Face mask protects us from the deadly virus. However, it is doing more than that. Like, it is breaking facial recognition algorithms as well. This is a completely unintentional role played by face masks.

Nothing ever comes with perfection. There are already 5% error chances associated with facial recognition algorithms in a normal situation. The error rates have jumped to 50% with a face mask on as it covers most parts of the face like nose, mouth, and chin. This was found in a study by the US National Institute of Standards and Technology (NIST).

There is more addition to the problem. The new 50% error rates depend on the color of face masks as well. Black colored masks are likely to catch up with the algorithms when compared to blue masks.

On the one hand, blue masks (surgical masks) protect you in a better way from the virus. On the other hand, it breaks the facial recognition algorithms. This in turn makes it more difficult to use this algorithm to unlock phones in a public place.

Nonetheless, it is established that the more part of the nose is covered with the mask, the more problematic it is for the programme to recognize a face. This makes it more difficult to use this unlock pattern in public places whilst protecting ourselves from the deadly virus.

As per Mei Ngan, as the pandemic came in the picture, the need for a detailed study of facial recognition algorithms has increased. There has developed the need for the study which will show how face masks are manipulating facial recognition technology.

Mei confirmed that they had been studying the algorithm before the pandemic began. But now as the subjects are wearing masks, the study might get disturbed. However, they are planning to have a study later in the summer which will study the veracity of algorithms that will be focused on masked faces.

The common study of NIST includes studying the distance between different facial features. Thereafter, different facial recognition algorithms are tried, tested, and verified. But with face-covering few facial features, the fidelity of the algorithms is reduced. However, there still are a few devices available that are functioning just fine. Lucky are those people who are not facing any issue with facial recognition algorithms. They are free from all the hassle in the public place.

iPhone is known for its security and protection of privacy. That is why its facial recognition algorithms are different from every other programme. It is designed in such a way that the system can’t be tricked by showing the picture of the person whose phone is to be unlocked.

NIST tests the facial algorithms of the most sensitive system around the world. Like the ones used during passport control at the border crossing and likewise systems. This system is called one-to-one matching and show pretty accurate results.

The post Study Claims Face Masks Are Breaking Facial Recognition Algorithm appeared first on The Nerd Series.