SPECTRUMRussellDRAFT
Visionary research
Growing up in Phoenix, Bryan Christopher Russell sometimes day-dreamed about getting help with chores: “I thought it would be cool to have a gadget with robot arms that could do the dishes or something,” he says.
Of course, there were no such machines, and still aren’t. But unlike his situation then, today Russell is in a position to help set the stage for that and other useful devices.
A grad student in MIT’s Computer Science and Artificial Intelligence Laboratory, Russell works on a capability vital to any truly versatile robot: vision. Specifically, he’s trying to craft artificial intelligence systems that meet what is, for them, a tough challenge, which is look at things, or images of things, and categorize them.
“If you ‘show’ a bunch of images to your system,” he says, “you want it to be able to characterize objects in those images — for example, chairs — without being told what a chair is or where to look for it.”
The field, he notes, is progressing. Today’s best AI systems can “look at,” and over time, learn to lump together objects like chairs or cars.
But even these systems don’t always work, especially if the objects are lit in an unconventional way or are partly obscured. And some living objects, particularly those with natural camouflage — leopards, tigers, zebras — all-but stymie the systems.
Rusell’s focus right now is enabling his group’s system to do something humans do without thinking about it: figure out the location of one part of an object like a chair in relation to the others.
“Current systems work on the assumption that each feature in an object is independent, so they don’t take spatial relations into account,” says Russell. “We’re trying to work on that.”
The student, a Dartmouth graduate, says attacking that problem will probably be the focus of his doctoral thesis. His plans after he gets his degree, meanwhile, are to be determined but will definitely involve teaching as well as research. “I like connecting with students,” he notes. “When they have that ‘Aha!’ moment, it’s a thrill.”
Russell says he’s grateful to have launched his MIT career with the award of a Norman B. Leventhal Presidential Fellowship — a program created in 2001 in honor of a leading MIT graduate by Jeffrey and Barbara Picower. The fellowship, he says, gave him flexibility to explore a range of interests in his first year.
He also says that after four years at MIT, he’s gained new insights into some of the Institute’s strengths. “What impresses me most,” he says, “is how passionate people are about their work.”