Multimedia Forensics: Biometric Presentation Attack Detection and Spoofing

This mock test covers the forensic science of biometric Presentation Attack Detection (PAD) -- the discipline of determining whether a biometric sample submitted to a sensor is from a live, genuine user or from a spoof artefact (Presentation Attack Instrument, PAI). Topics span the full PAD standards ecosystem (ISO/IEC 30107-1, 30107-2, 30107-3), fingerprint spoofing materials (gelatin gummy fingers from the landmark Matsumoto 2002 study, latex, silicone, wood glue PVA), iris spoofing (printed contact lenses, video replay, prosthetic eyes), face spoofing (printed photos, video replay, 3D masks, deepfake injection attacks), voice spoofing (replay attacks, text-to-speech synthesis, voice conversion, deepfake audio), and liveness detection methods (eye blink challenge-response, rPPG pulse detection, fingerprint sweat pore analysis, LBP texture analysis, deep CNN feature extraction). PAD error metrics -- APCER (Attack Presentation Classification Error Rate), BPCER (Bona-fide Presentation Classification Error Rate), and ACER (Average Classification Error Rate) as defined in ISO/IEC 30107-3 -- are tested with precision, as are multispectral imaging principles and challenge-response liveness limitations against real-time deepfakes.

Indian and benchmark context is directly integrated: UIDAI's Aadhaar Registered Device (RD) framework mandating on-device fingerprint and iris liveness detection (AUA/KUA circular, 2018) is examined, along with CDAC biometric research contributions. The test draws on the ASVspoof challenge series (2015, 2017, 2019) for voice anti-spoofing benchmarks and the LivDet competition series (2009 onwards) for fingerprint liveness detection evaluation protocols and PAI material categories (gelatine, latex, silicone, wood glue).

Topics covered:

• ISO/IEC 30107-1/2/3 framework, vocabulary, and testing standards
• APCER, BPCER, ACER -- definitions, error directions, and threshold trade-offs
• Fingerprint spoofing: Matsumoto gummy finger (gelatine), latex, silicone, wood glue (PVA)
• Iris spoofing: printed contact lens, video replay, challenge-response pupil dilation
• Face spoofing: Replay-Attack database, 3D masks, Apple Face ID structured light
• Voice spoofing: ASVspoof 2015/2017/2019 LA vs PA tracks, voice conversion artefacts
• Liveness detection: rPPG pulse, sweat pore, LBP micro-texture, multispectral subsurface
• UIDAI Aadhaar RD liveness mandate and AUA/KUA compliance requirements

Designed for UGC-NET Paper II Unit VIII (Multimedia Forensics), NFSU MSc Digital Forensics entrance, and FACT aspirants covering biometric system security. Allow 30 minutes.