Title: How do facial recognition systems evolve in the situation of COVID masks?
description: We are all aware of the hassle of unlocking phones with facial recognition while wearing a mask.
Unfortunately, there is no indication that we should wear masks. The masks have made it much harder to use facial recognition systems around the world.
It turns out, however, that wearing the mask is not necessarily impossible to identify the face.
Facial recognition has adapted well to the situation and we can see more advanced systems that will be able to identify a person with or without a mask. Algorithms developed for face analysis have difficulty identifying faces with masks.
The National Institute of Standards and Technology (NIST) thoroughly tested 89 commercial facial recognition systems and found a 5-50 percent error when matching masked faces to ordinary photos of the same person. So as you can see, the problem is considerable. We learned that the Japanese company NEC is working on a system for people wearing masks due to allergies, which has an efficiency of 99.9%.
The algorithm focuses on exposed areas such as the eyes, ears, and forehead to identify the person with the mask.
The company's goal is to sell the system to offices and other facilities. The proposed solution seems very interesting. We hope that this excellent solution will be very well received all over the world. It is worth adding that it will be a great convenience for smartphone users. When we receive more information on this subject, we will certainly provide it to you.
keywords: masks, difficulties, COVID, application, new solutions, problems