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

 

UK security industry should complete OneLogin adoption to save money

Gov.uk OneLogin should be in place for the UK’s the Security Industry Authority (SIA) by the fourth quarter of the…

 

Fraud intelligence software launched by Facephi, Feedzai

Facephi and Feedzai have introduced new fraud prevention products to complement their biometrics offerings. Smart Eye Technology and Resistant AI…

 

Documents aim to lay out use cases, standards for mobile driver’s licenses

The Secure Technology Alliance’s (STA) Identity and Access Forum has released a new resource on mobile driver’s license (mDL) use…

 

Brazil regulator demands details on stadium biometrics

Brazil’s National Data Protection Authority (ANPD) is asking for data protection impact assessment reports from 23 clubs that have deployed…

 

Clearview seeks refund for failed bulk purchase of SSNs, facial photos

Clearview AI is in the midst of a court battle to recover money it paid to a data broker for…

 

Vietnam PM urges all airports to use biometric authentication

The Prime Minister of Vietnam Pham Minh Chinh has requested all airports and border gates to use biometric authentication for…

Comments

Leave a Reply

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

Most Viewed This Week

Featured Company

Biometrics Insight, Opinion

Digital ID In-Depth

Biometrics White Papers

Biometrics Events