Skip navigation
Skip navigation

Domain adaptation by mixture of alignments of second- or higher-order scatter tensors

Koniusz, Piotr; Tas, Yusuf; Porikli, Fatih

Description

In this paper, we propose an approach to the domain adaptation, dubbed Second-or Higher-order Transfer of Knowledge (So-HoT), based on the mixture of alignments of second-or higher-order scatter statistics between the source and target domains. The human ability to learn from few labeled samples is a recurring motivation in the literature for domain adaptation. Towards this end, we investigate the supervised target scenario for which few labeled target training samples per category exist....[Show more]

dc.contributor.authorKoniusz, Piotr
dc.contributor.authorTas, Yusuf
dc.contributor.authorPorikli, Fatih
dc.contributor.editorO'Conner, Lisa
dc.coverage.spatialHonolulu USA
dc.date.accessioned2020-09-14T03:34:32Z
dc.date.createdJuly 21-26 2017
dc.identifier.isbn9781538604571
dc.identifier.urihttp://hdl.handle.net/1885/210115
dc.description.abstractIn this paper, we propose an approach to the domain adaptation, dubbed Second-or Higher-order Transfer of Knowledge (So-HoT), based on the mixture of alignments of second-or higher-order scatter statistics between the source and target domains. The human ability to learn from few labeled samples is a recurring motivation in the literature for domain adaptation. Towards this end, we investigate the supervised target scenario for which few labeled target training samples per category exist. Specifically, we utilize two CNN streams: the source and target networks fused at the classifier level. Features from the fully connected layers fc7 of each network are used to compute second-or even higher-order scatter tensors, one per network stream per class. As the source and target distributions are somewhat different despite being related, we align the scatters of the two network streams of the same class (within-class scatters) to a desired degree with our bespoke loss while maintaining good separation of the between-class scatters. We train the entire network in end-to-end fashion. We provide evaluations on the standard Office benchmark (visual domains) and RGB-D combined with Caltech256 (depth-to-rgb transfer). We attain state-of-the-art results.
dc.format.mimetypeapplication/pdf
dc.language.isoen_AU
dc.publisherIEEE
dc.relation.ispartof30th IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2017
dc.rights© 2017 IEEE
dc.sourceProceedings of the 30th IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2017
dc.titleDomain adaptation by mixture of alignments of second- or higher-order scatter tensors
dc.typeConference paper
local.description.notesImported from ARIES
local.description.refereedYes
dc.date.issued2017
local.identifier.absfor080104 - Computer Vision
local.identifier.ariespublicationa383154xPUB9047
local.publisher.urlhttps://www.ieee.org/
local.type.statusPublished Version
local.contributor.affiliationKoniusz, Piotr, College of Engineering and Computer Science, ANU
local.contributor.affiliationTas, Yusuf, College of Engineering and Computer Science, ANU
local.contributor.affiliationPorikli, Fatih, College of Engineering and Computer Science, ANU
local.description.embargo2037-12-31
local.bibliographicCitation.startpage7139
local.bibliographicCitation.lastpage7148
local.identifier.doi10.1109/CVPR.2017.755
local.identifier.absseo899999 - Information and Communication Services not elsewhere classified
dc.date.updated2020-06-23T00:53:03Z
local.identifier.scopusID2-s2.0-85041907871
CollectionsANU Research Publications

Download

File Description SizeFormat Image
01_Koniusz_Domain_adaptation_by_mixture_2017.pdf605.74 kBAdobe PDF    Request a copy


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