Dr. Saygin receives award from Calit2 to study human perception of artificial agents

‘What It Means to Be Human’

Saygin, an assistant professor who joined the cognitive science department last fall, will use functional magnetic resonance imaging (fMRI) to peer into the brains of test subjects as they interact with on-screen avatars. She is investigating “social cognition and interactive artificial agents.” Saygin hopes to provide a scientific basis for the “uncanny valley” – a phenomenon in robotics, computer graphics and animation.

As the theory goes, humans become more at ease interacting with avatars as they become more human-like. But when the avatars become too realistic (e.g., characters in the movie Polar Express), humans suddenly perceive the avatar as disturbing or negative – until the day when the anthropomorphism, thanks to improved computer graphics, becomes so complete that the brain returns to perceiving the avatar positively, as nearly human.

“We aim to improve our understanding of how the human brain enables social cognition, and to help engineers and designers develop interactive agents that are well-suited to their application domains as well as to the brains of their creators,” said Saygin, whose one-year project is budgeted at $54,000. “Thanks to the support from Calit2, we will be able to test how humans interact with a broad spectrum of avatars and to learn how the nervous system supports social cognition – an undeniably important skill and part of what it means to be human.”

And, according to Calit2 Associate Director Rajesh Gupta, who chaired the CSRO program committee, “understanding the critical man-machine interface is of crucial importance to better integration of emerging cyberinfrastructure to benefit society – a key aspect of Calit2’s charter.”

Reported by Doug Ramsey, Calit2

Read more at: