Liveness Detection

How It Works

Liveness systems fall into two main categories:

1. Active liveness — asks the user to perform an action (blink, smile, turn head). Effective but adds friction and fails at a distance or for non-cooperative subjects.
2. Passive liveness — analyzes a single frame or short clip without user action, looking for texture, reflection, depth, and micro-motion cues that distinguish real skin from printed or screen images.

Modern engines combine multiple signals: 3D depth (stereo or structured light), infrared, color-space analysis, and deep learning classifiers trained on spoofing attacks.

Why It Matters

Without liveness, face recognition is trivially bypassed:

• A printed photo can unlock a door.
• A phone screen displaying a social-media photo can clock into attendance.
• A high-resolution video replay can impersonate a VIP or employee.

Liveness closes this gap and is a requirement for any identity-critical deployment. IncoreSoft's face recognition module integrates liveness directly into its enrollment and matching pipeline so spoofing attempts are rejected automatically.

Use Cases

• Access control — doors, turnstiles, data centers, secure areas
• Attendance — prevent buddy-punching with photos of absent workers
• Remote onboarding — KYC for financial services and telecom
• Border control and airports — passenger identity verification
• Payment authorization — face-based approval for high-value transactions

Frequently Asked Questions

What's the difference between active and passive liveness?

Active liveness requires the user to perform an action; passive liveness works from a normal face capture with no cooperation required. Passive is smoother for users but technically harder.

Can liveness detect deepfakes?

High-quality passive liveness systems detect most deepfakes by analyzing inconsistencies in texture, depth, and temporal motion. Dedicated deepfake detection is an active research area, and state-of-the-art models are continuously updated.

Does liveness require special camera hardware?

Not necessarily. Passive 2D liveness runs on standard RGB cameras. Higher-security deployments use 3D depth (stereo or structured light) or IR sensors to defeat sophisticated masks.