Facial Recognition Succeeds in Defeating The COVID-Mask Challenge

Any individual who uses the feature of facial recognition to unlock their smartphone device has faced difficulty during the coronavirus pandemic as more than half of the face was covered by a protective Covid-mask as a preventive measure against the viral infection, which has taken the lives of tens of millions of people throughout the globe.

The issue with facial recognition unlocking amid the pandemic has proven to be very frustrating, as individuals became used to unlocking their smartphone device almost instantly without typing in the passcode and just keeping the phone at arm’s length from the face to unlock. Although face masks have thwarted the industry of facial recognition, technological advancement has also similarly adapted to the current needs of their users.

It might sound like a strange notion, but wearing a Covid-mask does not stop the computer technological innovations from identifying an individual through facial recognition.

Reduced accuracy

In the previous year, during the start of the coronavirus pandemic, people started wearing face masks to protect themselves from the deadly viral pathogen all around the world, but it presented as a huge challenge for the facial recognition software to identify individuals.

It is obvious that the algorithms of facial recognition, which have been designed to analyze an entire face for security purposes, would become less accurate if half of the face was covered. Various studies have been conducted to test the accuracy of facial recognition algorithms, which have backed up the reduced amount of accuracy.

According to the research conducted by the National Institute of Standards and Technology (NIST), in which they were able to test around 89 commercially developed algorithms for facial recognition, after which they were able to conclude that there is a 5-50% rate of possible errors in similar faces when a digital face mask is applied to cover part of their faces in photos of the same individual.

Although there are several facial recognition technological innovations present which is able to provide accurate results while an individual is wearing a face mask. In the month of January, the United States Department of Homeland Security has conducted a controlled-scenario test in which they were able to obtain a success rate of 96%; however, the results were variable between systems.

According to the statement given by the department, the report based on the conducted results shows that organizations that are utilizing facial recognition should also perform checking of photo ID for further security, and would also allow people to keep their masks continuously on their faces in public, which would limit the rate of transmission of coronavirus infection.

There are some police departments which are have started using less amount of facial recognition, including the London Metropolitan Police, as they have not conducted any facial recognition test in over a year since the start of the pandemic. However, it was reportedly used during the time of Black Lives Matter protests in the month of May last year, after the unlawful death of a black individual by a police officer as an act of oppression.

Touchless encounter

In the private sector, it is very hard to have a report regarding the usage of facial recognition over the past year during the coronavirus pandemic, as no record has been kept through which it would have been established regarding where and when it has been used.

It has been reported that Disney World is going to have a trial of facial recognition technology over the time duration of one month. The systems would capture the picture of the customer and convert it into a special series of numbers, which would be later associated with the admission form that is required to enter into the massive theme park.

This idea is to reduce the waiting time, which reduces the positive feedback of customers, while Disney is also looking forward to experiencing a touchless system through facial recognition amid the coronavirus pandemic. Most importantly, this technology would allow individuals to have their faces recognized while wearing face masks.

Even before the start of the coronavirus pandemic, which forced every person to wear a Covid-mask before leaving for a public place, there have been various ongoing researches which were looking to further innovate the algorithms of facial recognition which would allow the software to identify an individual while a part of their face is covered.

In countries like Japan, NEC has already been working on a technological system that would allow people to easily access while they wear face masks due to airborne allergies, while in South Korea, people tend to wear face masks due to increasing levels of fine dust in the atmospheric air.

Read also Using Virtual Reality Headsets For Workplace Meetings Could ‘Snowball’