Close
Metadata

Identity statement area
Reference TypeConference Proceedings
Sitesibgrapi.sid.inpe.br
Identifier8JMKD3MGPAW/3RRA45S
Repositorysid.inpe.br/sibgrapi/2018/09.16.01.52
Last Update2018:09.16.01.52.35 administrator
Metadatasid.inpe.br/sibgrapi/2018/09.16.01.52.35
Metadata Last Update2020:02.19.03.10.45 administrator
Citation KeyNazareCostMellPont:2018:EmAnUs
TitleColor quantization in transfer learning and noisy scenarios: an empirical analysis using convolutional networks
FormatOn-line
Year2018
DateOct. 29 - Nov. 1, 2018
Access Date2020, Dec. 04
Number of Files1
Size395 KiB
Context area
Author1 Nazare, Tiago S.
2 Costa, Gabriel B. Paranhos da
3 Mello, Rodrigo F. de
4 Ponti, Moacir A.
Affiliation1 University of São Paulo
2 University of São Paulo
3 University of São Paulo
4 University of São Paulo
EditorRoss, Arun
Gastal, Eduardo S. L.
Jorge, Joaquim A.
Queiroz, Ricardo L. de
Minetto, Rodrigo
Sarkar, Sudeep
Papa, João Paulo
Oliveira, Manuel M.
Arbeláez, Pablo
Mery, Domingo
Oliveira, Maria Cristina Ferreira de
Spina, Thiago Vallin
Mendes, Caroline Mazetto
Costa, Henrique Sérgio Gutierrez
Mejail, Marta Estela
Geus, Klaus de
Scheer, Sergio
e-Mail Addresstiagosn@usp.br
Conference NameConference on Graphics, Patterns and Images, 31 (SIBGRAPI)
Conference LocationFoz do Iguaçu, PR, Brazil
Book TitleProceedings
PublisherIEEE Computer Society
Publisher CityLos Alamitos
History2018-09-16 01:52:35 :: tiagosn@usp.br -> administrator ::
2020-02-19 03:10:45 :: administrator -> :: 2018
Content and structure area
Is the master or a copy?is the master
Document Stagecompleted
Document Stagenot transferred
Transferable1
Content TypeExternal Contribution
Tertiary TypeFull Paper
KeywordsDeep learning, transfer learning, convolutional neural networks, computer vision.
AbstractTransfer learning is seen as one of the most promising areas of machine learning. Lately, features from pre-trained models have been used to achieve state-of-the-art results in several machine vision problems. Those models are usually employed when the problem of interest does not have enough supervised examples to support the network training from scratch. Most applications use networks pre-trained on noise-free RGB image datasets, what is observed even when the target domain counts on grayscale images or when data is degraded by noise. In this paper, we evaluate the use of Convolutional Neural Networks (CNNs) on such transfer learning scenarios and the impact of using RGB trained networks on grayscale image tasks. Our results confirm that the use of networks trained using colored images on grayscale tasks hinders the overall performance when compared to a similar network trained on a quantized version of the original dataset. Results also show that higher quantization levels (resulting in less colors) increase the robustness of CNN features in the presence of noise.
source Directory Contentthere are no files
agreement Directory Content
agreement.html 15/09/2018 22:52 1.2 KiB 
Conditions of access and use area
Languageen
Target FileSIB_2018.pdf
User Grouptiagosn@usp.br
Visibilityshown
Allied materials area
Mirror Repositorysid.inpe.br/banon/2001/03.30.15.38.24
Next Higher Units8JMKD3MGPAW/3RPADUS
Host Collectionsid.inpe.br/banon/2001/03.30.15.38
Notes area
Empty Fieldsaccessionnumber archivingpolicy archivist area callnumber copyholder copyright creatorhistory descriptionlevel dissemination doi edition electronicmailaddress group holdercode isbn issn label lineage mark nextedition notes numberofvolumes orcid organization pages parameterlist parentrepositories previousedition previouslowerunit progress project readergroup readpermission resumeid rightsholder secondarydate secondarykey secondarymark secondarytype serieseditor session shorttitle sponsor subject tertiarymark type url versiontype volume

Close