Explaining face biometrics applications and adoption
Face is a biometric modality in widespread use today for both recognition and authentication. A computer system takes a two-dimensional or three-dimensional digital photo or video of a person’s face, and converts it to mathematical data based on specific details and nuances, such as the distance between the eyes and shape of the jaw. That information is compared to a database of one or more individuals with algorithms to produce a match result that identifies or verifies a person’s claim to an identity.
Facial recognition and authentication have grown into an enormous business, with market reports estimating it will reach close to $13 billion annually by 2028.
Rapid gains in accuracy have been made since the 2013 with the application of convolutional neural networks, even with faces partially occluded, such as by masks. Face biometrics systems are now capable of matching high-quality images within a fraction of a percent for one-to-one or one-to-many match scenarios.
Private and public sector organizations have employed face biometrics for various purposes ranging from unlocking consumer electronics to law enforcement and security. Consumer electronics like smartphones and laptops offer the option of facial authentication to open devices through their cameras, like Apple’s Face ID or Microsoft’s Windows Hello. Airports turn to facial recognition to expedite the boarding process by linking a face to a passport and ticket, thus eliminating traditional barriers that cause annoyance and delay. Its ease of use is explored by stores and automobile manufacturers to improve a customer’s shopping experience and start up a vehicle or monitor a driver, respectively.
The technology’s application in security is widespread in countries across the globe as it eases the identification process. It is used in closed-circuit television (CCTV) networks to identify perpetrators of crimes, compared against databases by law enforcement agencies to narrow down the identity of criminals and suspects, and in border crossings, immigration centers and airports to find out if the person should be denied the right to enter or travel. Retail stores implement facial recognition to recognize previously-identified shoplifters.
The fastest-growing type of face biometrics application is digital identity verification based on a match between an image submitted by a user and a reference image from a validated photo ID document, usually government-issued. Face biometrics based on selfies taken with mobile devices for identity verification, in combination with anti-spoofing technology like liveness detection, is now common in the financial services and online service sectors in many places as a way of satisfying know your customer (KYC) and anti-money laundering (AML) regulatory requirements.
Despite its benefits, facial recognition deployed in public and semi-public environments attracts a chorus of criticism from civil and human rights activists. Advocates say the technology is a threat to the right to privacy and freedom of expression, which is abused by governments to repress and control large numbers of people like the Muslim Uyghur minority in China. Opposition also takes aim at alleged failings in accountability, the inability for a person to freely consent to being surveilled and identified, and the introduction of bias by imperfect algorithms that deliver less accurate results for women or people of color. Critics question the efficacy of facial recognition in law enforcement and argue it can implicate the innocent in a crime. The source of images for training and reference databases has also become a point of common concern.
Facial analysis, in which biometrics are used to identify characteristics, rather than an individual, is also deployed for a variety of different tasks, prompting separate but related debates.
The ongoing debate has led to large companies halting facial recognition operations, and Meta, Facebook’s parent company, which decided to cancel the social media giant’s facial recognition tools deleting the biometric templates of over one billion users. Cities and states in the U.S. have signed facial recognition bans or restrictions into law and the European Union has proposed a ban.
Click here for more explainers on concepts in the field of biometrics.