Combining Synchrony and shape detection to sustain the robot focus of attention on a selected human partner

Caroline Grand (1) , Ghiles Mostafaoui (1) , Syed Khursheed Hasnain (2) , Philippe Gaussier (3)

1 ETIS - UMR 8051 - Equipes Traitement de l'Information et Systèmes
2 Neurocybernetic team
3 Neurocybernetique

In our previous work we took inspiration from human psychological and neurological data which suggest that synchrony is an important parameter for human-human interaction. We proposed synchrony as a way of interacting and presented a synchrony-based architecture capable of selecting the human partner and of locating the focus of attention. We extend here our approach to people recognition to track the synchronized partner in the context of an autonomous mobile robot. If an agent interacts with a frequency close to the robot's dynamic our architecture selects this agent as an interacting partner and focus on him. Therefor, a shape recognition model learns the shape of the interactant. Combining synchrony and shape strategies, we obtained a bio-inspired robust and efficient architecture to automatically initiate the interaction with a selected partner and to locate and track this preferred partner in real indoor environment using kinematic (synchrony) and form (shape) pathways.

Domaines: Robotique [cs.RO]

Caroline Grand, Ghiles Mostafaoui, Syed Khursheed Hasnain, Philippe Gaussier. Combining Synchrony and shape detection to sustain the robot focus of attention on a selected human partner. IEEE Joint conference on Development and learning and on Epigenetic Robotics, Aug 2013, Osaka, Japan. ⟨hal-00834548⟩