摘要: This invention includes a speech training system that allows student to enter any utterance be learned and have the articulatory model movements required produce displayed on CRT screen. The accepts typed utterance, breaking it down into set of units which could phonemes or syllables onset offset units. is sent synthesizer, produces parameters indicating acoustic characteristics utterance. are converted emphasizing frequency nasality each used generate tongue-palate contact patterns also formant synthesizer converts output. measures student's production then evaluates against for its similarity. Feedback similarity