Deep regularization for medical image analysisMS5

Recent advances have shown that learning regularization is a powerful tool to obtain customized regularizers for many low-level vision applications. In this presentation, we show our latest results on extending the Fields of Experts regularization model by deepening this structure using a series of convolutional and deconvolutional operators. The aim is to regularize not only low-level features but also mid- to high-level features to increase the reconstruction performance for medical imaging.

This presentation is part of Minisymposium “MS5 - Learning and adaptive approaches in image processing (2 parts)
organized by: Kostas Papafitsoros (Weierstrass Institute Berlin) , Michael Hintermüller (Humboldt University and Weierstrass Institute Berlin) .

Erich Kobler (Graz University of Technology)
Teresa Klatzer (Graz University of Technology)
Kerstin Hammernik (Graz University of Technology)
Thomas Pock (Graz University of Technology)
deep learning, image reconstruction, machine learning