ct 17(13): e4

Research Article

Beating-Time Gestures Imitation Learning for Humanoid Robots

Download841 downloads
  • @ARTICLE{10.4108/eai.8-11-2017.153335,
        author={Denis Amelynck and Pieter-Jan Maes and Jean-Pierre Martens and Marc Leman},
        title={Beating-Time Gestures Imitation Learning for Humanoid Robots},
        journal={EAI Endorsed Transactions on Creative Technologies},
        keywords={programming by demonstration, cubic spline regression, dynamical time warping, beating-time gestures},
  • Denis Amelynck
    Pieter-Jan Maes
    Jean-Pierre Martens
    Marc Leman
    Year: 2017
    Beating-Time Gestures Imitation Learning for Humanoid Robots
    DOI: 10.4108/eai.8-11-2017.153335
Denis Amelynck1, Pieter-Jan Maes1,*, Jean-Pierre Martens2, Marc Leman1
  • 1: IPEM, Department of Art, Music and Theatre Sciences, Ghent University, Ghent, Belgium
  • 2: DSSP-ELIS, Department of Electronics and Information Systems, Ghent University, Ghent, Belgium
*Contact email: pieterjan.maes@UGent.be


Beating-time gestures are movement patterns of the hand swaying along with music, thereby indicating accented musical pulses. The spatiotemporal configuration of these patterns makes it diÿcult to analyse and model them. In this paper we present an innovative modelling approach that is based upon imitation learning or Programming by Demonstration (PbD). Our approach - based on Dirichlet Process Mixture Models, Hidden Markov Models, Dynamic Time Warping, and non-uniform cubic spline regression - is particularly innovative as it handles spatial and temporal variability by the generation of a generalised trajectory from a set of periodically repeated movements. Although not within the scope of our study, our procedures may be implemented for the sake of controlling movement behaviour of robots and avatar animations in response to music.