Identixia
Request a Demo
Face Recognition SDK

Face Recognition SDK for 1:1 Verification and 1:N Identification

Match a live face against a reference image, or search a face against a gallery — on-device or server-side, deployed inside your own infrastructure.

Request Technical Demo View Documentation

Source available on GitHub. See pricing & licensing →

A selfie captured on a phone being matched against the portrait on an ID document, with a green checkmark showing a match
What it does

Compare and identify faces without leaving your infrastructure

The Face Recognition SDK extracts a mathematical representation of a face and compares it against one reference (1:1) or a gallery of many (1:N), returning a match decision your application controls.

1:1 Face Verification

Confirm a live face matches a specific reference image — the core of a selfie-to-ID check.

1:N Face Identification

Search a captured face against a gallery to find (or rule out) a match, without a pre-declared identity claim.

Face Matching & Comparison

Score similarity between any two face images for deduplication, watchlist screening, or repeat-user detection.

How it works

Capture, extract, compare — all inside your application

1

Capture

Face image or frame

2

Detect

Locate & align the face

3

Extract

Generate a face template

4

Compare

1:1 match or 1:N search

5

Decision

Match score returned to your app

Pairs naturally with the Face Liveness Detection SDK to confirm the face presented is a live person, not a photo or replay.

Key capabilities

  • ✓ 1:1 face verification
  • ✓ 1:N face identification against a gallery
  • ✓ Face matching and comparison scoring
  • ✓ On-device processing where supported
  • ✓ Server-side deployment for centralized workflows

On-device vs. server-side

Run matching directly on the device for the lowest latency and no network dependency, or deploy server-side when you need centralized galleries or high-throughput 1:N search. Both modes stay inside infrastructure you control — see on-premise deployment.

Supported platforms & languages

Integrate in the stack you already use

JavaKotlinSwift Objective-CC++Python FlutterReact Native Ionic CordovaIonic Capacitor.NET
AndroidiOSWindows LinuxServer / Docker
Use cases

Where teams use the Face Recognition SDK

Onboarding

Match a new user's selfie against their submitted ID photo during account creation.

Step-up authentication

Re-verify a returning user's identity for a high-risk action, without a password.

Duplicate account detection

Search new signups against an existing gallery to catch repeat or fraudulent accounts.

Licensing

Per App ID or per Machine — your choice

Mobile deployments license by App ID under a lifetime license; server deployments license per machine. Transaction-based tiers are also available for teams that prefer to license by verification volume.

See licensing details →
Layered diagram: your app, the Identixia SDK running inside it, and your infrastructure, with no call to a vendor cloud
FAQ

Frequently asked questions

Verification (1:1) confirms a face matches one specific reference. Identification (1:N) searches a face against a gallery of many candidates to find or rule out a match.
Yes, where supported by the target platform. On-device processing keeps face data on the device itself rather than requiring a network call.
In on-device deployment modes, yes — verification can complete without network access. Server-side deployments depend on connectivity to your own server, not to Identixia.
Yes. Most teams pair the Face Recognition SDK with the Face Liveness Detection SDK so a match is only trusted once the presented face is confirmed live.

Evaluate the Face Recognition SDK

Get a technical walkthrough with our engineering team.

Request Technical Demo View Documentation