The Department of Homeland Security (DSH) has been testing facial recognition systems that can recognize faces with masks on.
With mask-wearing mandates in effect in many parts of the country, and health professionals urging everyone to wear a mask, facial recognition technology has been one of the casualties. Traditional facial recognition relies on seeing and comparing the entire face. As more individuals wear masks, however, companies have had to adjust their algorithms to focus on the area around the eyes.
DHS is particularly interested in the technology, as facial recognition is playing an increasingly bigger role in airport security. DHS tested 60 different systems against 582 diverse volunteers, representing 60 countries.
Without masks, the median success rate was 93%, with the best system scoring 100%. With masks, the median success rate was considerably lower, coming in at 77%, although the best system scored and impressive 96%.
The testing shows there is quite a bit of variance from one facial recognition system to the next. However, at the upper end of the scale there is minimal difference between wearing a mask and going maskless.