Face Recognition and Matching

Connect an automated customer facial verification service. Verigram’s solution offers face recognition from photo or video, face matching to a reference photo from a document (1:1), face search and matching to your trusted database (1:N).

Our algorithms are tested by vendors at NIST and ranked in the top 1-20 for accuracy — over 99%. A powerful solution for fraud detection and protection. Part of the KYC process. Use this solution to simplify customer authorization on any device or for video analytics.

face recognition
Vector.svg

Easily confirmate identity

Verigram’s facial recognition is part of the identification system for fast and accurate remote verification. The solution uses a combination of face detection, capture and recognition functions. The customer's face is compared with a reference photo from an available source. The result is a confirmation of a person's identity.

Advanced fraud defense

Motion tracking, lip sync and texture analysis are used to verify that the user is physically present

User Experience

Reduce recognition time while increasing conversion rates

Fast onboarding

It takes only one simple step to convert applicants into paying customers

Stronger Authentication

It takes only one simple step to convert applicants into paying customers

1:1 Face recognition

1-to-1 face comparison module includes face detection, face descriptors extraction, and face-matching with a reference photo.

NIST independent test:

  • Processing time: 0.663 sec.
  • For 15 million transactions, only 1 error is possible
  • TOP 10 position in WILD
  • TOP 40 position in global algorithm ranking

1:N Face recognition

1-to-N module includes face recognition, extracting and searching for a similar face in the database, and matching faces from a reference to a photo in the database.

NIST independent test:

  • Processing time: from 0.418 sec.
  • TOP 1 position on 12M face dataset
  • False negative identification rate (FNIR) of 0.011% at a fixed false positive identification rate (FPIR) of 0.001% - the most stringent condition.

AI-powered Recognition technology

Verigram’s liveness detection uses AI, machine learning and biometrics to automate the verification process and help companies improve conversion rates and better detect fraud

  • Automatic face capture
  • 1 to 1 matching
  • Saas enabled
  • Face detection system

Use cases

All →