Dr. Jing Yuan (Ph.D., University Heidelberg, Germany)
Robarts Research Institute, Western University
Adjunct Research Professor
Department of Medical Biophysics, Schulich Medical School
email: cn.yuanjing AT gmail.com OR jyuan AT robarts.ca
I am working as the research scientist in Robarts Research Institute of Western University and as the adjunct research professor in the Department of Medical Biophysics of Schulich Medical School, Western University. I obtained my Ph.D degree with Excellence in Mathematics and Computer Science from Universitaet Heidelberg, Germany.
I constantly serve as the reviewer of the top journals and conferences: IEEE PAMI, IJCV, SIAM on Imaging Sciences, JMIV, CVPR, MICCAI, EMMCVPR, SSVM.
Please be free to contact me directly by email for discussions and collaborations in research and software development. I am happy to share from my experience, and I wish to be transparent about my studies and methods.
My research interests are high-performance optimization algorithms and associated optimization theories and analysis, with applications to image processing, medical imaging, computer vision and machine learning. During these two years, I especially focus on applying efficient convex optimization approaches to the challenging non-convex problems, especially with discrete-valued constraints, which can be mostly formulated by Markov Random Field (MRF) and information theory. Such problems widely distribute in computational imaging & vision, machine learning and many areas of medical imaging.
Under the perspective of convex optimization, my studies show that the proposed optimization methods not only provide elegant analytical tools in mathematical theory, but also result in efficient solvers in numerical practice. Promising mathematical results and algorithms were introduced to many different areas, e.g. mathematical image segmentation and labeling, fast level set techniques, optical flow estimation, medical image fusion and registration, clustering and machine learning etc, which lead to significant improvements over the state of art approaches.