While face masks are absolutely essential in our fight against coronavirus, they are also bringing forth a problem when it comes to technology. Face masks have been breaking down facial recognition algorithms, according to a new government study. NIST reported that if one were to wear a face mask in the way it has been prescribed, there is a 5-50% chance that the facial recognition algorithm will break down- leading to a crack in the surveillance system.
Black masks would be more of a problem than blue masks, and the more one sidled over the nose, the higher chance of inaccuracies in the system. “With the arrival of the pandemic, we need to understand how face recognition technology deals with masked faces,” said Mei Ngan, an author of the report and NIST computer scientist.
NIST uses algorithms that calculate the distance between the features of an individual’s face. With the mask put on, the chances of that get more and more inaccurate. And this is in complete agreement to several anecdotes prevailing about face masks thwarting facial recognition systems. Considering NIST is the de facto head of all things algorithm, it is pretty clear that the results are conclusive.