October 13, 2014
The China Academy of Science (CAS) is poised to launch a payment app next year that would use facial recognition for authentication. When made available, users will be able to install the app and link it to their bank accounts or credit cards
The accuracy of the authentication feature can be attributed to the biggest Asian face database ever set up by the research institute comprising of more than 50 million records. The database was created jointly by University of Illinois and National University of Singapore. Researchers from the institute have developed a unique data collection technology which gathers facial information from 91 different angles. The system has scored the highest accuracy at 99.8%.
To address further security concerns, the institute has incorporated special algorithms that easily detect whether the scanned faces are real ones. This has been done to detect difference between a user’s video or photo instead of the actual live face. The authentication system is also embedded with online learning capabilities. This enables the system to automatically study the changing facial characteristics of the user.
Apart from the payment app, the institute is also planning to work with retail outlets to set up face recognition counters at checkout. This would allow customers to complete transactions at point-of-sale by simply authenticating themselves as a person. Biometrics, which include facial, fingerprint and voice authentication technologies, are ideal alternatives to traditional authentication technologies when using mobile devices. The idea of using biometrics for payment authentication has been made popular by Apple with the recent launch of Apple Pay.
China is not the first to develop a face-recognition system for enabling payments. Uniqul, a Finland based company, developed a similar technology back in 2013 using military-grade algorithms. In September 2013, Paypal started a new trial project that introduced a payment system which recognizes customers by their first name and picture via a mobile application. In July 2013, at a conference in London, Diebold unveiled their ATM that worked by connecting to smartphones with the help of QR Code or Facial Recognition technology instead of credit or debit cards.
The app has already been tested for border control situations, but the financial