Blind Single Image Super-Resolution via Iterated Shared Prior Learning

Thomas Pinetz*, Erich Kobler, Thomas Pock, Alexander Effland

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review


In this paper, we adapt shared prior learning for blind single image super-resolution (SISR). From a variational perspective, we are aiming at minimizing an energy functional consisting of a learned data fidelity term and a data-driven prior, where the learnable parameters are computed in a mean-field optimal control problem. In the associated loss functional, we combine a supervised loss evaluated on synthesized observations and an unsupervised Wasserstein loss for real observations, in which local statistics of images with different resolutions are compared. In shared prior learning, only the parameters of the prior are shared among both loss functions. The kernel estimate is updated iteratively after each step of shared prior learning. In numerous numerical experiments, we achieve state-of-the-art results for blind SISR with a low number of learnable parameters and small training sets to account for real applications.

Original languageEnglish
Title of host publicationPattern Recognition - 44th DAGM German Conference, DAGM GCPR 2022, Proceedings
EditorsBjörn Andres, Florian Bernard, Daniel Cremers, Simone Frintrop, Bastian Goldlücke, Ivo Ihrke
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages15
ISBN (Print)9783031167874
Publication statusPublished - 2022
Event44th DAGM German Conference on Pattern Recognition: DAGM GCPR 2022 - Konstanz, Germany
Duration: 27 Sep 202230 Sep 2022

Publication series

NameLecture Notes in Computer Science
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference44th DAGM German Conference on Pattern Recognition
Abbreviated titleDAGM GCPR 2022

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'Blind Single Image Super-Resolution via Iterated Shared Prior Learning'. Together they form a unique fingerprint.

Cite this