Deep video deblurring

Handle URI:
http://hdl.handle.net/10754/626481
Title:
Deep video deblurring
Authors:
Su, Shuochen; Delbracio, Mauricio; Wang, Jue; Sapiro, Guillermo; Heidrich, Wolfgang ( 0000-0002-4227-8508 ) ; Wang, Oliver
Abstract:
Motion blur from camera shake is a major problem in videos captured by hand-held devices. Unlike single-image deblurring, video-based approaches can take advantage of the abundant information that exists across neighboring frames. As a result the best performing methods rely on aligning nearby frames. However, aligning images is a computationally expensive and fragile procedure, and methods that aggregate information must therefore be able to identify which regions have been accurately aligned and which have not, a task which requires high level scene understanding. In this work, we introduce a deep learning solution to video deblurring, where a CNN is trained end-to-end to learn how to accumulate information across frames. To train this network, we collected a dataset of real videos recorded with a high framerate camera, which we use to generate synthetic motion blur for supervision. We show that the features learned from this dataset extend to deblurring motion blur that arises due to camera shake in a wide range of videos, and compare the quality of results to a number of other baselines.
KAUST Department:
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division; Computer Science Program; Visual Computing Center (VCC)
Publisher:
arXiv
Issue Date:
25-Nov-2016
ARXIV:
arXiv:1611.08387
Type:
Preprint
Additional Links:
http://arxiv.org/abs/1611.08387v1; http://arxiv.org/pdf/1611.08387v1
Appears in Collections:
Other/General Submission; Other/General Submission; Computer Science Program; Visual Computing Center (VCC); Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorSu, Shuochenen
dc.contributor.authorDelbracio, Mauricioen
dc.contributor.authorWang, Jueen
dc.contributor.authorSapiro, Guillermoen
dc.contributor.authorHeidrich, Wolfgangen
dc.contributor.authorWang, Oliveren
dc.date.accessioned2017-12-28T07:32:12Z-
dc.date.available2017-12-28T07:32:12Z-
dc.date.issued2016-11-25en
dc.identifier.urihttp://hdl.handle.net/10754/626481-
dc.description.abstractMotion blur from camera shake is a major problem in videos captured by hand-held devices. Unlike single-image deblurring, video-based approaches can take advantage of the abundant information that exists across neighboring frames. As a result the best performing methods rely on aligning nearby frames. However, aligning images is a computationally expensive and fragile procedure, and methods that aggregate information must therefore be able to identify which regions have been accurately aligned and which have not, a task which requires high level scene understanding. In this work, we introduce a deep learning solution to video deblurring, where a CNN is trained end-to-end to learn how to accumulate information across frames. To train this network, we collected a dataset of real videos recorded with a high framerate camera, which we use to generate synthetic motion blur for supervision. We show that the features learned from this dataset extend to deblurring motion blur that arises due to camera shake in a wide range of videos, and compare the quality of results to a number of other baselines.en
dc.publisherarXiven
dc.relation.urlhttp://arxiv.org/abs/1611.08387v1en
dc.relation.urlhttp://arxiv.org/pdf/1611.08387v1en
dc.rightsArchived with thanks to arXiven
dc.titleDeep video deblurringen
dc.typePreprinten
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.contributor.departmentComputer Science Programen
dc.contributor.departmentVisual Computing Center (VCC)en
dc.eprint.versionPre-printen
dc.contributor.institutionThe Univerity of British Columbiaen
dc.contributor.institutionUniversidad de la Republicaen
dc.contributor.institutionAdobe Researchen
dc.contributor.institutionDuke Universityen
dc.identifier.arxividarXiv:1611.08387en
kaust.authorHeidrich, Wolfgangen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.