Research for big data acquisition for a behavior analysis and estimation(Fostering Joint International Research)

• Project/Area Number: 16KK0069
• Research Category: Fund for the Promotion of Joint International Research (Fostering Joint International Research)
• Allocation Type: Multi-year Fund
• Research Field: Sports science
• Research Institution: Iwate University
• Principal Investigator: Akashi Takuya 岩手大学, 理工学部, 准教授 (50403655)
• Project Period (FY): 2017 – 2019
• Project Status: Completed (Fiscal Year 2019)
• Budget Amount: ¥14,430,000 (Direct Cost: ¥11,100,000、Indirect Cost: ¥3,330,000)
• Keywords: スポーツ科学 / 行動解析 / マルチモーダル / 画像 / 人工知能 / 生体生命情報学 / 認知科学 / マルチモーダルインターフェース

Outline of Final Research Achievements

This research is interdisciplinary international joint research of the computer vision field and the neuroscience field, and obtained the following three results roughly.
1. The specific posture was used as a query and the technique of discovering the target posture efficiently from a human posture database was established. 2. The new framework of the artificial intelligence which combines different data, such as a sound and a image efficiently and learns was established. 3. The defect of spontaneous rhythm / tempo generation can be measured, and the new behavior experiment paradigm and electroencephalogram measurement paradigm, which can become a simple and objective diagnostic tool were established.
These 1 and 2 are results of taking take advantage of the knowledge from the neuroscience field. Moreover, 1 and 3 are applicable to the sports science fields in the near future.

Academic Significance and Societal Importance of the Research Achievements

基礎研究が非常に重要であるが，生活の質の向上，安心安全な持続可能な社会の実現，スポーツ科学への寄与などを考えると，学際的な領域の研究も重要である．本研究は，国際共同研究であるだけでなく，コンピュータビジョン分野とニューロサイエンス分野を融合させ，スポーツ科学分野に役立てるという学際的共同研究である．本研究を通じて，異分野からインスパイアされ，画像や音声データなど全く異なる種類のデータを効率よく利用する新たな人工知能を提案できた．この新しい人工知能は将来，ニューロサイエンス分野において，人間の脳の解明に役立つと考えられ，継続的な学際領域の研究発展に寄与しており，学術的かつ社会的意義は大きい．

Research Products

• [Int'l Joint Research] California Institute of Technology(米国) (2018)
  • Year and Date: 2018-06-15
• [Journal Article] Cross-Domain Deep Feature Combination for Bird Species Classification with Audio-Visual Data (2019)
  • Author(s): BOLD Naranchimeg、ZHANG Chao、AKASHI Takuya
  • Journal Title: IEICE Transactions on Information and Systems Volume: E102.D Issue: 10 Pages: 2033-2042
  • DOI: 10.1587/transinf.2018EDP7383
  • NAID: 130007722085
  • ISSN: 0916-8532, 1745-1361
  • Year and Date: 2019-10-01
  • Peer Reviewed / Open Access
• [Presentation] Free finger tapping tempo reflects intrinsic brain rhythms (1): assessed by EEG (2019)
  • Author(s): Sai Sun, Daw-An Wu, Takuya Akashi, Morimichi Furudate, Chao Zhang, Tetsuya Matsuda, Muneyoshi Takahashi, Shinsuke Shimojo
  • Organizer: 第42回日本神経科学大会-第62回日本神経化学会大会（NEURO2019）
• [Presentation] Free finger tapping tempo reflects intrinsic brain rhythms (2): assessed by resting-state fMRI (2019)
  • Author(s): Muneyoshi Takahashi, Takayuki Fujii, Sai Sun, Atsushi Miyazaki, Takuya Akashi, Morimichi Furudate, Chao Zhang, Toru Ishihara, Hiroki Tanaka, Haruto Takagishi, Shinsuke Shimojo, Tetsuya Matsuda
  • Organizer: 第42回日本神経科学大会-第62回日本神経化学会大会（NEURO2019）
• [Remarks] Smart Computer Vision Laboratory - 岩手大学
  • URL: http://www.scv.cis.iwate-u.ac.jp/