Impact of Demographic Factors on Performance of Biometric Systems


  1. Campbell, J.W.M.
Corporate Authors
Defence Research and Development Canada, Centre for Operational Research and Analysis, Ottawa ON (CAN);Bion Biometrics Inc., Ottawa ON (CAN)
Biometric systems using facial, fingerprint and iris recognition have been widely deployed by Canada and other governments. These systems are used to issue passports and visas, permit travelers to cross the border and to allow employees access to physical and digital resources. Biometric performance, in terms of the risk of non-matches and of false matches, is critical to the successful and secure operation of these systems. Recent research has uncovered significant performance differences in multiple biometric systems depending on the demographics (specifically age, sex and ethnicity) of the subjects whose biometric characteristics are being captured. In some cases, this simply leads to a bias that makes it more difficult for certain subjects to use the system. Existing studies show that this is the case for females with fingerprint recognition and facial recognition for older people with fingerprint recognition and iris recognition and for young children with all three modalities. In other cases, the performance difference leads to high rates of false matches and this can significantly undermine the security of the biometric system. So far this is known to be the case for certain ethnic groups and for young children when using facial recognition, but further studies are needed to investigate this phenomenon more thoroughly for all biometric modalities. Several governments are actively researching this problem and Canada need to do the same. Unless systems, especially those

Il y a un résumé en français ici.

Biometrics;Demographic Bias
Report Number
DRDC-RDDC-2018-C113 — Contract Report
Date of publication
01 Jun 2018
Number of Pages
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: