VerificationLiveness and face match
Liveness and face match A live person, a real camera.
Catch photos, masks, deepfakes and injected cameras. Then match the face to Aadhaar.
Prove a live person is in front of a real camera.
A live person on a real camera, in one take.
Every attack type
Photos, videos, masks, deepfakes and injected streams.
Instant, then final
An answer on the phone in one take, a verdict from the server.
Fails closed
When in doubt it denies or sends for review, never passes.
Trust the face only after you trust the camera.
A virtual camera or emulator is caught before the face is trusted.
- Live person
- Photo or screen
- Mask
- Deepfake
- Virtual camera
The face and the phone, read at the same moment.
- Photos, videos, masks, deepfakes and injected streams.
- An answer on the phone in one take, a verdict from the server.
- When in doubt it denies or sends for review, never passes.
Liveness checks
- Live person
- Photo or screen
- Mask
- Deepfake
- Virtual camera
ses_8f2c91abMatched to: Aadhaar portrait
- Verdict
- Server, fail closed
- Record
- Signed, replayable
Face search
Build the journey around risk.
- Run it only where the risk calls for it.
- Fall back to video kyc when it can’t finish.
- Change the order without an app release.
Start: take a selfie
Liveness and face matchLiveness, face match, face search
Next stepBind the phone and SIM
Approve
Every result, in one record.
- Sent to your app and your FRM.
- Each answer carries its reason code.
- Kept for your analysts and audit.
Returned to your app and FRMLIVE_PERSON
- Live person
- Clear
- Photo or screen
- Clear
- Mask
- Clear
- Deepfake
- Clear
- Virtual camera
- Clear
- Emulator
- Clear
Backed by the rest of DeepID.
Verification is one step. The same SDK reads the phone, links what it has seen before, and turns every check into one decision.
Device intelligenceThe phone is read at the same moment, so a virtual camera, an emulator or a cloned app is caught before anyone is trusted.Learn more Mule graphFaces, phones, PANs and accounts are linked across your customers, so one person behind many accounts shows.Learn more Journey builderChoose the checks, their order and the fallback for each market, and change them without an app release.Learn more DecisionEvery check becomes one answer with its reasons, sent to your app, your FRM and the audit log.Learn more
Questions banks ask.
Does liveness stop deepfakes and injected cameras?
It is built to. Liveness catches photos, videos, masks and deepfakes, and the device is read at the same time for a virtual camera, an emulator or hooking, which is how an injected stream reaches a selfie check. When in doubt it fails closed.
What does fail closed mean for our customers?
A doubtful check denies or goes to review; it never passes. A good customer gets an instant answer on the phone, and the server confirms it before the account opens, so most never notice the second step.
Is the face matched to Aadhaar?
To the portrait on Aadhaar or on the document the customer presented. Aadhaar face authentication, with the customer’s consent in the Aadhaar app, is a separate login route on the same SDK.
Can it find one person behind many accounts?
Yes. Face search runs across your customers, so the same live face on other accounts is flagged at sign-up. That is how a mule factory or a rented identity shows.
Is there a record we can show an auditor?
Every face check leaves a signed record that can be replayed. How long it is kept is agreed with your security team before the pilot.
Try one-tap login on your own number.
See the network confirm your number with nothing to type, and the next check take over when it can’t. Then plan a shadow proof on one journey, measured on your own traffic.