| Budget Amount *help |
¥16,250,000 (Direct Cost: ¥12,500,000、Indirect Cost: ¥3,750,000)
Fiscal Year 2018: ¥5,590,000 (Direct Cost: ¥4,300,000、Indirect Cost: ¥1,290,000)
Fiscal Year 2017: ¥5,330,000 (Direct Cost: ¥4,100,000、Indirect Cost: ¥1,230,000)
Fiscal Year 2016: ¥5,330,000 (Direct Cost: ¥4,100,000、Indirect Cost: ¥1,230,000)
|
|---|
| Outline of Final Research Achievements |
We developed methods of P1) gait-based age estimation, which have potential applications such as gait video retrieval of suspect candidates by witness on age group for criminal investigation and age group-dependent customer counting in a wide area shopping mall for marketing research. Specifically, we constructed the world largest gait video database with age labels, and developed methods of gait-based age estimation by age group clustering and manifold learning in addition to recent deep learning-based approaches to gait-based age estimation. We also developed a baseline and adversarial generative network-based approaches to P2) gait age progression/regression modelling, which are potentially applied to gait aging simulation systems to promote health and exercise.
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究で構築したOULP-Age は，2歳から90歳の幅広い年代の男女を含む合計63,846名（男性31,093名，女性32,753）の被験者の歩行映像並びに年齢・性別のラベルを含む，世界最大の歩行映像データベースであり，見えに基づく歩行映像解析の分野で代表的に用いられている歩容エネルギー画像と年齢・性別のラベルのセットとして公開していることから，歩行映像解析の研究分野の発展に貢献しており，学術的意義が大きい．また，世界で初めて歩行映像解析による歩容の経年変化モデリングによる研究を実施したこと，世界に先駆けて深層学習を用いた歩容年齢推定の研究を実施したことも，学術的に意義があると言える．
|
