Human Sensing and Ubiquitous Computing 

Mi |


Research in the Intelligent Systems and Ubiquitous Computing Laboratory focuses on the development of smart mobile, wearable and household systems embedded with sensing, computing, and communication capabilities that can sense and understand human activities in living environments. Our research interests and efforts span the areas of ubiquitous computing, mobile sensing, wearable computing, smart homes, activity recognition, and affective computing by applying expertise in sensor technologies, embedded systems, machine learning, statistical signal processing, and human-computer interaction. We are particularly interested in developing intelligent sensing and computing technologies with a special focus on healthcare and medical applications. This fits well with a fascinating research field referred to as Mobile Health (mHealth), which aims to realize the vision of patient-centric personalized health care.

Wearable Sensor/Smartphone-based System for Activity Recognition and its Health Applications

Wearable sensor/smartphone-based human activity monitoring and recognition is one of the most fundamental and important problems in the field of ubiquitous and mobile computing since it opens the door to a wide range of health applications. Our research in this area focused on developing wearable sensor/smartphone-based human activity monitoring systems and innovative automatic human activity recognition techniques using standalone wearable inertial sensor and smartphone-embedded accelerometer sensor. Moreover, we have applied these systems and techniques to a variety of health applications including physical fitness monitoring, fall detection for elderly people, and sleep quality assessment. Our work advances the state-of-the-art wearable sensor/smartphone-based human activity monitoring and recognition technologies and demonstrates the feasibility and validity of these technologies in a number of significant health applications.

Wearable Sensor System for Human Physiological Body Sounds Monitoring and Analysis

Non-speech human physiological body sounds such as coughing, breathing, clearing throat, eating and drinking contain invaluable information about an individual’s health condition. For example, coughing sounds are indicative of lung conditions and tracking eating and drinking sounds provides clues to diet-related disorders. Our research has contributed significantly in designing wearable microphone-based sensing and computing system to capture human physiological body sounds and developing acoustic signal processing and statistical pattern recognition algorithms to automatically recognize these body sounds in real life noisy environment. Our work pioneers the wearable microphone-based physiological body sounds monitoring and analysis technologies and demonstrates the feasibility and validity of these technologies in real world settings.

Smart Household System for Indoor Air Pollution Sensing and Analysis

Indoor air quality (IAQ) plays a significant role in our life because poor IAQ could pose significant risks to people’s health and is the leading cause of respiratory infections, chronic lung diseases such as asthma, and cancers. However, IAQ is often overlooked because many air pollutants are colorless, odorless, or too tiny to be seen, making them almost impossible to be detected by human beings. Our research has contributed significantly in developing a home-based intelligent IAQ sensing and analytic system that can continuously monitor indoor air pollutants including PM 2.5 and volatile organic compounds (VOCs), identify pollution sources, estimate personal exposure to indoor air pollutants, and provide actionable suggestions to help people manage IAQ at their homes. Our work pioneers the smart household indoor air pollution sensing and analytic technology and demonstrates the feasibility and validity of this technology in real-world settings.