Note Core dataset for iBeta Level 1 certification. 35,000+ videos covering paper-based presentation attacks (2D printouts, cutouts, eyeholes, wrapped/cylinder masks) and basic replay attacks (smartphone and PC display). Captured on iOS and Android with multi-ethnic balanced demographics. Ideal for eKYC, mobile onboarding, and entry-level liveness detection systems

Note Dataset for iBeta Level 2 PAD testing. Extends Level 1 coverage with more sophisticated 3D paper masks, wrapped/structural attacks, and advanced replay vectors (multiple brightness, distance, and angle variations). 25,000+ attack videos targeting banking, fintech, and government ID applications requiring stricter anti-spoofing compliance

Note Dataset for iBeta Level 3 — the strictest PAD certification tier (announced by iBeta in 2026). 10,000+ videos covering high-fidelity rubber and 3D resin masks, ultra-realistic 3D mask attacks, and the most advanced presentation attacks. For high-security use cases such as financial-grade authentication and regulated identity verification. Compliant with ISO/IEC 30107-3

Note Silicone mask presentation attack dataset — 12,500+ videos with 18 high-realism silicone masks for testing against the most advanced 3D spoofing vectors. Critical for iBeta Level 2 certification and high-security biometric applications (financial services, government ID, regulated KYC). Includes diverse skin tones and lighting conditions

Note Specialized 3D paper mask dataset for advanced anti-spoofing research. Includes printed photos with cylinder-based volume effects, structural facial elements (e.g., 3D noses), and multi-angle capture. Targets the gap between basic 2D paper attacks and full silicone-mask attacks - useful for iBeta Level 2 preparation and academic 3DMAD-aligned research

Note Display-based replay attack dataset capturing photo and video replays from monitors, laptops, and smartphone displays. Includes systematic variation across brightness levels, viewing distances, and angles to support robust replay-attack detection in face recognition systems. Required for iBeta Level 1 and Level 2 testing

Note Broader replay attack dataset covering multiple replay vectors — display, photo, and video replays - for face anti-spoofing and liveness detection systems. Complements the canonical academic Idiap Replay-Attack benchmark with modern smartphone capture conditions and broader demographic coverage. Sister dataset to Display Replay Attacks (specialized)