Presentation 2007/12/13
Efficient estimation method of scaling factors among probabilistic models in speech recognition
Tadashi EMORI, Yoshifumi ONISHI, Koichi SHINODA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose a new efficient method for estimating scaling factors among probabilistic models in speech recognition. Most speech recognition systems consist of more than one model, include an acoustic and a language model, and require scaling factors to balance probabilities among them. The scaling factors are conventionally optimized in preliminary recognition tests using data for development. In our proposed method, the scaling factors are regarded as parameters of a log-linear model, and they are estimated using a gradient-ascent method based on the maximum a posteriori probability criterion. Posterior probability is computed using word-lattices generated by a speech recognizer. We employ an iteration technique which repeats a word-lattice-generation/scaling-factor-estimation process, and the resulting scaling factor estimation is robust with respect to the changes in initial values. In an experimental evaluation of our method by LVCSR using Japanese dialogue speech data, estimated scaling factors were nearly identical to optimal values obtained in a greedy grid search. We have also confirmed that estimated scaling factors changed little with variations in initial values.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) speech recognition / scaling factor / log-linear model / word lattice
Paper # NLC2007-41,SP2007-104
Date of Issue

Conference Information
Committee NLC
Conference Date 2007/12/13(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Efficient estimation method of scaling factors among probabilistic models in speech recognition
Sub Title (in English)
Keyword(1) speech recognition
Keyword(2) scaling factor
Keyword(3) log-linear model
Keyword(4) word lattice
1st Author's Name Tadashi EMORI
1st Author's Affiliation NEC Corporation()
2nd Author's Name Yoshifumi ONISHI
2nd Author's Affiliation NEC Corporation
3rd Author's Name Koichi SHINODA
3rd Author's Affiliation Tokyo Institute of Technology
Date 2007/12/13
Paper # NLC2007-41,SP2007-104
Volume (vol) vol.107
Number (no) 405
Page pp.pp.-
#Pages 5
Date of Issue