CyberLink Reaches ‘Top Tier’ in NIST Facial Recognition Testing

“The WILD 1E-4 test is designed to assess the effectiveness of face scanning technologies in retail and smart city environments as well as surveillance scenarios. “

CyberLink Reaches 'Top Tier' in NIST Facial Recognition Testing

Taiwan-based CyberLink is celebrating an impressive showing in the National Institute of Standards and Technology’s Face Recognition Vendor Test (FRVT) in the WILD 1E-4 benchmark category.

The company attained an accuracy rating of 97.02 percent with its FaceMe facial recognition engine, ranking 12th among the test participants and ranking CyberLink “among the top tier”, as the company put it in a press release. (The top-ranked vendor in this particular round of NIST testing was Hikvision, a China-based firm.)

The WILD 1E-4 test is designed to assess the effectiveness of face scanning technologies in retail and smart city environments as well as surveillance scenarios. It’s a challenging set of circumstances entailing the identification of faces that are not necessarily well-positioned for cameras, but CyberLink’s enthusiasm over ranking 12th speaks to the overall sophistication of state-of-the-art facial recognition technologies.

Commenting on the results in a statement, CyberLink CEO Jau Huang framed it as the latest example of FaceMe’s “outstanding results in major facial recognition competitions, including NIST and the MegaFace Challenge,” and emphasized the technical difficulties entailed in the WILD 1E-4 test. “The WILD 1E-4 test reflects real-world usage scenarios,” he said. “The results of this test allow system integrators and developers to identify the most accurate and reliable facial recognition technology to implement in smart surveillance and smart retail scenarios.”