Technologies for visualizing social behaviors in multi-human motions with non-trivial behavioral rules
| Project/Area Number |
18K18116
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 61030:Intelligent informatics-related
|
|---|
| Research Institution | Nagoya University (2019-2020)
Institute of Physical and Chemical Research (2018) |
|---|
Principal Investigator |
Fujii Keisuke 名古屋大学, 情報学研究科, 准教授 (70747401)
|
|---|
| Project Period (FY) |
2018-04-01 – 2021-03-31
|
| Project Status |
Completed (Fiscal Year 2020)
|
| Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2019: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2018: ¥2,730,000 (Direct Cost: ¥2,100,000、Indirect Cost: ¥630,000)
|
|---|
| Keywords | 機械学習 / 集団運動 / スポーツ科学 / 時系列データ / 動的システム / 動的モード分解 / 身体運動 |
|---|
| Outline of Final Research Achievements |
The purpose of this project was to develop a technology for visualizing social behavior involving physical movement through data-driven modeling, and to create a basis for practical human use of the technology. As a result, we have developed a method that is useful for understanding multi-agent movements, even when using machine learning models with nonlinear structures that are generally difficult to interpret, such as by (1) extracting the mathematical structures behind them, (2) visualizing the learned representations, and (3) generating movements by modeling the components.
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究成果の学術的意義としては、様々な支配原理・法則が明確でない諸現象の中でも、より自由度の高い人間の集団運動を対象とするため、（1）背後の数学的な構造を抽出（2）学習した表現などを可視化（3）構成要素をモデル化して運動を生成する観点は、例えば人間以外の生物集団や、人工物の集団移動などにも応用可能である。社会的意義としては、本研究の研究対象である集団スポーツ解析はもちろん、子どもの集団遊びや、イベント時の移動軌跡などの上記の観点からの解析において役立つことが期待される。
|
Report
(4 results)
Research Products
(22 results)
