What is face verification?
Face verification compares a person's face with a reference portrait to check whether they are the same person.
For identity verification, a live selfie can be matched to the portrait in a passport, ID card or supported digital credential. Liveness detection adds a check that a real person is present.
Where you would use it
Remote onboarding
Compare a live selfie with the verified credential portrait before accepting a new customer.
Check the rightful holder
Help detect someone presenting another person’s genuine passport, ID card or supported digital credential.
A stronger verification step
Require face matching and liveness alongside credential verification when a journey needs evidence that the holder is physically present.
Connecting a credential to its holder
A genuine document establishes a trusted source for identity data. Face verification connects its portrait to the person presenting it, helping detect the use of another person's credential.
The reference matters. Reading the portrait from a verified chip provides the issuer's digital image directly. A wallet credential can provide a reference where it includes an issuer-signed portrait.
Learn how chip-based documents and digital identity wallets establish the origin and integrity of that data.
How selfie-to-document matching works
Read the reference portrait
Retrieve the credential's portrait and verify the signed identity data against a trusted issuer.
Capture a live selfie
Guide the person through camera capture, checking image quality and liveness.
Compare the faces
Compare facial features in the selfie and reference portrait, then evaluate the result alongside the document and liveness checks.
Face matching and liveness detection
Face matching asks whether two faces belong to the same person. Liveness detection checks for a person present during capture and helps detect attempts to present a photograph, screen or other substitute.
The checks complement each other. A photograph may resemble the reference portrait, while a live person may be presenting someone else's document. Combining matching, liveness and credential verification addresses these different questions.
NIST explanation of presentation attack detection and livenessUnderstanding the match result
Face comparison evaluates similarity. A threshold determines which comparisons are accepted as matches. A stricter threshold can reduce false matches while increasing the chance of rejecting the rightful holder.
Lighting, camera angle, image quality and changes in appearance can affect the comparison. Clear capture guidance and a way to retry help legitimate users complete the journey.
Your acceptance policy brings the evidence together: the trusted credential, its validity, the face comparison and the liveness outcome.
NIST research on one-to-one face verificationOne-to-one or one-to-many?
Selfie-to-document verification is a one-to-one comparison: one person against one reference. It answers whether the person matches the identity being presented.
Face search searches a collection of existing portraits. It helps recognize returning users and detect possible duplicate accounts, including across different documents.
Handling face data
Explain why a selfie is requested and how the verification result will be used. Limit access to portraits and results to the people and systems that need them.
Our Data practices and Privacy Policy describe verification-session retention and handling.
Add face verification to your flow
PassportReader combines chip-based identity verification with face matching and liveness detection. Enable face_verification when creating a session to require the selfie step.
Use the hosted verification flow or embed the journey in your iOS or Android app with our native SDKs. The face check forms part of the credential verification journey.
Face matching and liveness detection are included in our verification plans. See pricing for included usage and rates.
Try face verification in the playground or follow the integration guide.