didit-face-search

To detect duplicate user registrations, facial search performs a 1:N match against all previously verified sessions, returning ranked matches and similarity percentages to identify if the same face already exists in your identity workflows.

1:N face search works for blocklist risk detection by comparing a submitted user image against all previously verified sessions to flag blocklisted faces and prevent fraudulent account creations across identity workflows.

To perform a 1:N biometric face search via API, you need a valid DIDIT_API_KEY and must send a POST request containing a user_image file to the /v3/face-search/ endpoint to retrieve matching sessions.

Yes, you can use facial search ranking and similarity scores for automated decisioning in registration or identity workflows, as the API retrieves matching sessions with similarity percentages to support automated fraud prevention decisions.

You can prevent fraud by conducting a 1:N facial search that checks a new user image against all previously verified sessions, applying directly to registration deduplication and blocking blocklisted faces across identity workflows.

Face-search requires a POST request with a user_image file because the API needs the raw biometric image data to perform the 1:N facial comparison against verified sessions and accurately calculate similarity percentages for duplicate detection.