Yoti preps 1:1 face matching product for testing, ramps up for deepfake deluge

Yoti is aiming for the NIST leaderboard with its algorithms for biometric facial authentication. In a 2024 preview posted on its blog, the British biometrics and digital identity firm says it is awaiting imminent results of NIST testing on its facial age estimation technology, and is preparing to submit its new 1:1 face matching software to the standards organization for testing by the end of this month.

An explicit goal is to crack the top 100 on the NIST leaderboard of best facial recognition vendors. The post notes that, at present, Chinese, Japanese, South Korean and U.S. biometrics firms dominate the top 100. But Yoti is confident in its success: “We expect Yoti will soon be the only British company to rank highly on the NIST leaderboard,” says its blog.

Noting the looming risk to fair voting processes in a packed global election year, Yoti also continues to power up against deepfakes, with its MyFace NIST Level 2 liveness technology to detect spoofs, and patented secure image capture (SICAP) injection detection to verify footage sources.

The firm’s secret weapon, however, might be the most advanced technology of all – the human brain. Another of Yoti’s posts offers a glimpse into the world of Super Recognizers: people who have an innate ability to recognize faces. Only about 1-2 percent of the global population passes the test. Among those are the Yoti employees that provide 1:1 face matching and document verification, when automated checks “do not conclude this with high confidence, to help with trickier account creations, or when businesses want increased confidence about a person’s identity.”

“We can view hundreds of faces and documents in a day,” says Yoti Training Manager and Super Recogniser, Louise Bruder. “Because of this, we must take regular breaks and engage in other activities that use different parts of the brain. Our job demands a high level of visual acuity, stamina, and attention to detail.”

“Given the nature of our work and the fact that we want our customers to have complete trust in what we do, we must strive to be the best in all areas. This includes our human decision-making as well as our market-leading tech.”

The University of Greenwich hosts a free online test for those wishing to see if they qualify as a Super Recognizer.

Article Topics

age verification  |  biometric matching  |  biometric testing  |  deepfakes  |  facial recognition  |  NIST  |  research and development  |  Yoti