Segmenting Without Annotating: Crack Segmentation and Monitoring via Post-Hoc Classifier Explanations. Forest, F., Porta, H., Tuia, D., & Fink, O. In Proceedings of the 33rd European Safety and Reliability Conference (ESREL), pages 1392–1393, 2023. ![Segmenting Without Annotating: Crack Segmentation and Monitoring via Post-Hoc Classifier Explanations [link]](https://bibbase.org/img/filetypes/link.svg "link")
Link ![Segmenting Without Annotating: Crack Segmentation and Monitoring via Post-Hoc Classifier Explanations [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper doi abstract bibtex Monitoring the cracks in walls, roads and other types of infrastructure is essential to ensure the safety of a structure, and plays an important role in structural health monitoring. Automatic visual inspection allows an efficient, costeffective and safe health monitoring, especially in hard-to-reach locations. To this aim, data-driven approaches based on machine learning have demonstrated their effectiveness, at the expense of annotating large sets of images for supervised training. Once a damage has been detected, one also needs to monitor the evolution of its severity, in order to trigger a timely maintenance operation and avoid any catastrophic consequence. This evaluation requires a precise segmentation of the damage. However, pixel-level annotation of images for segmentation is labor-intensive. On the other hand, labeling images for a classification task is relatively cheap in comparison. To circumvent the cost of annotating images for segmentation, recent works inspired by explainable AI (XAI) have proposed to use the post-hoc explanations of a classifier to obtain a segmentation of the input image. In this work, we study the application of XAI techniques to the detection and monitoring of cracks in masonry wall surfaces. We benchmark different post-hoc explainability methods in terms of segmentation quality and accuracy of the damage severity quantification (for example, the width of a crack), thus enabling timely decision-making.
@inproceedings{forest2023segmenting,
title = {Segmenting {Without} {Annotating}: {Crack} {Segmentation} and {Monitoring} via {Post}-{Hoc} {Classifier} {Explanations}},
copyright = {All rights reserved},
isbn = {978-981-18807-1-1},
shorttitle = {Segmenting {Without} {Annotating}},
doi = {10.3850/978-981-18-8071-1_P290-cd},
abstract = {Monitoring the cracks in walls, roads and other types of infrastructure is essential to ensure the safety of a structure, and plays an important role in structural health monitoring. Automatic visual inspection allows an efficient, costeffective and safe health monitoring, especially in hard-to-reach locations. To this aim, data-driven approaches based on machine learning have demonstrated their effectiveness, at the expense of annotating large sets of images for supervised training. Once a damage has been detected, one also needs to monitor the evolution of its severity, in order to trigger a timely maintenance operation and avoid any catastrophic consequence. This evaluation requires a precise segmentation of the damage. However, pixel-level annotation of images for segmentation is labor-intensive. On the other hand, labeling images for a classification task is relatively cheap in comparison. To circumvent the cost of annotating images for segmentation, recent works inspired by explainable AI (XAI) have proposed to use the post-hoc explanations of a classifier to obtain a segmentation of the input image. In this work, we study the application of XAI techniques to the detection and monitoring of cracks in masonry wall surfaces. We benchmark different post-hoc explainability methods in terms of segmentation quality and accuracy of the damage severity quantification (for example, the width of a crack), thus enabling timely decision-making.},
booktitle = {Proceedings of the 33rd {European} {Safety} and {Reliability} {Conference} (ESREL)},
author = {Forest, Florent and Porta, Hugo and Tuia, Devis and Fink, Olga},
year = {2023},
pages = {1392--1393},
url_Link = {https://www.rpsonline.com.sg/proceedings/esrel2023/html/P290.html},
url_Paper = {https://www.rpsonline.com.sg/proceedings/esrel2023/pdf/P290.pdf},
}
Downloads: 0
{"_id":"atJQ3DfQvixw5AMb6","bibbaseid":"forest-porta-tuia-fink-segmentingwithoutannotatingcracksegmentationandmonitoringviaposthocclassifierexplanations-2023","author_short":["Forest, F.","Porta, H.","Tuia, D.","Fink, O."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Segmenting Without Annotating: Crack Segmentation and Monitoring via Post-Hoc Classifier Explanations","copyright":"All rights reserved","isbn":"978-981-18807-1-1","shorttitle":"Segmenting Without Annotating","doi":"10.3850/978-981-18-8071-1_P290-cd","abstract":"Monitoring the cracks in walls, roads and other types of infrastructure is essential to ensure the safety of a structure, and plays an important role in structural health monitoring. Automatic visual inspection allows an efficient, costeffective and safe health monitoring, especially in hard-to-reach locations. To this aim, data-driven approaches based on machine learning have demonstrated their effectiveness, at the expense of annotating large sets of images for supervised training. Once a damage has been detected, one also needs to monitor the evolution of its severity, in order to trigger a timely maintenance operation and avoid any catastrophic consequence. This evaluation requires a precise segmentation of the damage. However, pixel-level annotation of images for segmentation is labor-intensive. On the other hand, labeling images for a classification task is relatively cheap in comparison. To circumvent the cost of annotating images for segmentation, recent works inspired by explainable AI (XAI) have proposed to use the post-hoc explanations of a classifier to obtain a segmentation of the input image. In this work, we study the application of XAI techniques to the detection and monitoring of cracks in masonry wall surfaces. We benchmark different post-hoc explainability methods in terms of segmentation quality and accuracy of the damage severity quantification (for example, the width of a crack), thus enabling timely decision-making.","booktitle":"Proceedings of the 33rd European Safety and Reliability Conference (ESREL)","author":[{"propositions":[],"lastnames":["Forest"],"firstnames":["Florent"],"suffixes":[]},{"propositions":[],"lastnames":["Porta"],"firstnames":["Hugo"],"suffixes":[]},{"propositions":[],"lastnames":["Tuia"],"firstnames":["Devis"],"suffixes":[]},{"propositions":[],"lastnames":["Fink"],"firstnames":["Olga"],"suffixes":[]}],"year":"2023","pages":"1392–1393","url_link":"https://www.rpsonline.com.sg/proceedings/esrel2023/html/P290.html","url_paper":"https://www.rpsonline.com.sg/proceedings/esrel2023/pdf/P290.pdf","bibtex":"@inproceedings{forest2023segmenting,\n\ttitle = {Segmenting {Without} {Annotating}: {Crack} {Segmentation} and {Monitoring} via {Post}-{Hoc} {Classifier} {Explanations}},\n\tcopyright = {All rights reserved},\n\tisbn = {978-981-18807-1-1},\n\tshorttitle = {Segmenting {Without} {Annotating}},\n\tdoi = {10.3850/978-981-18-8071-1_P290-cd},\n\tabstract = {Monitoring the cracks in walls, roads and other types of infrastructure is essential to ensure the safety of a structure, and plays an important role in structural health monitoring. Automatic visual inspection allows an efficient, costeffective and safe health monitoring, especially in hard-to-reach locations. To this aim, data-driven approaches based on machine learning have demonstrated their effectiveness, at the expense of annotating large sets of images for supervised training. Once a damage has been detected, one also needs to monitor the evolution of its severity, in order to trigger a timely maintenance operation and avoid any catastrophic consequence. This evaluation requires a precise segmentation of the damage. However, pixel-level annotation of images for segmentation is labor-intensive. On the other hand, labeling images for a classification task is relatively cheap in comparison. To circumvent the cost of annotating images for segmentation, recent works inspired by explainable AI (XAI) have proposed to use the post-hoc explanations of a classifier to obtain a segmentation of the input image. In this work, we study the application of XAI techniques to the detection and monitoring of cracks in masonry wall surfaces. We benchmark different post-hoc explainability methods in terms of segmentation quality and accuracy of the damage severity quantification (for example, the width of a crack), thus enabling timely decision-making.},\n\tbooktitle = {Proceedings of the 33rd {European} {Safety} and {Reliability} {Conference} (ESREL)},\n\tauthor = {Forest, Florent and Porta, Hugo and Tuia, Devis and Fink, Olga},\n\tyear = {2023},\n\tpages = {1392--1393},\n\turl_Link = {https://www.rpsonline.com.sg/proceedings/esrel2023/html/P290.html},\n\turl_Paper = {https://www.rpsonline.com.sg/proceedings/esrel2023/pdf/P290.pdf},\n}\n\n","author_short":["Forest, F.","Porta, H.","Tuia, D.","Fink, O."],"key":"forest2023segmenting","id":"forest2023segmenting","bibbaseid":"forest-porta-tuia-fink-segmentingwithoutannotatingcracksegmentationandmonitoringviaposthocclassifierexplanations-2023","role":"author","urls":{" link":"https://www.rpsonline.com.sg/proceedings/esrel2023/html/P290.html"," paper":"https://www.rpsonline.com.sg/proceedings/esrel2023/pdf/P290.pdf"},"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://florentfo.rest/files/publications.bib","dataSources":["2puawT8ZAQyYRypA3","6rNfa4Kp6dL5sGmf5","xH8ySTsEPTLou9gyR","pBkCjKbyeirr5jeAd","DgnR6pzJ98ZEp97PW"],"keywords":[],"search_terms":["segmenting","without","annotating","crack","segmentation","monitoring","via","post","hoc","classifier","explanations","forest","porta","tuia","fink"],"title":"Segmenting Without Annotating: Crack Segmentation and Monitoring via Post-Hoc Classifier Explanations","year":2023}