Glossary · Technology
Liveness detection
Liveness detection is the check that a face presented to a camera belongs to a living person who is physically present, rather than a photo, a screen, a mask or a deepfake stream. It is part of identity verification and video-call security, not of face search, but it is the main technical answer to the face swaps and replayed images that face search cannot detect.
Updated November 6, 2026 · 2 min read
What problem does it solve?
Face recognition answers "is this the same face". It does not ask whether the face is real. A printed photo held to a camera, a video played on a phone, a silicone mask or a real-time deepfake injected into a video stream can all pass a plain face match. Liveness detection, called presentation attack detection in the standards, is the layer that rejects those. It matters for bank onboarding, exam proctoring, SIM registration and, increasingly, for the video calls that scammers now fake.
Active or passive?
| Approach | What the user does | What the system checks | Trade-off |
|---|---|---|---|
| Active | Turns the head, blinks, reads digits, follows a dot | The response matches the challenge and arrives with natural timing | Harder to spoof, more friction, still beatable by live face swaps that track the user's motion |
| Passive | Nothing; a short selfie or clip | Texture, depth cues, reflections, screen moiré, blood-flow signals, 3D structure from a moving camera | Smooth, depends on image quality and the model's training data |
| Injection detection | Nothing | Whether the camera feed itself was replaced by software | Needed against deepfake streams on video calls, outside the ISO scope |
Is there a standard?
ISO/IEC 30107-3 defines how presentation attack detection is tested, and accredited laboratories (iBeta is the best known) certify systems at Level 1 (paper and screen attacks) and Level 2 (masks and more elaborate props). Certification says a system resisted a defined set of attacks in a lab; it does not cover software injection or attacks invented after the test.
Does face search use liveness detection?
No. A reverse face search compares an uploaded photo with photos on public pages; nothing is live, so liveness is not a concept there. Where it meets face search is in verifying someone you met online: a live video call with simple active checks (turn your head, touch your nose) is the practical defence against a profile whose photos a search cannot trace, and it defeats most consumer face-swap tools, which smear at the boundary when the head turns fast.
Questions
Can a deepfake pass liveness detection?
Some can, especially when the attacker injects a synthetic stream rather than holding a screen to the camera. Vendors add injection detection and device attestation for that reason. Treat liveness as one layer, not a verdict.
Does FaceSearch or PimEyes check liveness?
No face search engine does. They compare still images with public pages. Liveness belongs to identity verification and video security products.
Sources
- 1.ISO/IEC 30107-3:2023, Biometric presentation attack detection, Part 3: Testing and reporting (checked November 6, 2026)
- 2.NIST, Face Technology Evaluations (FRTE/FRVT) overview (checked November 6, 2026)
Entry reviewed November 6, 2026. Found an error? Tell us.
Related entries
- Term Deepfake Deepfake defined: faces, voices and video synthesised or swapped by machine learning, how they are made and detected, and the laws that cover them.
- Term Face recognition Face recognition defined: comparing faces to decide whether they match, the difference between verification, identification and search, and how it is regulated.
- Concept Deepfakes and synthetic faces How deepfakes and AI-generated faces are made, why a face search cannot find a face that never existed, the tells that still work, and the laws that now apply.
- Concept Accuracy, bias and testing How face recognition accuracy is measured, what NIST FRTE reports found on error rates and demographic differences, and why marketing claims cannot be checked.
Cite this entry
FaceSearch Wiki editors. "Liveness detection". FaceSearch Wiki, updated November 6, 2026. https://facesearch.wiki/glossary/liveness-detection/