M26. Variational Methods for Inverse Problems in Imaging

Variational methods nowadays rank among the most powerful and flexible approaches to tackle inverse problems in imaging. These include real-world applications like magnetic resonance imaging, positron emission tomography, computer tomography, transmission electron microscopy and many other recovery problems in medicine, engineering, and life sciences. Further, variational methods have proven themselves highly valuable for data pre-processing and image post-processing such as denoising, deblurring and inpainting. The recent years have seen several new developments in this area, coming from different mathematical backgrounds and being applicable for a variety of inverse problems. These developments include, for instance, deep learning approaches, tensorial lifting strategies, as well as novel optimal-transport- and total-variation-based regularizers. On the other hand, new techniques such as deep neural networks for inverse problems are inspired by variational methods. The aim of this minisymposium is to bring together experts with various backgrounds to discuss these recent achievements in the context of inverse problems in imaging, and to initiate potential new research directions and collaborations.

Organizers:
Robert Beinert, Technische Universität Berlin, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.@tu-berlin.de
Kristian Bredies, University of Graz, Austria, This email address is being protected from spambots. You need JavaScript enabled to view it.

Invited Speakers:
Robert Beinert, Technische Universität Berlin, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.@tu-berlin.de
Robust PCA via Regularized REAPER and Matrix-Free Proximal Algorithms

Benjamin Berkels, RWTH Aachen University, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Joint exit wave reconstruction and image registration as a least-squares problem

Kristian Bredies, University of Graz, Austria, This email address is being protected from spambots. You need JavaScript enabled to view it.
Optimal-transport-based approaches for dynamic image reconstruction

Tatiana Bubba, University of Bath, UK, This email address is being protected from spambots. You need JavaScript enabled to view it.
Deep neural networks for inverse problems with pseudodifferential operators: An application to limited angle tomography

Vincent Duval, INRIA Paris, France, This email address is being protected from spambots. You need JavaScript enabled to view it.
Towards off-the-grid algorithms for total variation regularized inverse problems

Richard Huber, University of Graz, Austria, This email address is being protected from spambots. You need JavaScript enabled to view it.
Pixel-driven projection methods: Analysis and the Gratopy toolbox

Dirk Lorenz, TU Braunschweig, Germany, This email address is being protected from spambots. You need JavaScript enabled to view it.
Denoising of image gradients and total generalized variation

Ségolène Martin, University Paris-Saclay, France, segolene.martin@ centralesupelec.fr
Penalized methods for solving constrained variational problems in image recovery

Rudolf Stollberger, TU Graz, Austria, This email address is being protected from spambots. You need JavaScript enabled to view it.
Variational methods for functional and quantitative MRI

Robert Tovey, National Institute for Research in Computer Science and Control, INRIA, France, This email address is being protected from spambots. You need JavaScript enabled to view it.
Accelerating the solution of sparse dynamic inverse problems using tools from dynamical programming