Scientific Computing Center (SCC)
The Scientific Computing Center (SCC) is the Information Technology Center of KIT. The junior research group “Robust and Efficient AI” at SCC develops scalable, energy-efficient machine learning methods for data-intensive applications in the natural sciences, such as weather forecasting and medical diagnostics. Our research combines modern AI with high-performance computing (HPC) to enable robust analysis of extremely large datasets.We are seeking a Postdoctoral Researcher to join the BMFTR-funded project MorphoSphere, which focuses on AI-based analysis of ultra-high-resolution 3D imaging data generated at large-scale neutron and synchrotron facilities (e.g. DESY, ESRF). These datasets enable advances in materials science and biomedical research, but they easily reach petabyte scale and cannot be processed with standard deep learning approaches. Within this project, you will develop and implement machine learning approaches that enable processing these large 3D image at full resolution.Your contributions will include:Developing parallel and distributed algorithms for processing ultra-large 3D images across multiple GPUsDesigning and adapting scalable deep learning models for segmentation and landmark detection (e.g. U-Net variants, Transformers)Extending and optimizing existing approaches within the HeAT software frameworkImplementing solutions using GPU-based HPC systems and modern supercomputing architecturesApplying and validating methods on real-world synchrotron and neutron imaging dataContributing to research software engineering (RSE) best practices, including maintainable, reproducible, and high-quality research codePublishing results in peer-reviewed venues and collaborating with domain scientists
Salary
Salary category 13 TV-L, depending on the fulfillment of professional and personal requirements.
Contract duration
limited up to 3 years
Application up to
16.03.2026
Contact person in line-management
For further information, please contact Dr. Charlotte Debus, charlotte.debus@kit.edu.
