Performance Design of Violin-playing Robot for Expressing Timbre
| Project/Area Number |
17K00513
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Entertainment and game informatics 1
|
|---|
| Research Institution | Ryukoku University |
|---|
Principal Investigator |
Shibuya Koji 龍谷大学, 理工学部, 教授 (20287973)
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥4,420,000 (Direct Cost: ¥3,400,000、Indirect Cost: ¥1,020,000)
Fiscal Year 2019: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2018: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2017: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
|
|---|
| Keywords | バイオリン演奏 / 演奏設計 / 表情付け / 強化学習 / ロボット / ビブラート / 音量 |
|---|
| Outline of Final Research Achievements |
We are aiming at building an algorithm which determines the violin performance for an anthropomorphic robot. We analyzed trained violinists’ performance focusing on the change pattern of sound pressure and tempo, and we found typical sound pressure patterns. We also revealed the following design principles to convey the bright or dark impressions: (1) the same tempo change pattern should be used while the different tempo value should be used, (2) the sound pressure change pattern should be different while the same sound pressure value should be used. Based on the results, we designed the violin performance for the robot and realized the performance. Based on a preliminary experiment, we confirmed that listeners were able to distinguish the bright performances from the dark when the both the tempo and the sound pressure were changed. In addition, we succeeded in building an algorithm for determining the violin performance using reinforcement learning and neural network.
|
| Academic Significance and Societal Importance of the Research Achievements |
楽器演奏には演奏動作という要因があり，すべての楽曲を好きなテンポ・音量で演奏できるわけではない．すなわち，ロボットの運動能力が選択するテンポや音圧に影響を与えているのである．そして，テンポと音圧の変化パターンと音色との関係は明らかではなかった．本研究成果は，人間の動作分析を軸にこの問題にチャレンジし，簡単な楽曲におけるテンポと音圧の設計指針の一端を明らかにした．また，また，強化学習という手法を導入し，演奏動作を自動的に決定するアルゴリズムも構築したことは，より汎用性のあるアルゴリズム構築に道を開いたといえる．これらの成果は，よりよい人間・ロボット間コミュニケーションに貢献するといえる．
|
Report
(4 results)
Research Products
(10 results)
