Bio-signal Authentication Algorithm
Your identity information can be inferred from other biosignals.
Developed a biometric authentication system leveraging the Body Channel Coupling Effect, analyzing human channel path loss under varying stimulation frequencies. Designed and implemented deep learning algorithms for precise feature extraction and pattern recognition, achieving a False Rejection Rate (FRR) of 1.25% and a False Acceptance Rate (FAR) of 0%.
To address open environments and multi-user verification scenarios, proposed a multi-stage hybrid deep learning architecture, enabling both high efficiency and robust security. Conducted hardware deployment evaluations to ensure the algorithm remains lightweight and capable of real-time execution on microprocessors, demonstrating a seamless integration of advanced signal analysis, AI, and embedded system design.