What happens in the brain when we talk to a Furhat robot?
As robots become more and more human-like, they also need to exhibit more and more social behaviour. But what is it exactly that makes the social competence of a robot? And how can it be measured?
The interdisciplinary research group at Aix Marseille University in France is investigating these very questions by contrasting neural activity in the human brain in human-human conversations vs. human-robot conversations. They used functional Magnetic Resonance Imaging (fMRI) to record the brain activities of 21 participants while talking to a human in comparison to talking to a Furhat robot. The participants were unaware of the experiment’s real purpose and the Furhat robot was operated in a Wizard of Oz procedure. In addition to the fMRI data, even physiological (blood oxygen level-dependent, respiration and peripheral blood flow) and behavioural data (recorded speech from all interlocutors, eye tracking from the scanned participant, face recordings of the human and robot) have been collected.
Furhat’s human-like face can exhibit many social traits during a conversation with a human. This makes Furhat an invaluable tool to assess the social competence of a robot in the present fMRI-based research experiments.
The interdisciplinary research group at the Aix-Marseille University in France consists of the following three labs.
The Institute of Neuroscience of Timone (INT) is a world class research centre for fundamental and clinical approaches to neuroscience research. Thierry Chaminade and his colleagues are interested in understanding the neural substrates of social cognition using interactions of humans with computer animations and robots.
Researchers at the Laboratoire Parole et Langage (LPL) study the mechanisms involved in speech and language production, perception and understanding through a wide range of methodologies. Laurent Prévot is a professor in language sciences and the director of LPL. His recent research interests include conversational feedback, interpersonal dynamics in conversation as well as neural bases of conversation.
The Laboratoire d’informatique et systèmes (LIS) has a strong disciplinary identity and focuses on the fundamental and applied activities in the fields of computer science, automation, signal and image. Magalie Ochs is doing research on integrating social and emotional intelligence into interactive, human-like systems.