Liveness and Anti-Spoofing

How is Liveness and Face Matching delivered

Liveness detection is a crucial step in facial biometrics to ensure that the user present is a live individual and not a spoof, such as a photograph, video, or mask. Our One-Time Verification provides advanced facial recognition capabilities service, which includes liveness detection features to enhance security and user verification processes.

This guide explains how liveness detection works and how we have integrated it into our system.

Key Features of Liveness Detection

• Anti-Spoofing Mechanisms: Detects and prevents the use of fake representations such as photos, videos, or masks.

• Real-Time Analysis: Performs live checks to confirm the presence of a real person during authentication.

How it works

Liveness detection and facial matching are integral components of our One-Time Verification process, detailed as follows:

1. Initial Capture

   • The user is prompted to position their face within the camera’s view.

   • High-quality images or video frames are captured for analysis within milliseconds.

2. Facial Feature Analysis

   • Our Face service analyzes the captured data to detect and map facial features such as eyes, nose, mouth, and contours.

3. Liveness Verification

   • Our system performs checks to ensure the presented face is live. Techniques include:

     • Blink Detection: Confirms natural eye movements.

     • 3D Face Mapping: Analyzes depth and spatial information to distinguish between 2D images and a real face.

     • Texture Analysis: Examines skin texture to identify fake surfaces.

4. AI-Powered Decision Making

   • Our AI model evaluates the data and determines if the face is live or spoofed.

   • Results are returned with a confidence score indicating the likelihood of liveness.

5. Authentication or Rejection

   • If liveness is confirmed, the user proceeds with authentication.

   • If a spoof is detected, the process is terminated, and the user is rejected.