Liveness Detection / Anti-Spoofing for Face Verification

Build an anti-spoofing module that distinguishes a real human face on camera from a printed photo, a phone-screen replay attack, or a 3D mask. Combines passive cues (texture, depth-from-focus, micro-expression) + active cues (random eye-blink prompt, head-turn prompt). Critical infrastructure piece — Aarakshak relies on liveness to prevent identity fraud in exams + banking KYC.

Course Learning Outcomes (CLOs):
CLO1: Analyze the threat model for biometric spoofing (print / replay / mask attacks).
CLO2: Apply CNN-based passive anti-spoofing methods on a benchmark dataset.
CLO3: Design an active-liveness flow (eye-blink / head-turn challenge).
CLO4: Combine passive + active signals into a unified decision policy.
CLO5: Evaluate using Half Total Error Rate (HTER) and report failure modes.

Industry/societal relevance: Liveness is mandatory in Indian banking KYC (RBI master direction). Aarakshak uses liveness for exam proctoring + bank KYC. Direct placement track at IDfy, HyperVerge, Signzy, jocata, AuthBridge.