Türkçe
English
РусскийBiometric face matching
The selfie is compared with the ID and chip photos; the false acceptance rate is 0.0001% and the false rejection rate is 0.87%.
Features
Face Verification and Liveness
Performs biometric matching with a NIST-listed face recognition model and applies active and passive liveness detection.
Active liveness
Random tasks are requested from a pool of eye, mouth and horizontal–vertical head movements.
Passive liveness
An AI model analyses liveness in the image without requiring any extra movement.
Deepfake and screen attack detection
AI-generated faces and recordings replayed from a screen are detected.
Multiple face detection
If more than one face is detected on screen, the session is flagged.
Age and gender consistency
The user’s estimated age and gender are checked for consistency with the ID data.
How It Works
Results in Four Steps
-
01
Automatic face capture
The face is captured automatically once it enters the frame.
-
02
Liveness
Active and passive liveness checks run with 1, 3 or 5 second options.
-
03
Matching
The face is compared with the ID and chip photos.
-
04
Risk check
Deepfake, multiple face and blacklist checks are added to the result.
Regulatory Compliance
Designed for Compliance
This solution works in line with the remote identification regulations your sector is subject to. See our security page for certifications and article-by-article compliance.
- MASAKCommuniqué No. 19All obliged parties; real persons and trade-registered legal entities
- MASAKCommuniqué No. 32Non-Turkish real persons and foreign representatives of legal entities
- BDDKBanksBanks and participation banks
- SPKIII-42.1Brokerages, portfolio management companies and crypto-asset service providers
Frequently Asked Questions
Your Questions, Answered
What is the difference between active and passive liveness?
Active liveness asks the user for random facial tasks such as eye, mouth or head movements. Passive liveness uses an AI model to analyse the image without asking for any extra movement. Prove ID can use both methods together.
Has face verification accuracy been independently tested?
Yes. Techsign’s face recognition model is listed in the NIST FRTE evaluation, and Techsign is the only Turkish company with a score above 90% at NIST. Face verification accuracy in banking is 99.13%.
Does liveness detection protect against deepfakes?
Yes. The passive AI model detects deepfakes and screen attacks, and random tasks that change every session prevent the use of pre-recorded videos.
Related Solutions
Explore all solutions
End-to-end digital onboarding
Remote Identity Verification (KYC)
Combines OCR, NFC, face matching, liveness and video call in a single flow for end-to-end remote identity verification.
ExploreVideo Call & AI Agent
Runs the BDDK steps end to end with an avatar-based digital assistant and customer representative screens.
ExploreNFC Chip Verification
Runs ID card and NFC passport chips through active and passive authentication and a certificate check.
ExploreSee Prove ID in your own flow
Let our team show you in a live demo how Prove ID adapts to your processes.