An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information
In this paper, we focus on the problem of unsupervised image-sentence matching. Existing research explores to utilize document-level structural information to sample positive and negative instances for model training. Although the approach achieves positive results, it introduces a sampling bias and...
Gespeichert in:
Hauptverfasser: | , , , , |
---|---|
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | |
container_volume | |
creator | Li, Zejun Wei, Zhongyu Fan, Zhihao Shan, Haijun Huang, Xuanjing |
description | In this paper, we focus on the problem of unsupervised image-sentence
matching. Existing research explores to utilize document-level structural
information to sample positive and negative instances for model training.
Although the approach achieves positive results, it introduces a sampling bias
and fails to distinguish instances with high semantic similarity. To alleviate
the bias, we propose a new sampling strategy to select additional
intra-document image-sentence pairs as positive or negative samples.
Furthermore, to recognize the complex pattern in intra-document samples, we
propose a Transformer based model to capture fine-grained features and
implicitly construct a graph for each document, where concepts in a document
are introduced to bridge the representation learning of images and sentences in
the context of a document. Experimental results show the effectiveness of our
approach to alleviate the bias and learn well-aligned multimodal
representations. |
doi_str_mv | 10.48550/arxiv.2104.02605 |
format | Article |
fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2104_02605</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2104_02605</sourcerecordid><originalsourceid>FETCH-LOGICAL-a675-31ec2a1d25240a9d067cea7e902b6a3a09e2d8960ec36d8d552018f68151c7d13</originalsourceid><addsrcrecordid>eNotj7tugzAYhb10qNI-QKf6BaC-YAMjSm9IVB1IZvTX_kksgUHmovbtG9Iu5wznIn2EPHAWJ5lS7AnCt1tjwVkSM6GZuiWu8PTop2XEsLoJLa2hHzvnT7QYxzCAOdN2CLTs4YRRjX5Gb5B-wGzOW-k4bfo8mKW_ZFGFK3a0nsNi5iVAR0t_Wfcwu8HfkZsWugnv_31HDq8vh_17VH2-lfuiikCnKpIcjQBuhRIJg9wynRqEFHMmvjRIYDkKm-WaoZHaZlYpwXjW6owrblLL5Y48_t1eWZsxuB7CT7MxN1dm-QtXXFKw</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information</title><source>arXiv.org</source><creator>Li, Zejun ; Wei, Zhongyu ; Fan, Zhihao ; Shan, Haijun ; Huang, Xuanjing</creator><creatorcontrib>Li, Zejun ; Wei, Zhongyu ; Fan, Zhihao ; Shan, Haijun ; Huang, Xuanjing</creatorcontrib><description>In this paper, we focus on the problem of unsupervised image-sentence
matching. Existing research explores to utilize document-level structural
information to sample positive and negative instances for model training.
Although the approach achieves positive results, it introduces a sampling bias
and fails to distinguish instances with high semantic similarity. To alleviate
the bias, we propose a new sampling strategy to select additional
intra-document image-sentence pairs as positive or negative samples.
Furthermore, to recognize the complex pattern in intra-document samples, we
propose a Transformer based model to capture fine-grained features and
implicitly construct a graph for each document, where concepts in a document
are introduced to bridge the representation learning of images and sentences in
the context of a document. Experimental results show the effectiveness of our
approach to alleviate the bias and learn well-aligned multimodal
representations.</description><identifier>DOI: 10.48550/arxiv.2104.02605</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Multimedia</subject><creationdate>2021-03</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2104.02605$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2104.02605$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Li, Zejun</creatorcontrib><creatorcontrib>Wei, Zhongyu</creatorcontrib><creatorcontrib>Fan, Zhihao</creatorcontrib><creatorcontrib>Shan, Haijun</creatorcontrib><creatorcontrib>Huang, Xuanjing</creatorcontrib><title>An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information</title><description>In this paper, we focus on the problem of unsupervised image-sentence
matching. Existing research explores to utilize document-level structural
information to sample positive and negative instances for model training.
Although the approach achieves positive results, it introduces a sampling bias
and fails to distinguish instances with high semantic similarity. To alleviate
the bias, we propose a new sampling strategy to select additional
intra-document image-sentence pairs as positive or negative samples.
Furthermore, to recognize the complex pattern in intra-document samples, we
propose a Transformer based model to capture fine-grained features and
implicitly construct a graph for each document, where concepts in a document
are introduced to bridge the representation learning of images and sentences in
the context of a document. Experimental results show the effectiveness of our
approach to alleviate the bias and learn well-aligned multimodal
representations.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Computer Science - Multimedia</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj7tugzAYhb10qNI-QKf6BaC-YAMjSm9IVB1IZvTX_kksgUHmovbtG9Iu5wznIn2EPHAWJ5lS7AnCt1tjwVkSM6GZuiWu8PTop2XEsLoJLa2hHzvnT7QYxzCAOdN2CLTs4YRRjX5Gb5B-wGzOW-k4bfo8mKW_ZFGFK3a0nsNi5iVAR0t_Wfcwu8HfkZsWugnv_31HDq8vh_17VH2-lfuiikCnKpIcjQBuhRIJg9wynRqEFHMmvjRIYDkKm-WaoZHaZlYpwXjW6owrblLL5Y48_t1eWZsxuB7CT7MxN1dm-QtXXFKw</recordid><startdate>20210321</startdate><enddate>20210321</enddate><creator>Li, Zejun</creator><creator>Wei, Zhongyu</creator><creator>Fan, Zhihao</creator><creator>Shan, Haijun</creator><creator>Huang, Xuanjing</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20210321</creationdate><title>An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information</title><author>Li, Zejun ; Wei, Zhongyu ; Fan, Zhihao ; Shan, Haijun ; Huang, Xuanjing</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a675-31ec2a1d25240a9d067cea7e902b6a3a09e2d8960ec36d8d552018f68151c7d13</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Computer Science - Multimedia</topic><toplevel>online_resources</toplevel><creatorcontrib>Li, Zejun</creatorcontrib><creatorcontrib>Wei, Zhongyu</creatorcontrib><creatorcontrib>Fan, Zhihao</creatorcontrib><creatorcontrib>Shan, Haijun</creatorcontrib><creatorcontrib>Huang, Xuanjing</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Li, Zejun</au><au>Wei, Zhongyu</au><au>Fan, Zhihao</au><au>Shan, Haijun</au><au>Huang, Xuanjing</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information</atitle><date>2021-03-21</date><risdate>2021</risdate><abstract>In this paper, we focus on the problem of unsupervised image-sentence
matching. Existing research explores to utilize document-level structural
information to sample positive and negative instances for model training.
Although the approach achieves positive results, it introduces a sampling bias
and fails to distinguish instances with high semantic similarity. To alleviate
the bias, we propose a new sampling strategy to select additional
intra-document image-sentence pairs as positive or negative samples.
Furthermore, to recognize the complex pattern in intra-document samples, we
propose a Transformer based model to capture fine-grained features and
implicitly construct a graph for each document, where concepts in a document
are introduced to bridge the representation learning of images and sentences in
the context of a document. Experimental results show the effectiveness of our
approach to alleviate the bias and learn well-aligned multimodal
representations.</abstract><doi>10.48550/arxiv.2104.02605</doi><oa>free_for_read</oa></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | DOI: 10.48550/arxiv.2104.02605 |
ispartof | |
issn | |
language | eng |
recordid | cdi_arxiv_primary_2104_02605 |
source | arXiv.org |
subjects | Computer Science - Artificial Intelligence Computer Science - Computer Vision and Pattern Recognition Computer Science - Multimedia |
title | An Unsupervised Sampling Approach for Image-Sentence Matching Using Document-Level Structural Information |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-08T11%3A07%3A39IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=An%20Unsupervised%20Sampling%20Approach%20for%20Image-Sentence%20Matching%20Using%20Document-Level%20Structural%20Information&rft.au=Li,%20Zejun&rft.date=2021-03-21&rft_id=info:doi/10.48550/arxiv.2104.02605&rft_dat=%3Carxiv_GOX%3E2104_02605%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |