Stochastic segment modelling using the estimate-maximize algorithm [speech recognition]

Author

Roucos, Salim ; Ostendorf, Mari ; Gish, Herbert ; Derr, Alan

Author_Institution

BBN Lab. Inc., Cambridge, MA, USA

fYear

1988

fDate

11-14 Apr 1988

Firstpage

127

Abstract

A probabilistic model called the stochastic segment model is introduced that describes the statistical dependence of all the frames of a speech segment. The model uses a time-warping transformation to map the sequence of observed frames to the appropriate frames of the segment model. The joint density of the observed frames is then given by the joint density of the selected model frames. The automatic training and recognition algorithms are discussed and a few preliminary recognition results are presented

Keywords

probability; speech recognition; stochastic processes; automatic training algorithms; estimate-maximise algorithms; joint density; observed frames; probabilistic model; recognition algorithms; speech recognition; speech segment; statistical dependence; stochastic segment model; time-warping transformation; Automatic speech recognition; Character recognition; Hidden Markov models; Pairwise error probability; Performance evaluation; Sampling methods; Speech recognition; Stochastic processes; Time frequency analysis; Vocabulary;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

Conference_Location

New York, NY

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1988.196528

Filename

196528