CHAMPAIGN, Ill., Sept. 25 (UPI) -- U.S. programmers say they have developed software that can fairly accurately estimate a person's age by examining his or her face.
"Age-estimation software is useful in applications where you don't need to specifically identify someone, such as a government employee, but would like to know their age," said University of Illinois Professor Thomas Huang, the study's leader.
Huang said age-recognition algorithms might be used to stop underage drinkers from entering bars, preventing minors from purchasing tobacco products from vending machines or denying children access to adult Web sites.
Consisting of three modules -- face detection, discriminative manifold learning and multiple linear regression -- the researchers' age-estimation software was trained on a database containing photos of 1,600 faces.
Huang said the program can estimate ages from 1 year to 93 years, with about 50 percent accuracy when estimating ages to within five years. The program achieves more than 80 percent accuracy when estimating ages to within 10 years.
"All of this can be done without violating anyone's privacy," Huang said. "Our software does not identify specific individuals. It just estimates their ages."
The researchers published their findings in IEEE Transactions on Multimedia and IEEE Transactions on Image Processing.
