Research Scientist

Position overview

KU is seeking an experienced and highly motivated research scientist to work on devising efficient and adaptive machine learning for vision data analytics. This research project focuses on developing machine learning algorithms and techniques that can efficiently and effectively handle the challenges posed by big data, including the scarcity of annotated data, class imbalance, and online learning. It involves developing active learning strategies that can select the most informative samples for annotation, developing robust and scalable algorithms that can handle class imbalance, and devising online learning algorithms that can learn from data streams while adapting to changing environments and coping with data drift, model stability, bias, and fairness. Agile usage of the developed learning framework algorithm will also be accounted for by investigating proper real-time implementation models.

This research project aims to improve the scalability and accuracy of machine learning for visual data analytics by addressing the aforementioned challenges. This research will undergo validation through two testbed projects that are closely related to critical areas in healthcare and security. The first project will focus on the early detection of pathology in medical imaging, while the second project will involve the detection of threat items in baggage scans.

As part of the project, the researcher will have access to state-of-the-art research equipment and facilities at the C2PS and will get an opportunity to work alongside researchers and graduate students from various backgrounds as part of a wider team.

Position requirements

• In-depth knowledge in deep learning models for object/scene detection, classification, and recognition
• Strong record in publications in high venues
• Adhere to the university's information security and confidentiality policies and procedures, and report breaches or other security risks accordingly
• Perform any other tasks assigned by the line manager