Motion capture based human motion recognition and imitation by direct marker control

Motion capture based human motion recognition and imitation by direct marker control. Ott, C., Lee, D., & Nakamura, Y. In Proceedings of the IEEE/RAS International Conference on Humanoid Robots, pages 399--405, 2008.
doi abstract bibtex

This paper deals with the imitation of human motions by a humanoid robot based on marker point measurements from a 3D motion capture system. For imitating the humanpsilas motion, we propose a Cartesian control approach in which a set of control points on the humanoid is selected and the robot is virtually connected to the measured marker points via translational springs. The forces according to these springs drive a simplified simulation of the robot dynamics, such that the real robot motion can finally be generated based on joint position controllers effectively managing joint friction and other uncertain dynamics. This procedure allows to make the robot follow the marker points without the need of explicitly computing inverse kinematics. For the implementation of the marker control on a humanoid robot, we combine it with a center of gravity based balancing controller for the lower body joints. We integrate the marker control based motion imitation with the mimesis model, which is a mathematical model for motion learning, recognition, and generation based on hidden Markov models (HMMs). Learning, recognition, and generation of motion primitives are all performed in marker coordinates paving the way for extending these concepts to task space problems and object manipulation. Finally, an experimental evaluation of the presented concepts using a 38 degrees of freedom humanoid robot is discussed.

@InProceedings{Ott2008,
  Title                    = {Motion capture based human motion recognition and imitation by direct marker control},
  Author                   = {Ott, C. and Lee, D. and Nakamura, Y.},
  Booktitle                = {Proceedings of the IEEE/RAS International Conference on Humanoid Robots},
  Year                     = {2008},
  Pages                    = {399--405},

  Abstract                 = {This paper deals with the <span class='snippet'>imitation</span> of <span class='snippet'>human</span> <span class='snippet'>motions</span> <span class='snippet'>by</span> a humanoid robot <span class='snippet'>based</span> on marker point measurements from a 3D <span class='snippet'>motion</span> <span class='snippet'>capture</span> system. For imitating the humanpsilas <span class='snippet'>motion</span>, we propose a Cartesian control approach in which a set of control points on the humanoid is selected <span class='snippet'>and</span> the robot is virtually connected to the measured marker points via translational springs. The forces according to these springs drive a simplified simulation of the robot dynamics, such that the real robot <span class='snippet'>motion</span> can finally be generated <span class='snippet'>based</span> on joint position controllers effectively managing joint friction <span class='snippet'>and</span> other uncertain dynamics. This procedure allows to make the robot follow the marker points without the need of explicitly computing inverse kinematics. For the implementation of the marker control on a humanoid robot, we combine it with a center of gravity <span class='snippet'>based</span> balancing controller for the lower body joints. We integrate the marker control <span class='snippet'>based</span> <span class='snippet'>motion</span> <span class='snippet'>imitation</span> with the mimesis model, which is a mathematical model for <span class='snippet'>motion</span> learning, <span class='snippet'>recognition</span>, <span class='snippet'>and</span> generation <span class='snippet'>based</span> on hidden Markov models (HMMs). Learning, <span class='snippet'>recognition</span>, <span class='snippet'>and</span> generation of <span class='snippet'>motion</span> primitives are all performed in marker coordinates paving the way for extending these concepts to task space problems <span class='snippet'>and</span> object manipulation. Finally, an experimental evaluation of the presented concepts using a 38 degrees of freedom humanoid robot is discussed.},
  Doi                      = {10.1109/ICHR.2008.4755984},
  Keywords                 = {ECE780, Control},
  Review                   = {This paper uses motion from human motion capture, and fits the motion to a robot. They do so by applying virtual springs that applies attraction forces between the motion capture data and a simplified robot model, and insert the joint trajectories into the robot. These trajectories are than encoded by HMMs. They note that this is motivated by mirror neurons in the brain. 

For simplicity, they focus only on upper body motions. The lower body was set to maintain overall stability by keeping ZMP within the support polygon.},
  Timestamp                = {2011.02.09}
}

Downloads: 0

{"_id":"JepJXt6swYoJ6ynBx","bibbaseid":"ott-lee-nakamura-motioncapturebasedhumanmotionrecognitionandimitationbydirectmarkercontrol-2008","downloads":0,"creationDate":"2017-09-14T16:34:36.901Z","title":"Motion capture based human motion recognition and imitation by direct marker control","author_short":["Ott, C.","Lee, D.","Nakamura, Y."],"year":2008,"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/jfslin/jfslin.github.io/master/jf2lin.bib","bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Motion capture based human motion recognition and imitation by direct marker control","author":[{"propositions":[],"lastnames":["Ott"],"firstnames":["C."],"suffixes":[]},{"propositions":[],"lastnames":["Lee"],"firstnames":["D."],"suffixes":[]},{"propositions":[],"lastnames":["Nakamura"],"firstnames":["Y."],"suffixes":[]}],"booktitle":"Proceedings of the IEEE/RAS International Conference on Humanoid Robots","year":"2008","pages":"399--405","abstract":"This paper deals with the imitation of human motions by a humanoid robot based on marker point measurements from a 3D motion capture system. For imitating the humanpsilas motion, we propose a Cartesian control approach in which a set of control points on the humanoid is selected and the robot is virtually connected to the measured marker points via translational springs. The forces according to these springs drive a simplified simulation of the robot dynamics, such that the real robot motion can finally be generated based on joint position controllers effectively managing joint friction and other uncertain dynamics. This procedure allows to make the robot follow the marker points without the need of explicitly computing inverse kinematics. For the implementation of the marker control on a humanoid robot, we combine it with a center of gravity based balancing controller for the lower body joints. We integrate the marker control based motion imitation with the mimesis model, which is a mathematical model for motion learning, recognition, and generation based on hidden Markov models (HMMs). Learning, recognition, and generation of motion primitives are all performed in marker coordinates paving the way for extending these concepts to task space problems and object manipulation. Finally, an experimental evaluation of the presented concepts using a 38 degrees of freedom humanoid robot is discussed.","doi":"10.1109/ICHR.2008.4755984","keywords":"ECE780, Control","review":"This paper uses motion from human motion capture, and fits the motion to a robot. They do so by applying virtual springs that applies attraction forces between the motion capture data and a simplified robot model, and insert the joint trajectories into the robot. These trajectories are than encoded by HMMs. They note that this is motivated by mirror neurons in the brain. For simplicity, they focus only on upper body motions. The lower body was set to maintain overall stability by keeping ZMP within the support polygon.","timestamp":"2011.02.09","bibtex":"@InProceedings{Ott2008,\n Title = {Motion capture based human motion recognition and imitation by direct marker control},\n Author = {Ott, C. and Lee, D. and Nakamura, Y.},\n Booktitle = {Proceedings of the IEEE/RAS International Conference on Humanoid Robots},\n Year = {2008},\n Pages = {399--405},\n\n Abstract = {This paper deals with the imitation of human motions by a humanoid robot based on marker point measurements from a 3D motion capture system. For imitating the humanpsilas motion, we propose a Cartesian control approach in which a set of control points on the humanoid is selected and the robot is virtually connected to the measured marker points via translational springs. The forces according to these springs drive a simplified simulation of the robot dynamics, such that the real robot motion can finally be generated based on joint position controllers effectively managing joint friction and other uncertain dynamics. This procedure allows to make the robot follow the marker points without the need of explicitly computing inverse kinematics. For the implementation of the marker control on a humanoid robot, we combine it with a center of gravity based balancing controller for the lower body joints. We integrate the marker control based motion imitation with the mimesis model, which is a mathematical model for motion learning, recognition, and generation based on hidden Markov models (HMMs). Learning, recognition, and generation of motion primitives are all performed in marker coordinates paving the way for extending these concepts to task space problems and object manipulation. Finally, an experimental evaluation of the presented concepts using a 38 degrees of freedom humanoid robot is discussed.},\n Doi = {10.1109/ICHR.2008.4755984},\n Keywords = {ECE780, Control},\n Review = {This paper uses motion from human motion capture, and fits the motion to a robot. They do so by applying virtual springs that applies attraction forces between the motion capture data and a simplified robot model, and insert the joint trajectories into the robot. These trajectories are than encoded by HMMs. They note that this is motivated by mirror neurons in the brain. \n\nFor simplicity, they focus only on upper body motions. The lower body was set to maintain overall stability by keeping ZMP within the support polygon.},\n Timestamp = {2011.02.09}\n}\n\n","author_short":["Ott, C.","Lee, D.","Nakamura, Y."],"key":"Ott2008","id":"Ott2008","bibbaseid":"ott-lee-nakamura-motioncapturebasedhumanmotionrecognitionandimitationbydirectmarkercontrol-2008","role":"author","urls":{},"keyword":["ECE780","Control"],"downloads":0},"search_terms":["motion","capture","based","human","motion","recognition","imitation","direct","marker","control","ott","lee","nakamura"],"keywords":["ece780","control"],"authorIDs":[],"dataSources":["iCsmKnycRmHPxmhBd"]}