Direct application of linguistic variable to the speech segment description in isolated word recognizer
Abstract
A simple model of speech recognition based on identification of broad articulatory classes is presented. Direct application of the notion of linguistic variable to description of acoustic parameters evolutions and recognition of broad articulatory classes is discussed. The recognition algorithm is based on a fuzzy automatic network. An example of application of the described method to isolated word recognition is given.References
[1] W. BEZDEL, Some problems in man-machine communication using voice, Int. Journ. Man–Machine Stud. 2, 257-269 (1970).
[2] M. R. SAMBUR, L. R. RABINER, A speaker-independent digit recognition system, Bell Syst. Techn. Journ. 54, 1, (1975).
[3] C. J. WEINSTEIN, S. S. Mc CANDLESS, L. F. MONDSHEIN, Y. W. ZUE, A system for acoustic-phonetic analysis of continuous speech, IEEE ASSP 23, 1, 54-67 (1975).
[2] M. R. SAMBUR, L. R. RABINER, A speaker-independent digit recognition system, Bell Syst. Techn. Journ. 54, 1, (1975).
[3] C. J. WEINSTEIN, S. S. Mc CANDLESS, L. F. MONDSHEIN, Y. W. ZUE, A system for acoustic-phonetic analysis of continuous speech, IEEE ASSP 23, 1, 54-67 (1975).
