Recognizing and remembering individuals: online and unsupervised face recognition for humanoid robot

Individual recognition is a widely reported phenomenon in the animal world, where it contributes to successful maternal interaction, parental care, group breeding, cooperation, mate choice, etc. This work addresses the question of how one may implement such social competence in a humanoid robot. We argue that the robot must be able to recognize people and learn about their various characteristics through embodied social interaction. Thus, we proposed an initial implementation of an online and unsupervised face recognition system for Kismet, our sociable robotic platform. We show how specific features of this particular application drove our decision and implementation process, challenged by the difficulty of the face recognition problem, which has so far been explored in the supervised manner. Experimental results are reported to illustrate what was solved and the lessons learned from the current implementation.