Show simple item record

dc.contributor.authorPardo, Alejandro
dc.contributor.authorAlwassel, Humam
dc.contributor.authorHeilbron, Fabian Caba
dc.contributor.authorThabet, Ali Kassem
dc.contributor.authorGhanem, Bernard
dc.date.accessioned2021-06-16T06:39:19Z
dc.date.available2019-12-18T11:23:14Z
dc.date.available2021-06-16T06:39:19Z
dc.date.issued2021
dc.identifier.citationPardo, A., Alwassel, H., Heilbron, F. C., Thabet, A., & Ghanem, B. (2021). RefineLoc: Iterative Refinement for Weakly-Supervised Action Localization. 2021 IEEE Winter Conference on Applications of Computer Vision (WACV). doi:10.1109/wacv48630.2021.00336
dc.identifier.isbn978-1-6654-4640-2
dc.identifier.issn2472-6737
dc.identifier.doi10.1109/WACV48630.2021.00336
dc.identifier.urihttp://hdl.handle.net/10754/660670
dc.description.abstractVideo action detectors are usually trained using datasets with fully-supervised temporal annotations. Building such datasets is an expensive task. To alleviate this problem, recent methods have tried to leverage weak labeling, where videos are untrimmed and only a video-level label is available. In this paper, we propose RefineLoc, a novel weaklysupervised temporal action localization method. RefineLoc uses an iterative refinement approach by estimating and training on snippet-level pseudo ground truth at every iteration. We show the benefit of this iterative approach and present an extensive analysis of five different pseudo ground truth generators. We show the effectiveness of our model on two standard action datasets, ActivityNet v1.2 and THUMOS14. RefineLoc shows competitive results with the stateof-the-art in weakly-supervised temporal localization. Additionally, our iterative refinement process is able to significantly improve the performance of two state-of-the-art methods, setting a new state-of-the-art on THUMOS14.
dc.description.sponsorshipThis work is supported the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under Award No. OSRCRG2017-3405.
dc.publisherIEEE
dc.relation.urlhttps://ieeexplore.ieee.org/document/9423165/
dc.relation.urlhttps://openaccess.thecvf.com/content/WACV2021/papers/Pardo_RefineLoc_Iterative_Refinement_for_Weakly-Supervised_Action_Localization_WACV_2021_paper.pdf
dc.rightsArchived with thanks to IEEE.
dc.titleRefineLoc: Iterative Refinement for Weakly-Supervised Action Localization
dc.typeConference Paper
dc.contributor.departmentKing Abdullah University of Science and Technology (KAUST)
dc.contributor.departmentComputer Science Program
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
dc.contributor.departmentVisual Computing Center (VCC)
dc.contributor.departmentElectrical Engineering Program
dc.conference.date3-8 Jan. 2021
dc.conference.name2021 IEEE Winter Conference on Applications of Computer Vision (WACV)
dc.conference.locationWaikoloa, HI, USA
dc.eprint.versionPost-print
dc.contributor.institutionAdobe Research
dc.identifier.arxivid1904.00227
kaust.personPardo, Alejandro
kaust.personAlwassel, Humam
kaust.personThabet, Ali Kassem
kaust.personGhanem, Bernard
kaust.grant.numberOSRCRG2017-3405
dc.relation.issupplementedbygithub:HumamAlwassel/RefineLoc
refterms.dateFOA2019-12-18T11:24:16Z
display.relations<b>Is Supplemented By:</b><br/> <ul><li><i>[Software]</i> <br/> Title: HumamAlwassel/RefineLoc: RefineLoc: Iterative Refinement for Weakly-Supervised Action Localization (WACV 2021). Publication Date: 2021-01-03. github: <a href="https://github.com/HumamAlwassel/RefineLoc" >HumamAlwassel/RefineLoc</a> Handle: <a href="http://hdl.handle.net/10754/669741" >10754/669741</a></a></li></ul>
kaust.acknowledged.supportUnitOffice of Sponsored Research (OSR)


Files in this item

Thumbnail
Name:
Pardo_RefineLoc_Iterative_Refinement_for_Weakly-Supervised_Action_Localization_WACV_2021_paper.pdf
Size:
2.059Mb
Format:
PDF
Description:
Accepted Manuscript
Thumbnail
Name:
Pardo_RefineLoc_Iterative_Refinement_WACV_2021_supplemental.pdf
Size:
1.596Mb
Format:
PDF
Description:
Supplementary Material

This item appears in the following Collection(s)

Show simple item record

VersionItemEditorDateSummary

*Selected version