• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Physically consistent human motion generation based on fusion theory of motion optimization and deep learning

Research Project

Project/Area Number 18H03315
Research Category

Grant-in-Aid for Scientific Research (B)

Allocation TypeSingle-year Grants
Section一般
Review Section Basic Section 61050:Intelligent robotics-related
Research InstitutionNational Institute of Advanced Industrial Science and Technology

Principal Investigator

Ayusawa Ko  国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 主任研究員 (60649086)

Co-Investigator(Kenkyū-buntansha) 吉安 祐介  国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 主任研究員 (10712234)
Project Period (FY) 2018-04-01 – 2021-03-31
Project Status Completed (Fiscal Year 2020)
Budget Amount *help
¥17,160,000 (Direct Cost: ¥13,200,000、Indirect Cost: ¥3,960,000)
Fiscal Year 2020: ¥3,770,000 (Direct Cost: ¥2,900,000、Indirect Cost: ¥870,000)
Fiscal Year 2019: ¥4,940,000 (Direct Cost: ¥3,800,000、Indirect Cost: ¥1,140,000)
Fiscal Year 2018: ¥8,450,000 (Direct Cost: ¥6,500,000、Indirect Cost: ¥1,950,000)
Keywords人運動解析 / ヒューマノイド / 運動最適化 / 深層学習 / 人型システム運動生成
Outline of Final Research Achievements

In this study, we proposed a new methodology that can generate physically consistent human movements by combining motion optimization theory for physical consistency and deep learning based on a motion dataset. We have developed an efficient method for gradient computation in motion optimization which can handle various physical quantities, a method for constructing a deep neural network considering kinematics consistency of the human skeletal system, and a motion generation method that can evaluate mechanical properties such as velocities, accelerations, and external forces. As application examples, we have developed a real-time estimation of human musculoskeletal motion by using measurement devices such as 2D video images and inertia sensors, and a motion retargeting method from humans motion to humanoid robots.

Academic Significance and Societal Importance of the Research Achievements

運動最適化と機械学習を融合させることで、簡易的な計測から力学的整合性を持つ人の運動生成を実現させる方法論を開発した。最適化理論の観点からは学習により冗長性問題を解決し、機械学習的観点からは運動最適化により力学的整合性を実現させた。これにより、人の運動に纏わる力学的整合性・冗長性問題を同時に扱えるという学術的意義を示せた。また、ビデオ画像や慣性センサなどの計測デバイスによる人の実時間筋骨格運動解析技術は、スポーツ・リハビリなどの特定の場面だけでなく、人の運動モニタリングによる健康・労務管理などの日常生活や労働環境において活用できることが期待されるため、社会的・産業的な波及効果も期待できる。

Report

(4 results)
  • 2020 Annual Research Report   Final Research Report ( PDF )
  • 2019 Annual Research Report
  • 2018 Annual Research Report
  • Research Products

    (9 results)

All 2020 2019 2018

All Journal Article (8 results) (of which Int'l Joint Research: 1 results,  Peer Reviewed: 8 results,  Open Access: 1 results) Presentation (1 results)

  • [Journal Article] Simultaneous Control Framework for Humanoid Tracking Human Movement With Interacting Wearable Assistive Device2020

    • Author(s)
      Ito Takahiro、Ayusawa Ko、Yoshida Eiichi、Kobayashi Hiroshi
    • Journal Title

      IEEE Robotics and Automation Letters

      Volume: 5 Issue: 2 Pages: 3604-3611

    • DOI

      10.1109/lra.2020.2979663

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Motion Synthesis Using Low-Dimensional Feature Space and Its Application to Inverse Optimal Control2020

    • Author(s)
      Shimizu Soya、Ayusawa Ko、Venture Gentiane
    • Journal Title

      Proceedings of the 23rd CISM IFToMM Symposium

      Volume: 601 Pages: 503-511

    • DOI

      10.1007/978-3-030-58380-4_60

    • ISBN
      9783030583798, 9783030583804
    • Related Report
      2020 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Fusion of Multiple Motion Capture Systems for Musculoskeletal Analysis2020

    • Author(s)
      Samy Vincent、Ayusawa Ko、Yoshiyasu Yusuke、Sagawa Ryusuke、Yoshida Eiichi
    • Journal Title

      Proc. of the IEEE/SICE International Symposium on System Integration

      Volume: - Pages: 295-299

    • DOI

      10.1109/sii46433.2020.9025818

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Predictive Inverse Kinematics: optimizing Future Trajectory through Implicit Time Integration and Future Jacobian Estimation2019

    • Author(s)
      Ayusawa Ko、Suleiman Wael、Yoshida Eiichi
    • Journal Title

      Proc. of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems

      Volume: - Pages: 566-573

    • DOI

      10.1109/iros40897.2019.8968110

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] Musculoskeletal Estimation Using Inertial Measurement Units and Single Video Image2019

    • Author(s)
      Samy Vincent、Ayusawa Ko、Yoshiyasu Yusuke、Sagawa Ryusuke、Yoshida Eiichi
    • Journal Title

      Proc. of the IEEE Int. Workshop on Advanced Robotics and its Social Impacts

      Volume: - Pages: 39-44

    • DOI

      10.1109/arso46408.2019.8948820

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Skeleton Transformer Networks: 3D Human Pose and Skinned Mesh from Single RGB Image2019

    • Author(s)
      Yusuke Yoshiyasu, Ryusuke Sagawa, Ko Ayusawa, Akihiko Murai
    • Journal Title

      In proceedings of Asian Conference on Computer Vision

      Volume: - Pages: 1-16

    • DOI

      10.1007/978-3-030-20870-7_30

    • ISBN
      9783030208691, 9783030208707
    • Related Report
      2018 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Online Center of Mass and Momentum Estimation for a Humanoid Robot Based on Identification of Inertial Parameters2018

    • Author(s)
      Mori Kenya、Ayusawa Ko、Yoshida Eiichi
    • Journal Title

      Proc. of the IEEE-RAS Int. Conf. on Humanoid Robots

      Volume: 1 Pages: 257-263

    • DOI

      10.1109/humanoids.2018.8625037

    • Related Report
      2018 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Optimization Framework of Humanoid Walking Pattern for Human Motion Retargeting2018

    • Author(s)
      Masuda Shimpei、Ayusawa Ko、Yoshida Eiichi
    • Journal Title

      Proc. of the IEEE-RAS Int. Conf. on Humanoid Robots

      Volume: 1 Pages: 725-731

    • DOI

      10.1109/humanoids.2018.8625060

    • Related Report
      2018 Annual Research Report
    • Peer Reviewed
  • [Presentation] Motion Optimization Framework of Humanoid Systems toward Human Ergonomic Analysis2020

    • Author(s)
      Ayusawa Ko
    • Organizer
      The 6th Symposium on Theoretical and Applied Mechanics
    • Related Report
      2020 Annual Research Report

URL: 

Published: 2018-04-23   Modified: 2022-01-27  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi