| Project/Area Number |
18K11525
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 62010:Life, health and medical informatics-related
|
|---|
| Research Institution | Kanazawa University |
|---|
Principal Investigator |
Satou Kenji 金沢大学, 融合科学系, 教授 (10215783)
|
|---|
| Project Period (FY) |
2018-04-01 – 2023-03-31
|
| Project Status |
Completed (Fiscal Year 2022)
|
| Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2020: ¥390,000 (Direct Cost: ¥300,000、Indirect Cost: ¥90,000)
Fiscal Year 2019: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2018: ¥2,600,000 (Direct Cost: ¥2,000,000、Indirect Cost: ¥600,000)
|
|---|
| Keywords | 生物行動 / 深層学習 / 位置推定 / 姿勢推定 / 動作認識 |
|---|
| Outline of Final Research Achievements |
Thanks to the rapid growth of deep learning, researches on the accurate recognition of human and object in images have made rapid strides. In contrast, researches about the recognition of animal's pose and behavior were relatively fewer and based on the motion capture with markers. In this research, it was shown that by using various machine learning algorithms including deep learning, time-series data of animal's location and pose can be accurately measured from video. In addition, based on the measured data, we developed methods for accurate recognition of specified behavior of animals, extraction of behaviors specific to animal individuals, and extraction of repetitive and/or characteristic behavior of animal and human by unsupervised learning.
|
| Academic Significance and Societal Importance of the Research Achievements |
人間の動画像を対象とした各種の認識技術や計測技術は近年急速に発展したが、動物の行動を自動的に認識し計量する研究は比較的少数に留まっていた。本研究により開発された各種の手法は、主に医学や薬学の分野で動物を用いて行われる行動観察実験の精度を高めるのに貢献するだけでなく、近年需要が高まりつつあるペット産業分野や畜産業および水産養殖業で、動画像を用いて動物の行動を計量し異常検出を行うなど、様々な応用が期待される点で社会的意義がある。また、繰り返し行動および特徴的行動の自動検出や、単色生物の動き検出の高精度化など、従来殆ど研究されていなかった分野で成果があったことに学術的意義がある。
|
