Covid-19 pandemic has affected the entire world, and the best source for its prevention is a face mask. The pandemic has led to huge demand and the use of face masks among all. However, a face mask’s continuous use might have been breaking the facial recognition algorithm, say a new government study. Since a facial mask covers the entire mouth and nose area on the face, it leads to errors on the most widely used facial algorithms, which is as high as 50%.
The study on the same was conducted by the US National Institute of Standards and Technology (NIST). The more of the nose area covered by the mask, the harder it becomes for the facial recognition algorithm to identify them.
Also, they discovered that the black colored masks created more errors than the blue colored ones. It is why new algorithms for many companies and software are being created, identifying a person with the face mask. Currently, the algorithm development is in progress, however, creating one to recognize with a mask will be difficult and might take a long time.
The algorithm to detect faces with masks will be tested later this summer. The algorithm in development by NIST will measure the distance between the face features. It will remove most of the features from the older algorithm but will continue to use some.
The new facial recognition software ties tested by NIST only tested a type of facial recognition known as one-to-one matching. It is the procedure used in border crossings and passport control scenarios, where the algorithm checks to see if the target’s face matches their ID. NIST is deeply concerned about the security threats that might occur due to the mask’s widespread use. As per them, it could hamper a lot of security systems that use facial recognition software.