Server-Side Biometry
The Liveness check ensures that an online user is the right person, real person, and authenticating in real-time, effectively guarding against imposters and cyber threats. User’s face is compared with the provided trusted image, which is by default extracted from the verified document.
- Right person
- Real person
- Right now
Detection Modes
Verification can be performed in two modes: Express Liveness or Dynamic Liveness. Selected mode can be configured per use-case.
Express Liveness: Verifies it’s a real person in front of a camera, not someone attempting to spoof the system. Individuals simply align their face in the oval on the screen for a near-instant face capture. The imagery is analyzed for signs of spoofs presented to the camera, such as 2D and 3D masks, photos, and presented videos and deepfakes. The device is analyzed for signs of spoofs injected into the system, such as virtual cameras, emulators, code modifications, and Man-in-the-Middle attacks (MiTM).
Dynamic Liveness: Verifies it’s a real person and they’re authenticating in real time, delivering the highest level of identity assurance and mission-critical deepfake defense. The advanced security levels are achieved by patented Flashmark™ technology which illuminates the device screen during the face capture process. The reflection of light on the individual’s face confirms they’re genuinely present and delivers advanced mitigation of AI-based attacks, like deepfakes and face swaps.
Liveness Check Process
Liveness check process requires three sequential steps.
- Upload a trusted image
- Scan face during biometric session.
- Complete the liveness check and get the result
Trusted Image Upload
Trusted image is a photo of the user’s face that will be used for face matching. By default, we use the photo extracted from the verified document in the previous step. However, clients can use their own database of trusted images if they have one.
Trusted image requirements
- Supported media types: PNG, JPEG, and JP2
- Max image size: 2 MB
- The image should conform to: ISO/IEC 19794 Information technology — Biometric data interchange formats (Part 5 Face image data)
- Minimum distance between the eyes: 85 pixels
A face photo extracted from the ‘Document Verification’ step fulfils those requirements.
Face Scan
This is a process in which the user scans their face using a provided Liveness Check SDK. The results, together with the feedback codes, are available immediately after the scan via the mobile SDK, enabling the mobile application to react accordingly and help achieve a higher conversion rate in the event of errors relating to the conditions.
Read more about Detection modes.
Liveness Check Result
The verification status is either ‘SUCCESS’ or ‘FAILURE’.
The verification result also contains a frame of the user’s face from the biometrics session, which can be used as evidence.