Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects

An image is not just a collection of objects, but rather a graph where each object is related to other objects through spatial and semantic relations. Using relational reasoning modules, such as the non-local module \cite{wang2017non}, can therefore improve object detection. Current schemes apply su...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Levi, Hila, Ullman, Shimon
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 Levi, Hila
Ullman, Shimon
description An image is not just a collection of objects, but rather a graph where each object is related to other objects through spatial and semantic relations. Using relational reasoning modules, such as the non-local module \cite{wang2017non}, can therefore improve object detection. Current schemes apply such dedicated modules either to a specific layer of the bottom-up stream, or between already-detected objects. We show that the relational process can be better modeled in a coarse-to-fine manner and present a novel framework, applying a non-local module sequentially to increasing resolution feature maps along the top-down stream. In this way, information can naturally passed from larger objects to smaller related ones. Applying the module to fine feature maps further allows the information to pass between the small objects themselves, exploiting repetitions of instances of the same class. In practice, due to the expensive memory utilization of the non-local module, it is infeasible to apply the module as currently used to high-resolution feature maps. We redesigned the non local module, improved it in terms of memory and number of operations, allowing it to be placed anywhere along the network. We further incorporated relative spatial information into the module, in a manner that can be incorporated into our efficient implementation. We show the effectiveness of our scheme by improving the results of detecting small objects on COCO by 1-2 AP points over Faster and Mask RCNN and by 1 AP over using non-local module on the bottom-up stream.
doi_str_mv 10.48550/arxiv.1811.12152
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1811_12152</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1811_12152</sourcerecordid><originalsourceid>FETCH-LOGICAL-a672-41e4f6bc76f0597fbc0616b0731a99e2614556858113f4377fe0e22bdeb458323</originalsourceid><addsrcrecordid>eNotz71OwzAUBWAvDKjwAEz4BRx8_ZuMKLSAFOhA98h2r4VRGiPHIHh7SmE60hmOzkfIFfBGtVrzG1e-0mcDLUADArQ4J8M6xhQSzpX22ZUFWc1sk2akz3lmQw5uok95_zEhjbnQ-or0DiuGmvJMc6QvBzdNdOvfjtVyQc6imxa8_M8V2W3Wu_6BDdv7x_52YM5YwRSgisYHayLXnY0-cAPGcyvBdR0KA0pr0-rjSxmVtDYiRyH8Hr3SrRRyRa7_Zk-c8b2kgyvf4y9rPLHkD-7lRgk</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects</title><source>arXiv.org</source><creator>Levi, Hila ; Ullman, Shimon</creator><creatorcontrib>Levi, Hila ; Ullman, Shimon</creatorcontrib><description>An image is not just a collection of objects, but rather a graph where each object is related to other objects through spatial and semantic relations. Using relational reasoning modules, such as the non-local module \cite{wang2017non}, can therefore improve object detection. Current schemes apply such dedicated modules either to a specific layer of the bottom-up stream, or between already-detected objects. We show that the relational process can be better modeled in a coarse-to-fine manner and present a novel framework, applying a non-local module sequentially to increasing resolution feature maps along the top-down stream. In this way, information can naturally passed from larger objects to smaller related ones. Applying the module to fine feature maps further allows the information to pass between the small objects themselves, exploiting repetitions of instances of the same class. In practice, due to the expensive memory utilization of the non-local module, it is infeasible to apply the module as currently used to high-resolution feature maps. We redesigned the non local module, improved it in terms of memory and number of operations, allowing it to be placed anywhere along the network. We further incorporated relative spatial information into the module, in a manner that can be incorporated into our efficient implementation. We show the effectiveness of our scheme by improving the results of detecting small objects on COCO by 1-2 AP points over Faster and Mask RCNN and by 1 AP over using non-local module on the bottom-up stream.</description><identifier>DOI: 10.48550/arxiv.1811.12152</identifier><language>eng</language><subject>Computer Science - Computer Vision and Pattern Recognition</subject><creationdate>2018-11</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/1811.12152$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1811.12152$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Levi, Hila</creatorcontrib><creatorcontrib>Ullman, Shimon</creatorcontrib><title>Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects</title><description>An image is not just a collection of objects, but rather a graph where each object is related to other objects through spatial and semantic relations. Using relational reasoning modules, such as the non-local module \cite{wang2017non}, can therefore improve object detection. Current schemes apply such dedicated modules either to a specific layer of the bottom-up stream, or between already-detected objects. We show that the relational process can be better modeled in a coarse-to-fine manner and present a novel framework, applying a non-local module sequentially to increasing resolution feature maps along the top-down stream. In this way, information can naturally passed from larger objects to smaller related ones. Applying the module to fine feature maps further allows the information to pass between the small objects themselves, exploiting repetitions of instances of the same class. In practice, due to the expensive memory utilization of the non-local module, it is infeasible to apply the module as currently used to high-resolution feature maps. We redesigned the non local module, improved it in terms of memory and number of operations, allowing it to be placed anywhere along the network. We further incorporated relative spatial information into the module, in a manner that can be incorporated into our efficient implementation. We show the effectiveness of our scheme by improving the results of detecting small objects on COCO by 1-2 AP points over Faster and Mask RCNN and by 1 AP over using non-local module on the bottom-up stream.</description><subject>Computer Science - Computer Vision and Pattern Recognition</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2018</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz71OwzAUBWAvDKjwAEz4BRx8_ZuMKLSAFOhA98h2r4VRGiPHIHh7SmE60hmOzkfIFfBGtVrzG1e-0mcDLUADArQ4J8M6xhQSzpX22ZUFWc1sk2akz3lmQw5uok95_zEhjbnQ-or0DiuGmvJMc6QvBzdNdOvfjtVyQc6imxa8_M8V2W3Wu_6BDdv7x_52YM5YwRSgisYHayLXnY0-cAPGcyvBdR0KA0pr0-rjSxmVtDYiRyH8Hr3SrRRyRa7_Zk-c8b2kgyvf4y9rPLHkD-7lRgk</recordid><startdate>20181129</startdate><enddate>20181129</enddate><creator>Levi, Hila</creator><creator>Ullman, Shimon</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20181129</creationdate><title>Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects</title><author>Levi, Hila ; Ullman, Shimon</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a672-41e4f6bc76f0597fbc0616b0731a99e2614556858113f4377fe0e22bdeb458323</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2018</creationdate><topic>Computer Science - Computer Vision and Pattern Recognition</topic><toplevel>online_resources</toplevel><creatorcontrib>Levi, Hila</creatorcontrib><creatorcontrib>Ullman, Shimon</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Levi, Hila</au><au>Ullman, Shimon</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects</atitle><date>2018-11-29</date><risdate>2018</risdate><abstract>An image is not just a collection of objects, but rather a graph where each object is related to other objects through spatial and semantic relations. Using relational reasoning modules, such as the non-local module \cite{wang2017non}, can therefore improve object detection. Current schemes apply such dedicated modules either to a specific layer of the bottom-up stream, or between already-detected objects. We show that the relational process can be better modeled in a coarse-to-fine manner and present a novel framework, applying a non-local module sequentially to increasing resolution feature maps along the top-down stream. In this way, information can naturally passed from larger objects to smaller related ones. Applying the module to fine feature maps further allows the information to pass between the small objects themselves, exploiting repetitions of instances of the same class. In practice, due to the expensive memory utilization of the non-local module, it is infeasible to apply the module as currently used to high-resolution feature maps. We redesigned the non local module, improved it in terms of memory and number of operations, allowing it to be placed anywhere along the network. We further incorporated relative spatial information into the module, in a manner that can be incorporated into our efficient implementation. We show the effectiveness of our scheme by improving the results of detecting small objects on COCO by 1-2 AP points over Faster and Mask RCNN and by 1 AP over using non-local module on the bottom-up stream.</abstract><doi>10.48550/arxiv.1811.12152</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.1811.12152
ispartof
issn
language eng
recordid cdi_arxiv_primary_1811_12152
source arXiv.org
subjects Computer Science - Computer Vision and Pattern Recognition
title Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-11T11%3A19%3A14IST&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=Efficient%20Coarse-to-Fine%20Non-Local%20Module%20for%20the%20Detection%20of%20Small%20Objects&rft.au=Levi,%20Hila&rft.date=2018-11-29&rft_id=info:doi/10.48550/arxiv.1811.12152&rft_dat=%3Carxiv_GOX%3E1811_12152%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