• DocumentCode
    670246
  • Title

    Learning new skills by a humanoid robot through imitation

  • Author

    Boboc, R.G. ; Toma, Madalina-Ioana ; Panfir, Alina Ninett ; Talaba, D.

  • Author_Institution
    Dept. of Robot. & Virtual Reality, Transilvania Univ., Brasov, Romania
  • fYear
    2013
  • fDate
    19-21 Nov. 2013
  • Firstpage
    515
  • Lastpage
    519
  • Abstract
    Today´s humanoid robots are used in many successful research topics. According to users needs these robots must be able to perform new tasks, and to adapt in a workspace. They have a similar appearance with human body. Thus the meaningful features of an humanoid robots are: the capability to imitate the user behavior, and the ability to learn new skills. In this paper, we propose to develop a robot system based on an humanoid robot that has abilities to learn new user skills by imitating human beings. To achieve this goal, we needed to design and to implement a new framework, which is able to teach an humanoid robot via user demonstration and to imitate movements of the human body. Therefore, the movements of the human body are automatically tracked using the Kinect sensor, and they teleoperate a whole body of a Nao robot. The results of experiments suggest that the feasibility of the proposed system is verified.
  • Keywords
    human-robot interaction; humanoid robots; image sensors; user interfaces; Kinect sensor; Nao robot; humanoid robots; new user skills; research topics; robot system; user behavior; user demonstration; Humanoid robots; Joints; Robot sensing systems; Service robots; Tracking;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computational Intelligence and Informatics (CINTI), 2013 IEEE 14th International Symposium on
  • Conference_Location
    Budapest
  • Print_ISBN
    978-1-4799-0194-4
  • Type

    conf

  • DOI
    10.1109/CINTI.2013.6705251
  • Filename
    6705251