Research on a speech generating device utilizing remaining physical function for speech disorders.

2016 Fiscal Year Final Research Report

• Project/Area Number: 26750219
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Rehabilitation science/Welfare engineering
• Research Institution: The University of Tokyo
• Principal Investigator: Yabu Ken-ichiro 東京大学, 高齢社会総合研究機構, 特任研究員 (50626215)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 音声生成 / ホルマント / 構音障害 / 発話障害 / 声質 / スマートフォン

Outline of Final Research Achievements

We improved and proposed algorithms step by step with a prototype of a device which substitutes "motion of a finger" for "motion of a tongue" and in which continuous speeches are generated in real time like a musical instrument by the input with a pen or a finger on an operating panel plane.
First, we designed and improved an algorithm of the device with which a velocity of pen-input operation is complemented so that beginner users could also generate formant transitions well. And, we proposed and implemented an algorithm by which any continuous vowel sounds are generated with vocal quality similar to the short vowel sound for several seconds recorded by any speakers. Furthermore, we proposed the system which makes clear an inarticulate voice which remains in a person only with articulation disorder by inputting a motion of a pen, uttering voice. And we showed that the prototype system could generate voices similar to a speaker, keeping original prosody.

Free Research Field

福祉工学