0583748 20250317090929.520240306235959.9 85191338505 10.5220/0012397600003660 7 s. E cav_un_epca*0583747978-989-758-679-82184-4321559-566SetúbalSciTePress2024RadevaPetiaFurnariAntoninoBouatouchKadiSousaA. Augusto Blind Image Deconvolution Deep Image Prior No-Blur Variational Bayes cav_un_auth*0464277 Brožová Antonie UTIA-B Adaptivní systémy Department of Adaptive Systems AS AS CZ Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0101207 Šmídl Václav UTIA-B Adaptivní systémy Department of Adaptive Systems AS AS Ústav teorie informace a automatizace AV ČR, v. v. i. http://library.utia.cas.cz/separaty/2024/AS/brozova-0583748.pdf https://www.scitepress.org/Link.aspx?doi=10.5220/0012397600003660 GA20-27939S GA ČR cav_un_auth*0391986 GA24-10400S GA ČR CZ cav_un_auth*0464279 Blind image deconvolution (BID) is a severely ill-posed optimization problem requiring additional information, typically in the form of regularization. Deep image prior (DIP) promises to model a naturally looking image due to a well-chosen structure of a neural network. The use of DIP in BID results in a significant perfor-mance improvement in terms of average PSNR. In this contribution, we offer qualitative analysis of selected DIP-based methods w.r.t. two types of undesired solutions: blurred image (no-blur) and a visually corrupted image (solution with artifacts). We perform a sensitivity study showing which aspects of the DIP-based algorithms help to avoid which undesired mode. We confirm that the no-blur can be avoided using either sharp image prior or tuning of the hyperparameters of the optimizer. The artifact solution is a harder problem since variations that suppress the artifacts often suppress good solutions as well. Switching to the structural similarity index measure fro m L 2 norm in loss was found to be the most successful approach to mitigate the artifacts. cav_un_auth*0464278 International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024) /19./ 20240227 20240229 Roma IT IN 10000 10200 10201 2025 2 PR RVO:67985556 https://hdl.handle.net/11104/0353240 S 2024 85191338505 SCOPUS PUBMED WOS cav_un_epca*0583747 Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024) SciTePress 2024 Setúbal 559 566 978-989-758-679-8 2184-4321 Radeva Petia 340 Furnari Antonino 340 Bouatouch Kadi 340 Sousa A. Augusto 340