Please use this identifier to cite or link to this item: https://hdl.handle.net/11147/7559
Full metadata record
DC FieldValueLanguage
dc.contributor.authorOrhan, Semih-
dc.contributor.authorBaştanlar, Yalın-
dc.date.accessioned2020-01-06T11:22:06Z
dc.date.available2020-01-06T11:22:06Z
dc.date.issued2018-04en_US
dc.identifier.citationOrhan, S., and Baştanlar, Y. (2018). Training CNNs with image patches for object localisation. Electronics Letters, 54(7), 424-426. doi:10.1049/el.2017.4725en_US
dc.identifier.issn0013-5194
dc.identifier.issn0013-5194-
dc.identifier.urihttps://doi.org/10.1049/el.2017.4725
dc.identifier.urihttps://hdl.handle.net/11147/7559
dc.description.abstractRecently, convolutional neural networks (CNNs) have shown great performance in different problems of computer vision including object detection and localisation. A novel training approach is proposed for CNNs to localise some animal species whose bodies have distinctive patterns such as leopards and zebras. To learn characteristic patterns, small patches which are taken from different body parts of animals are used to train models. To find object location, in a test image, all locations are visited in a sliding window fashion. Crops are fed into trained CNN and their classification scores are combined into a heat map. Later on, heat maps are converted to bounding box estimates for varying confidence scores. The localisation performance of the patch-based training approach is compared with Faster R-CNN – a state-of-the-art CNN-based object detection and localisation method. Experimental results reveal that the patch-based training outperforms Faster R-CNN, especially for classes with distinctive patterns.en_US
dc.description.sponsorshipTUBITAK (115E918)en_US
dc.language.isoenen_US
dc.publisherInstitution of Engineering and Technologyen_US
dc.relation.ispartofElectronics Lettersen_US
dc.rightsinfo:eu-repo/semantics/openAccessen_US
dc.subjectObject detectionen_US
dc.subjectComputer visionen_US
dc.subjectObject recognitionen_US
dc.subjectConvolutional Neural Networksen_US
dc.subjectAnimal speciesen_US
dc.titleTraining CNNs with image patches for object localisationen_US
dc.typeArticleen_US
dc.authorid0000-0002-3774-6872en_US
dc.departmentIzmir Institute of Technology. Computer Engineeringen_US
dc.identifier.volume54en_US
dc.identifier.issue7en_US
dc.identifier.startpage424en_US
dc.identifier.endpage426en_US
dc.identifier.wosWOS:000428477900012
dc.identifier.scopusSCOPUS:2-s2.0-85044624717
dc.relation.tubitakinfo:eu-repo/grantAgreement/TUBITAK/EEEAG/115E918
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.identifier.doi10.1049/el.2017.4725-
dc.relation.doi10.1049/el.2017.4725en_US
dc.coverage.doi10.1049/el.2017.4725en_US
item.fulltextWith Fulltext-
item.openairecristypehttp://purl.org/coar/resource_type/c_18cf-
item.openairetypeArticle-
item.languageiso639-1en-
item.cerifentitytypePublications-
item.grantfulltextopen-
crisitem.author.deptDepartment of Computer Engineering-
Appears in Collections:Computer Engineering / Bilgisayar Mühendisliği
Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection
WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection
Files in This Item:
File Description SizeFormat 
orhan2018.pdfMakale (Article)589.75 kBAdobe PDFThumbnail
View/Open
Show simple item record

CORE Recommender

SCOPUSTM   
Citations

7
checked on Oct 23, 2021

WEB OF SCIENCETM
Citations

7
checked on Oct 23, 2021

Page view(s)

46
checked on Oct 23, 2021

Download(s)

46
checked on Oct 23, 2021

Google ScholarTM

Check

Altmetric


Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.