Cognitec tops the leaderboard of average ranking for the accuracy of age estimation and verification technologies, according to test results from the U.S. National Institute of Standards and Technology (NIST).
The report, NIST FATE: Age Estimation and Verification, includes test results of age estimation accuracy globally and across demographic groups (gender and geographic region). Averaging the rankings across two genders and all six regions, Cognitec achieves the best rank overall.
Ordering the algorithms by the average of the Mean Absolute Error (MAE) rates over all genders and regions also puts Cognitec’s algorithm in the top position.
NIST also measured algorithmic accuracy gains since 2014 by running the current algorithms on the exact same dataset used in 2014. Compared to the best-performing algorithm in 2014, which was a Cognitec algorithm submitted in 2013, the current Cognitec algorithm reached the best error rate reduction, and the highest fraction of images with better accuracy.
Governments worldwide are issuing new regulations that ask for effective age assurance technology to keep underage users from accessing prohibited digital content, or from illegally purchasing goods on digital shopping channels. Immigration agencies conduct age assessments of asylum-seeking individuals without birthdate evidence.
Technology vendors are providing age estimation algorithms that can support age assurance methods by using facial analysis to verify that a person is above age 18, 21, or other key ages. The ongoing NIST test currently evaluates 30 algorithm submissions from 18 different developers. For all test results, please visit the NIST website.
The report, NIST FATE: Age Estimation and Verification, includes test results of age estimation accuracy globally and across demographic groups (gender and geographic region). Averaging the rankings across two genders and all six regions, Cognitec achieves the best rank overall.
Ordering the algorithms by the average of the Mean Absolute Error (MAE) rates over all genders and regions also puts Cognitec’s algorithm in the top position.
NIST also measured algorithmic accuracy gains since 2014 by running the current algorithms on the exact same dataset used in 2014. Compared to the best-performing algorithm in 2014, which was a Cognitec algorithm submitted in 2013, the current Cognitec algorithm reached the best error rate reduction, and the highest fraction of images with better accuracy.
Governments worldwide are issuing new regulations that ask for effective age assurance technology to keep underage users from accessing prohibited digital content, or from illegally purchasing goods on digital shopping channels. Immigration agencies conduct age assessments of asylum-seeking individuals without birthdate evidence.
Technology vendors are providing age estimation algorithms that can support age assurance methods by using facial analysis to verify that a person is above age 18, 21, or other key ages. The ongoing NIST test currently evaluates 30 algorithm submissions from 18 different developers. For all test results, please visit the NIST website.
