Full metadata record

DC Field Value Language
dc.contributor.author 정희철 ko
dc.contributor.author 최민국 ko
dc.contributor.author 김준광 ko
dc.contributor.author 권순 ko
dc.contributor.author 정우영 ko
dc.date.accessioned 2021-01-22T07:24:38Z -
dc.date.available 2021-01-22T07:24:38Z -
dc.date.created 2020-09-11 -
dc.date.issued 2020-08 -
dc.identifier.citation 대한임베디드공학회논문지, v.15, no.4, pp.197 - 204 -
dc.identifier.issn 1975-5066 -
dc.identifier.uri http://hdl.handle.net/20.500.11750/12761 -
dc.description.abstract The ImageNet dataset is a large scale dataset and contains various natural scene images. In this paper, we propose a convolutional neural network (CNN)-based weighted ensemble technique for the ImageNet classification task. First, in order to fuse several models, our technique uses weights for each model, unlike the existing average-based ensemble technique. Then we propose an algorithm that automatically finds the coefficients used in later ensemble process. Our algorithm sequentially selects the model with the best performance of the validation set, and then obtains a weight that improves performance when combined with existing selected models. We applied the proposed algorithm to a total of 13 heterogeneous models, and as a result, 5 models were selected. These selected models were combined with weights, and we achieved 3.297% Top-5 error rate on the ImageNet test dataset. -
dc.language Korean -
dc.publisher 대한임베디드공학회 -
dc.title 대용량 이미지넷 인식을 위한 CNN 기반 Weighted 앙상블 기법 -
dc.title.alternative CNN-based Weighted Ensemble Technique for ImageNet Classification -
dc.type Article -
dc.identifier.doi 10.14372/IEMEK.2020.15.4.197 -
dc.type.local Article(Domestic) -
dc.type.rims ART -
dc.description.journalClass 2 -
dc.identifier.kciid ART002618602 -
dc.contributor.localauthor 최민국 -
dc.contributor.localauthor 김준광 -
dc.contributor.localauthor 권순 -
dc.contributor.localauthor 정우영 -
dc.identifier.citationVolume 15 -
dc.identifier.citationNumber 4 -
dc.identifier.citationStartPage 197 -
dc.identifier.citationEndPage 204 -
dc.identifier.citationTitle 대한임베디드공학회논문지 -
dc.description.isOpenAccess N -
dc.subject.keywordAuthor Deep learning -
dc.subject.keywordAuthor ImageNet -
dc.subject.keywordAuthor ILSVRC -
dc.subject.keywordAuthor Data augmentation -
dc.subject.keywordAuthor Ensemble -
dc.subject.keywordAuthor Weighted ensemble -
dc.subject.keywordAuthor Image classification -

qrcode mendeley

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