Signal-to-string conversion based on high likelihood regions using embedded dynamic programming

Author

Gong, Yifan ; Haton, Jean-Paul

Author_Institution

CRIN/INRIA-Lorraine, Vandoevre, France

Volume

13

Issue

3

fYear

1991

fDate

3/1/1991 12:00:00 AM

Firstpage

297

Lastpage

302

Abstract

A method of signal-to-string conversion based on embedded dynamic programming (DP) which can adapt its search to the variation of the input signal is proposed. The optimizing process is guided by high-valued portions of the likelihood function of symbols composing the string and is solved by two embedded dynamic programming processes. Algorithms in a Pascal-like language relating to the solution are given. When applied to continuous speech recognition on a 100-word vocabulary using the phoneme as the basic recognition unit, the method is shown to achieve a 4% improvement in the recognition rate compared to a classical DP-based method

Keywords

dynamic programming; pattern recognition; search problems; speech recognition; 100-word vocabulary; Pascal-like language; continuous speech recognition; embedded dynamic programming; high likelihood regions; search; signal-to-string conversion; Dynamic programming; Image converters; Pattern matching; Pattern recognition; Signal mapping; Signal processing; Speech recognition; Time factors; Vocabulary;

fLanguage

English

Journal_Title

Pattern Analysis and Machine Intelligence, IEEE Transactions on

Publisher

ieee

ISSN

0162-8828

Type

jour

DOI

10.1109/34.75518

Filename

75518