Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.dojah.io/llms.txt

Use this file to discover all available pages before exploring further.

Face Match is a biometric comparison check that determines whether a selfie or live image matches the photo on a government-issued ID. It’s a critical step in confirming document ownership and detecting impersonation attempts during onboarding, account recovery, or high-trust actions. Dojah’s face match engine compares facial embeddings from two sources:
  • A selfie or liveness-captured image
  • A document photo (from ID, passport, or license)
The system returns a match result, confidence score, and fallback signals to guide your decision flow.

​
How it Works

1

Image Capture

You capture two images:
  • A selfie (either passive image or from a liveness session)
  • A document photo, usually auto-extracted from a submitted ID
2

Face Matching

Both images are compared using deep-learning models trained to detect fraud, obstructions, and minor facial changes (e.g., hairstyle, glasses).
3

Scoring & Result

The system returns a boolean match result and a confidence score between 0 and 100. The confidence threshold defines what is considered an acceptable match.

​
Integration Options

API Lookup

Send two image files (selfie and document image) via API. You’ll receive a match result, confidence score, and optional metadata.

SDK/Libraries

Face Match is bundled into the EasyOnboard flow. The SDK handles selfie capture with auto-focus, and give the required results.