Email: xiaowei.xu@uni-leipzig.de
Office: D03.41 · Uferstraße 11, Leipzig, Germany
Structure-Aware Machine Learning & Medical AI
ScaDS.AI · Leipzig University
We are looking for motivated doctoral researchers to develop machine learning methods that use spatial, temporal, and distributional structures. Topics include medical image analysis, multimodal learning, robustness under distribution shifts, and efficient learning.
Research Environment: Founded in 1409, Leipzig University is one of Germany's oldest research universities. ScaDS.AI is one of Germany's five national AI competence centers, offering an excellent interdisciplinary research environment and access to powerful HPC infrastructure, including over 800 NVIDIA H100/A100 GPUs through shared computing facilities.
Research Focus: Structure-Aware Machine Learning, including representation learning, computer vision, spatial-temporal modeling, domain generalization, and medical AI.
Positions: 2 fully funded PhD positions (100% full-time, TV-L E13, subject to final confirmation).
Applicants with a background in computer science, machine learning, electrical engineering, mathematics, or a related field are welcome. Experience in deep learning and programming is valuable; an interest in methodological research and real-world applications is essential.
How to apply: Please email your CV and a brief description of your research interests, together with relevant publications or project links, if available, to xiaowei.xu@uni-leipzig.de. Suggested subject: PhD Application – Your Name.
Apply by email →I lead a Junior Research Group for Machine Learning at ScaDS.AI, Leipzig University. My research focuses on Structure-Aware Machine Learning: using spatial, temporal, and distributional structures as inductive biases to improve the scalability, robustness, and interpretability of learning systems.
Our group studies how structure in objects, processes, populations, and learned representations can guide machine learning. We connect methodological research in segmentation, multimodal learning, learning under distribution shifts, and efficient AI with applications in medical imaging and healthcare.
Previously, I was a Humboldt Research Fellow at the Leibniz-Institut für Analytische Wissenschaften – ISAS in Dortmund. At Guangdong Provincial People's Hospital, I worked as an Associate Professor and Principal Investigator, and co-directed the Laboratory of Artificial Intelligence and 3D Technologies for Cardiovascular Diseases. This experience shaped my research on AI for cardiovascular diagnosis, surgical planning, and clinical workflows.
From 2016 to 2019, I was a postdoctoral researcher in the Department of Computer Science and Engineering at the University of Notre Dame, working with Prof. Yiyu Shi on medical AI and hardware-efficient machine learning. I received my Ph.D. (2016) and B.S. (2011) from Huazhong University of Science and Technology. During my Ph.D. studies, I also visited the University of Alberta and the University at Buffalo.
I value close collaboration between machine learning researchers and clinicians, as well as reproducible research. Our datasets and code are available through GitHub and Kaggle. Please get in touch if you are interested in collaborating or joining the group.
