报告题目：Machine Learning and Its Applications in Image Processing

报告时间：2024年10月31日（星期四）19:20-21:30

报告地点：机械馆J3

报告人：刘宗华（罗伯特·戈登大学（英国））

报告人简介：

Dr. Zonghua Liu is now a Lecturer in Electronic and Electrical Engineering in the School of Computing, Engineering and Technology at the Robert Gordon University, UK. He is a member of IEEE and IET. He is also a committee member of the IEEE OES UKRI Chapter. He received his B.Eng. degree in Applied Physics, M.SC. degree in Laser Technology degrees in China, and Ph.D. degree in Engineering in the University of Aberdeen, UK. He has much research experience in underwater sensors, underwater optical imaging and sensing, and machine-learning-based image processing. He has a good track record and is often involved in national/international collaborations (researcher in a UK-Japan project and an EU project, Co-I in two Japan projects and a Turing Ecosystem Leadership Award). He worked with a UK group and a Japanese group to develop a novel subsea camera which first combined optical holography and Raman spectroscopy into a single subsea device. His work has been highly respected proven by invitations as a speaker at international conferences/workshops. He has published in international journals and conference proceedings of high reputation, such as IEEE Journal of Oceanic Engineering, Optical Express, Journal of the Optical Society of America, and IEEE OES/MTS OCEANS Conferences. He is an invited Reviewer for IEEE OES/MTS Oceans. He also often reviews manuscripts for IEEE Journal of Oceanic Engineering, Artificial Intelligence Review, Journal of the Optical Society of America A, Applied Optics, The Visual Computer, and IEEE International Conference on Robotics and Automation.

His current research interests include but are not limited to robotics (e.g., underwater and medical), underwater/medical sensors, optical fibre sensors, computer vision, machine-learning-based signal and image processing, water/environment quality monitoring, real-time reporting systems using AI and cloud computing.

报告内容简介：

Machine learning (ML) is a subfield of artificial intelligence (AI) focusing on building algorithmic systems that learn from data and improve their performance over time without being explicitly programmed for specific tasks. ML has rapidly developed over the past decade due to the dramatic development of computational resources (e.g. CPUs and GPUs) and availability of large data. It has a wide range of applications across industry and academia, including various research. Therefore, ML is becoming an essential technique and skill for researchers.

This presentation will briefly introduce ML, its types and corresponding learning algorithms and applications. Afterwards, four cases based on my research will be shown where ML was used for particle classification, feature extraction, image segmentation, and object detection and identification. Lastly, a short discussion will be made about the future directions of ML.