AI-Driven Road Maintenance Inspection v2: Reducing Data Dependency & Quantifying Road Damage

Road infrastructure maintenance inspection is typically a labor-intensive and critical task to ensure the safety of all road users. Existing state-of-the-art techniques in Artificial Intelligence (AI) for object detection and segmentation help automate a huge chunk of this task given adequate annota...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Iqbal, Haris, Chawla, Hemang, Varma, Arnav, Brouns, Terence, Badar, Ahmed, Arani, Elahe, Zonooz, Bahram
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 Iqbal, Haris
Chawla, Hemang
Varma, Arnav
Brouns, Terence
Badar, Ahmed
Arani, Elahe
Zonooz, Bahram
description Road infrastructure maintenance inspection is typically a labor-intensive and critical task to ensure the safety of all road users. Existing state-of-the-art techniques in Artificial Intelligence (AI) for object detection and segmentation help automate a huge chunk of this task given adequate annotated data. However, annotating videos from scratch is cost-prohibitive. For instance, it can take an annotator several days to annotate a 5-minute video recorded at 30 FPS. Hence, we propose an automated labelling pipeline by leveraging techniques like few-shot learning and out-of-distribution detection to generate labels for road damage detection. In addition, our pipeline includes a risk factor assessment for each damage by instance quantification to prioritize locations for repairs which can lead to optimal deployment of road maintenance machinery. We show that the AI models trained with these techniques can not only generalize better to unseen real-world data with reduced requirement for human annotation but also provide an estimate of maintenance urgency, thereby leading to safer roads.
doi_str_mv 10.48550/arxiv.2210.03570
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2210_03570</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2210_03570</sourcerecordid><originalsourceid>FETCH-LOGICAL-a670-2992c3786ee04fc494d6d5a5e7766ddf3a905aeefed651de90794483eff7644f3</originalsourceid><addsrcrecordid>eNotj0tLw0AURmfjQqo_wJWzcpc6nWfGXWl8BCpi6VIIl5k7ZcDehjQN5t9rU1cfnA8OHMbuFmKuS2PEI3Q_eZhL-QeEMk5cs69lXVRdHpD45gCRv0OmHgkoIK_p2GLo84H4IJ_4BuMpZNrxCnrgFbZIESmM_IF_noD6nMbzO2kq2MMOb9hVgu8j3v7vjG1fnrert2L98VqvlusCrBOF9F4G5UqLKHQK2utoowGDzlkbY1LghQHEhNGaRUQvnNe6VJiSs1onNWP3F-2U17Rd3kM3NufMZspUv1jWTLc</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>AI-Driven Road Maintenance Inspection v2: Reducing Data Dependency &amp; Quantifying Road Damage</title><source>arXiv.org</source><creator>Iqbal, Haris ; Chawla, Hemang ; Varma, Arnav ; Brouns, Terence ; Badar, Ahmed ; Arani, Elahe ; Zonooz, Bahram</creator><creatorcontrib>Iqbal, Haris ; Chawla, Hemang ; Varma, Arnav ; Brouns, Terence ; Badar, Ahmed ; Arani, Elahe ; Zonooz, Bahram</creatorcontrib><description>Road infrastructure maintenance inspection is typically a labor-intensive and critical task to ensure the safety of all road users. Existing state-of-the-art techniques in Artificial Intelligence (AI) for object detection and segmentation help automate a huge chunk of this task given adequate annotated data. However, annotating videos from scratch is cost-prohibitive. For instance, it can take an annotator several days to annotate a 5-minute video recorded at 30 FPS. Hence, we propose an automated labelling pipeline by leveraging techniques like few-shot learning and out-of-distribution detection to generate labels for road damage detection. In addition, our pipeline includes a risk factor assessment for each damage by instance quantification to prioritize locations for repairs which can lead to optimal deployment of road maintenance machinery. We show that the AI models trained with these techniques can not only generalize better to unseen real-world data with reduced requirement for human annotation but also provide an estimate of maintenance urgency, thereby leading to safer roads.</description><identifier>DOI: 10.48550/arxiv.2210.03570</identifier><language>eng</language><subject>Computer Science - Computer Vision and Pattern Recognition</subject><creationdate>2022-10</creationdate><rights>http://creativecommons.org/licenses/by-nc-sa/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2210.03570$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2210.03570$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Iqbal, Haris</creatorcontrib><creatorcontrib>Chawla, Hemang</creatorcontrib><creatorcontrib>Varma, Arnav</creatorcontrib><creatorcontrib>Brouns, Terence</creatorcontrib><creatorcontrib>Badar, Ahmed</creatorcontrib><creatorcontrib>Arani, Elahe</creatorcontrib><creatorcontrib>Zonooz, Bahram</creatorcontrib><title>AI-Driven Road Maintenance Inspection v2: Reducing Data Dependency &amp; Quantifying Road Damage</title><description>Road infrastructure maintenance inspection is typically a labor-intensive and critical task to ensure the safety of all road users. Existing state-of-the-art techniques in Artificial Intelligence (AI) for object detection and segmentation help automate a huge chunk of this task given adequate annotated data. However, annotating videos from scratch is cost-prohibitive. For instance, it can take an annotator several days to annotate a 5-minute video recorded at 30 FPS. Hence, we propose an automated labelling pipeline by leveraging techniques like few-shot learning and out-of-distribution detection to generate labels for road damage detection. In addition, our pipeline includes a risk factor assessment for each damage by instance quantification to prioritize locations for repairs which can lead to optimal deployment of road maintenance machinery. We show that the AI models trained with these techniques can not only generalize better to unseen real-world data with reduced requirement for human annotation but also provide an estimate of maintenance urgency, thereby leading to safer roads.</description><subject>Computer Science - Computer Vision and Pattern Recognition</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj0tLw0AURmfjQqo_wJWzcpc6nWfGXWl8BCpi6VIIl5k7ZcDehjQN5t9rU1cfnA8OHMbuFmKuS2PEI3Q_eZhL-QeEMk5cs69lXVRdHpD45gCRv0OmHgkoIK_p2GLo84H4IJ_4BuMpZNrxCnrgFbZIESmM_IF_noD6nMbzO2kq2MMOb9hVgu8j3v7vjG1fnrert2L98VqvlusCrBOF9F4G5UqLKHQK2utoowGDzlkbY1LghQHEhNGaRUQvnNe6VJiSs1onNWP3F-2U17Rd3kM3NufMZspUv1jWTLc</recordid><startdate>20221007</startdate><enddate>20221007</enddate><creator>Iqbal, Haris</creator><creator>Chawla, Hemang</creator><creator>Varma, Arnav</creator><creator>Brouns, Terence</creator><creator>Badar, Ahmed</creator><creator>Arani, Elahe</creator><creator>Zonooz, Bahram</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20221007</creationdate><title>AI-Driven Road Maintenance Inspection v2: Reducing Data Dependency &amp; Quantifying Road Damage</title><author>Iqbal, Haris ; Chawla, Hemang ; Varma, Arnav ; Brouns, Terence ; Badar, Ahmed ; Arani, Elahe ; Zonooz, Bahram</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a670-2992c3786ee04fc494d6d5a5e7766ddf3a905aeefed651de90794483eff7644f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Computer Science - Computer Vision and Pattern Recognition</topic><toplevel>online_resources</toplevel><creatorcontrib>Iqbal, Haris</creatorcontrib><creatorcontrib>Chawla, Hemang</creatorcontrib><creatorcontrib>Varma, Arnav</creatorcontrib><creatorcontrib>Brouns, Terence</creatorcontrib><creatorcontrib>Badar, Ahmed</creatorcontrib><creatorcontrib>Arani, Elahe</creatorcontrib><creatorcontrib>Zonooz, Bahram</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Iqbal, Haris</au><au>Chawla, Hemang</au><au>Varma, Arnav</au><au>Brouns, Terence</au><au>Badar, Ahmed</au><au>Arani, Elahe</au><au>Zonooz, Bahram</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>AI-Driven Road Maintenance Inspection v2: Reducing Data Dependency &amp; Quantifying Road Damage</atitle><date>2022-10-07</date><risdate>2022</risdate><abstract>Road infrastructure maintenance inspection is typically a labor-intensive and critical task to ensure the safety of all road users. Existing state-of-the-art techniques in Artificial Intelligence (AI) for object detection and segmentation help automate a huge chunk of this task given adequate annotated data. However, annotating videos from scratch is cost-prohibitive. For instance, it can take an annotator several days to annotate a 5-minute video recorded at 30 FPS. Hence, we propose an automated labelling pipeline by leveraging techniques like few-shot learning and out-of-distribution detection to generate labels for road damage detection. In addition, our pipeline includes a risk factor assessment for each damage by instance quantification to prioritize locations for repairs which can lead to optimal deployment of road maintenance machinery. We show that the AI models trained with these techniques can not only generalize better to unseen real-world data with reduced requirement for human annotation but also provide an estimate of maintenance urgency, thereby leading to safer roads.</abstract><doi>10.48550/arxiv.2210.03570</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2210.03570
ispartof
issn
language eng
recordid cdi_arxiv_primary_2210_03570
source arXiv.org
subjects Computer Science - Computer Vision and Pattern Recognition
title AI-Driven Road Maintenance Inspection v2: Reducing Data Dependency & Quantifying Road Damage
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-26T07%3A23%3A41IST&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=AI-Driven%20Road%20Maintenance%20Inspection%20v2:%20Reducing%20Data%20Dependency%20&%20Quantifying%20Road%20Damage&rft.au=Iqbal,%20Haris&rft.date=2022-10-07&rft_id=info:doi/10.48550/arxiv.2210.03570&rft_dat=%3Carxiv_GOX%3E2210_03570%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