Self-supervised Blur Detection from Synthetically Blurred Scenes

Simple item page
dc.contributor.authorAlvarez-Gila, Aitor
dc.contributor.authorGaldran, Adrian
dc.contributor.authorGarrote, Estibaliz
dc.contributor.authorVan de Weijer, Joost
dc.contributor.institutionTecnalia Research & Innovation
dc.contributor.institutionVISUAL
dc.contributor.institutionQuantum
dc.date.issued2019-12
dc.descriptionPublisher Copyright: © 2019 Elsevier B.V.
dc.description.abstractBlur detection aims at segmenting the blurred areas of a given image. Recent deep learning-based methods approach this problem by learning an end-to-end mapping between the blurred input and a binary mask representing the localization of its blurred areas. Nevertheless, the effectiveness of such deep models is limited due to the scarcity of datasets annotated in terms of blur segmentation, as blur annotation is labour intensive. In this work, we bypass the need for such annotated datasets for end-to-end learning, and instead rely on object proposals and a model for blur generation in order to produce a dataset of synthetically blurred images. This allows us to perform self-supervised learning over the generated image and ground truth blur mask pairs using CNNs, defining a framework that can be employed in purely self-supervised, weakly supervised or semi-supervised configurations. Interestingly, experimental results of such setups over the largest blur segmentation datasets available show that this approach achieves state of the art results in blur segmentation, even without ever observing any real blurred image.en
dc.description.statusPeer reviewed
dc.format.extent27777255
dc.identifier.citationAlvarez-Gila , A , Galdran , A , Garrote , E & Van de Weijer , J 2019 , ' Self-supervised Blur Detection from Synthetically Blurred Scenes ' , Image and Vision Computing , vol. unknown , 103804 . https://doi.org/10.1016/j.imavis.2019.08.008
dc.identifier.doi10.1016/j.imavis.2019.08.008
dc.identifier.issn1872-8138
dc.identifier.otherresearchoutputwizard: 11556/754
dc.identifier.urlhttp://www.scopus.com/inward/record.url?scp=85074354732&partnerID=8YFLogxK
dc.language.isoeng
dc.relation.ispartofImage and Vision Computing
dc.rightsinfo:eu-repo/semantics/openAccess
dc.subject.keywordsBlur
dc.subject.keywordsDefocus
dc.subject.keywordsMotion
dc.subject.keywordsDeep learning
dc.subject.keywordsSelf-supervised learning
dc.subject.keywordsSynthetic
dc.subject.keywordsBlur
dc.subject.keywordsDefocus
dc.subject.keywordsMotion
dc.subject.keywordsDeep learning
dc.subject.keywordsSelf-supervised learning
dc.subject.keywordsSynthetic
dc.subject.keywordsBlur detection
dc.subject.keywordsDefocus blur
dc.subject.keywordsMotion blur
dc.subject.keywordsSignal Processing
dc.subject.keywordsComputer Vision and Pattern Recognition
dc.subject.keywordsFunding Info
dc.subject.keywordsThis research was partially funded by the Basque Government’s Industry Department under the ELKARTEK program’s project ONKOIKER under agreement KK2018/00090. We thank the Spanish project TIN2016- 79717-R and mention Generalitat de Catalunya CERCA Program.
dc.subject.keywordsThis research was partially funded by the Basque Government’s Industry Department under the ELKARTEK program’s project ONKOIKER under agreement KK2018/00090. We thank the Spanish project TIN2016- 79717-R and mention Generalitat de Catalunya CERCA Program.
dc.titleSelf-supervised Blur Detection from Synthetically Blurred Scenesen
dc.typejournal article
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
AAG_.pdf
Size:
26.49 MB
Format:
Adobe Portable Document Format
Description:
Download