A Chinese company says its facial recognition software can now identify people that are wearing masks to protect against the coronavirus.
Hanwang Technology used a sample database of around 6 million unmasked faces and a much smaller database of masked faces to create the system.
The Beijing-based firm, which also goes by the English name Hanvon, began to develop the tech in January, as people in China began donning face masks in their droves. The system was rolled out just one month later.
Hanwang Vice President Huang Lei says the system’s recognition rate reached about 95% when people wore a mask — still some way below its regular success rate of 99.5%.
[Read: 7 best coronavirus dashboards to map the spread of COVID-19]
Huang told Reuters:
If connected to a temperature sensor, it can measure body temperature while identifying the person’s name, and then the system would process the result, say, if it detects a temperature over 38 degrees.
Why wear a mask?
The World Health Organization advises healthy people to only wear masks if they’re taking care of someone who might have the infection and warns that shortages are already leaving healthcare workers ill-equipped to care for patients.
Nonetheless, millions of people in China are already donning them, as well as western celebrities such as Gwyneth Paltrow and Selena Gomez.
Their attempts to escape infection may be futile, but they may have another motivation for wearing the masks: avoiding facial recognition.
Hanwang’s tech means this may no longer be possible — and the company is not the only one claiming its tech can unmask you. China’s SenseTime, the world’s most valuable AI startup, announced in February that it had also adapted its product to identify people wearing masks. Such developments have led critics to claim that the coronavirus is being used as an excuse to ramp up surveillance.
In the case of Hanwang, there is still one way to hide from its system: wearing the fashionable combination of both a face mask and sunglasses.
“In this situation, all of the key facial information is lost,” Huang said. “In such cases recognition is tough.”
Get the TNW newsletter
Get the most important tech news in your inbox each week.