Euijoon Ahn
- euijoon.ahn@jcu.edu.au
- https://orcid.org/0000-0001-7027-067X
- Lecturer, Information Technology
Projects
5
Publications
29
Awards
4
Biography
I am a Lecturer at the College of Science and Engineering, James Cook University. Prior to this, I was a postdoctoral research fellow at the School of Computer Science, The University of Sydney.
I obtained my PhD degree in Computer Science (medical image analysis) from The University of Sydney in 2020. I received B. IT degree from The University of Newcastle, Australia, 2009 and M. IT (2014) and MPhil (2016) degree from The University of Sydney.
I have produced top-tier publications in the area of computer vision and medical image computing, including papers in IEEE T-MI, T-BME, JBHI, MedIA, PR, CVPR, AAAI and MICCAI. I am a regular reviewer for IEEE T-PAMI, T-MI, Nature Communications, CVPR, MICCAI and ISBI.
I am always looking for highly-motivated undergraduate (Honours) and postgraduate students (MPhil and PhD). Please send me your CV and transcripts.
Available / Current Projects
Unsupervised / Self-supervised learning for Image Classification and Object Recognition
The availability of large-scale annotated image datasets and recent advances in supervised deep learning methods have made important breakthroughs in natural scene analysis, visual object detection / classification and medical image analysis. These deep learning approaches, however, require large labelled training data (e.g., ImageNet archive with over 1 million images) and the labelling must be done manually which is costly, slow, and can be subjective / prone to errors if specialised skills are required. This project aims to study various techniques of unsupervised / self-supervised learning representation learning, and to develop new approaches for various applications (e.g., computer-aided diagnosis) in the medical imaging domain (X-ray, Ultrasound, microscopy, MRI, CT, PET-CT).
A Deep Learning Framework for Hidden Pattern Discovery
This project aims to develop a machine learning framework, which is not dependent on labelled data, to discover hidden patterns within data. These ‘hidden patterns’ represent meaningful new knowledge. Existing machine learning methods learn data features that are directly associated with labels and require large amounts of labelled data. This project expects to develop an unsupervised deep learning framework that will learn features that are independent of labels from large unlabelled image data archives. The expected outcome is a novel image analytic framework that can be applied to a range of scientific applications.
Telehealth – AI-enabled Remote Patient Monitoring
Remote patient monitoring (RPM) has seen tremendous uptake in usage due to the outbreak of COVID-19 and transition to remote patient care, and with it, a rapid rise in the development of new technologies. In an RPM setting, accurate remote measurement of vital signs such as heart rate (HR), blood pressure (BP) and temperature are essential in accurate diagnosis and screening of diseases. Current measurement methods, however, are constrained by wearable sensors or inflatable cuff-based devices that could be inconvenient, uncomfortable and are subject to variability. Development of a system that can measure such vital signs conveniently and comfortably, in a contactless and COVID-safe manner, is therefore of important requirement for use in RPM. This project aims to develop AI-enabled deep learning techniques that estimate vital signs using facial videos.
Research
Research Interests
Keywords: Deep Learning, Machine Learning, Data Science, Medical Image Analysis, Health Informatics, Telehealth
My research focus is on the development of Machine Learning and Deep Learning, Computer Vision and, more specifically, unsupervised and self-supervised deep learning models for biomedical image analysis, for improving image segmentation, retrieval, quantification and classification without relying on labelled data.
I also work at the coalface of translational health technology researches, e.g., health data analytics and telehealth.