Paper Info
Title
Automatic set-up for speech recognition engines based on merit optimization
Abstract
We propose an automatic method to set-up the several parameters that define the behavior and performance of a typical speech recognition engine. Such parameters include weights and beam widths among others. Our method is based on the definition of a merit function. Here, merit is understood as an intuitive notion of recognition performance based on both recognition accuracy and computation time. A convenient definition of merit allows an optimization procedure to be applied to define a convenient set-up for the recognizer with little human intervention. The method is applied to adjust the recognition parameters of two different LVCSR (large vocabulary continuous speech recognition) applications, one in American English and another in Japanese.
Year
DOI
Venue
2003
10.1109/ICASSP.2003.1198750
ICASSP '03). 2003 IEEE International Conference
Keywords
DocType
Volume
fuzzy logic,optimisation,speech recognition,American English,Japanese,automatic parameter set-up,beam widths,fuzzy logic,large vocabulary continuous speech recognition,merit function,merit optimization,recognition parameters,speech recognition engines,weights
Conference
1
ISSN
ISBN
Citations 
1520-6149
0-7803-7663-3
2
PageRank 
References 
Authors
0.45
1
7
Name
Order
Citations
PageRank
Gustavo Hernández Ábrego182.75
Xavier Menéndez-Pidal26210.26
Thomas Kemp324630.93
Katsuki Minamino420.79
Helmut Lucke51617.07
Hernandez-Abrego, G.620.45
Menendez-Pidal, X.720.45