Yoti updated transparency report on biometric age verification shows further accuracy gains

biometric facial recognition

Yoti has improved the accuracy of its biometrics-based age estimation algorithm for the important 13 to 15 year age group, according to internal benchmarking.

The company updated its transparency efforts for its age scanning accuracy, publishing a summary version of its Age Estimation white paper to its website.

The white paper provides details on Yoti’s efforts to build up biometric training data for its model, information on how people can opt out of research and development projects, and reviews the transparency efforts the company makes with the onboarding and opt-out screens of the Yoti app.

Compared to Yoti’s previous age estimation transparency report, the company’s algorithm has reduced the mean average error (MAE) rate for 13-to-15 year-olds from 1.60 to 1.35, and its overall MAE rate across all age groups from 2.28 to 2.19.

Yoti recommends setting the age threshold higher than the age of interest, in order to create a ‘buffer’ for “an acceptably low false positive rate.”

The company’s biometric age estimation technology was recently deployed for a pilot by several UK supermarkets to prevent underage alcohol purchases.

Related Posts

Article Topics

 |   |   |   |   |   |   |   |   | 

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Most Read This Week

Featured Company

Biometrics Research

Biometrics White Papers

Biometrics Events

Explaining Biometrics