Unlocking a smartphone while wearing a mask is a difficult task that can often fail. But the pandemic seems to be forcing facial recognition technology to evolve, with developers increasingly adapting algorithms to deal with this issue.
In July of this year, an American study revealed the difficulties that technology had in keeping up with this new "trend." The research results showed that the error rates of 89 facial recognition systems increased when the mouth and nose or the lower part of the face were covered.
The reality now seems to be different. A new study by the agency revealed that there has been a race among companies to update their algorithms to be more effective with masked faces. This was particularly noticeable in solutions where facial recognition is used to confirm a person's identity, such as authorizing access to a secure building.
A year after the first cases of Covid-19 emerged, the data shows that the error rates were ten times better than those recorded in the July study.
SOURCE:
THAT IS
