The use of face masks have been beneficial in curbing the spread COVID-19 but as masks becomes a regular thing, businesses and authorities experience new issues as face-recognition dealers race to update algorithms that can’t operate with masks.
Before covid-19, mask were controversial and the Hong Kong authorities prohibited their wearing by protestors who were wearing masks to avert being caught by artificial intelligence (AI)-powered surveillance cameras.
Presently, the ongoing analysis by the US National Institute of Standards and Technology (NIST) has revealed the usefulness of regular face-recognition algorithms which is now famous amongst even smartphones users.
False match rate (FMR) figures in the Ongoing Face Recognition Vendor Test (OFRVT) rose tenfold or more when 138 various face-recognition algorithms were assessed against the NIST data set of more than 6.2 million photographs.
Different digitally applied mask shapes, colours, and nose coverage was assessed, and submitted algorithms indicated wildly different results as authors attempted – with varying levels of success – to modify their algorithms for increasingly accepted face-mask wearing.
FMR rates of approximately 0.3 per cent rose to a standard of around 5 per cent when masks were used, with some algorithms failing to authorize up to half of the images they were presented with.
However, one assessment of the testing concluded, full-face-width masks, such as widely-used disposable surgical masks, obstructed twice as many recognition trials as rounder N95-type masks, which don’t conceal as much of the face.
Black masks were more inclined to obstruct face-recognition trials than light blue masks, the justifications for this is not known to the researchers.
In the US, reports implied that security-conscious businesses, like banks and casinos, were obliging customers to raise their face masks for recognition by security cameras.
The Commonwealth Bank, also implores customers to “please wear a mask and maintain physical distancing” if they visit a branch but the NAB notifies that customers are “welcome to come into our branches with a face mask, but we may ask you to remove it for identification purposes.”
Competitors like ANZ and Westpac give no particular tactic on mask wearing. However, the issue is much more than banks because related situations can correlate in any public place, as accentuated in a 2017 bill recommended by Senator Jacqui Lambie.
“Full face coverings conceal the identity of the wearer, disrupting the authorities’ ability to track down a perpetrator in the event of a crime,” notes the bill, which would have criminalised those who “intimidate or force an adult into wearing identity concealing garments”.
Now, after three years, millions of Australians have to wear identity-concealing masks everyday.
Firms like Tryolabs are working on refining surveillance techniques to automatically survey whether people are adapting with public-health guidance or not.