音声応答式ＣＡＩシステムの開発Ⅱ 音声合成装置の改良

概要

論文の詳細を見る
On the basis of the voice-wave-segment-edition method of speech synthesis reported earlier, a new synthesizer has been developed. This transforms an input text into speech with pitch and stress accents and duration specified by the input for each phoneme. Pitch control giving pitch accent is made by changing the time interval with which digital (wave form) data are output onto a D/A converter. The volume control giving stress accent is made by changing the control voltage for an electronic attenuator placed between the LPF and the amplifier. The control voltage is given by the CPU through a parallel output port. Duration of each phoneme is also controlled by the timer interrupt, where a phoneme is elongated essentially by prolonging the vowel part of the phoneme. Pseudo-natural speech has been obtained by this method.
奈良教育大学の論文
1984-03-16