Questions About Passive Liveness Detection

Active liveness requires the user to perform an action, such as blinking, smiling, or turning their head. Passive liveness detects the involuntary physiological signals of a living human without any user cooperation. Circadify reads rPPG blood flow patterns that are always present in a real person and cannot be faked by spoofing media. This eliminates user friction while providing a fundamentally stronger detection signal.

Deepfake generators synthesize realistic visual appearances but do not reproduce the subtle physiological signals present in real human faces. Circadify extracts the rPPG blood flow signal from the video feed and identifies when the expected cardiac-driven color fluctuations are absent, synthetic, or temporally inconsistent. This approach catches deepfakes regardless of their visual quality.

Circadify detects printed photo attacks, digital screen replay attacks, 3D mask attacks, and AI-generated deepfake video attacks. Each attack type lacks the genuine rPPG physiological signature that only a living human produces. The system returns both a liveness score and an attack type classification for forensic analysis.

Circadify provides a RESTful API with native SDKs for iOS, Android, and web platforms. You send a short video sequence from your existing camera capture step, and receive a liveness determination with confidence score and attack classification. Most identity platform teams complete integration within days. On-premise deployment is also available for air-gapped environments.

Circadify passive liveness is designed to align with ISO 30107 presentation attack detection standards and supports the biometric liveness requirements present in eKYC regulations across financial services and government identity frameworks. Our team works directly with compliance officers to document how the technology maps to specific regulatory requirements in your jurisdiction.

Circadify supports both cloud and on-premise deployment models. In cloud mode, video frames are processed in transit and are not stored after liveness determination. In on-premise mode, all processing occurs within your own infrastructure and no data leaves your environment. Both models are designed for organizations with strict data sovereignty and privacy requirements.