Mouna Gharbi

Welcome!

I am a postdoctoral researcher currently employed at University of Graz in Austria. I am part of the Department of Mathematics and scientific computing, working in the research group Inverse problems and Mathematical Imaging. In terms of research I am primarily interested in the intersection of model-based approaches and data-driven methods to develop mathematically sound, robust, convergent, efficient models to solve inverse problems. I am also eager to ensure the scalability and applicability of devised approaches to solve real-world problems. My research interests involve and are not restricted to:

• Monotone operator theory and convex optimization – I study variational and operator-theoretic methods for solving inverse problems in imaging, with applications ranging from chemical spectroscopy to medical imaging.
• Unrolling optimization algorithms – During my PhD, I investigated unrolled variants of classical iterative schemes, including half-quadratic majorization-minimization, primal-dual methods, and iterative soft-thresholding.
• Plug-and-play regularization – My current postdoctoral work focuses on plug-and-play regularization, exploring the integration of learned and model-based priors.
• Deep equilibrium models and implicit networks – I am increasingly interested in implicit models for inverse problems, particularly in bridging optimization algorithms and deep learning through equilibrium formulations.