New NIST Data Details Facial Recognition Improvements for Masked Faces

The pandemic that has ravaged our planet for pretty much this entire year has caused a lot of changes to how we do things and has posed quite a few problems that are out there. One rather strange problem that ended up developing was that the algorithms that govern facial recognition often struggled to find any way to recognize faces that were wearing masks. Mask wearing is a key component of preventing the virus from having an even greater impact on the world, so making users take their masks off to make use of facial ID is the sort of thing that would be rather counterproductive.

With all of that having been said and now out of the way, it is important to note that many vendors of facial recognition algorithms have done a lot of research in order to be able to solve this issue. The National Institute for Standards and Technology received a lot of submissions from vendors for their Facial Recognition Vendor Test, and these tests showed that while initially the algorithms struggled to function with mask wearing users improvements to the technology have been made and now many algorithms are able to detect faces even if they are wearing masks.

Now, this is a great thing for people use facial recognition in order to unlock their phones, but at the same time it could have certain negative implications when it comes to various kinds of law enforcement and the like. One of the main uses for facial recognition technology has to do with cataloging people’s faces for the purposes of law enforcement, and this is the sort of thing that many people don’t like which is why they often end up wearing masks. If facial recognition can work regardless of masks, this poses a potential privacy risk for a lot of users out there.


Read next: Revealed: Most valuable American is worth $66 MILLION!

No comments:

Post a Comment