Avoiding Undesirable Solutions of Deep Blind Image Deconvolution

0583748 - ÚTIA 2025 RIV PT eng C - Conference Paper (international conference)
Brožová, Antonie - Šmídl, Václav
Avoiding Undesirable Solutions of Deep Blind Image Deconvolution.
Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024). Setúbal: SciTePress, 2024 - (Radeva, P.; Furnari, A.; Bouatouch, K.; Sousa, A.), s. 559-566. ISBN 978-989-758-679-8. ISSN 2184-4321.
[International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024) /19./. Roma (IT), 27.02.2024-29.02.2024]
R&D Projects: GA ČR GA20-27939S; GA ČR(CZ) GA24-10400S
Institutional support: RVO:67985556
Keywords : Blind Image Deconvolution * Deep Image Prior * No-Blur * Variational Bayes
OECD category: Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
http://library.utia.cas.cz/separaty/2024/AS/brozova-0583748.pdf

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.
Permanent Link: https://hdl.handle.net/11104/0353240