Face Recognition & Biometrics

Prove the person is real, and the right one.

Tie a live, present person to their verified identity document. IDV Pacific pairs face matching with IDV Liveness and deepfake defence, with an Australian-hosted option.

Active and passive liveness
Deepfake and injection aware
Verify (1:1) and recognise (1:N)
Australian-hosted option
<2s*
Selfie to verified result
1:1 + 1:N
Verify and recognise
30107-3*
ISO PAD-tested liveness
1M+*
Faces per watchlist
NIST-tested recognition engines and IDV Liveness, with an Australian-hosted deployment option.*
The Problem We Solve

A document check alone is not enough

A valid ID proves the document is real. It does not prove the person presenting it is real, present, and the rightful holder. That gap is where modern fraud lives.

Stolen or borrowed IDs

A genuine document in the wrong hands passes a document-only check.

Printed and screen spoofs

A photo, a screen replay or a mask held to the camera can fool a weak selfie check.

Deepfakes and injection

Off-the-shelf tools now feed synthetic video straight past the camera. This is the fast-growing 2026 threat.

No proof of a live person

Without liveness, there is no evidence a real human was actually present at capture.

How It Works

From selfie to verified person in seconds

One flow captures a face, confirms it belongs to a living person, and matches it to a trusted identity.

Step 1

Capture

A selfie on the user’s device, or a face captured on camera at a counter or gate.

Step 2

Liveness

Confirm a real, living person is present. Passive by default, active where you need it.

Step 3

Match

Compare the face to the ID portrait or a trusted record. This is the 1:1 check.

Step 4

Verified

Return a clear pass or fail with a confidence score, ready for your workflow.

Passive or active liveness Deepfake and injection aware Pass or fail with a score
Liveness & Anti-Spoofing

Two kinds of liveness, two kinds of attack

IDV Liveness proves a real person is present. Anti-spoofing stops the fakes, both the ones held to the camera and the ones injected past it.

Active liveness

User acts

The user follows a short prompt, such as a turn or a guided capture. Strong assurance for higher-risk steps.

  • Guided, on-screen prompt
  • Best for high-value actions
  • Clear evidence of presence

Passive liveness

Runs invisibly

Runs on a single selfie or frame with no extra effort from the user. Lowest friction, built for scale.

  • No user effort
  • Single frame or selfie
  • Highest completion rates

Presentation attacks

Printed photos, screen replays and 3D masks held up to the camera. Countered by presentation-attack detection (PAD), the discipline measured by ISO/IEC 30107-3 and certified at iBeta levels.*

PrintScreen replayMasksISO 30107-3

Injection & deepfake attacks

Synthetic or pre-recorded video fed straight into the capture layer through virtual-camera tools, bypassing the camera. This needs its own defence, separate from PAD.*

Virtual cameraDeepfakesReplay injection
PAD alone does not stop injection. A complete 2026 defence covers both threat classes. We can walk you through the certification level and coverage of the engine matched to your use case.*
Verify vs Recognise

One face, two questions

Face biometrics answer two different questions. Knowing which you need shapes the whole design.

1:1

Verify

Is this the document holder?

Match one live selfie to one reference: the portrait on an ID, an e-passport chip, or a trusted enrolled record. The safe starting point for onboarding and access.

Onboarding & KYC Account recovery Age assurance
1:N

Recognise

Have we seen this face before?

Search one face against many: an enrolled gallery or a watchlist. Used to spot duplicate accounts, known persons and repeat offenders.

Duplicate detection Watchlist Investigations
Enterprise Platform

Enterprise face recognition and monitoring

For security and government teams, our enterprise face-recognition platform adds 1:N recognition, alerting and access control. NIST-tested and deployable on-premise.*

Identity check

Cross-check a live face against an identity document for high-assurance verification.

Access control

Grant entry with liveness, so only the enrolled person is admitted.

Live watch

Monitor a live camera feed and raise a real-time alert on a watchlist match.

Video search

Search recorded video or a database for a known person after the fact.

Simultaneous multi-face Real-time alerting* 1M+ face galleries* Developer SDK Multiple cameras On-premise deployment
Use Cases

Where face biometrics earn their place

From remote onboarding to the front door, the same platform covers verify and recognise across your business.

Onboarding & KYC

Match a selfie to the customer’s ID at sign-up, with liveness, so a remote applicant is proven to be the genuine holder.

Account recovery & step-up

Re-verify a returning user with a quick face check before a password reset or a high-risk action.

Age assurance

Confirm a real person matches an age-bearing document for age-restricted products and venues. No age-assurance method is perfectly accurate, so results are best used with an age buffer and a fallback check near the threshold.*

Time & attendance

Contactless sign-on for staff and contractors, with a clear record of who was present and when.

Site & building access

Grant entry only to enrolled people, with liveness at the gate to stop photos and replays.

Watchlist & investigations

Alert on a person of interest in real time, or search recorded footage after an event.

Trusted across industries
Financial services Gaming, clubs & pubs Government Workforce & HR Property & access Telecommunications
Built for Australian Compliance

Ready for the 2026 identity landscape

Australia’s AML/CTF reforms moved to a risk-based model in 2026, with stronger controls, including biometrics, for higher-risk customers. Face verification is how you meet that bar without adding friction for everyone.

Regulatory references are general guidance, not legal advice. Confirm the exact obligations for your business with your compliance team.*

AML/CTF ready

Add biometric assurance for risk-based customer due diligence under the 2026 reforms.

Consent-first

Capture and record consent for biometric verification, with a fallback path where consent is withheld.*

Australian-hosted, configurable retention

On-premise or Australian hosting, with configurable retention and no-store options.*

Digital ID & FVS aware

Designed to sit alongside government Digital ID and the Face Verification Service as they open to commercial use.*

Accuracy & Proof

Independent accuracy, real-world results

The engines are measured on the NIST Face Recognition Technology Evaluation, the independent benchmark for 1:1 and 1:N accuracy. Every check returns a confidence score you can tune to your risk.*

NIST
FRTE-tested engines*
<2s*
Selfie to result
1M+*
Faces per watchlist
Multi
Faces matched at once*
FAQs

Frequently asked questions

The essentials on accuracy, liveness, deepfakes and privacy.

Accuracy depends on the engine and the capture conditions. We use engines measured on the NIST Face Recognition Technology Evaluation (FRTE), the independent benchmark for 1:1 and 1:N accuracy, and every result comes with a confidence score you can set a threshold on. We will confirm the current, dated figures for your chosen engine before go-live.*

IDV Liveness confirms a real, living person is present rather than a photo or a video. Passive liveness runs invisibly on a single selfie for the lowest friction. Active liveness asks the user to follow a short prompt for stronger assurance on high-risk steps. You can use either, or both.

There are two threat classes. Presentation attacks are spoofs held to the camera (photos, screens, masks), countered by presentation-attack detection measured to ISO/IEC 30107-3. Injection attacks feed synthetic or deepfake video past the camera through virtual-camera tools, and need their own defence. A complete 2026 setup covers both. We will confirm the certification level and coverage of the engine matched to your use case.*

Retention is configurable, including no-store options where the check runs and the image is discarded. We help you capture and record consent for biometric verification and provide a fallback path where consent is withheld, in line with regulator guidance. Exact data-handling is confirmed for your deployment.*

Yes. Our platform supports flexible on-premise deployment, and an Australian-hosted option is available for cloud. That keeps sensitive biometric data where your policy needs it.*

Australia’s 2026 AML/CTF reforms use a risk-based model where higher-risk customers need stronger controls, including biometrics. Face verification adds that assurance, and is designed to sit alongside government Digital ID and the Face Verification Service. It pairs naturally with IDV Pacific mDL, document validation and VEVO visa checks.*

Ready When You Are

See a live face match on your own flow

Tell us your use case and we will show liveness, face matching and deepfake defence working end to end. No obligation.

  • Verify (1:1) and recognise (1:N) in one platform
  • Active and passive liveness, deepfake and injection aware
  • Australian-hosted or on-premise deployment
  • Pairs with mDL, FoxID, eNRoL and VEVO checks
Request a call
A specialist will be in touch shortly.

    This site is protected by reCAPTCHA. Google Privacy Policy and Terms of Service apply.