ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations

State-of-the-art empirical work has shown that visual representations learned by deep neural networks are robust in nature and capable of performing classification tasks on diverse datasets. For example, CLIP demonstrated zero-shot transfer performance on multiple datasets for classification tasks i...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2022-11
Hauptverfasser: Grover, Chanda, Mastan, Indra Deep, Gupta, Debayan
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Grover, Chanda
Mastan, Indra Deep
Gupta, Debayan
description State-of-the-art empirical work has shown that visual representations learned by deep neural networks are robust in nature and capable of performing classification tasks on diverse datasets. For example, CLIP demonstrated zero-shot transfer performance on multiple datasets for classification tasks in a joint embedding space of image and text pairs. However, it showed negative transfer performance on standard datasets, e.g., BirdsNAP, RESISC45, and MNIST. In this paper, we propose ContextCLIP, a contextual and contrastive learning framework for the contextual alignment of image-text pairs by learning robust visual representations on Conceptual Captions dataset. Our framework was observed to improve the image-text alignment by aligning text and image representations contextually in the joint embedding space. ContextCLIP showed good qualitative performance for text-to-image retrieval tasks and enhanced classification accuracy. We evaluated our model quantitatively with zero-shot transfer and fine-tuning experiments on CIFAR-10, CIFAR-100, Birdsnap, RESISC45, and MNIST datasets for classification task.
doi_str_mv 10.48550/arxiv.2211.07122
format Article
fullrecord <record><control><sourceid>proquest_arxiv</sourceid><recordid>TN_cdi_arxiv_primary_2211_07122</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2736483903</sourcerecordid><originalsourceid>FETCH-LOGICAL-a952-61dcecc80ac527a1cb5842fdeb73363010d4dedf779ba6882fc86a8f5daee4fa3</originalsourceid><addsrcrecordid>eNotj01rg0AQhpdCoSHND-ipCz1rd2ddd9NbkH4IQnvwLqPuBoNRu6sh_ffVJKdhmOcd3oeQJ87CSEvJXtGdm1MIwHnIFAe4IysQggc6AnggG-8PjDGIFUgpViRP-m405zHJ0p83elsmbOmubfbd0XQj7S1Nj7g3QT6f6ICN87Tv6JKgp8YvsDODM36GcWz6zj-Se4utN5vbXJP84z1PvoLs-zNNdlmAWwlBzOvKVJVmWElQyKtSzh1tbUolRCwYZ3VUm9oqtS0x1hpspWPUVtZoTGRRrMnz9e1FuRhcc0T3VyzqxUV9Jl6uxOD638n4sTj0k-vmTgUoEUdabJkQ_wNOXNA</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2736483903</pqid></control><display><type>article</type><title>ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations</title><source>arXiv.org</source><source>Free E- Journals</source><creator>Grover, Chanda ; Mastan, Indra Deep ; Gupta, Debayan</creator><creatorcontrib>Grover, Chanda ; Mastan, Indra Deep ; Gupta, Debayan</creatorcontrib><description>State-of-the-art empirical work has shown that visual representations learned by deep neural networks are robust in nature and capable of performing classification tasks on diverse datasets. For example, CLIP demonstrated zero-shot transfer performance on multiple datasets for classification tasks in a joint embedding space of image and text pairs. However, it showed negative transfer performance on standard datasets, e.g., BirdsNAP, RESISC45, and MNIST. In this paper, we propose ContextCLIP, a contextual and contrastive learning framework for the contextual alignment of image-text pairs by learning robust visual representations on Conceptual Captions dataset. Our framework was observed to improve the image-text alignment by aligning text and image representations contextually in the joint embedding space. ContextCLIP showed good qualitative performance for text-to-image retrieval tasks and enhanced classification accuracy. We evaluated our model quantitatively with zero-shot transfer and fine-tuning experiments on CIFAR-10, CIFAR-100, Birdsnap, RESISC45, and MNIST datasets for classification task.</description><identifier>EISSN: 2331-8422</identifier><identifier>DOI: 10.48550/arxiv.2211.07122</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Alignment ; Artificial neural networks ; Computer Science - Computer Vision and Pattern Recognition ; Datasets ; Embedding ; Image classification ; Image enhancement ; Machine learning ; Representations ; Retrieval ; Robustness ; Visual observation</subject><ispartof>arXiv.org, 2022-11</ispartof><rights>2022. This work is published under http://creativecommons.org/licenses/by-sa/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><rights>http://creativecommons.org/licenses/by-sa/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,784,885,27924</link.rule.ids><backlink>$$Uhttps://doi.org/10.48550/arXiv.2211.07122$$DView paper in arXiv$$Hfree_for_read</backlink><backlink>$$Uhttps://doi.org/10.1145/3571600.3571653$$DView published paper (Access to full text may be restricted)$$Hfree_for_read</backlink></links><search><creatorcontrib>Grover, Chanda</creatorcontrib><creatorcontrib>Mastan, Indra Deep</creatorcontrib><creatorcontrib>Gupta, Debayan</creatorcontrib><title>ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations</title><title>arXiv.org</title><description>State-of-the-art empirical work has shown that visual representations learned by deep neural networks are robust in nature and capable of performing classification tasks on diverse datasets. For example, CLIP demonstrated zero-shot transfer performance on multiple datasets for classification tasks in a joint embedding space of image and text pairs. However, it showed negative transfer performance on standard datasets, e.g., BirdsNAP, RESISC45, and MNIST. In this paper, we propose ContextCLIP, a contextual and contrastive learning framework for the contextual alignment of image-text pairs by learning robust visual representations on Conceptual Captions dataset. Our framework was observed to improve the image-text alignment by aligning text and image representations contextually in the joint embedding space. ContextCLIP showed good qualitative performance for text-to-image retrieval tasks and enhanced classification accuracy. We evaluated our model quantitatively with zero-shot transfer and fine-tuning experiments on CIFAR-10, CIFAR-100, Birdsnap, RESISC45, and MNIST datasets for classification task.</description><subject>Alignment</subject><subject>Artificial neural networks</subject><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Datasets</subject><subject>Embedding</subject><subject>Image classification</subject><subject>Image enhancement</subject><subject>Machine learning</subject><subject>Representations</subject><subject>Retrieval</subject><subject>Robustness</subject><subject>Visual observation</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><sourceid>GOX</sourceid><recordid>eNotj01rg0AQhpdCoSHND-ipCz1rd2ddd9NbkH4IQnvwLqPuBoNRu6sh_ffVJKdhmOcd3oeQJ87CSEvJXtGdm1MIwHnIFAe4IysQggc6AnggG-8PjDGIFUgpViRP-m405zHJ0p83elsmbOmubfbd0XQj7S1Nj7g3QT6f6ICN87Tv6JKgp8YvsDODM36GcWz6zj-Se4utN5vbXJP84z1PvoLs-zNNdlmAWwlBzOvKVJVmWElQyKtSzh1tbUolRCwYZ3VUm9oqtS0x1hpspWPUVtZoTGRRrMnz9e1FuRhcc0T3VyzqxUV9Jl6uxOD638n4sTj0k-vmTgUoEUdabJkQ_wNOXNA</recordid><startdate>20221114</startdate><enddate>20221114</enddate><creator>Grover, Chanda</creator><creator>Mastan, Indra Deep</creator><creator>Gupta, Debayan</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20221114</creationdate><title>ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations</title><author>Grover, Chanda ; Mastan, Indra Deep ; Gupta, Debayan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a952-61dcecc80ac527a1cb5842fdeb73363010d4dedf779ba6882fc86a8f5daee4fa3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Alignment</topic><topic>Artificial neural networks</topic><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Datasets</topic><topic>Embedding</topic><topic>Image classification</topic><topic>Image enhancement</topic><topic>Machine learning</topic><topic>Representations</topic><topic>Retrieval</topic><topic>Robustness</topic><topic>Visual observation</topic><toplevel>online_resources</toplevel><creatorcontrib>Grover, Chanda</creatorcontrib><creatorcontrib>Mastan, Indra Deep</creatorcontrib><creatorcontrib>Gupta, Debayan</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection><collection>arXiv Computer Science</collection><collection>arXiv.org</collection><jtitle>arXiv.org</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Grover, Chanda</au><au>Mastan, Indra Deep</au><au>Gupta, Debayan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations</atitle><jtitle>arXiv.org</jtitle><date>2022-11-14</date><risdate>2022</risdate><eissn>2331-8422</eissn><abstract>State-of-the-art empirical work has shown that visual representations learned by deep neural networks are robust in nature and capable of performing classification tasks on diverse datasets. For example, CLIP demonstrated zero-shot transfer performance on multiple datasets for classification tasks in a joint embedding space of image and text pairs. However, it showed negative transfer performance on standard datasets, e.g., BirdsNAP, RESISC45, and MNIST. In this paper, we propose ContextCLIP, a contextual and contrastive learning framework for the contextual alignment of image-text pairs by learning robust visual representations on Conceptual Captions dataset. Our framework was observed to improve the image-text alignment by aligning text and image representations contextually in the joint embedding space. ContextCLIP showed good qualitative performance for text-to-image retrieval tasks and enhanced classification accuracy. We evaluated our model quantitatively with zero-shot transfer and fine-tuning experiments on CIFAR-10, CIFAR-100, Birdsnap, RESISC45, and MNIST datasets for classification task.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><doi>10.48550/arxiv.2211.07122</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2022-11
issn 2331-8422
language eng
recordid cdi_arxiv_primary_2211_07122
source arXiv.org; Free E- Journals
subjects Alignment
Artificial neural networks
Computer Science - Computer Vision and Pattern Recognition
Datasets
Embedding
Image classification
Image enhancement
Machine learning
Representations
Retrieval
Robustness
Visual observation
title ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-12T11%3A13%3A02IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_arxiv&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=ContextCLIP:%20Contextual%20Alignment%20of%20Image-Text%20pairs%20on%20CLIP%20visual%20representations&rft.jtitle=arXiv.org&rft.au=Grover,%20Chanda&rft.date=2022-11-14&rft.eissn=2331-8422&rft_id=info:doi/10.48550/arxiv.2211.07122&rft_dat=%3Cproquest_arxiv%3E2736483903%3C/proquest_arxiv%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2736483903&rft_id=info:pmid/&rfr_iscdi=true