Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation

As a new embodied vision task, Instance ImageGoal Navigation (IIN) aims to navigate to a specified object depicted by a goal image in an unexplored environment. The main challenge of this task lies in identifying the target object from different viewpoints while rejecting similar distractors. Existi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lei, Xiaohan, Wang, Min, Zhou, Wengang, Li, Li, Li, Houqiang
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 Lei, Xiaohan
Wang, Min
Zhou, Wengang
Li, Li
Li, Houqiang
description As a new embodied vision task, Instance ImageGoal Navigation (IIN) aims to navigate to a specified object depicted by a goal image in an unexplored environment. The main challenge of this task lies in identifying the target object from different viewpoints while rejecting similar distractors. Existing ImageGoal Navigation methods usually adopt the simple Exploration-Exploitation framework and ignore the identification of specific instance during navigation. In this work, we propose to imitate the human behaviour of ``getting closer to confirm" when distinguishing objects from a distance. Specifically, we design a new modular navigation framework named Instance-aware Exploration-Verification-Exploitation (IEVE) for instance-level image goal navigation. Our method allows for active switching among the exploration, verification, and exploitation actions, thereby facilitating the agent in making reasonable decisions under different situations. On the challenging HabitatMatterport 3D semantic (HM3D-SEM) dataset, our method surpasses previous state-of-the-art work, with a classical segmentation model (0.684 vs. 0.561 success) or a robust model (0.702 vs. 0.561 success)
doi_str_mv 10.48550/arxiv.2402.17587
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2402_17587</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2402_17587</sourcerecordid><originalsourceid>FETCH-LOGICAL-a677-61244b3aa5e74017312d7525f8dfd96ba7ed8f86c9a6fcac1000c109c10caf413</originalsourceid><addsrcrecordid>eNo1j82qwjAQhbNxIeoDuDIvkJqk-etSRL0FuW7EbRnTRAK1lVj8efur0buYMxzOzIEPoSmjmTBS0jnER7hlXFCeMS2NHqJD2V57aK0jcIfo8OpxaboIfehacnAx-GA_JgWhTwb7LuL_R1ye4eQ2HTT4F27hlC7GaOChubrJd4_Qfr3aL3_Idrcpl4stAaU1UYwLccwBpNOCMp0zXmvJpTe1rwt1BO1q442yBShvwTJK6UuK11jwguUjNPvUJrDqEsMZ4rN6A1YJMP8DBOdM2Q</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation</title><source>arXiv.org</source><creator>Lei, Xiaohan ; Wang, Min ; Zhou, Wengang ; Li, Li ; Li, Houqiang</creator><creatorcontrib>Lei, Xiaohan ; Wang, Min ; Zhou, Wengang ; Li, Li ; Li, Houqiang</creatorcontrib><description>As a new embodied vision task, Instance ImageGoal Navigation (IIN) aims to navigate to a specified object depicted by a goal image in an unexplored environment. The main challenge of this task lies in identifying the target object from different viewpoints while rejecting similar distractors. Existing ImageGoal Navigation methods usually adopt the simple Exploration-Exploitation framework and ignore the identification of specific instance during navigation. In this work, we propose to imitate the human behaviour of ``getting closer to confirm" when distinguishing objects from a distance. Specifically, we design a new modular navigation framework named Instance-aware Exploration-Verification-Exploitation (IEVE) for instance-level image goal navigation. Our method allows for active switching among the exploration, verification, and exploitation actions, thereby facilitating the agent in making reasonable decisions under different situations. On the challenging HabitatMatterport 3D semantic (HM3D-SEM) dataset, our method surpasses previous state-of-the-art work, with a classical segmentation model (0.684 vs. 0.561 success) or a robust model (0.702 vs. 0.561 success)</description><identifier>DOI: 10.48550/arxiv.2402.17587</identifier><language>eng</language><subject>Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Robotics</subject><creationdate>2024-02</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>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2402.17587$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2402.17587$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Lei, Xiaohan</creatorcontrib><creatorcontrib>Wang, Min</creatorcontrib><creatorcontrib>Zhou, Wengang</creatorcontrib><creatorcontrib>Li, Li</creatorcontrib><creatorcontrib>Li, Houqiang</creatorcontrib><title>Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation</title><description>As a new embodied vision task, Instance ImageGoal Navigation (IIN) aims to navigate to a specified object depicted by a goal image in an unexplored environment. The main challenge of this task lies in identifying the target object from different viewpoints while rejecting similar distractors. Existing ImageGoal Navigation methods usually adopt the simple Exploration-Exploitation framework and ignore the identification of specific instance during navigation. In this work, we propose to imitate the human behaviour of ``getting closer to confirm" when distinguishing objects from a distance. Specifically, we design a new modular navigation framework named Instance-aware Exploration-Verification-Exploitation (IEVE) for instance-level image goal navigation. Our method allows for active switching among the exploration, verification, and exploitation actions, thereby facilitating the agent in making reasonable decisions under different situations. On the challenging HabitatMatterport 3D semantic (HM3D-SEM) dataset, our method surpasses previous state-of-the-art work, with a classical segmentation model (0.684 vs. 0.561 success) or a robust model (0.702 vs. 0.561 success)</description><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Computer Science - Robotics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNo1j82qwjAQhbNxIeoDuDIvkJqk-etSRL0FuW7EbRnTRAK1lVj8efur0buYMxzOzIEPoSmjmTBS0jnER7hlXFCeMS2NHqJD2V57aK0jcIfo8OpxaboIfehacnAx-GA_JgWhTwb7LuL_R1ye4eQ2HTT4F27hlC7GaOChubrJd4_Qfr3aL3_Idrcpl4stAaU1UYwLccwBpNOCMp0zXmvJpTe1rwt1BO1q442yBShvwTJK6UuK11jwguUjNPvUJrDqEsMZ4rN6A1YJMP8DBOdM2Q</recordid><startdate>20240225</startdate><enddate>20240225</enddate><creator>Lei, Xiaohan</creator><creator>Wang, Min</creator><creator>Zhou, Wengang</creator><creator>Li, Li</creator><creator>Li, Houqiang</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20240225</creationdate><title>Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation</title><author>Lei, Xiaohan ; Wang, Min ; Zhou, Wengang ; Li, Li ; Li, Houqiang</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a677-61244b3aa5e74017312d7525f8dfd96ba7ed8f86c9a6fcac1000c109c10caf413</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Computer Science - Robotics</topic><toplevel>online_resources</toplevel><creatorcontrib>Lei, Xiaohan</creatorcontrib><creatorcontrib>Wang, Min</creatorcontrib><creatorcontrib>Zhou, Wengang</creatorcontrib><creatorcontrib>Li, Li</creatorcontrib><creatorcontrib>Li, Houqiang</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Lei, Xiaohan</au><au>Wang, Min</au><au>Zhou, Wengang</au><au>Li, Li</au><au>Li, Houqiang</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation</atitle><date>2024-02-25</date><risdate>2024</risdate><abstract>As a new embodied vision task, Instance ImageGoal Navigation (IIN) aims to navigate to a specified object depicted by a goal image in an unexplored environment. The main challenge of this task lies in identifying the target object from different viewpoints while rejecting similar distractors. Existing ImageGoal Navigation methods usually adopt the simple Exploration-Exploitation framework and ignore the identification of specific instance during navigation. In this work, we propose to imitate the human behaviour of ``getting closer to confirm" when distinguishing objects from a distance. Specifically, we design a new modular navigation framework named Instance-aware Exploration-Verification-Exploitation (IEVE) for instance-level image goal navigation. Our method allows for active switching among the exploration, verification, and exploitation actions, thereby facilitating the agent in making reasonable decisions under different situations. On the challenging HabitatMatterport 3D semantic (HM3D-SEM) dataset, our method surpasses previous state-of-the-art work, with a classical segmentation model (0.684 vs. 0.561 success) or a robust model (0.702 vs. 0.561 success)</abstract><doi>10.48550/arxiv.2402.17587</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2402.17587
ispartof
issn
language eng
recordid cdi_arxiv_primary_2402_17587
source arXiv.org
subjects Computer Science - Computer Vision and Pattern Recognition
Computer Science - Robotics
title Instance-aware Exploration-Verification-Exploitation for Instance ImageGoal Navigation
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-02T09%3A28%3A48IST&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=Instance-aware%20Exploration-Verification-Exploitation%20for%20Instance%20ImageGoal%20Navigation&rft.au=Lei,%20Xiaohan&rft.date=2024-02-25&rft_id=info:doi/10.48550/arxiv.2402.17587&rft_dat=%3Carxiv_GOX%3E2402_17587%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