| Project/Area Number |
18K11444
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 61030:Intelligent informatics-related
|
|---|
| Research Institution | Tokyo Denki University |
|---|
Principal Investigator |
Satoshi Suzuki 東京電機大学, 未来科学部, 教授 (20328537)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
雨宮 由紀枝 日本女子体育大学, 体育学部, 教授 (40366802)
|
|---|
| Project Period (FY) |
2018-04-01 – 2021-03-31
|
| Project Status |
Completed (Fiscal Year 2020)
|
| Budget Amount *help |
¥3,770,000 (Direct Cost: ¥2,900,000、Indirect Cost: ¥870,000)
Fiscal Year 2020: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2019: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2018: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
|
|---|
| Keywords | 行動識別 / 粗大運動 / 深層学習 / 運動発育 / 特別支援教育 / 機械学習 |
|---|
| Outline of Final Research Achievements |
Regarding activity recognition (AR) for physical motor function assessment of children, we studied a gross motor AR, an assessment AR, and a motor development AI. These steps were achieved by enhancement of robustness (improvement of recognition abilities) and an AI coding of motor development assessment process (realization of evaluation function).
The specific results are as follows: on-site measurement and database maintenance according to TGMD-3, development of various iOS apps to support an AI data-set creation, various AR deep learning algorithms for child gross motors, and deep-network design methods for visualization of identification reasons / inadequate body motions using Grad-CAM / VAE.
|
| Academic Significance and Societal Importance of the Research Achievements |
従来の行動識別(AR)では個別に扱われていた「身体動作の識別・評価・分析」をシームレスに行うAR理論とその実践法を示せたことが本成果の学術的意義である．特に身体動作不具合箇所の動画可視化法は，従来ARの課題であった，学習済識別器の"意味解釈の困難"の軽減に寄与した．
社会的意義は，属人的判断に傾倒しがちな子どもの身体発達評価において，掌AIサーバとタブレットアプリによる客観評価ツールの具現化により，利用関係者間の情報共有利便性が向上することと，AR学が運動発達アセスメント体制の礎になり得ることを示せた点である．
|
