Show simple item record

dc.contributor.authorIoannou, Y
dc.contributor.authorRobertson, D
dc.contributor.authorCipolla, R
dc.contributor.authorCriminisi, A
dc.date.accessioned2018-03-22T16:04:33Z
dc.date.available2018-03-22T16:04:33Z
dc.date.issued2017
dc.identifier.isbn9781538604571
dc.identifier.issn1063-6919
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/274245
dc.description.abstractWe propose a new method for creating computationally efficient and compact convolutional neural networks (CNNs) using a novel sparse connection structure that resembles a tree root. This allows a significant reduction in computational cost and number of parameters compared to state-of-the-art deep CNNs, without compromising accuracy, by exploiting the sparsity of inter-layer filter dependencies. We validate our approach by using it to train more efficient variants of state-of-the-art CNN architectures, evaluated on the CIFAR10 and ILSVRC datasets. Our results show similar or higher accuracy than the baseline architectures with much less computation, as measured by CPU and GPU timings. For example, for ResNet 50, our model has 40% fewer parameters, 45% fewer floating point operations, and is 31% (12%) faster on a CPU (GPU). For the deeper ResNet 200 our model has 25% fewer floating point operations and 44% fewer parameters, while maintaining state-of-the-art accuracy. For GoogLeNet, our model has 7% fewer parameters and is 21% (16%) faster on a CPU (GPU).
dc.description.sponsorshipMicrosoft Research PhD Scholarship
dc.publisherIEEE
dc.titleDeep roots: Improving CNN efficiency with hierarchical filter groups
dc.typeConference Object
prism.endingPage5986
prism.publicationDate2017
prism.publicationNameProceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017
prism.startingPage5977
prism.volume2017-January
dc.identifier.doi10.17863/CAM.21352
dcterms.dateAccepted2017-02-27
rioxxterms.versionofrecord10.1109/CVPR.2017.633
rioxxterms.versionVoR
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserved
rioxxterms.licenseref.startdate2017-11-06
dc.contributor.orcidIoannou, Yani [0000-0002-9797-5888]
dc.contributor.orcidCipolla, Roberto [0000-0002-8999-2151]
dc.publisher.urlhttp://dx.doi.org/10.1109/CVPR.2017.633
rioxxterms.typeConference Paper/Proceeding/Abstract
cam.issuedOnline2017-05-20
pubs.conference-nameThe IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
pubs.conference-start-date2017-07-21
cam.orpheus.successThu Nov 05 11:58:27 GMT 2020 - The item has an open VoR version.
pubs.conference-finish-date2017-07-26
rioxxterms.freetoread.startdate2100-01-01


Files in this item

Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record