Abstract

Abstract: The ubiquity of interconnected systems has given rise to a world enriched with omnipresent computing. The proliferation of devices, embedded with sophisticated sensors, generates data at an unprecedented scale, presenting both opportunities and challenges for artificial intelligence (AI) systems in personal health monitoring. Self-supervised learning and decentralized AI are emerging as effective ways to harness the power of distributed data and computing resources, enabling the development of predictive models that form the bases of next-generation embedded intelligence. This talk will highlight our recent work on addressing the challenges associated with data-centric federated learning, developing foundation models, and unifying large language and sensory models for modalities, such as electrocardiograms, for comprehensive health analysis.

Biography: Aaqib Saeed is an Assistant Professor of On-body Sensing and Edge AI at Eindhoven University of Technology (TU/e) and the recipient of the prestigious AiNed Fellowship Grant. Prior to this, he served as a Research Scientist in AI at Philips Research. His research interests include deep learning, self-supervision, federated learning, and audio understanding for personal health. He received his Ph.D. cum laude from Eindhoven University of Technology. During his PhD, Aaqib also collaborated with renowned researchers at Google Brain through research internships. Currently, at TU/e, he is leading a Decentralized AI Research Lab focusing on developing self-learning systems with applications in healthcare.