Can facial recognition work when people wear masks?

Facial recognition technology is increasingly becoming part of our everyday lives. You might not think you use it, but the chances are you probably do. Ask yourself this – how do you unlock your smartphone? And when travelling abroad, do you use the automated border control gates at airports?

But with millions of us having to wear masks in public at the moment, is this technology effectively redundant? After all, you can’t scan what you can’t see.

Well, not quite. Facial recognition tech is evolving at such a rapid pace that some companies have even worked out how to scan your features when only around half of your face is showing.

We dig more into this here.

 

So how does facial recognition work with masks?

Facial recognition software studies facial features such as the eyes, nose, mouth and ears to identify the individual. A face mask, which typically covers the nose and mouth naturally makes this trickier.

So what experts have taught the software to do is to firstly understand whether the person being scanned is wearing a mask, and if so, will in effect ‘crop’ the area of the face being identified – making a judgement based on that, rather than assuming the mask is part of the face.

 

Why is this important?

Being able to understand when people are wearing masks is important as countries decide on their citizens’ compliance with laws around wearing facial coverings.

Teaching facial recognition software to decide whether someone is wearing a mask can give decision-makers valuable data about how many people are actually complying with the guidance to do so. They can then decide whether stricter measures are needed to come in to promote the wearing of masks and thus limit the spread of the virus.

It’s worth noting that in this case, the data is anonymised – so while information used in this way can give policymakers trends, it can’t be used to pinpoint individuals.

 

What’s next for facial recognition in a masked world?

The technology behind FR is evolving all the time. While everyday uses of facial recognition still have a way to go before the can seamlessly identify you with a mask on, companies are now adapting their models to understand that masks are likely to become a default part of our day-to-day lives.

Last year, Apple released an update in iOS 13 that would automatically default to the passcode when it recognised a mask was being worn. The Department of Homeland Security (DHS) in the United States is also already piloting facial recognition technologies that can see through masks with a “promising” level of accuracy.

Interested to find out more about facial recognition and how it is evolving? We have posted a few more articles on the topic below: