ARCHIVES

Deepfake technology and AI-generated facial simulations pose major security challenges for modern identity verification systems. Traditional authentication methods rely on static biometric data such as facial images, videos, or voice recordings, which can be easily manipulated through spoofing attacks like replay attacks and screen-based impersonation. To address these issues, this project proposes an edge-based hardware architecture for privacy-preserving anti-deepfake physical identity verification. The system integrates sensors such as a camera module, infrared/depth sensor, and ambient light sensor with an embedded edge processor for real-time analysis. By detecting facial behaviors like eye blinking, micro-expressions, and depth variations, the system verifies the presence of a live human and identifies spoofing attempts.