Skip navigation
Skip navigation

Kernel Support Vector Machines and Convolutional Neural Networks

Jiang, Shihao; Hartley, Richard; Fernando, Basura

Description

Convolutional Neural Networks (CNN) have achieved great success in various computer vision tasks due to their strong ability in feature extraction. The trend of development of CNN architectures is to increase their depth so as to increase their feature extraction ability. Kernel Support Vector Machines (SVM), on the other hand, are known to give optimal separating surfaces by their ability to automatically select support vectors and perform classification in higher dimensional spaces. We...[Show more]

dc.contributor.authorJiang, Shihao
dc.contributor.authorHartley, Richard
dc.contributor.authorFernando, Basura
dc.date.accessioned2021-04-14T05:15:08Z
dc.identifier.isbn978-1-5386-6602-9
dc.identifier.urihttp://hdl.handle.net/1885/229863
dc.description.abstractConvolutional Neural Networks (CNN) have achieved great success in various computer vision tasks due to their strong ability in feature extraction. The trend of development of CNN architectures is to increase their depth so as to increase their feature extraction ability. Kernel Support Vector Machines (SVM), on the other hand, are known to give optimal separating surfaces by their ability to automatically select support vectors and perform classification in higher dimensional spaces. We investigate the idea of combining the two such that best of both worlds can be achieved and a more compact model can perform as well as deeper CNNs. In the past, attempts have been made to use CNNs to extract features from images and then classify with a kernel SVM, but this process was performed in two separate steps. In this paper, we propose one single model where a CNN and a kernel SVM are integrated together and can be trained end-to-end. In particular, we propose a fully-differentiable Radial Basis Function (RBF) layer, where it can be seamless adapted to a CNN environment and forms a better classifier compared to the normal linear classifier. Due to end-to-end training, our approach allows the initial layers of the CNN to extract features more adapted to the kernel SVM classifier. Our experiments demonstrate that the hybrid CNN-kSVM model gives superior results to a plain CNN model, and also performs better than the method where feature extraction and classification are performed in separate stages, by a CNN and a kernel SVM respectively.
dc.format.mimetypeapplication/pdf
dc.languagehttps://www.ieee.org/publications/rights/author-posting-policy.html..."The policy reaffirms the principle that authors are free to post their own version of their IEEE periodical or conference articles on their personal Web sites, those of their employers, or their funding agencies for the purpose of meeting public availability requirements prescribed by their funding agencies" from the publisher site (as at 19/04/2021).
dc.language.isoen_AU
dc.publisherIEEE
dc.relation.ispartof2018 Digital Image Computing: Techniques and Applications (DICTA)
dc.rights© 2018 IEEE
dc.titleKernel Support Vector Machines and Convolutional Neural Networks
dc.typeConference paper
dc.date.issued2018
local.publisher.urlhttps://www.ieee.org/
local.type.statusAccepted Version
local.contributor.affiliationJiang, S., ANU College of Engineering & Computer Science, The Australian National University
local.contributor.affiliationHartley, R., ANU College of Engineering & Computer Science, The Australian National University
local.contributor.affiliationFernando, B., ANU College of Engineering & Computer Science, The Australian National University
local.bibliographicCitation.startpage1
local.bibliographicCitation.lastpage7
local.identifier.doi10.1109/DICTA.2018.8615840
dcterms.accessRightsOpen Access
CollectionsANU Research Publications

Download

File Description SizeFormat Image
2018 Kernel Support Vector Machines and Convolutional Neural Networks_Richard Hartley.pdf340.64 kBAdobe PDF


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

Updated:  19 May 2020/ Responsible Officer:  University Librarian/ Page Contact:  Library Systems & Web Coordinator