FB pixel

Apple patents for a secure transition of enrolling biometric templates with every software update

Apple patents for a secure transition of enrolling biometric templates with every software update
 

Apple has potentially addressed security concerns when it comes to safeguarding sensitive biometric data, such as facial images or fingerprints, by storing processed data that are derived from raw data templates. This measure minimizes the risk of exposing biometric information in the event of a device breach.

Additionally, software updates present a challenge as pre-existing enrollment biometric data may not be compatible with the new version. However, Apple has proposed a solution that allows for a transition of the enrollment templates from the old algorithm to the updated one.

On Tuesday, the U.S. patent office granted Apple patent 11935327, “On the Fly Enrollment for Facial Recognition.” It describes an enrollment process that eliminates the re-enrollment of biometric information with every software update.

Although the existing enrollment data may not be compatible with the latest software version, the software updates often come with improvements to the algorithms and models, including neural networks that can assess and compare biometric data.

These techniques may include algorithms that can update templates or transform them into a format that works with the new software version, all while maintaining the security and accuracy of the authentication process.

During a software update, the device utilizes two facial recognition processes simultaneously: the current neural network and a new neural network. The current neural network manages authentication tasks, while the new neural network operates in a virtual mode, running in the background without interfering with the authentication process.

As time progresses, the device gathers data on the performance of both the current and new neural networks in recognizing biometric information. This evaluation assesses accuracy, speed, and security.

By analyzing performance, it can be determined whether to switch operational control from the current neural network to the new one. It’s important to transition between models gradually to maintain user trust and system integrity.

Related Posts

Article Topics

 |   |   |   |   | 

Latest Biometrics News

 

Digital sovereignty, infrastructure and credential formats colliding

Digital credential formats are on the verge of being sucked into a global debate over the nature of sovereignty with…

 

Digital ID for alcohol sales is coming to the UK: hear what it means for the identity sector

For about two years now, leaders in the UK have been promising that, soon, people will be able to use…

 

Ant International, Visa, Mastercard work to make agentic protocols interoperable 

Ant International, Mastercard and Visa are collaborating on a Know-Your-Agent (KYA) interoperability framework, which a release says is “designed to…

 

Unico acquires biometrics provider Valida, gaining entry into Argentina market

Unico has announced it signed a definitive agreement to acquire Valida, “an Argentine biometric identity verification platform with an established…

 

KPMG investment in Reality Defender backs hiring spree to fight deepfakes

U.S.-based KPMG LLP has taken a minority stake in deepfake detection developer Reality Defender and plans to integrate its fraud-protecting…

 

Australia expands eSafety powers, doubles penalties in platform accountability push

Australia’s parliament has formally passed the Online Safety Amendment (Strengthening Enforcement for the Social Media Minimum Age) Bill 2026, granting…

Comments

Leave a Reply

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

Market Intelligence

Featured Company

Biometric Update Podcast

Most Read This Week

White Papers

Latest Webinars

Biometrics Industry Events