Title :
A syntax aided isolated word recognition system for a subset of FORTRAN
Author :
Keeler, L.O. ; Fu, K.S.
Author_Institution :
Hughes Aircraft co., Fullerton, California
Abstract :
This paper describes an isolated word recognition system for a subset of FORTRAN in which the syntax of a formal language is used to aid a simple acoustic level word recognizer. The acoustic level recognizer makes use of the fast Fourier transform to estimate a short-time spectrum which is coded with a set of 24 binary values. Words to be recognized are compared with patterns stored in a reference vocabulary. The syntax of the language allows the number of comparisons that must be made to be reduced significantly, and also aids in directing the recognition procedure. Test results indicate that inclusion of syntactic constraints increases the per cent of correct recognition by approximately 25%.
Keywords :
Aircraft; Fast Fourier transforms; Formal languages; Pattern recognition; Real time systems; Signal processing; Speech processing; Testing; Vocabulary;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '77.
DOI :
10.1109/ICASSP.1977.1170345