Top 5 Trends in Facial Recognition That Will Dominate in 2020

By Junaid Dar•Updated on May 5, 2020

Facial recognition technologies – the biometric ability to identify and verify a human face from a digital image – are being rapidly deployed across a diverse range of business industries. Notwithstanding the obvious security benefits, facial recognition technologies’ ever improving accuracy, speed of response and ubiquitous rollout has provoked a degree of unease in some quarters – especially in terms of consumer privacy.

Facial recognition technologies, as the name itself suggests, is a process whereby an individual can be identified by capturing, analysing and comparing patterns on the individual’s face. It’s a specific kind of image recognition service that consists of the following:

• Face detection – identifying and locating human faces in videos and images
• Face capture – transforming the analogue information of a face into data, according to the facial features (i.e. spacing of the eyes, contour of the lips, chin, bridge of the nose)
• Face match – verifying whether or not two faces belong to the same individual

The established global technology giants are vying for the #1 spot: Google, Apple, Amazon, Facebook and Microsoft are the recognised key players in this biometric innovation.

• In 2014, Facebook launched DeepFace – a program that can determine with an accuracy rate of 97.25% whether two faces belong to the same person
• In 2015, Google launched FaceNet – an accurate facial recognition system that is used by Google Photos to identify, sort, and tag people in pictures
• In 2018, Ars Technica highlighted that Amazon’s facial recognition software ‘Rekognition’ can identify up to one-hundred people in a single image and quickly cross-check with databases to deliver precise results
• To understand better where the technology is heading, we’ve looked at the top 5 trends in the future of facial recognition services for 2020.

According to a recent research report on the facial recognition market by Markets and Markets™, the global facial recognition market which in 2019 was worth $3.2 billion is expected to grow to $7 billion by 2024 at an annual compounded growth rate of 16.6%. The key drivers of this projected phenomenal growth are an ever-increasing user base, evolving government security strategies, and the increasing need for fraud detection and mobile device proliferation.

Where once facial recognition software was a technology advancement used only by militaries and intelligence agencies, the wider spread of facial recognition – and the availability of free software with facial recognition capabilities – means that it’s use by the general public is bound to transform how we understand everyday transactions.

• Health

Facial recognition has an essential role to play in the health sector. Did you know that facial recognition systems can help to accurately track the use of medication by parents, support pain management procedures as well as help detect genetic diseases?

Especially in the field of public health, this ability to track patients and their routines becomes ever more relevant – especially if they’re carriers of contagious diseases. That’s why for epidemiologists, the wider spread of facial recognition systems may eventually help them contain the threat of an epidemic. Although still in its infancy, the technology is set to rapidly transform the health sector over the coming decade.

• Marketing and Retail

If shopping experiences can be tailored to suit individual customers’ needs more readily, consumers – and companies – can benefit. Amazon and Facebook are 2 of the largest companies using facial recognition technology in retail today.

Through the use of facial recognition systems, these companies use collected data to pitch relevant products and services as well as analyse shopping behaviour in real time to make their offerings ever more customer-centric.

This particular ability of facial recognition systems is bolstered by advances in technology. Most prominently, the ability of these systems to also be capable of identifying features like age group or gender, just from a picture of a face, makes targeted marketing ever easier to achieve. For instance, facial recognition systems in vending machines can share with providers what products are selling – and who’s buying them.

What’s more astonishing is the concept of selfie payment. Since 2017, KFC – the American fast food company – and Alibaba – the Chinese multi-national conglomerate – have been testing a facial recognition system for self-payments.

This means that rather than credit or debit cards, financial transactions may just be validated through facial recognition systems – imagine walking into any ATM, and being identified as an account holder with just your own face.

In previous years, the advent of card-based transactions made those optimistic about the future of finance and banking believe that in their lifetimes, cash-based transactions would eventually become obsolete. Now, the possibility is not just of a cashless society, but one without cards too. Face ID might just be the requirement needed to validate a financial transaction.

Artificial intelligence and deep learning – a specific branch of machine learning whereby a system continuously learns from data to improve incrementally – together make up two of the most relevant emerging technologies for the future of facial recognition systems.

These technologies have been pivotal to the growth of facial recognition in 2020, and are essential for facial tracking, facial detection and facial matching. As such, significant improvements in facial recognition systems are predicted in the near future.

A 2018 report by NIST highlights that in the 5 year period leading up to the report, there was a considerable increase in the accuracy of facial recognition systems that was substantially higher than the improvements made in the period 2010-2013.

The next 2-3 years, the report goes on to say, will be even more important for the development of facial recognition technologies through advancements made on the back of artificial neural network algorithms – networks that after a learning phase become capable of giving a correct output value (or Result) after processing various input values.

Given that deep learning also entails these recognition systems will be programmed to improve with more and more exposure and experience, rapid improvements in the way these systems work is likely.

China and India are experiencing rapid growth in the use of facial recognition technologies. Today, the U.S is home to the biggest facial recognition applications market.

According to Reuters in 2018, security officials in Beijing, China tested smart glasses that used facial recognition technology to identify suspects in real-time. Building on these findings, the New York Times reports that China is working with a number of revolutionarily A.I companies including SenseTime, Yitu, and CloudWalk to set up and perfect a facial recognition camera and video surveillance system nationwide.

In India, the Adhaar project – the biggest biometric database in the world – is now being further upgraded as the feature of facial authentication is included. This could mean that close to 1.2 billion people- the entire population of India- will be facially recognised by systems, devices and software.

Unfortunately, this dimension of facial recognition systems has been subject to severe criticism, especially in the ways governments around the world have used such systems to persecute citizens within their countries. At most risk are undocumented migrants, or even refugees – as few nations have opted to naturalize people who enter their countries on a “refugee” status, this would mean governments now have more power to keep checks on who is and who isn’t a national of their countries. This aspect of facial recognition may also empower governments to pursue ethnically discriminatory policies, although as of yet, only China seems to have used facial recognition to such unusual extremes.

While facial recognition-based logins and facial recognition for online services are becoming well established, the use of facial recognition for detecting and in turn, preventing crime is regarded as the most important application of this emerging technology.

From being used at borders and high-risk locations such as government buildings, airports and nuclear power plants to fast-becoming the integral component of the security detail of commonplace buildings everywhere be it a local business, multi-national supermarket or library, facial recognition is a recognised deterrent designed to help boost security. Even in office buildings that require clearance for certain sections, facial recognition systems are being introduced to further protect corporate secrets.

Through facial recognition, law enforcement agencies can recognise individuals with a past criminal record as well as identify those looking to engage in suspicious activities that could lead to an unwanted security breach. This in turn allows organisations to speedily undertake necessary actions to effectively protect the safety and security of its people as well as its physical assets.

In the United States, schools are mulling using facial recognition services to alert authorities whenever expelled students or known criminals enter such facilities. Agencies all over the world have also started using such software to track people who are reported “missing”.

With the combined effectiveness of deep machine learning and the expansion of facial recognition services for the market in general, this kind of software is primed to also function within social media and the internet as a “moderating” force.

This would mean that with facial recognition software, webpages can quickly catch images that are violent or inappropriate on their pages. Besides just ensuring that the internet remains a safer place for everyone, it also acts as an inhibitor for those that wish to use social media accounts to incite violent or dangerous activity.

As one social media giant, Tumblr, has already been forced to remove all adult content from their page because of the abundance of child pornography on the website, facial recognition ensures such circumstances are less likely to arise in the future. That’s also a testament to the power of face ID within these image recognition services, with the capability of determining the “graphic” nature of photos and developing an estimate of both the age and gender of the subjects photographed.

With the evolution of facial recognition technologies and their diffusion throughout a wider stratum of society, such systems are slated to become a common aspect of everyday life. Science-fiction films may have been the first to pioneer their usage as verification devices, but now that even smartphones have built-in facial recognition capabilities, every face is effectively also a source of personal data for all future machines.

It’s fair to ponder over how these systems will compromise people’s privacy. If databases owned by corporations, governments, and professional organizations all carry information about people’s facial information, much more information about the average individual would be recorded. Since there’s little precedent of the amount of consumer privacy that would become part of the market, digital rights activists and their concerns are well-grounded.

What’s also worth remembering is that individuals already upload vast amounts of personal data to social media accounts. It’s data, that like facial features, would have once been conceived as private information, but is now considered vital by businesses in their attempts to target specific kinds of audiences. Already in the way the world is structured, consumer privacy is increasingly being undermined as a concept.

In the face of the huge security and law enforcement advantages facial recognition can privilege its users with, its threats to consumer privacy are difficult to use compellingly against their usage.

It’s difficult to surmise whether this is ultimately for the worse, but the advantages that these kinds of services can result in are very real. Only the future will tell if this kind of technology will ultimately empower people and businesses.