Action understanding and imitation learning in a robot–human task.pdf

Action Understanding and Imitation Learning in a Robot-Human Task Wolfram Erlhagen1 1 2 3 , Albert Mukovskiy , Estela Bicho , Giorgio Panin , 3 3 4 4 Csaba Kiss , Alois Knoll , Hein van Schie , and Harold Bekkering 1 Dept. Mathematics for Science and Technology, University of Minho, 4800-058 Guimaraes, Portugal {wolfram.erlhagen, am}@mct.uminho.pt 2 Dept. Industrial Electronics, University of Minho, 4800-058 Guimaraes, Portugal estela.bicho@dei.uminho.pt 3 Informatics, Chair for Robotics and Embedded Systems, Technical University Munich, 85748 Garching, Germany {panin, knoll, kiss}@in.tum.de 4 Nijmegen Institute for Cognition and Information, Radboud University Nijmegen, 6500 HE Nijmegen, The Netherlands {H.vanschie, H.bekkering}@nici.ru.nl Abstract. We report results of an interdisciplinary project which aims at endowing a real robot system with the capacity for learning by goal- directed imitation. The control architecture is biologically inspired as it reflects recent experimental findings in action observation/execution studies. We test its functionality in variations of an imitation paradigm in which the artefact has to reproduce the observed or inferred end state of a grasping-placing sequence displayed by a human model. 1 Introduction In robotics research imitation has attracted a lot of attention in recent years since it is considered a promising learning mechanism to transfer knowledg