De’Aira Bryant wants to change the lives of the people who most deserve it. A doctoral student in computer science and intelligent systems at Georgia Tech, her main focus has been exploring the potential for interactive communication between children and robots.
Bryant believes that significant improvements in emotion recognition must be made in order for robots to perform successfully in a variety of environments, especially with children.
A native of small-town Estill, South Carolina (population 2,000), Bryant didn’t know exactly what computer science was, even as an undergraduate. But she knew she wanted to “make websites look pretty,” so she signed up for a computer science course at the University of South Carolina.
During her first year, Bryant’s professor Jenay Beer, who holds an M.S. and Ph.D. in engineering psychology from Georgia Tech, invited her to work in a robotics lab as a research assistant.
Beer also suggested that Bryant sign up to do summer research at another university. This led her to apply to the Distributed Research Experience for Undergraduates (DREU) program, which matched her with Ayanna Howard, the Linda J. and Mark C. Smith Professor and chair of the School of Interactive Computing at Georgia Tech. And she discovered that she loved the work.
When Bryant started delving into artificial intelligence and machine learning, she realized that many of these systems didn’t perform well with children. Most systems relied on input solely from adults, who also generally express emotions differently from children.
“I want to cater what I’m making specifically for my audience so that I know when I place the robot in front of a child, it’s going to perform as it should,” Bryant explained.
The robots will need to function in places where engineers will not always be present, which means they will have to be adaptive. Bryant sees robots being utilized in places like physical therapy, schools, and hospitals — places that often need increased staffing to better serve children.
Improving emotion recognition in robotics begins with large data sets of annotated images. When armed with numerous annotated images referencing emotions, the robot is able to make a sophisticated guess about the emotional state of the child they are interacting with. Bryant is currently annotating hundreds of images, solely of children, to ensure her robots perform at an optimum level. Using cameras, the robots can focus on different points on the face to best determine how to analyze the subjects’ emotions.
“The projects I’m working on — using artificial intelligence, machine learning, and robotics to actually change the lives of people who deserve to have this technology — are very important to me. I’m glad I was able to find that here at Georgia Tech,” she said.
Bryant has a long way to go, but she’s confident in her progress. And she hopes to move from facial recognition alone to analyzing sound, body posture, and potentially incorporating cameras to track heat, smartwatches to determine heart rate, and other physiological measures to more accurately detect emotion.
Story by Evan Atkinson, Institute Communications