Brown CS News

Practitioners in fields as diverse as education, leadership development, communication, and process improvement are likely to share one of the key elements of Stefanie Tellex's recent research: feedback is essential. Stefanie, a faculty member of Brown University's Department of Computer Science, has just won a Defense Advanced Research Projects Agency (DARPA) Young Faculty Award (YFA) to test her hypothesis that a robot can improve speed and accuracy at interpreting a person's speech and gesture requests by incrementally processing a person's language and gesture and by producing its own real-time language and gesture feedback, creating a social feedback loop between the human and the robot. 

"People from all walks of life can benefit from robotic help with physical tasks," Stefanie explains, "ranging from assisting a disabled veteran in his home by fetching objects to a captain coordinating with a robotic assistant on a search-and-rescue mission."

The objective of the YFA program is to identify and engage rising research stars in junior faculty positions at American academic institutions. It provides funding, mentoring, and industry and DoD contacts to awardees early in their careers, focusing on untenured faculty and emphasizing those without prior DARPA funding. The long-term goal of the YFA program is to develop the next generation of academic scientists, engineers and mathematicians.

"This project," Stefanie says, "will lead to robots that infer helpful actions more quickly and accurately compared to baseline methods, by incrementally interpreting a person's language and gesture, and actively gathering additional information. We aim to improve interaction speed up to 2x and increase accuracy up to 8x compared to baseline approaches that do not use feedback, consistent with gains in human performance in the presence of language and gesture feedback."