The Trends That Will Define Facial Recognition Tech in 2020 and Beyond

When the facial recognition tech ban in San Francisco came to light last year, it was not only good news for those that had fought tooth and nail for it. It was also a big win for everyone concerned about the use of technology as a surveillance tool to invade the privacy of the very citizens that the state has sworn to protect.

One of the places where the news was met with great acclaim is Canada. While many people look to the US as the place where many of these issues occur, it would be interesting to note that the largest police force in Canada – the ones that find their home in Toronto – have been using this tech for over a year.

However, a report on the matter also claimed that the police are not using the tech to make arrests alone, but also generating potential faces for investigation. This is only one of the many examples of controversies around this technology.

Looking forward, here are the major trends that will shape the use of this tech in the years to come.

Trend #1 – Security and Privacy

The supporters of this technology claim that it helps generate a better security profile for the community. Truth be told, a case can be made for this argument.

Between March and December 2018, it was confirmed that the police in Toronto got 60% matches for around 1500 faces run through the system. The best part of the news is that they were able to make arrests of criminal defenders with the information they got from facial recognition tech up to 80% of the time.

With that in tow, it can be argued that facial recognition will help make our societies a better place. The fact that they are installed could even deter some crimes from happening in the first place. The gray area here, though, is that the system cannot choose who to target and who to leave out.

Thus, just going to work, getting something from the grocery store, or other types of day-to-day activities are now being closely monitored by the government. 

Some may say that people should not be concerned if they don’t have anything to hide. However, no one should be stripped of their privacy in the first place. Finally, the fact that the system is used to trace people who could be potential criminals speaks a lot of how it could be abused too. Think of how minority groups can now become the subject of extreme targeting and surveillance, and more.

Trend #2 – Consumer Acceptance

While some forms of facial recognition might not be favorable, the rate at which the consumer-grade market accepts other forms will also determine the growth of the entire facial tech industry.

We have seen many facial recognition options deployed on personal levels such as passcodes on our phones. This offers a faster, contactless, and unique way of signing into devices and logging into accounts compared to other forms of biometric support.

The only downside to this is that the system might not be as secure as we want it to be.

For mobile-grade users, the Face ID technology developed by Apple is widely adjudged to be the most advanced facial recognition tech in the market. To back that up, the Cupertino-based company has stated that the chances of a wrong face fooling their system is about 1 in 1,000,000.

That is a high claim to make for a company whose tech that has been found wanting on this form of tech.

After all, a Chinese woman demanded a refund of her money when her iPhone was unlocked by another coworker. The funny thing about this issue was that Apple sent her a replacement unit, and the same thing happened again. 

Of course, she got a third replacement – and the company has now improved its facial recognition model. Still, the question remains of how secure it can get.

Truth be told, this is better than what we have seen with other providers. Some other phones can have their face unlock tech spoofed by a picture of the person, a twin, lookalike, or another family member. That is less than ideal for data privacy and security.

Trend #3 – Artificial Intelligence and Machine Learning

We cannot possibly talk about the future of facial recognition without bringing in artificial intelligence and machine learning. How these systems are used to better the tech could determine whether they could be allowed to operate more in the future.

The facial recognition system is trained with facial models from different people so that it can ‘learn’ how a face is unique from the other. Most times, this is determined by the geometrical measurements of the distance between the forehead and the chin, the length between both eyes and the likes. 

With this data, a facial signature is created. When a similar face is run by the system again, it matches this face against all other facial signatures in its database before it can throw a match.

From the basis of operation, it can be seen that the machine needs to be trained with a lot of different facial models so that it can be accurate to a strong degree. This is to avoid issues like false identification, which is unfortunately quite common within communities of color, especially women. 

That might not sound like much of a thing to fuss about till you think about the consequences of the system generating a false positive or false negative.

In the case of false positives, a facial tech could wrongly identify someone to be a person of interest to law enforcement. 

Depending on how much of a threat the actual crime has been deemed, this could lead to serious bodily harm, mortal danger, litigations, and more. That is not to mention the stigma that comes with being wrongly accused of a crime. 

False negatives can also be harmful.  

Think about being locked out of work just because the facial tech would not pick up your face. Think about being arrested for impersonating another traveler at the airport – just because the system does not pick up your face right. Think about all those possibilities, and you suddenly start seeing why AI and ML have to step up their game if they are to save the widespread usage of this tech.

Trend #4 – Regulations

Regulations will be the make or break for facial recognition systems from all around the world.

The ban of these systems in the San Francisco area is not an isolated event. Other areas like Hampshire, Oregon, and California have already kicked against the use of facial recognition tech in police body cameras. They are being followed closely by states such as New York and New Jersey that also do not want anything to do with this system at all.

What that tells us is that the spotlight is starting to shine on facial recognition systems. They have been left to run amok for several years now, but the tech has matured to gain public attention and the need for regulation.

This is not necessarily a bad time for the advocates and developers of these systems. Only if they can work within the confines of moral deployment, ensuring that they can get the public consent before they start fielding their facial data, can they make headway.

Final Words

The turn of the decade is going to be an interesting one in the battle for who allows facial recognition tech to stay, and who kicks it out onto the streets. 

The biggest concern for the people is how their facial data is being captured, stored, and used – all of which happens without their consent. Everyone also has a right to their privacy, and convenience should not come before that.