Statistical Music Modelling and Automation Techniques for Expressive Performance and Accompaniment Generation Capturing Performers' Musicality

2016 Fiscal Year Final Research Report

• Project/Area Number: 15K16054
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Intelligent informatics
• Research Institution: Kyoto University
• Principal Investigator: Nakamura Eita 京都大学, 情報学研究科, 特別研究員(PD) (10707574)
• Research Collaborator: YOSHII Kazuyoshi 京都大学, 大学院情報学研究科, 講師 (20510001) ; KANEKO Hitomi 桐朋学園大学, 音楽学部, 教授 (00408949) ; WATANABE Kenji 東京芸術大学, 音楽学部, 教授 ; SAITO Yasuyuki 木更津工業高等専門学校, 情報工学科, 准教授 (40331996) ; KATAYOSE Haruhiro 関西学院大学, 理工学部, 教授 (70294303) ; HASHIDA Mitsuyo 相愛大学, 音楽学部, 講師 (20421282)
• Project Period (FY): 2015-04-01 – 2017-03-31
• Keywords: 音楽情報処理 / 自動伴奏 / 自動作曲 / 音楽アラインメント / 演奏モデル

Outline of Final Research Achievements

Statistical music performance models have been constructed to capture various skill levels and characteristics of performance expressions and applied to information processing techniques. Performance elements such as tempo variation, onset and offset time fluctuation, and performance errors have been incorporated in the constructed performance models, and they have been successfully applied to multiple techniques including automatic music accompaniment, music alignment, and music transcription. In addition, performance data with accessible references to the score information, which are necessary in the research fields of expressive performance generation, music transcription, and music alignment, have been collected and set up as database, which were made available for future studies. Achievements include four journal papers, 15 refereed international conference papers, and 38 other conference presentations.

Free Research Field

音楽情報学