trained model and the MMIE trained model are used for model combination. The combination criterion is based on the ratio of the inter-variance to the intra-variance of each model. Besides we also propose a ldquoclusterrdquo method for the model to choose its confusion models in order to get the accuracy information for the combination weight calculation. The experiments demonstrate that the new model can get better performance than any of the single MPE and MMIE trained models.

Keywords

speech recognition; acoustic model training; cluster method; discriminative model; model combination; speech recognition; Computational intelligence; Computer errors; Error correction; Hidden Markov models; Laboratories; Maximum likelihood estimation; Pattern recognition; Security; Speech recognition; Training data; MMIE; MPE; model combination; speech training;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Security, 2008. CIS '08. International Conference on

Conference_Location

Suzhou

Print_ISBN

978-0-7695-3508-1

Type

conf

DOI

10.1109/CIS.2008.128

Filename

4724641