BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Sparsity-Inspired Regularization for Image Reconstruction
DTSTART:20231205T161500
DTEND:20231205T171500
DTSTAMP:20260406T170127Z
UID:84e3e1a07a7745b584a05a139818d18e2ebf9cb70f3f65e15c0fa1e1
CATEGORIES:Conferences - Seminars
DESCRIPTION:Sébastian Neumayer (EPFL)\nIn this talk\, I will introduce a 
 generic framework for learning filter-based regularization functionals fro
 m image data. If we pursue a variational reconstruction ansatz for solving
  inverse problems\, these can be deployed to a variety of different imagin
 g modalities (universality). Further\, this ansatz ensures data consistenc
 y and we are able to derive some stability guarantees. Obeying with such p
 aradigms is very important when working in critical applications such as m
 edical imaging\, since false diagnosis can have fatal consequences. After 
 introducing the baseline architecture\, I will discuss an improvement of t
 his architecture via conditioning on the data. In the last part of the tal
 k\, I will present numerical results for denoising and MRI. These indicate
  that even relatively restricted architectures can be able to achieve high
 ly competitive performance.\n 
LOCATION:GA 3 21 https://plan.epfl.ch/?room==GA%203%2021
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR