Chinese firm performs face recognition through masks
18 March 2020 14:10 GMT

Chinese electronics manufacturer Hanwang has detailed how its face recognition technology can see "through" the masks needed to protect populations from the Covid-19 virus.

The company has told the FT that its system is based on existing technologies but involved adding a collection of about 6 million unmasked faces and a much smaller collection of masked faces.

The company now says its masked facial recognition program has reached 95 percent accuracy in lab tests, and even claims that it is more accurate in real life, where its cameras take multiple photos of a person if the first attempt to identify them fails.

“The problem of masked facial recognition is not new, but belongs to the family of facial recognition with occlusion,” Huang Lei, the company’s chief technical officer, told the FT. He added that his company had first encountered similar issues with people with beards in Turkey and Pakistan, as well as with northern Chinese customers wearing winter clothing that covered their ears and face.

Hanwang is now selling “single channel” recognition, which is designed to be used at the entrances to buildings, and a “multi-channel” recognition system that uses groups of surveillance cameras.

Huang Lei, told Reuters that the multi-channel system can identify individuals in a crowd of up to 30 people “within a second.”

“When wearing a mask, the recognition rate can reach about 95 percent, which can ensure that most people can be identified,” Huang said. He added that the system’s success rate for people not wearing a mask is about 99.5 percent.

Industry Events



connect:ID 17-18 Mar 21