Object-based color constancy in a deep neural network
Color constancy refers to our capacity to see consistent colors under different illuminations. In computer vision and image processing, color constancy is often approached by explicit estimation of the scene's illumination, followed by an image correction. In contrast, color constancy in human...
Gespeichert in:
Veröffentlicht in: | Journal of the Optical Society of America. A, Optics, image science, and vision Optics, image science, and vision, 2023-03, Vol.40 (3), p.A48-A56 |
---|---|
Hauptverfasser: | , |
Format: | Artikel |
Sprache: | eng |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | A56 |
---|---|
container_issue | 3 |
container_start_page | A48 |
container_title | Journal of the Optical Society of America. A, Optics, image science, and vision |
container_volume | 40 |
creator | Heidari-Gorji, Hamed Gegenfurtner, Karl R |
description | Color constancy refers to our capacity to see consistent colors under different illuminations. In computer vision and image processing, color constancy is often approached by explicit estimation of the scene's illumination, followed by an image correction. In contrast, color constancy in human vision is typically measured as the capacity to extract color information about objects and materials in a scene consistently throughout various illuminations, which goes beyond illumination estimation and might require some degree of scene and color understanding. Here, we pursue an approach with deep neural networks that tries to assign reflectances to individual objects in the scene. To circumvent the lack of massive ground truth datasets labeled with reflectances, we used computer graphics to render images. This study presents a model that recognizes colors in an image pixel by pixel under different illumination conditions. |
doi_str_mv | 10.1364/JOSAA.479451 |
format | Article |
fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_2809003937</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2809003937</sourcerecordid><originalsourceid>FETCH-LOGICAL-c291t-ccef93de854025e56a043c10e443090461986cd74d8744397d6fc10ee8cb09d83</originalsourceid><addsrcrecordid>eNo9kE1PwzAMhiMEYmNw44x65ECH0yRNcpwmPjVpB-AcpYkrbXTtSFqh_XsyNrjYlv3olfUQck1hSlnJ71-Xb7PZlEvNBT0hYyoKyJVgxWmaQfFcikKPyEWMawDgpZLnZMQkZQyAjYlYVmt0fV7ZiD5zXdOFVNvY29btslWb2cwjbrMWh2Cb1PrvLnxekrPaNhGvjn1CPh4f3ufP-WL59DKfLXJXaNrnzmGtmUclOBQCRWmBM0cBOWeg0zNUq9J5yb2SaaWlL-v9GZWrQHvFJuT2kLsN3deAsTebVXTYNLbFboimUCkGmGYyoXcH1IUuxoC12YbVxoadoWD2osyvKHMQlfCbY_JQbdD_w39m2A9lcWFl</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2809003937</pqid></control><display><type>article</type><title>Object-based color constancy in a deep neural network</title><source>Optica Publishing Group Journals</source><creator>Heidari-Gorji, Hamed ; Gegenfurtner, Karl R</creator><creatorcontrib>Heidari-Gorji, Hamed ; Gegenfurtner, Karl R</creatorcontrib><description>Color constancy refers to our capacity to see consistent colors under different illuminations. In computer vision and image processing, color constancy is often approached by explicit estimation of the scene's illumination, followed by an image correction. In contrast, color constancy in human vision is typically measured as the capacity to extract color information about objects and materials in a scene consistently throughout various illuminations, which goes beyond illumination estimation and might require some degree of scene and color understanding. Here, we pursue an approach with deep neural networks that tries to assign reflectances to individual objects in the scene. To circumvent the lack of massive ground truth datasets labeled with reflectances, we used computer graphics to render images. This study presents a model that recognizes colors in an image pixel by pixel under different illumination conditions.</description><identifier>ISSN: 1084-7529</identifier><identifier>EISSN: 1520-8532</identifier><identifier>DOI: 10.1364/JOSAA.479451</identifier><identifier>PMID: 37133003</identifier><language>eng</language><publisher>United States</publisher><ispartof>Journal of the Optical Society of America. A, Optics, image science, and vision, 2023-03, Vol.40 (3), p.A48-A56</ispartof><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c291t-ccef93de854025e56a043c10e443090461986cd74d8744397d6fc10ee8cb09d83</citedby><cites>FETCH-LOGICAL-c291t-ccef93de854025e56a043c10e443090461986cd74d8744397d6fc10ee8cb09d83</cites><orcidid>0000-0001-5390-0684</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,780,784,3258,27924,27925</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/37133003$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Heidari-Gorji, Hamed</creatorcontrib><creatorcontrib>Gegenfurtner, Karl R</creatorcontrib><title>Object-based color constancy in a deep neural network</title><title>Journal of the Optical Society of America. A, Optics, image science, and vision</title><addtitle>J Opt Soc Am A Opt Image Sci Vis</addtitle><description>Color constancy refers to our capacity to see consistent colors under different illuminations. In computer vision and image processing, color constancy is often approached by explicit estimation of the scene's illumination, followed by an image correction. In contrast, color constancy in human vision is typically measured as the capacity to extract color information about objects and materials in a scene consistently throughout various illuminations, which goes beyond illumination estimation and might require some degree of scene and color understanding. Here, we pursue an approach with deep neural networks that tries to assign reflectances to individual objects in the scene. To circumvent the lack of massive ground truth datasets labeled with reflectances, we used computer graphics to render images. This study presents a model that recognizes colors in an image pixel by pixel under different illumination conditions.</description><issn>1084-7529</issn><issn>1520-8532</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><recordid>eNo9kE1PwzAMhiMEYmNw44x65ECH0yRNcpwmPjVpB-AcpYkrbXTtSFqh_XsyNrjYlv3olfUQck1hSlnJ71-Xb7PZlEvNBT0hYyoKyJVgxWmaQfFcikKPyEWMawDgpZLnZMQkZQyAjYlYVmt0fV7ZiD5zXdOFVNvY29btslWb2cwjbrMWh2Cb1PrvLnxekrPaNhGvjn1CPh4f3ufP-WL59DKfLXJXaNrnzmGtmUclOBQCRWmBM0cBOWeg0zNUq9J5yb2SaaWlL-v9GZWrQHvFJuT2kLsN3deAsTebVXTYNLbFboimUCkGmGYyoXcH1IUuxoC12YbVxoadoWD2osyvKHMQlfCbY_JQbdD_w39m2A9lcWFl</recordid><startdate>20230301</startdate><enddate>20230301</enddate><creator>Heidari-Gorji, Hamed</creator><creator>Gegenfurtner, Karl R</creator><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0001-5390-0684</orcidid></search><sort><creationdate>20230301</creationdate><title>Object-based color constancy in a deep neural network</title><author>Heidari-Gorji, Hamed ; Gegenfurtner, Karl R</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c291t-ccef93de854025e56a043c10e443090461986cd74d8744397d6fc10ee8cb09d83</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Heidari-Gorji, Hamed</creatorcontrib><creatorcontrib>Gegenfurtner, Karl R</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>Journal of the Optical Society of America. A, Optics, image science, and vision</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Heidari-Gorji, Hamed</au><au>Gegenfurtner, Karl R</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Object-based color constancy in a deep neural network</atitle><jtitle>Journal of the Optical Society of America. A, Optics, image science, and vision</jtitle><addtitle>J Opt Soc Am A Opt Image Sci Vis</addtitle><date>2023-03-01</date><risdate>2023</risdate><volume>40</volume><issue>3</issue><spage>A48</spage><epage>A56</epage><pages>A48-A56</pages><issn>1084-7529</issn><eissn>1520-8532</eissn><abstract>Color constancy refers to our capacity to see consistent colors under different illuminations. In computer vision and image processing, color constancy is often approached by explicit estimation of the scene's illumination, followed by an image correction. In contrast, color constancy in human vision is typically measured as the capacity to extract color information about objects and materials in a scene consistently throughout various illuminations, which goes beyond illumination estimation and might require some degree of scene and color understanding. Here, we pursue an approach with deep neural networks that tries to assign reflectances to individual objects in the scene. To circumvent the lack of massive ground truth datasets labeled with reflectances, we used computer graphics to render images. This study presents a model that recognizes colors in an image pixel by pixel under different illumination conditions.</abstract><cop>United States</cop><pmid>37133003</pmid><doi>10.1364/JOSAA.479451</doi><orcidid>https://orcid.org/0000-0001-5390-0684</orcidid></addata></record> |
fulltext | fulltext |
identifier | ISSN: 1084-7529 |
ispartof | Journal of the Optical Society of America. A, Optics, image science, and vision, 2023-03, Vol.40 (3), p.A48-A56 |
issn | 1084-7529 1520-8532 |
language | eng |
recordid | cdi_proquest_miscellaneous_2809003937 |
source | Optica Publishing Group Journals |
title | Object-based color constancy in a deep neural network |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T21%3A31%3A33IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Object-based%20color%20constancy%20in%20a%20deep%20neural%20network&rft.jtitle=Journal%20of%20the%20Optical%20Society%20of%20America.%20A,%20Optics,%20image%20science,%20and%20vision&rft.au=Heidari-Gorji,%20Hamed&rft.date=2023-03-01&rft.volume=40&rft.issue=3&rft.spage=A48&rft.epage=A56&rft.pages=A48-A56&rft.issn=1084-7529&rft.eissn=1520-8532&rft_id=info:doi/10.1364/JOSAA.479451&rft_dat=%3Cproquest_cross%3E2809003937%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2809003937&rft_id=info:pmid/37133003&rfr_iscdi=true |