Curated by UCL

BLOG: New, powerful face recognition could mean checking out with just your face

Imagine just walking out of a store with items that you’ve paid for – without having to queue, wait for your card to be accepted, or count individual notes and coins. Developments in the technology behind facial recognition that make it faster and more secure might make this pipe dream a reality.

Currently, among British consumers, card payment is the most common way of paying.[1] Although cards have proven to be incredibly useful and efficient, they are notoriously easy for thieves to steal. According to Financial Fraud Action UK, over 230,000 cards were stolen in 2016,[2] with that number predicted to rise every year. The financial loss caused by this is devastating: over £96 million. In addition to their insecurity, cards can take a long time to process, making queues at stores unbearably long.[3] Furthermore, the inconvenience of having to carry so many different cards has become a cause for concern over time.

To make checkout faster, banks have introduced contactless cards, which have soared in popularity since they were first issued. The number of transactions made by contactless cards has more than doubled over the past year,[4] proving high adoption rates and ease of access. However, fraud as an issue has grown even more, with contactless credit card fraud reaching £7 million in 2016.[5] It has also done nothing to prevent card theft, which remains a prominent issue.

Seeking to provide the world with a new, radically different way to pay, the founder of Chinese company Face++, Qin Yi[6] set out to create a facial recognition system so secure that it would be able to authorise transactions based on your face. In 2017, the company unveiled Face++ Cognitive Services – which realised the vision of its founder. It advertises itself as “an all new platform, with our prestigious computer-vision algorithms at the core, even faster, simpler, and better” on its website,[7] with large and powerful clients backing their claim.

The technology works by comparing a scan of the user’s face to an existing database of faces. This compare returns a similarity result with a certain level of confidence that the two faces are the same.[8] Face++ also overcomes the traditional problem of scanners being fooled by 2D photographs by scanning the face at different angles to create a 3D model.[9] The comparing algorithm is the system’s secret: with competitors appearing far and wide in China, including search behemoth Baidu,[10] the compare algorithm is what’s giving Face++ the cutting-edge, and allowing it to retain its users. One client, AliPay,[11] has implemented the technology to allow customers at certain restaurants to pay just using their face. This shows not only that the technology is feasible, but also that there are customers willing to use it.

As of now, China is at the forefront of development, thanks to its lax privacy laws and more embracing consumer culture.[12] A call for a similar technology in the past, led by American retail giant Walmart, was attacked for being too invasive. “Facial recognition and biometric data is highly sensitive and personal and Walmart is significantly over-reaching to use it,” said Mark Ryski, CEO of HeadCount Corporation, a data analytics firm.[13] For the face recognition to be fully embraced and implemented globally, cultural opposition and hurdles need to be overcome.

In spite of cultural differences, western companies are slowly embracing the different biometric securities behind the technology of their Chinese competitors: with industry leaders such as Microsoft, with their Hello service;[14] and Apple, with their FaceID[15] service, pushing facial recognition onto international consumers. Masses of consumers are now adopting facial recognition as a means of security, and the reluctance to trust the technology — from both corporations and consumers — is slowly fading. In a historic move, Lloyd’s Bank, in partnership with Microsoft, is becoming the first British bank to use facial recognition to tighten bank security,[16] proving the technology’s reliability.

Looking ahead, it can be expected that more and more of our devices will support facial recognition as a means of security, and that it will be more widely accepted as a means to authenticate payments. Security companies around the world, including Cubic Transportation Systems (the company behind the famous Oyster card) have already begun trialing similar works.

This new technology is equally important for the government. In one application example, the technology could be used to automatically detect the faces of wanted criminals caught on CCTV and report their position. With so much money being poured into the development of this technology by massive international firms, the implementation of this technology globally is almost inevitable, despite privacy risks and possible security concerns.

References

  1. Cash no longer king as contactless payments soar in UK stores, The Guardian, 2017.
  2. Financial Fraud Action UK, Fraud the Facts 2017, Financial Fraud Action UK, 2017.
  3. British people spend a STAGGERING amount of time queuing up in their lifetime, The Mirror, 2017.
  4. Contactless Payments - Key Statistics & Figures, The UK Cards Association, 2017.
  5. Contactless payment fraud soars, BBC News, 2017.
  6. Megvii Technology Inc.: Private Company Information, Bloomberg.
  7. What is Face++?, Face++.
  8. Face Comparing, Face++.
  9. H. Bejaoui, H. Ghazouani, and W. Barhoumi, Fully automated facial expression recognition using 3D morphable model and mesh-local binary pattern, ACIVS, 2017.
  10. How China is quickly embracing facial recognition, Forbes, 2017.
  11. Alibaba debuts ‘smile to pay’ facial recognition payments at KFC in China, Tech Crunch, 2017.
  12. How China's facial recognition technology is changing daily life, ITV News, 2017.
  13. Walmart's facial recognition tech would overstep boundaries, Forbes, 2017.
  14. Windows Hello, Microsoft.
  15. iPhone X, Apple.
  16. Lloyds says ‘Hello’ to facial recognition banking, Financial Times, 2017.

About the Author

Overenthusiastic computer scientist, with an undying passion for writing and technology.

Articles

Gold rush for lignin conversion
Gilding with graphene
Putting quantum noise to work