Theses: Master’s and Bachelor’s Thesis Projects

Frequent supervisor in master’s and bachelor’s thesis projects

I supervise computer vision and deep learning thesis projects with a focus on different learning paradigms. Key topics include semi-supervised and data-efficient segmentation of medical images, domain adaptation, and image-to-image translation. I also support research on large language models and visual in-context learning, with an emphasis on model efficiency and interactivity. Students gain hands-on experience developing advanced AI methods for real-world applications and improving model generalization with limited data. I frequently publish papers with my students at international conferences.

In case you are interested in working on your thesis with me feel free to send me a mail, let’s discuss interesting topics!

Topics and alumni:

• Visual in-Context Learning via Grid-based Image Inpainting with Diffusion Models, Master’s Thesis, Tim Le Large, 2026
• Benchmarking Visual in-Context Learning Methods and the Effect of Test-Time Fine-Tuning, Master’s Thesis, Pradnya Halady, 2026
• Parameter Efficient Visual In-context Learning on General Visual Imagery, Master’s Thesis, Sunil Khatri, 2026 (joint project with Department of Civil Engineering, Geo and Environmental Sciences)
• Explain by example: In-Context Learning with Vision Transformers for Retinal Image Analysis, Master’s Thesis, Benjamin-Philipp Roth, 2025
• On Compressing Large Language Models via Pruning, Master’s Thesis, Calvin Tanama, 2025 (joint supervision with Saquib Sarfraz)
• Exploring Interactions for Visual in-Context Learning, Master’s Thesis, Carlos Schmidt, 2025
• Synthetic Cataract Dataset: Generation and Evaluation for Deep Learning-based Scene Understanding in Ophthalmic Surgery, Bachelor’s Thesis, Erik Wu, 2025 (external project with Rebekka Peter from Carl Zeiss AG)
• An Exploration of Deep Learning for Transcatheter Aortic Valve Implantation based on CTs and Medical Guidelines, Bachelor’s Thesis, Cedric Zöllner, 2024
• Visual Task Representation Learning using Contrastive Training, Master’s Thesis, Jonas Muth, 2024
• Towards General Retinal Optical Coherence Tomography Analysis through in-Context Learning, Master’s Thesis, Alessio Negrini, 2024
• Image-to-Image Translation for Medical Microscopy with Deep Neural Networks, Master’s Thesis, Philipp Marquardt, 2024 (joint project with HS Analysis)
• Deep Learning Based Muscle Activation Analysis in Human Motion, Master’s Thesis, Marco Kugler, 2024 (joint supervision with David Schneider)
• Unsupervised and Weakly Supervised Domain Adaptation for Semantic Cell Organelle Segmentation in Electron Microscopy Images, Master’s Thesis, Dmitrii Seletkov, 2023
• Data-efficient Representation Learning for Cell Organelle Segmentation in Electron Microscopy Images, Master’s Thesis, Johannes Jestram, 2023
• Semi-supervised Retinal Fluid Segmentation in Optical Coherence Tomography Scans via Image-informed Pseudo-labeling, Master’s Thesis, Paul Kaiser, 2022