Towards General Intelligence Sensing Technologies

Speaker: Dr. Zhenyu Yan
Chinese University of Hong Kong

Title: Towards General Intelligence Sensing Technologies

Date: Monday, 16 March 2026

Time: 4:00pm to 5:00pm

Venue: Lecture Theater F
(Leung Yat Sing Lecture Theater), near lift 25/26, HKUST

Abstract:

The rapid proliferation of IoT devices has generated massive amounts of multi-modal sensory data, yet bridging the gap between physical sensing and general-purpose intelligence remains a challenge. This talk explores the design of intelligent systems that leverage machine learning algorithms and and Foundation Models (FMs) to enhance real-world sensing applications.

We first discuss a novel sonar-based system for drowning detection that overcomes environmental noise through physics-informed filtering. Next, we introduce a framework that enables open-set recognition on resource-constrained edge devices by distilling knowledge from cloud-based models. Moving beyond individual tasks, the talk presents methods utilizing LLM reasoning to proactively coordinate heterogeneous sensor networks and adapt to complex user intents. Finally, we examine an LLM-empowered diagnostic system integrating clinical guidelines with wearable data, alongside a smart-glass assistant for real-time social interaction support.

Biography:

Zhenyu Yan is an Assistant Professor at The Chinese University of Hong Kong. Dr. Yan has extensive experience in sensing systems, signal and information processing, cyber-physical systems, and machine learning in IoT systems. His works have been published in top international conferences and journals, such as MobiCom, SenSys, IPSN, IEEE Transactions on Mobile Computing, and ACM Transactions on Sensor Networks. He is the recipient of the Rising Star Award (Early Career Award) from ACM SIGBED China. His papers also received the Best Paper Award at ACM MobiCom 2025, the Best Community Contributions Award at ACM MobiCom 2023, the Best Paper Award Runner-up at ACM MobiCom 2022, and the Best Artifact Award Runner-up at ACM/IEEE IPSN 2021.