Home Office identifies areas for improvement in biometric self-enrollment systems

0

Thumbs and bezels are notable challenges for biometric self-enrollment solutions, the UK Home Office found and informed the public at the latest meeting of the European Association for Biometrics (e-bike) an event.

Graham Camm, head office technical architecture lead for biometrics, and Jordan Webster, senior policy advisor for future biometrics, presented the results of the first phase of biometric self-enrollment trials for the processes of immigration during the EAB lunch.

The visa-free visitor program, slated to launch next year, is similar to other electronic travel authorization systems around the world, Webster says. The system is expected to eventually be used to process between 27 and 30 million people each year and go beyond facial biometrics to include fingerprints.

Remote biometric self-enrollment is at the center of the project, mainly with smartphones but also with unsupervised terminals.

Persons requiring a visa are considered to be at higher risk and will therefore likely have their biometrics recorded in a supervised environment.

Results of feasibility tests

The 2021 feasibility trials for biometric face and fingerprint matching and presentation attack detection with volunteers from the public were divided into six technology categories, Camm explained. Idemia was the only vendor to submit technologies for all six. Half a dozen tech vendors have been involved in testing the fingerprint PAD on a personal device, though it’s “newer” than some other tasks, Camm said.

Linking facial biometrics to an ID document seemed to work quite well, but including fingerprints in this linking process proved more difficult. Camm says only one vendor “really got the problem space right.” Even this vendor acknowledged that the technology to link fingerprints to an ID document is in its infancy.

The kiosks submitted by vendors were mostly similar in design, according to Camm, but there was more variance between the different mobile solutions offered. Many mobile solution providers come from the identity verification space, Camm notes, and the results indicate the significant difference between this app and biometric enrollment.

The trial results showed that all solution providers on both types of capture devices met the enrollment efficiency target of three minutes or less. User satisfaction was also high for all kiosks. For mobile solutions capturing facial biometrics, the user satisfaction standard of 95% was achieved by best-in-class, but generally lower. For mobile fingerprint capture, user satisfaction was lowest, with particular difficulty noted for fingerprint capture.

The efficiency of capture technology was also generally rated as unsatisfactory, with too many registration attempts triggering review processes, such as for image quality. A high number of reviews would increase costs and ultimately render the system unsustainable.

A combination of Face PAD solutions helped identify all attack species for kiosks and mobile devices, and like capturing on kiosks, Camm speculates improvements in tuning could yield better results. The effectiveness of fingerprint PAD was found to be insufficient.

No provider was able to guarantee that the face and fingerprints linked to an ID document were from the same person, Camm notes.

The kiosks met the biometric matching accuracy target, but only the best-in-class mobile solution met the face target, and none for fingerprints.

Glasses were the most common cause of low scores, and apps didn’t remind people to remove their hats or masks, which negatively impacted some attempts.

Matching scores, determined using the latest available Idemia face-matching algorithm, were slightly higher for kiosks than for mobile solutions, with the best of the former achieving an TPIR of 99.1% at 0%. of FPIR, while the last class was led by a score of 98.5. percentage of TPIR at the same threshold.

Some of the contactless fingerprint capture technologies have provided impressions that appear indistinguishable from impressions captured with a contact scanner, Camm says. Others, however, have distorted or even introduced (false) minutiae.

Camm outlined some of the remaining challenges in contactless fingerprint capture based on Home Office observations for EAB audiences.

The Home Office also considered latent brand search performance, demographic differentials, and usability.

Next steps

The next phase of testing was conducted separately for kiosks and mobile devices, due to differences in maturity between the two.

The kiosk test started in early 2022 and the fingerprint application test will be launched in late 2022.

The Home Office sees the biggest risk for the second round of feasibility trials is the possibility that they will not yield new information, which would then require another round of trials in late 2023 or early 2024. To mitigate the expense these would impose on both the Home Office and participating vendors, smaller “benchmark tests” are planned to assess the maturity of vendor solutions.

Once the benchmarking exercises have been successfully completed, a larger trial including subjects as young as five years old will be conducted.

Future testing may consider palm biometric capture, subject knowledge-assisted smartphone enrollment, and focus more on usability.

Article topics

biometric enrollment | biometric matching | biometrics | eBike | facial biometrics | fingerprint biometrics | kiosk | mobile device | presentation attack detection | research and development | UK

Share.

Comments are closed.