DC Field | Value | Language |
---|---|---|
dc.contributor.author | Lim, Sungjun | ko |
dc.contributor.author | Park, Hyoungjun | ko |
dc.contributor.author | Lee, Sang-Eun | ko |
dc.contributor.author | Chang, Sunghoe | ko |
dc.contributor.author | Sim, Byeongsu | ko |
dc.contributor.author | Ye, Jong Chul | ko |
dc.date.accessioned | 2021-03-26T02:50:44Z | - |
dc.date.available | 2021-03-26T02:50:44Z | - |
dc.date.created | 2020-08-10 | - |
dc.date.created | 2020-08-10 | - |
dc.date.issued | 2020-07 | - |
dc.identifier.citation | IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING, v.6, pp.1127 - 1138 | - |
dc.identifier.issn | 2333-9403 | - |
dc.identifier.uri | http://hdl.handle.net/10203/281972 | - |
dc.description.abstract | Deconvolution microscopy has been extensively used to improve the resolution of the wide-field fluorescent microscopy, but the performance of classical approaches critically depends on the accuracy of a model and optimization algorithms. Recently, the convolutional neural network (CNN) approaches have been studied as a fast and high performance alternative. Unfortunately, the CNN approaches usually require matched high resolution images for supervised training. In this article, we present a novel unsupervised cycle-consistent generative adversarial network (cycleGAN) with a linear blur kernel, which can be used for both blind- and non-blind image deconvolution. In contrast to the conventional cycleGAN approaches that require two deep generators, the proposed cycleGAN approach needs only a single deep generator and a linear blur kernel, which significantly improves the robustness and efficiency of network training. We show that the proposed architecture is indeed a dual formulation of an optimal transport problem that uses a special form of the penalized least squares cost as a transport cost. Experimental results using simulated and real experimental data confirm the efficacy of the algorithm. | - |
dc.language | English | - |
dc.publisher | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC | - |
dc.title | CycleGAN With a Blur Kernel for Deconvolution Microscopy: Optimal Transport Geometry | - |
dc.type | Article | - |
dc.identifier.wosid | 000552269500004 | - |
dc.identifier.scopusid | 2-s2.0-85089338985 | - |
dc.type.rims | ART | - |
dc.citation.volume | 6 | - |
dc.citation.beginningpage | 1127 | - |
dc.citation.endingpage | 1138 | - |
dc.citation.publicationname | IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING | - |
dc.identifier.doi | 10.1109/TCI.2020.3006735 | - |
dc.contributor.localauthor | Ye, Jong Chul | - |
dc.contributor.nonIdAuthor | Lim, Sungjun | - |
dc.contributor.nonIdAuthor | Park, Hyoungjun | - |
dc.contributor.nonIdAuthor | Lee, Sang-Eun | - |
dc.contributor.nonIdAuthor | Chang, Sunghoe | - |
dc.description.isOpenAccess | N | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordAuthor | Deconvolution microscopy | - |
dc.subject.keywordAuthor | unsupervised learning | - |
dc.subject.keywordAuthor | generative adversarial network (GAN) | - |
dc.subject.keywordAuthor | cycle consistency | - |
dc.subject.keywordAuthor | optimal transport | - |
dc.subject.keywordAuthor | penalized least squares | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.