Over the past decades, first-order operator splitting methods have become ubiquitous for many fields including signal/image processing and inverse problems owing to their simplicity and efficiency. In recent years, with the increasing of model complexity and data size, the needs for fast optimisation methods is becoming increasingly stronger. The aim of this mini-symposium is to highlight the recent advances in the acceleration of optimisation methods. The main topics of the mini-symposium will cover: inertial and acceleration schemes, preconditioning techniques, half quadratic regularisation, Krylov subspace, quasi-Newton method and other related ones.
- Proximal Interior Point Algorithm For Large Scale Image Processing Problems
- Marie-Caroline Corbineau (CentraleSupélec, Université Paris Saclay, Gif-sur-Yvette)
- Preconditioned Proximal-Point Methods for Imaging Applications
- Tuomo Valkonen (University of Liverpool)
- Adaptive Fista
- Peter Ochs (Saarland University)
- Inertial Proximal ADMM for Linearly Constrained Separable Convex Optimization
- Raymond H. Chan (Department of Mathematics, The Chinese University of Hong Kong)
- Accelerated Alternating Descent Methods for Dykstra-like Problems
- Pauline Tan (CMLA, École normale supérieure Paris-Saclay)
- Preconditioning and Acceleration Techniques for the Douglas-Rachford Iteration
- Kristian Bredies (Universität Graz)
- FSI Schemes: Fast Semi-Iterative Solvers for PDEs and Optimisation Methods
- Joachim Weickert (Saarland University)
- Imaging by Krylov Methods
- Silvia Gazzola (University of Bath)
- Make FISTA Faster Again
- Jingwei Liang (University of Cambridge)
- Organizers:
-
Jingwei Liang (University of Cambridge)
-
Mila Nikolova (CMLA - CNRS ENS Cachan, University Paris-Saclay)
-
Carola-Bibiane Schönlieb (University of Cambridge)
- Keywords:
- acceleration technique, half quadratic regularisation, image processing, inertial and acceleration schemes, inverse problems, krylov subspace, non-smooth optimisation, nonlinear optimization, numerical linear algebra, preconditioning techniques, quasi-newton method
