Towards autonomous synthesis of empathetic speech for a information system to obtain users' confidence

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K00272
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Human interface and interaction
• Research Institution: Tsuyama National College of Technology (2016-2017); Nara Institute of Science and Technology (2015)
• Principal Investigator: Kawanami Hiromichi 津山工業高等専門学校, 総合理工学科, 特命准教授 (80335489)
• Co-Investigator(Kenkyū-buntansha): 神原 誠之 奈良先端科学技術大学院大学, 先端科学技術研究科, 准教授 (10346306)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: ヒューマンインタフェイス / 対話システム / 韻律 / デジタルデバイド

Outline of Final Research Achievements

This research aims to contribute to reduce mental barrier to an information system for those who are unfamiliar to such systems. We investigated two approaches to generate empathetic speech of a spoken dialogue system autonomously: (1)Synthesizing a response speech to mimic its user's speech especially prosody, (2)Generation and timing control of hierarchical responses consist of a back channel, a repetition of a user utterance and selected comments from SNS(Social Network Service). These methods are applied to a chat robot and the experimental result indicates that the methods encourage the users' to use the system continuously.

Free Research Field

音声情報処理