Name: Amir Hosein Afandizadeh Zargari
Chair: Prof. Fadi Kurdahi
Date: June 5, 2024
Time: 11 AM
Location: EH 3206
Committee: Prof. Peter Tseng, Prof. Amir Rahmani
Title: Advanced Machine Learning and AI Techniques for Enhancing Wearable Health Monitoring Systems
Abstract: This research advances wearable technology for real-time health monitoring using advanced machine learning (ML) and innovative hardware solutions. Key contributions include a novel ML-based model for accurate tracking of in-mouth nutrient sensors, employing a wide range of frequencies and advanced algorithms to enhance measurement reliability. An innovative use of Cycle Generative Adversarial Networks (CycleGAN) addresses motion artifacts in photoplethysmography (PPG) signals, transforming noisy data into clean, artifact-free signals without relying on accelerometers. The development of smart clothing systems with passive sensors and copper coils, using Near Field Communication (NFC) and energy harvesting technologies, ensures wireless connectivity and power in various environments, including underwater. AI-Coach, a personalized wearable system, integrates multiple sensors and a BERT model for real-time workout monitoring, adapting to individual needs for accurate activity tracking. The research also addresses security vulnerabilities threatening health monitoring systems with robust anomaly detection systems and adversarial training methods. These advancements aim to improve the accuracy, power efficiency, and security of wearable health devices, with significant implications for individual and public health monitoring.
Short Bio: Amir Hosein Afandizadeh Zargari is a Ph.D. candidate in Computer Engineering at the University of California, Irvine. He earned his B.Sc. in Computer Engineering from Sharif University of Technology in Iran in 2018 and his M.S. in Computer Engineering from UC Irvine in 2020. Throughout his master’s and Ph.D. studies, Amir has published multiple papers in prestigious journals, including Natures, ACMs, IEEEs, and conferences like ESWEEK. His research focuses on advancing wearable technology for real-time health monitoring using machine learning and artificial intelligence.