World's most accurate facial recognition system developed in HK
Updated: 2014-08-08 09:40
A research team at Chinese University of Hong Kong (CUHK) has developed an automated facial recognition system that achieved the highest accuracy rate in the world, according to Hong Kong China News Agency.
The system reached an accuracy rate of 99.15 percent in recognizing faces in the Labeled Faces in the Wild, a database of more than 13,000 photographs designed for studying the problem of unconstrained face recognition.
The accuracy rate also set a record in a computed system being more accurate than human sight.
Professor Wang Xiaogang, one of the team members, said the hardest part for automated recognition was to deflate the variety of one person in different pictures and magnify the tiny differences among people. They have coped with the two detections effectively and enhanced the accuracy, Wang added.
Wang also said the improved system can help law enforcement and security services recognize targets in massive crowds.