| Project/Area Number |
18K11598
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 62040:Entertainment and game informatics-related
|
|---|
| Research Institution | Morioka College (2019-2020)
Iwate University (2018) |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
吉田 等明 岩手大学, 教育学部, 特命教授 (00220666)
劉 忠達 石巻専修大学, 理工学部, 助教 (00782533)
|
|---|
| Project Period (FY) |
2018-04-01 – 2021-03-31
|
| Project Status |
Completed (Fiscal Year 2020)
|
| Budget Amount *help |
¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2020: ¥260,000 (Direct Cost: ¥200,000、Indirect Cost: ¥60,000)
Fiscal Year 2019: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2018: ¥520,000 (Direct Cost: ¥400,000、Indirect Cost: ¥120,000)
|
|---|
| Keywords | 感性情報処理 / 楽器演奏者 / 楽譜 / テンポ / 機械学習 / 人工知能 / データセット / 被験者実験 / 感性情報 / 音楽 |
|---|
| Outline of Final Research Achievements |
The data set (melody set), which is the basis for collecting the Kansei information of the musical instrument player, was constructed by extracting from the full score. This musical score set is played by a musical instrument player who has a certain level of reading ability and playing ability. The performance is the result of Kansei information processing, and Kansei information is obtained. But, In the experiment to obtain Kansei information, the experiment was restricted due to the countermeasure of "COVID-19".
As a result of computer experiments, it is shown that depending on the machine learning method, it is necessary to use an appropriate method, such as unknown data for evaluation, especially recognition of slow-paced classes being 0% even if learning is 100% successful. Was done. MLP and CNN gave better results than SVM.
|
| Academic Significance and Societal Importance of the Research Achievements |
楽譜を人間が処理した結果である演奏（音）ではなく，楽譜についての研究であり特異である．感性情報を収集しデータセットを構築するための旋律（楽譜）セットを構築したことは，感性情報を得る上でも意義が大きい．
この旋律（楽譜）セットを用いて，楽器演奏者が吹奏時のテンポなどの感性パラメータを収集し，楽譜の感性データセット構築の基礎を築いた．楽譜と対になる感性データセットも特異である．
楽譜の感性情報のうちテンポのクラス分類問題の試行では，機械学習手法によっては学習が100%成功しても未知の評価データの正答率が50%となるものもあり，MLPやCNNを用いると97%程度の認識率となることが示された．
|
