Understand human behaviours with IoT sensors : from physical to physiological sensing

Abstract

This thesis explores the design of Internet of Things (IoT) sensing systems to understand human behaviours, in the physical and physiological aspects. The objectives of this thesis are three-fold: (i) to study and showcase new sensing principles to capture physical and physiological human behaviours, (ii) to design innovative algorithms and techniques to analyse the physical and physiological indicators, and (iii) to demonstrate various applications where behavioural sensing can be enabled with IoT sensors. These objectives are fulfilled with five research attempts, where each attempt presents the complete problem formulation, technical design, system implementation, application showcase and empirical study development cycle. The first two attempts focus on physical sensing and propose two novel eyewear interaction methods enabled by facial and rim gestures, which are achieved with miniature ultrasonic and vibrational sensors. The last three attempts focus on physiological sensing and present a series of new home-based pulmonary health management systems using ubiquitous speakers, earphones, and depth cameras.

Date of Award	2024
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Qian ZHANG (Supervisor) & Jin Zhang (Supervisor)