Communications of the ACM
Monkey Mustaches and Beards Help Algorithm Recognize Faces
Researchers at New York University have developed an algorithm that can correctly identify the guenon variety of monkey.
Credit: William Allen
New York University (NYU) researchers have developed an algorithm that can correctly identify colorful monkeys called guenons, which have mustaches, nose spots, and ear tufts, by their faces.
The researchers analyzed 541 photographs of 110 monkeys of 12 different guenon species, based on the eigenface technique, which helps computers recognize human faces.
The researchers say future scientists might be able to use algorithms such as this one to study which factors are important in evolution. However, they found the algorithm could not determine the monkeys' ages or sex from their faces, suggesting guenon faces have not evolved to convey this information.
"If communicating sex was a key aim of guenon faces, males and females should look different from their facial appearance, but for most species they don't," says NYU professor James Higham.
The algorithm works on the principle that variations among faces can be described numerically, with each individual face scored based on how it relates to a set of general faces.
"We sought to test a computer's ability to do something close to what a guenon viewing other guenons' faces would do," says study author William Allen, who completed the work while at NYU, but who is now a postdoctoral researcher at the University of Hull in the U.K.
From LiveScience
View Full Article
Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA
No entries found