About

I am a PhD student at HTW Berlin, working in the DFG-funded project Berlin Initiative for Applied Foundation Model Research (Appl-FM). My research asks how visual foundation models, pre-trained on natural images, can be adapted to specialised scientific domains such as microscopy, where labelled data is scarce. I work on fine-tuning with little data, synthetic data generation, and benchmarks for machine learning on scientific data.

I hold a degree in mathematics from Freie Universität Berlin and bring more than ten years of hands-on R&D experience in machine learning to my research.

PhD thesis · working title

Adaptation Techniques for Visual Foundation Models for Continuous Quantification in Microscopy

Foundation models transfer poorly to microscopy because of large domain gaps. The thesis studies fine-tuning on limited domain data and synthetic microscopy data to reach human-grade quantification with far less manual annotation.

Highlights

Selected publications

All publications →

Hover, tap or focus a card.