Standards 

IDEMIA is an active contributor to many international, national, and industry standards bodies that publish Identity Proofing related standards. IDEMIA brings customer requirements and extensive operations experience to standards work groups to improve the applicability and utility of the standards development effort.

Industry and government stakeholders develop standards to achieve common, desired outcomes. Standards provide a common vocabulary and a set of best practices.

Standards carry out consistency and fitness for purpose of products and services. In terms of official government-issued credentials, standards permit users to know: who authoritative issuing authorities are, what the credentials look like, how to confirm the credentials are genuine, and how to use them.

Industry standards 

The IDEMIA Identity Proofing platform is designed to permit organizations to create applications and procedures that conform to international standards such as ISO/IEC 29115:2013 Entity authentication assurance framework, US NIST 800-63-3, and EU eIDAS CIR 2015/1502.

IDEMIA is also committed to adhering to all relevant data privacy regulations such as the EU General Data Protection Regulation (GDPR) and the California Privacy Rights Act (CPRA). You can read IDEMIA's Privacy Policy, Privacy Policy for Identity Proofing Platform, and Data Protection Agreement for Identity Proofing Platform on the IDEMIA developer portal.

Standard assessment 

IDEMIA uses external assessors to determine conformity to important standards and seeks product certifications such as:

  • Kantara Initiative Trust Mark for conformance to NIST SP 800-63-3 (in progress)

  • iBeta Certification for ISO/IEC 30107-3 at Level 2

Both of IDEMA liveness detection solutions have been awarded Level 1 and Level 2 compliance by iBeta, an independent third-party tester, in accordance with the ISO/IEC 30107-3 standards. The testing method simulated the enrollment of a user in a biometric authentication system, where legitimate users consented to photos and videos of their likeness being used in attempts to spoof the system (Level 1 testing), and also the use of more sophisticated artifacts like silicone, latex, and plastic masks (Level 2 testing). In both cases, testers were unable to gain access to the system using any of the presentation attacks -meaning that IDEMIA’s liveness detection technologies achieved an attack presentation classification error rate (APCER) of 0%.NIST Facial Recognition Vendor Test

NIST (National Institute of Standards and Technology) has recognized the significant improvement in facial recognition technology in recent years. The FRVT 2013 (Face Recognition Vendor Test), which ended in May 2014, was the previous large-scale test of facial recognition technology. The independent test used a large, real-world dataset, and it provides an objective and quantitative comparison between vendors.

Results are publicly available on the NIST page Face Recognition Vendor Test (FRVT) Ongoing.

The test addresses two common use cases from two different operational datasets:

  • Law enforcement images, mostly mugshots with some webcam images

  • Identity document images from a smaller Visa database.

A future test will address video sequences.

FRVT 2013 confirmed the outstanding performance of IDEMIA’s technology. IDEMIA’s algorithm achieves the second best result across all tests, and performs particularly well at the lowest false alarm levels, which corresponds to the operational setting for a multi-million gallery and limited resources for visual inspection.

IDEMIA was one of the only vendors who demonstrated a miss rate lower than 10%, and ranked first, for a 1.4 million-record mugshot dataset.

The accuracy advantage of IDEMIA technology has also been shown to increase as the database grows, as demonstrated by a limited loss in accuracy accompanied by a decreased False Acceptance Rate (FAR). With a larger database, the probability of finding look-alike people increases, leading to an increase in the FAR. FRVT shows that even with a ten-fold population size increase (from 160,000 to 1.6 million), the FAR increases by only small factors, with IDEMIA achieving the smallest increase (1.1).

Another important aspect is threshold stability. The threshold is the value of the similarity score between two portraits, above which they are considered to be of the same person. This value is usually set with the aim of obtaining a predictable FAR. The threshold often needs to be adjusted as the size of the database increases, which is not possible in real life systems. IDEMIA technology displays stable scoring that allows users to set reliable thresholds that do not require adjustment as the database grows.

Ongoing FRVT 

NIST's March 2021 FRVT evaluation of facial recognition technologies identified IDEMIA as the best identification system on the market. Unlike previous evaluations, FRVT Ongoing is conducted on an ongoing basis and remains open indefinitely so that developers can submit their algorithms to NIST whenever they are ready. This approach more closely aligns evaluation with development schedules. The evaluation uses very large sets of facial imagery to measure the performance of facial recognition algorithms developed in commercial and academic communities worldwide. Multiple evaluation tracks relevant to facial recognition are conducted under this test.

NIST’s latest report found that IDEMIA:

  • has the best technology on a combination of datasets used for ranking, namely VISA and MUGSHOT.

  • ranked first for VISA, WEBCAM, and SELFIE scenarios and third for MUGSHOT

IDEMIA submitted two algorithms:

  • The first submission generates the smallest template file size

  • While being the most accurate, IDEMIA second submission’s template is still very small, which is key for a suitable trade-off between cost and performance

This NIST report shows that IDEMIA has the best performing algorithms for facial recognition in many applications.