16:00-18:00 p.m., August 28, 2023, GMT+8
Room 1501, Science Building #1 (Yanyuan)
In this talk, I will present machine learning (ML) based health assessment using wearable sensors. First, I will briefly introduce my previous works on ML-based computational behaviour analysis, after that I will focus on how to use these techniques (and wearable sensors) for automated health monitoring and assessment. Given a medical data analysis task, depending on the resources available (annotation, interpretability requirement, clinicians’ input, expected outcome, etc.), one should decide what is the best strategy to model the problem. Taking both accuracy and interpretability into account, here I will present a general processing pipeline, based on which some of our most recent sensor-based health assessment works will be introduced, including Perinatal Stroke Screening for infants, Parkinson’s disease classification, Remote Stroke Rehabilitation Monitoring, Automated Sleep stage classification, and fatigue assessment. This pipeline can effectively facilitate the collaboration between clinicians and data scientists, and at last I will also give some future works.
Dr. Yu Guan (Associate Professor, University of Warwick) received his PhD in 2015 from the University of Warwick. Then he worked as a Postdoc research associate and lecturer in Newcastle University, before joining Warwick University as an Associate Professor in 2022. His research interests include machine learning, activity/behaviour analytics, biometrics/forensics, computer vision, and ubiquitous/wearable computing. He has published more than 70 peer-reviewed papers including top venues like IEEE Trans PAMI, IEEE Trans IP, IEEE CVPR, ECCV, ACM IMWUT, ACM Multimedia, etc. Currently, he is the associate editor of ACM IMWUT and also the associate editor of Frontiers in Computer Science. Yu was the principle investigator of Cambridge-Newcastle Open Movement Collaboration (£100K), and also the co-investigator in several projects, e.g., EPSRC DERC: Digital Economy Research Centre (£4.1M), EPSRC Centre for digital citizens (£3.8M), IMI IDEA-FAST: Identifying Digital Endpoints to Assess Fatigue, Sleep, and Activities daily living in Neurodegenerative disorders and Immune-mediated inflammatory diseases (21M Euros).
School of Computer Sciences