A little more than a year ago, after several unsuccessful attempts, Craig Federighi, vice-president of Apple’s software engineering branch, revealed to the public that the company had managed to create a facial recognition software that would let you unlock your cell phone simply by looking at it. Hailed as a revolution, the technological advance is now deeply embedded in the lives of a vast number of users. Naturally, its use has become widespread, graduating beyond the universe of smartphones.
Generalizing facial recognition
But let’s return to Apple’s invention. Called Face ID, the biometric technology integrated into the iPhone X depends on a group of sensors using the technology True Depth, as Futura Sciences explains. A projector in front of the phone shoots more than 30,000 invisible points onto a user’s face. They’re analyzed by a camera and an infrared light, guaranteeing it to work even in dim light. True Depth makes a 3D model of the person’s face, which is then analyzed by the smartphone using a processor for apps using artificial intelligence. This A11 Bionic processor then compares the generated image to a reference image, digitalized beforehand by the user.
Integrating 3D allows one to avoid hijacking the facial recognition system using a simple photo of the smartphone’s owner. As another element guaranteeing further security, a user has to look at his or her telephone and keep their eyes open during the duration of the milliseconds long process. The app was even designed to take into account eventual changes in appearance (glasses, makeup).
From the start, this technology was not only intended to protect the smartphone by guaranteeing that its user would be the only person able to unlock it. Apple gave it the ability, for example, to validate and secure purchases via Apple Pay, but also to connect to apps with facial recognition. The ease of using the system has seduced most users and convinced designers to integrate facial recognition into their own products. Take for example Google’s Pixel 2, Samsung’s Galaxy Note 9 and Motorola’s Moto G6. At the same time, many companies and labs are working on developing new uses for the technology, although its reliability has sometimes been called into question, as Wired reports.
Potential breach of privacy
In an article on the future of facial recognition, Wired highlights some of the possible future uses of the technology. Clare Garvie, a lawyer specializing in privacy issues, explains that we should be wary of some of the uses of facial recognition, “like surveillance by law enforcement.” She points to the risk of habituation and addiction to facial recognition technologies, leading to accepted widespread use. Even if this addiction isn’t necessarily harmful with the iPhone X, it might not be so harmless in other contexts, she warns.
The police, for example, could use it for public security ends. This particular element is, according to Clare Garvive, the worst-case scenario. A recent example reported by The Verge more or less justifies this fear. In the US, police departments in Orlando, Florida and in Oregon’s Washington County have already tested a pilot program for facial recognition in order to identify suspects in photos and videos. The program, called Rekognition, is provided by Amazon and is intended to search images of individuals in a database of mugshots. Many – starting with America’s most powerful civil liberties organization, the ACLU, but also Amazon employees themselves – have denounced the risks that come with the technology, including those of mass surveillance and identification of protesters.
Adding to these concerns, “hiccups” led to a premature cessation of the pilot program in Orlando. Les Echos reports that the system is still imprecise, particularly with regard to black people. Still, Clare Garive believes that people today simply haven’t become habituated enough to the technology for these kinds of uses to be comfortably accepted.
A technology to save lives?
Police use of facial recognition is far from its only future. Other uses, also intended for security, have been put forth, inciting the same debates. The telecommunications company RealNetworks developed a tool allowing schools to couple their video surveillance systems with a facial recognition system to prevent possible break-ins. Called SAFR, it aims to prevent mass murders in American school buildings. Detractors of the technology fear certain inherent potentials, such as the accentuation of racial discrimination. This argument is supported by a study at the Electronic Frontier Foundation.
The health sector could also benefit from facial recognition to better protect its patients. The digital security company Gemalto explains in a study of the technology how facial recognition can make it possible to track a patient’s drug intake. Another possible use is the detection of genetic illnesses that have physical consequences, such as DiGeorge syndrome (in this case, the technology is 96.6% accurate).
Beyond health and security, there are also possible uses for facial recognition in commercial sectors, which could explode in the next few years. Many companies understand that there are huge benefits to it. The San Francisco-based startup Zippin offers a visual recognition system (not facial, for the moment) letting people create cashless stores. In an article, the site Usine Digitale explains the technology: customers scan a QR code as they enter the store, then are identified and tracked by a network of cameras that detect which products they buy. Once their errands are done, customers leave the store and are automatically billed by an app. The concept has also been developed by Amazon (Amazon Go), Microsoft in association with Walmart, and several startups (Aipoloy, AiFi, Trigo Vision and Standard Cognition). The goal is to make it easier for customers to buy goods, particularly in downtown and densely populated areas.
As usual, this technology has generated plenty of suspicion. Despite the fact that it tends to make people’s lives easier – be they customers, patients, doctors or law enforcement – some are still reticent to accept widespread use of facial recognition. Civil liberties organizations and certain politicians worry about the potential and eventual uses and inaccuracy of it. Others, particularly businesses, see a means for better fulfilling their missions, all while increasing profits.