I am a postdoctoral researcher at Technical University Dortmund, where I work on data-driven modelling of biophysical systems — particularly focusing on cell mechanics and cell motion. My research combines physics-based insight with modern computational approaches to understand complex biological processes.
I earned my PhD in Biophysics from the University of Cologne, specialising in advanced data analysis. During my doctoral studies, I explored innovative methods for traction force microscopy and the data-driven discovery of physical laws.
After completing my PhD, I joined Helmholtz-Zentrum Hereon as a postdoctoral researcher, developing symmetrical and physically constrained neural models for solving partial differential equations. Later, at Ludwig Maximilian University of Munich, I investigated stochastic automatic differentiation techniques for Monte Carlo processes.
My current research interests revolve around the quantification of uncertainty in inverse ill-posed problems — especially in image reconstruction and the identification of governing equations (ODEs, PDEs, and SDEs). To tackle these challenges, I integrate concepts from deep learning, Bayesian inference, optimisation algorithms, and information theory.
Beyond research, I am passionate about interdisciplinary collaboration and advancing reproducible computational science.
PhD in Biophysics, 2021
University of Cologne, Germany
MSc in Solid Mechanics, 2015
École Polytechnique, France
My research lies at the interface of machine learning and physics. I develop deep-learning and Bayesian methods that build physical structure, such as symmetries, conservation laws and sparsity, into the model, to solve ill-posed inverse problems and learn dynamical systems from data with quantified uncertainty.
Traction force microscopy infers the forces that cells exert on an elastic substrate, an ill-posed inverse problem. I developed Bayesian regularization methods released as open-source software, and now build hybrid Vision Transformer + U-Net networks that use cell-type metadata and generalize across scales and noise levels.
Neural surrogates accelerate numerical solvers but tend to accumulate errors over long rollouts. I design group-equivariant layers for staggered grids that enforce symmetries and conservation of mass and momentum as hard constraints, giving accurate and stable predictions of fluid flows and real ocean currents.
Identifying governing equations turns data into interpretable models. My sparse Bayesian learning method (ATSBL) selects the relevant terms of ODEs, PDEs and SDEs with uncertainty on every coefficient, and an active-learning feedback loop makes the inference of stochastic dynamics more data-efficient.
Peer-reviewed papers, preprints and theses on machine learning for physics and biophysics, from traction force microscopy to neural PDE surrogates and equation discovery.
No publications match.
Invited talks, conference presentations and posters at machine-learning and physics venues.
Teaching in machine learning for the physical sciences at the University of Hamburg.
Teaching assistant for the exercise sessions · University of Hamburg
Teaching assistant for the exercise sessions · University of Hamburg