Interpersonal Robot Learning

Developing machines able to engage humans in rich and natural interpersonal interactions requires capturing social signals and their dynamics at the individual and interpersonal levels. In this talk, we will review basic concepts of interpersonal interaction and show how social signal processing techniques could be exploited to capture behavioral dynamics. Examples in psychopathology and social robotics will be detailed. In particular, we will show that interpersonal interaction models could improve robot learning in specific contexts such as imitation, joint attention or social guidance.