Introduction

They say a picture is worth a thousand words, but in the tech world, a single picture can be worth more than that. Thanks to facial recognition technology, an image of your face can be  used for everything from law enforcement to access control. Yes, facial recognition has finally come of age, and companies and governments across the world are not just exploring but implementing the capabilities that facial-recognition systems offer.

Facial recognition can be viewed as the ultimate biometric: after all, your face is a crucial representation of you as a person, and it makes both psychological and technical sense to use it in matters of identity. Facial recognition systems combine a number of variables based on facial characteristics to determine the identity of an individual. And facial recognition is big business, too, with the market worth expected to be around $7.76 billion by 2022.

But facial-recognition technology is no stranger to controversy. Recently, the Amazon facial-recognition platform, Rekognition, was found to have incorrectly identified 28 Congress members as known criminals. And facial recognition, by definition, tends to raise privacy concerns: to some, it smacks of the totalitarian “Big Brother” from the book 1984. This sort of view is strengthened by research from Georgetown University, which found that one in two American adults is included in a law enforcement facial-recognition system.

In this article, I’ll take a look at the sort of facial-recognition applications we are seeing that are a positive force in the world of technology. Conversely, I’ll look at why we perhaps need to be cautious when we use powerful and intimate technologies like facial recognition.

Where is Facial Recognition Being Used?

Facial-recognition technology is being used as an access-control measure for a number of services. This includes banking access, with large banks such as HSBC integrating Apple’s iPhoneX Face ID system with their banking app. Facial recognition is also getting state backing across the world for tasks such as law enforcement and personalization of services. Some examples of facial recognition technology being implemented on a mass-scale include:

China

The Chinese government is adding facial recognition alongside artificial intelligence (AI) to over 170 million CCTV cameras across the country. AI will be used to match captured faces to those held on police databases. Examples of the application of the system are already being highlighted as successes. For example, at a festive market in Haizhu, two suspects were quickly identified using the AI/facial recognition system. Around 4000 people have been arrested using the technology since 2016.

United Kingdom

Several U.K. police forces are rolling out facial-recognition pilots. The technology was trialed at the Notting Hill Festival this year. The move has been highly controversial, and a UK newspaper has revealed the technology has a 98% false positive result, although this is disputed by the police.

United States

The Department of Homeland Security is to implement a facial recognition system on the Mexico border from August 2018. The system, known as the Vehicle Face System (VFS) will use AI and cross-reference every person crossing the border with government databases. The system is able to recognize faces even behind car windscreens.

Russia

As well as using facial recognition for law enforcement, Russia is offering facial recognition to travelers for personalization of journeys. The technology offered by the company, VisionLabs, will be offered to business-class passengers to improve their journeys through airports and to check into hotels more quickly.

The Problems with Facial Recognition

There are several useful aspects to facial recognition technology, especially when coupled with artificial intelligence, but nothing is without drawbacks. So where do the issues with facial recognition lie?

False Results

UCLA conducted a test of Amazon’s Rekognition platform. They found that the system was prone to high false match rates, particularly for people of color. The survey concluded: “Nearly 40 percent of Rekognition’s false matches in our test were of people of color, even though they make up only 20 percent of Congress.”

Privacy Issues

Big Brother really is watching when you apply facial recognition to a problem. There are a number of issues that concern the privacy of such intimate biometrics. The obvious being that digital assistants, like Amazon Echo, now offers camera versions which always watch you to build up behavioral profiles to “better serve you.” In the era of constant privacy breaches and identity thefts, the idea isn’t exactly reassuring.

The privacy watchdog group epic.org have looked at a number of privacy issues around implementation of facial recognition. One such case is the collation of a mass biometric database by the FBI. Epic is arguing that there are multiple errors in the database and issues around accuracy, relevance and transparency which may lead to data exposure.

Racist Facial Recognition?

The Apple iPhoneX applies facial recognition as a biometric for phone access. The system, known as Face ID, was slammed by a number of Chinese users as being “racist” and was said to have been tested using “white faces only.” One woman was offered a refund when her phone was accessed by a work colleague using the facial recognition system.

Facing the Future

When we are presented with technology like facial recognition which has both positive and negative aspects to it, we have to weigh these up carefully and decide if we truly need to have this in our lives. The movement towards using facial recognition alongside artificial intelligence is compelling, especially for law enforcement. It is also very useful for solving the password fatigue problem, giving us a neat way to login to banking and other services.

But one of the issues we have is that it is also open to abuse. The same application in law enforcement could, under an authoritarian government, be used to control activists. And facial recognition can, and is, being used for social control – a recent use of the technology in a school in China highlights this, with children being rated on attentiveness based on real-time feedback by the AI/facial-recognition system.

Human beings love tools, but we need to use them with caution and under advice. Hopefully, this new world that facial recognition is creating will not become a dystopian future.

 

Sources

Facial Recognition Market Worth $7.76 Billion USD by 2022, Markets and Markets

Unregulated Police Face Recognition in America, Georgetown Law

China’s all-seeing social control network brings an end to fugitives’ festive fun, South China Morning Post

Metropolitan Police’s facial recognition technology 98% inaccurate, figures show, The Independent

DHS to deploy facial recognition AI at US/Mexico border, The Next Web

VisionLabs face recognition technology introduced to hotel check-in systems, Skolkovo

Amazon’s Face Recognition Falsely Matched 28 Members of Congress With Mugshots, UCLA

Spotlight on Surveillance – December 2013, EPIC

Chinese woman offered refund after facial recognition allows colleague to unlock iPhone X, South China Morning Post

Facial-recognition technology used to monitor student engagement in Chinese school, IAPP