Show simple item record

dc.contributor.authorGiancola, Silvio
dc.contributor.authorAmine, Mohieddine
dc.contributor.authorDghaily, Tarek
dc.contributor.authorGhanem, Bernard
dc.date.accessioned2019-03-13T11:56:08Z
dc.date.available2018-04-19T10:45:32Z
dc.date.available2019-03-13T11:56:08Z
dc.date.issued2018-12-18
dc.identifier.citationGiancola S, Amine M, Dghaily T, Ghanem B (2018) SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Available: http://dx.doi.org/10.1109/CVPRW.2018.00223.
dc.identifier.doi10.1109/CVPRW.2018.00223
dc.identifier.urihttp://hdl.handle.net/10754/627570
dc.description.abstractIn this paper, we introduce SoccerNet, a benchmark for action spotting in soccer videos. The dataset is composed of 500 complete soccer games from six main European leagues, covering three seasons from 2014 to 2017 and a total duration of 764 hours. A total of 6,637 temporal annotations are automatically parsed from online match reports at a one minute resolution for three main classes of events (Goal, Yellow/Red Card, and Substitution). As such, the dataset is easily scalable. These annotations are manually refined to a one second resolution by anchoring them at a single timestamp following well-defined soccer rules. With an average of one event every 6.9 minutes, this dataset focuses on the problem of localizing very sparse events within long videos. We define the task of spotting as finding the anchors of soccer events in a video. Making use of recent developments in the realm of generic action recognition and detection in video, we provide strong baselines for detecting soccer events. We show that our best model for classifying temporal segments of length one minute reaches a mean Average Precision (mAP) of 67.8%. For the spotting task, our baseline reaches an Average-mAP of 49.7% for tolerances d ranging from 5 to 60 seconds. Our dataset and models are available at https://silviogiancola.github.io/SoccerNet.
dc.description.sponsorshipThis work was supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research.
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)
dc.relation.urlhttps://ieeexplore.ieee.org/document/8575386
dc.rightsArchived with thanks to 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
dc.titleSoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos
dc.typeConference Paper
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
dc.contributor.departmentElectrical Engineering Program
dc.contributor.departmentVCC Analytics Research Group
dc.contributor.departmentVisual Computing Center (VCC)
dc.identifier.journal2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
dc.conference.date2018-06-18 to 2018-06-22
dc.conference.name31st Meeting of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2018
dc.conference.locationSalt Lake City, UT, USA
dc.eprint.versionPost-print
dc.identifier.arxividarXiv:1804.04527
kaust.personGiancola, Silvio
kaust.personAmine, Mohieddine
kaust.personDghaily, Tarek
kaust.personGhanem, Bernard
refterms.dateFOA2018-06-14T05:54:32Z
dc.date.published-online2018-12-18
dc.date.published-print2018-06
dc.date.posted2018-04-12


Files in this item

Thumbnail
Name:
1804.04527.pdf
Size:
3.382Mb
Format:
PDF
Description:
Accepted Manuscript

This item appears in the following Collection(s)

Show simple item record

VersionItemEditorDateSummary

*Selected version