bibtype	J - Journal Article
ARLID	0474858
utime	20240103214110.1
mtime	20170529235959.9
SCOPUS	85018507914
WOS	000399396400034
DOI	10.1109/TIP.2017.2676981
title (primary) (eng)	Blind Deconvolution With Model Discrepancies
specification	page_count 12 s. media_type P
serial	ARLID cav_un_epca*0253235 ISSN 1057-7149 title IEEE Transactions on Image Processing volume_id 26 volume 5 (2017) page_num 2533-2544 publisher name Institute of Electrical and Electronics Engineers
keyword	blind deconvolution
keyword	variational Bayes
keyword	automatic relevance determination
author (primary)	ARLID cav_un_auth*0293863 name1 Kotera name2 Jan full_dept (cz) Zpracování obrazové informace full_dept (eng) Department of Image Processing department (cz) ZOI department (eng) ZOI institution UTIA-B full_dept Department of Image Processing country CZ fullinstit Ústav teorie informace a automatizace AV ČR, v. v. i.
author	ARLID cav_un_auth*0101207 name1 Šmídl name2 Václav full_dept (cz) Adaptivní systémy full_dept Department of Adaptive Systems department (cz) AS department AS institution UTIA-B full_dept Department of Adaptive Systems fullinstit Ústav teorie informace a automatizace AV ČR, v. v. i.
author	ARLID cav_un_auth*0101209 name1 Šroubek name2 Filip full_dept (cz) Zpracování obrazové informace full_dept Department of Image Processing department (cz) ZOI department ZOI institution UTIA-B full_dept Department of Image Processing fullinstit Ústav teorie informace a automatizace AV ČR, v. v. i.
source	url http://library.utia.cas.cz/separaty/2017/ZOI/kotera-0474858.pdf
cas_special	project ARLID cav_un_auth*0292734 project_id GA13-29225S agency GA ČR project ARLID cav_un_auth*0314467 project_id GA15-16928S agency GA ČR abstract (eng) Blind deconvolution is a strongly ill-posed problem comprising of simultaneous blur and image estimation. Recent advances in prior modeling and/or inference methodology led to methods that started to perform reasonably well in real cases. However, as we show here, they tend to fail if the convolution model is violated even in a small part of the image. Methods based on variational Bayesian inference play a prominent role. In this paper, we use this inference in combination with the same prior for noise, image, and blur that belongs to the family of independent non-identical Gaussian distributions, known as the automatic relevance determination prior. We identify several important properties of this prior useful in blind deconvolution, namely, enforcing non-negativity of the blur kernel, favoring sharp images over blurred ones, and most importantly, handling non-Gaussian noise, which, as we demonstrate, is common in real scenarios. The presented method handles discrepancies in the convolution model, and thus extends applicability of blind deconvolution to real scenarios, such as photos blurred by camera motion and incorrect focus. RIV JD FORD0 20000 FORD1 20200 FORD2 20206 reportyear 2018 num_of_auth 3 mrcbC52 4 A hod 4ah 20231122142442.8 inst_support RVO:67985556 permalink http://hdl.handle.net/11104/0271794 mrcbC64 1 Department of Adaptive Systems UTIA-B 10201 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE mrcbC64 1 Department of Image Processing UTIA-B 10201 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE confidential S mrcbC86 2 Article Computer Science Artificial Intelligence|Engineering Electrical Electronic mrcbC86 2 Article Computer Science Artificial Intelligence|Engineering Electrical Electronic mrcbC86 2 Article Computer Science Artificial Intelligence|Engineering Electrical Electronic mrcbT16-e COMPUTERSCIENCEARTIFICIALINTELLIGENCE|ENGINEERINGELECTRICALELECTRONIC mrcbT16-j 1.817 mrcbT16-s 1.374 mrcbT16-B 91.212 mrcbT16-D Q1* mrcbT16-E Q1* arlyear 2017 mrcbTft \nSoubory v repozitáři: kotera-0474858.pdf mrcbU14 85018507914 SCOPUS mrcbU24 28278468 PUBMED mrcbU34 000399396400034 WOS mrcbU63 cav_un_epca*0253235 IEEE Transactions on Image Processing 1057-7149 1941-0042 Roč. 26 č. 5 2017 2533 2544 Institute of Electrical and Electronics Engineers