Machine-learning algorithms better at recognising masked faces

As everyone is now wearing masks on a daily basis, it has made it more difficult for facial recognition software to identify faces.

 

Indeed, the NIST studies the impact of mask-wearing on the technology and found out that lots of vendors struggled with it. After a few months, the tech has now been improved to fit the current circumstances.

 

A NIST scientist declared that the new algorithms from developers showed better results and fewer error rates. These machine learning algorithms are now more accurate than ever.

 

A report published by NIST revealed their findings, however, the test images used were photographs of people with digital masks and not people wearing actual masks, which might represent an issue.

 

More
articles

Menu