Functional speech design framework based on auditory information representation

2015 Fiscal Year Final Research Report

• Project/Area Number: 26540087
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Perceptual information processing
• Research Institution: Wakayama University
• Principal Investigator: KAWAHARA Hideki 和歌山大学, 学内共同利用施設等, 名誉教授 (40294300)
• Co-Investigator(Kenkyū-buntansha): IRINO Toshio 和歌山大学, システム工学部, 教授 (20346331) ; MATSUI Toshie 和歌山大学, システム工学部, 助教 (10510034) ; MORISE Masanori 山梨大学, 総合研究部, 助教 (60510013)
• Project Period (FY): 2014-04-01 – 2016-03-31
• Keywords: 音声情報処理 / 音声分析 / 音声変換 / パラ言語情報 / 信号処理 / 音声生成 / 音声知覚

Outline of Final Research Achievements

We introduced a new foundation of interactive investigation of functional speech sounds, based on our inventions, STRAIGHT, a speech analysis, modification and resynthesis framework, and a generalised morphing framework, which can morph arbitrarily many voices with individual and temporal control of constituent attributes. We also introduced a set of realtime interactive tools for providing understanding of speech production physics, physiology of voicing and articulatory organs, auditory perception and engineering representations. We also made the latter tools open sourced. Finally, we implemented a system to add glowl-like impression to any singing voices as an example of our proposed framework.

Free Research Field

音声情報処理