Communications of the ACM
Computer Shows Human-Like Learning For Writing Characters
Credit: ABC
U.S. and Canadian researchers have developed a computer program that teaches a computer to learn to recognize handwritten characters such as letters of the alphabet after seeing just one example of each. That's something humans, even children, can easily do. But not computers.
The research could lead to computers that are much better at speech recognition — especially recognizing uncommon words — or classifying objects and behavior for businesses or the military.
"It has always been very difficult to build machines that required as little data as humans, especially when it comes down to learning a new concept that goes beyond simple recognition or classification tasks," says Ruslan Salakhutdinov, assistant professor of computer science at the University of Toronto, who describes the research in "Human-Level Concept Learning Through Probabilistic Program Induction," co-authored with Brenden M. Lake and Joshua B. Tenenbauman, and published in the journal Science.
"I believe that our computational model . . . takes a first step toward this one-shot learning ability," Salakhutdinov says.
From CBC News
View Full Article
No entries found