It’s surprising how quickly public opinion can change. Winding the clocks back 12 months, a lot of us would have checked out a masked individual in public with suspicion.
Now, some countries have enshrined face mask use in law. They’ve also been made compulsory in Victoria and are advisable in several other states.
One consequence of that is that facial recognition systems in place for security and crime prevention may now not find a way to fulfil their purpose. In Australia, most agencies are silent in regards to the use of facial recognition.
But documents leaked earlier this yr revealed Australian Federal Police and state police in Queensland, Victoria and South Australia all use Clearview AI, a industrial facial recognition platform. New South Wales police also admitted using a biometrics tool called PhotoTrac.
What is facial recognition?
Facial recognition involves using computing to discover human faces in images or videos, after which measuring specific facial characteristics. This can include the gap between eyes, and the relative positions of the nose, chin and mouth.
This information is combined to create a facial signature, or profile. When used for individual recognition – equivalent to to unlock your phone – a picture from the camera is in comparison with a recorded profile. This strategy of facial “verification” is comparatively easy.
However, when facial recognition is used to discover faces in a crowd, it requires a major database of profiles against which to match the essential image.
These profiles might be legally collected by enrolling large numbers of users into systems. But they’re sometimes collected through covert means.
teguhjatipras/pixabay
The problem with face masks
As facial signatures are based on mathematical models of the relative positions of facial expression, anything that reduced the visibility of key characteristics (equivalent to the nose, mouth and chin) interferes with facial recognition.
There are already some ways to evade or interfere with facial recognition technologies. Some of those evolved from techniques designed to evade number plate recognition systems.
Although the coronavirus pandemic has escalated concerns across the evasion of facial recognition systems, leaked US documents show these discussions happening back in 2018 and 2019, too.
And while the talk on the use and legality of facial recognition continues, the main focus has recently shifted to the challenges presented by mask-wearing in public.
On this front, the US National Institute of Standards and Technology (NIST) coordinated a major research project to guage how masks impacted the performance of varied facial recognition systems used across the globe.
Its report, published in July, found some algorithms struggled to appropriately discover mask-wearing individuals as much as 50% of the time. This was a major error rate in comparison with when the identical algorithms analysed unmasked faces.
Some algorithms even struggled to locate a face when a mask was covering an excessive amount of of it.
Finding ways around the issue
There are currently no usable photo data sets of mask-wearing people who might be used to coach and evaluate facial recognition systems.
The NIST study addressed this problem by superimposing masks (of varied colors, sizes and positions) over images of faces, as seen here:
While this will not be a sensible portrayal of an individual wearing a mask, it’s effective enough to review the consequences of mask-wearing on facial recognition systems.
It’s possible images of real masked people would allow more details to be extracted to enhance recognition systems – perhaps by estimating the nose’s position based on visible protrusions within the mask.
Many facial recognition technology vendors are already preparing for a future where mask use will proceed, and even increase. One US company offers masks with customers’ faces printed on them, so that they can unlock their smartphones without having to remove it.
Growing incentives for wearing masks
Even before the coronavirus pandemic, masks were a standard defence against air pollution and viral infection in countries including China and Japan.
Political activists also wear masks to evade detection on the streets. Both the Hong Kong and Black Lives Matter protests have reinforced protesters’ desire to dodge facial recognition by authorities and government agencies.
As experts forecast a future with more pandemics, rising levels of air pollution, persisting authoritarian regimes and a projected increase in bushfires producing dangerous smoke – it’s likely mask-wearing will develop into the norm for a minimum of a proportion of us.
Facial recognition systems might want to adapt. Detection will likely be based on features that remain visible equivalent to the eyes, eyebrows, hairline and general shape of the face.
Such technologies are already under development. Several suppliers are offering upgrades and solutions that claim to deliver reliable results with mask-wearing subjects.
For those that oppose using facial recognition and need to go undetected, a plain mask may suffice for now. But in the long run they could have to think about alternatives, equivalent to a mask printed with a fake computer-generated face.
This article was originally published at theconversation.com
