Probabilistic Noise2Void: Unsupervised Content-Aware Denoising

Simple item page
dc.contributor.authorKrull, Alexandercs
dc.contributor.authorVičar, Tomášcs
dc.contributor.authorPrakash, Mangalcs
dc.contributor.authorLalit, Manancs
dc.contributor.authorJug, Floriancs
dc.coverage.issue5cs
dc.coverage.volume2cs
dc.date.issued2020-02-19cs
dc.description.abstractToday, Convolutional Neural Networks (CNNs) are the leading method for image denoising. They are traditionally trained on pairs of images, which are often hard to obtain for practical applications. This motivates self-supervised training methods, such as Noise2Void (N2V) that operate on single noisy images. Self-supervised methods are, unfortunately, not competitive with models trained on image pairs. Here, we present Probabilistic Noise2Void (PN2V), a method to train CNNs to predict per-pixel intensity distributions. Combining these with a suitable description of the noise, we obtain a complete probabilistic model for the noisy observations and true signal in every pixel. We evaluate PN2V on publicly available microscopy datasets, under a broad range of noise regimes, and achieve competitive results with respect to supervised state-of-the-art methods.en
dc.formattextcs
dc.format.extent1-9cs
dc.format.mimetypeapplication/pdfcs
dc.identifier.citationFrontiers in Computer Science. 2020, vol. 2, issue 5, p. 1-9.en
dc.identifier.doi10.3389/fcomp.2020.00005cs
dc.identifier.issn2624-9898cs
dc.identifier.orcid0000-0002-9136-7873cs
dc.identifier.other159778cs
dc.identifier.researcheridC-6006-2018cs
dc.identifier.scopus57202426072cs
dc.identifier.urihttp://hdl.handle.net/11012/193231
dc.language.isoencs
dc.publisherFrontiers Media SAcs
dc.relation.ispartofFrontiers in Computer Sciencecs
dc.relation.urihttps://www.frontiersin.org/articles/10.3389/fcomp.2020.00005/fullcs
dc.rightsCreative Commons Attribution 4.0 Internationalcs
dc.rights.accessopenAccesscs
dc.rights.sherpahttp://www.sherpa.ac.uk/romeo/issn/2624-9898/cs
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/cs
dc.subjectdenoisingen
dc.subjectCAREen
dc.subjectdeep learningen
dc.subjectmicroscopy dataen
dc.subjectprobabilisticen
dc.titleProbabilistic Noise2Void: Unsupervised Content-Aware Denoisingen
dc.type.driverarticleen
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen
sync.item.dbidVAV-159778en
sync.item.dbtypeVAVen
sync.item.insts2025.02.03 15:39:43en
sync.item.modts2025.01.17 16:54:20en
thesis.grantorVysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. Ústav biomedicínského inženýrstvícs
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
fcomp0200005.pdf
Size:
2.96 MB
Format:
Adobe Portable Document Format
Description:
fcomp0200005.pdf
Download
Collections
Ústav biomedicínského inženýrství