Socure reveals Selfie Reverification with Deepfake Detection

Self Reverification features

The Selfie Reverification capability compares a recently submitted selfie with the original ID headshot photo from onboarding, ensuring a 99.9% accurate match rate. It also detects deepfake signs and age disparities between the photo and the credential to confirm the user's identity reliably.

Socure's Selfie Reverification upgrades the company's top-tier Document Verification (DocV) tool. When onboarding, verifying photo IDs is important. The system boasts a 99% initial success rate in capturing selfie photos, along with 98% accuracy in liveness detection, and an impressive true acceptance rate of 98.2%. Considering the increase in identity attacks, even verified identities are vulnerable. For better security during significant events like high-value transactions or online identity verification, Selfie Reverification offers a swift and precise method to establish trust. It incorporates 'liveness' detection to thwart advanced spoofing techniques and confirm the physical presence of the person.

Selfie Reverification is versatile for various industries (financial services, online gaming, ecommerce etc.), ensuring a positive customer experience, reducing fraud, and lowering operational costs through automated risk assessment.

Selfie reverification offers a range of applications:

• Preventing account takeovers by triggering reverification for unusual changes.

• Protecting high-risk transactions, like large financial transfers, with additional security measures.

• Providing instant recovery for locked-out accounts or suspected fraud cases.• Offering quick identity verification for various needs like rideshare driver checks or hotel check-ins.

• Adding security for government services delivery by ensuring benefits reach the rightful recipients.

How does it work? 

The system prompts the user to take a selfie and provides real-time feedback on the quality. Next, a liveness check is performed to prevent fraud, ensuring compliance with NIST PAD and detecting deepfakes. Following this, biometric analytics assess over 80 facial landmarks, such as eye spacing, nose width, jawline shape, and emotional expressions, to establish and validate a facial map for precise identification.