Vol.1, No 4, 1999 pp. 27 - 37
UDC 534.843.5:519.6
INTELLIGIBILITY SPEECH ESTIMATION
USING FUZZY LOGIC INFERENCING
Miloš Manić, Dragan Cvetković, Momir Praščević
University of Niš, Faculty of Occupational Safety, Čarnojevića 10a,
18000 Niš, Serbia, Yugoslavia, e-mail: misko@ieee.org
Abstract. Ambiental information intelligibility represents an ever
growing problem because of an evident growing acoustic load in case of
various halls. Fuzzy eXPert Systems represent an extension of existing
expert systems and are featuring some qualitative improvements, tending
to solve problems, previously approached only in a crisp manner. The example
given in this paper refers to an speech recognition estimation, by application
of proposed mechanisms.
Key words: fuzzy logic, soft computing, intelligibility speech,
knowledge base
PROCENA RAZUMLJIVOSTI GOVORA
KORIŠĆENJEM FAZI LOGIČKOG ZAKLJUČIVANJA
Razumljivost ambijentalnih informacija predstavlja rastući problem zbog
evidentno sve većeg akustičkog opterećenja u slučaju različitih hala. Fazi
ekspertni sistemi predstavljaju proširenje postojećih ekspertnih sistema
i karakterišu se nekim kvalitativnim poboljšanjima, koja nastoje da reše
probleme kojima se do tada pristupalo na kvantitativan, numerički način.
Primer dat u ovom radu, odnosi se na procenu prepoznavanja govora primenom
predloženih mehanizama.