The development of efficient optimization methods applicable to large scale image processing tasks is an important and current research topic as it leads to improvements in speed and stability or the ability of revealing hidden image properties. The goal of this minisymposium is to discuss and compare image and video processing tasks addressed by means of powerful and efficient optimization algorithms, with particular attention devoted to non-smooth or non convex cost functions. The presentations in this minisymposium discuss both theoretical aspects as well as concrete applications of state-of-the-art optimisation methods relevant to modern mathematical imaging.
- Inexact forward-backward and primal-dual methods for applied inverse problems
- Julian Rasch (Westfälische Wilhelms-Universität Münster)
- Non-smooth non-convex Bregman minimization: unification and new algorithms
- Peter Ochs (Saarland University)
- A block coordinate proximal algorithm for N-th order tensor decomposition
- Caroline Chaux (Aix-Marseille Université)
- TV-based Poisson image restoration by IRLS and gradient projection methods
- Daniela di Serafino (University of Campania "L. Vanvitelli")
- A variational Bayesian approach for image restoration with Poisson-Gaussian noise
- Jean-Christophe Pesquet (Université Paris-Saclay)
- Proximity operator of a sum of functions: Application to image segmentation
- Nelly Pustelnik (CNRS, Laboratoire de Physique de l'ENS de Lyon)
- Fast algorithms for nonlocal myriad filtering
- Gabriele Steidl (University of Kaiserslautern)
- A fast algorithm for non-convex optimization in highly under-sampled MRI
- Fabiana Zama (Dept. Mathematics, University of Bologna)
- Organizers:
-
Ignace Loris (Université Libre de Bruxelles)
-
Marco Prato (University of Modena and Reggio Emilia)
- Keywords:
- image reconstruction, inverse problems, nonlinear optimization