Facia is now iBeta Level 2 Compliant with ISO 30107-3 Presentation Attack Detection Protocols

Facia, a leading UK-based identity verification provider, proudly announces its achievement of iBeta Level 2 compliance. This compliance aligns with ISO 30107-3 Presentation Attack Detection standards, highlighting Facia’s commitment to providing clients with the most advanced and secure facial recognition technology.

This achievement builds upon Facia’s previous iBeta Level 1 compliance and reinforces the company’s dedication to maintaining the highest quality standards within the identity verification industry. Facia’s robust suite of solutions includes 3D liveness detection, face search and matching, age verification, and iris detection.

Key Takeaways

• Facia’s facial recognition technology achieved a perfect score with a 0% APCER on both Android and iOS devices. This means it successfully identified ALL presentation attacks as fakes.
• The company demonstrates a focus on ethical and unbiased AI through diverse datasets.
• Facia’s commitment to security extends to combating deepfake threats with its cutting-edge Facia Morpheus 2.0 algorithm.
• The company aims to play a leading role in global identity verification, targeting the verification of 8 billion identities by 2030.

Facia’s leadership envisions a digitally secure world where businesses can seamlessly onboard genuine customers. The company currently assists a broad spectrum of industries, including banking, KYC, cryptocurrency, airports, and government-level security agencies. 

“At Facia, we are driven by a core mission: to transform the landscape of digital security by empowering businesses with the most secure and reliable facial recognition technology available,”  Stated Mujadad Naeem, CEO of Facia. 

“Achieving iBeta Level 2 compliance is a testament to our unwavering commitment to excellence. This rigorous compliance process validates our ability to safeguard our clients from sophisticated fraud attempts and ensure a seamless onboarding experience for legitimate users.”

Rigorous iBeta testing was conducted on Facia’s Android SDK v1.0.3 application (OnePlus Nord 200 running Android 12) and iOS SDK (Apple iPhone 12 Pro Max running iOS 16).  A remarkable 1500 presentation attacks were performed – 750 on each device – with Facia’s technology successfully preventing every attempt. This result reinforces Facia’s proficiency in maintaining a 0% APCER (Attack Presentation Classification Error Rate).

Confirmation of iBeta Level 2 Compliance

See Confirmation Letter: Facia PAD iBeta Level 2 Confirmation Letter

Facia solidifies its position as a premier identity verification solution with its exceptional 0% False Acceptance Rate (FAR) and sub-1% False Rejection Rate (FRR) during iBeta Level 2 testing. Driven by AI-powered detection, Facia’s technology aims to exceed the capabilities of competing providers. The organisation prioritises diverse datasets to guarantee 100% accuracy and eliminate any potential for ethical or racial bias within its facial recognition systems.

Facia’s dedication to innovation is exemplified by its state-of-the-art algorithm, Facia Morpheus 2.0. This algorithm is specifically tailored to counter the evolving threat of deepfake technology, safeguarding businesses from various forms of fraud.  

Facia’s bold vision extends globally, aiming to play a pivotal role in verifying 8 billion identities by 2030.

Facia is on a mission to build trust with people digitally

What is ISO 30107-3 Presentation Attack Detection?

ISO 30107-3 is an international standard that provides the framework for evaluating the effectiveness of Presentation Attack Detection (PAD) within biometric systems. iBeta, as an accredited testing lab, applies the standard in the following ways:

• Test Levels: iBeta defines testing Levels (Level 1 and Level 2) specifying:
  • Time allowed to create presentation attacks (e.g., 8 hours for Level 1)
  • Material cost limits for attacks (e.g., $30 for Level 1, $300 for Level 2)
  • Expertise required to carry out attacks (none for Level 1, moderate for Level 2)
  • Success thresholds for vendors (0% penetration for Level 1, 1% for Level 2)

• Standardized Attack Types (PAIS):  iBeta uses 6 types of presentation attacks (e.g., 3D masks, 2D photos, replayed videos) that are consistent across vendors and tailored to challenge the specific biometric method and liveness detection.
• Cooperative Subjects: iBeta uses real, cooperative people to provide high-quality biometric samples for creating the attacks.  This makes their tests more rigorous and conservative.
• Time Constraints: Tests are time-bound (e.g., 8 hours for Level 1) to maintain uniformity and efficiency.
• Metrics: iBeta focuses on core PAD metrics:
  • APCER: How often a fake biometric is wrongly accepted?
  • BPCER: How often a real biometric is wrongly rejected?
• Reproducibility: iBeta records all details of the vendor’s solution and the device used for testing. This lets others recreate the test to verify the results.
• Focus on Security, Not Just Usability: iBeta sets thresholds for acceptable false rejection rates (BPCER), but the primary focus is to ensure a strong defence against fraud (low APCER).

Why this matters: iBeta’s approach to ISO 30107-3  combined with their NIST/NVLAP accreditation ensures their evaluations are fair, transparent, and data-driven, and puts a strong emphasis on the security of biometric systems against presentation attacks.