ارائه روشهايي براي آموزش تلفظ و نواي گفتار زبان خارجي به كمك فناوري پردازش گفتار

Presentation of Methods for a Foreign Language Pronunciation and Prosody Teaching Using Speech Processing Technology

This paper concerns Computer-Assisted Language Learning. Speech processing technology was used to develop techniques to teach correct phoneme and intonation pronunciation to those who want to learn a new language. Evaluation of correct pronunciation is done in phoneme level. For this purpose, phoneme models are trained using speech data from speakers of the destination language. Speech data from a student is compared to the phoneme models and the goodness of pronunciation of each phoneme is scored. The assessment is made at phoneme levels of the learnersʹ speech, it will be possible to advise them of the exact errors. Fundamental frequency contour and fluency parameters are used to evaluate the goodness of intonation. Fluency is evaluated using speech temporal features, speech rate, rate of uttered phonemes and number of inserted pauses. Phoneme pronunciation and intonation scores are also used to evaluate the amount of nativeness of speakers and to evaluate how good a speaker utters a new language. The proposed techniques can be applied to any other languages.