@techreport{oai:ipsj.ixsq.nii.ac.jp:00234758,
 author = {高力, あまね and 伊藤, 仁 and Amane, Koriki and Masashi, Ito},
 issue = {71},
 month = {Jun},
 note = {手足の動作を入力として，音高，音量，音韻などのパラメータを即時的に制御する音声合成システムについて，様々な研究が続けられている．本稿では，これまで我々が開発してきた音声合成システムを紹介し，操作に用いる身体部位による制御精度の変化，視覚フィードバックの影響，身体動作の速度限界などについて調べた結果を示す．また，これらの結果に基づいて設計した片手で把持できる声道を模擬した入力デバイスは，正確かつ高速な操作ができると考えられ，サーボモーターを用いた動作の呈示機構により操作の習熟時間も削減できると期待できる．このデバイスの試作結果と現在の課題について述べ，失声者の代替発声装置としての応用を目指す．, Several studies have been made on real-time speech synthesizers in which users’ arm or hand motions are immediately converted to speech parameters. We have realized several types of the motion-to-speech systems. Experimental results show that portion of body for input motion and visual feedback can affect the control accuracy of the synthesizer. Further, it was revealed to be difficult to realize formant transitions for glide syllables such as /ya/ or /wa/ by using the synthesizer because the input arm motion was slower than the utterance. On the basis of these studies, a new synthesizer is designed of which input device simulates vocal-tract shape, and can be operated by fingers. Using servo motors, the system can produce ideal finger motions corresponded to speech. This mechanism can be useful to reduce training period of the users.},
 title = {手指で声道断面積を操作する音声合成システム},
 year = {2024}
}