Tech5 introduces flexible biometric template protection for its ABIS

Tech5 has developed biometric template protection technology that it says meets the criteria set out in the ISO/IEC 30136 standard. The technology is now available as a feature of the multi-modal T5-OmniMatch ABIS. It can be used in both 1:1 and 1:N modes and across all the fingerprint, face and iris biometrics modalities the system supports.

ISO 30136 establishes performance testing definitions and metrics for biometric template protection schemes. The four main characteristics required for a system to comply with the standard are template cancelability, unlinkability, irreversibility and accuracy.

Tech5 notes the importance of ensuring that even if hackers are able to get ahold of biometric templates, they are not able to glean personally identifiable information (PII) from them. In addition to the risk to consumers and the reputation of any business that leaks their personal information, template protection is being built into privacy regulations like the EU’s GDPR, under which biometric templates are considered sensitive data.

Tech5 says its template protection technology uses layers of AI, machine learning and traditional methods to protect the entire database and individual templates. Customers can therefore implement it on either centralized or decentralized authentication systems, according to the announcement.

If a biometric database is breached, new protected templates can be deployed for all subjects immediately, and for individual compromised templates, new protected templates can be issued by the system and accessed with a user PIN or passphrase.

“Continuous innovation to achieve inclusion without compromising the security and privacy of the users is one of the founding principles at Tech5,” says company Co-founder, Chairman and CTO Rahul Parthe. “Our biometric template protection technology prioritizes privacy while maintaining high accuracy to ensure both inclusion and security. This achievement is made possible through the exceptional collaboration of research scientists with expertise in biometrics and machine learning.”

Researchers with the Idiap Research Institute showed how biometric templates can be compromised in a paper last year, and suggested that they can be protected from the vulnerability through hashing.

Article Topics

ABIS  |  biometric template protection  |  biometrics  |  data protection  |  research and development  |  TECH5