We're delighted to announce the release of a new face biometric matching model as part of our commitment to continuously improve the security of user authentication and identity verification solutions.

Our authentication and identity verification solutions rest on a top-tier face authentication API – we called it YouFace. At our company, we are committed to providing the best possible face authentication algorithm to our customers. We understand that security is a top priority for many businesses and individuals, and we believe that our technology can help to address these concerns.

As such, we invest a significant amount of time and resources into improving our face authentication API, through ongoing research and development, as well as rigorous testing and evaluation.

The new version of YouFace include some more additional features such as:

• New face detection algorithm that can handle faces in any orientation.
• An extended list of face quality metadata attributes for verification of ISO 19794-5 compliance face images.

A major leap in performance... For any application scenario!

Certification is crucial for biometric algorithms as it provides a standardized evaluation of their performance and accuracy. This ensures that the algorithms meet the required standards for security and reliability and helps to build trust among customers and users of the technology. Therefore, we submitted a new NIST evaluation with major improvements.

Concerning identity verification in Know Your Customer scenarios (that shares similarities to the VISABORDER FRVT evaluation track), we have decreased the error rates of our solution 20x. We can now ensure that for a 1 in 1M false acceptance, only 0.4% of the users are rejected on the first try.

Similarly, a major performance increase was achieved in the cooperative live vs live scenario (BORDER evaluation track), applicable to user passwordless login or border control. Here we have decreased the error rates 30x, enabling false rejection of 0.7% valid attempts for 1 in 1M false acceptance.

A face authentication API with a single model approach

Unlike others that build specialized models for specific NIST evaluation tracks, we dedicated time to developing a single model fit all to build such amazing performance. This means YouFace face biometric matching is truly general and can be applied in any application with ensures of great performance.

💡 The face matching algorithm YouFace can be applied to a variety of scenarios across different industries. Its versatility makes it a valuable tool for enhancing security, streamlining processes, and improving customer experiences across various industries.

YouFace is now available in offline mobile authentication

In parallel to the development of great new face-matching algorithms, we have also developed mobile authentication that can be performed fully offline (without an internet connection). Our commitment is user security and ownership. When implementing mobile authentication all the processing is performed locally on the device, and it can be performed offline.  Solutions are available for both iOS and Android platforms.

I am thrilled with the results achieved in the latest NIST report. It is very hard to see a startup that performs on par with global multinational enterprises working in the face recognition field. In this submission, we have leveraged a new proprietary technology for data handling and generation that has the potential to set us apart from the competition. The new and highly cost-effective data preparation technology has unlimited potential to build a scalable biometric solution and deliver better and better performance.

Miguel Lourenço, Youverse CPO

Get started and test it for free

The new version is available to test for free upon request. Please contact us via our discord channel or send an email to our support team. We’re happy to facilitate testing and bring your biometric solution to the next level.