{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/sequential-score-adaptation-with-extreme","title":"Sequential Score Adaptation with Extreme Value Theory for Robust Railway Track Inspection","arxiv_id":"1510.05822","date":"2015-10-20","proceeding":null,"authors":["Xavier Gibert","Vishal M. Patel","Rama Chellappa"],"abstract":"Periodic inspections are necessary to keep railroad tracks in state of good\nrepair and prevent train accidents. Automatic track inspection using machine\nvision technology has become a very effective inspection tool. Because of its\nnon-contact nature, this technology can be deployed on virtually any railway\nvehicle to continuously survey the tracks and send exception reports to track\nmaintenance personnel. However, as appearance and imaging conditions vary,\nfalse alarm rates can dramatically change, making it difficult to select a good\noperating point. In this paper, we use extreme value theory (EVT) within a\nBayesian framework to optimally adjust the sensitivity of anomaly detectors. We\nshow that by approximating the lower tail of the probability density function\n(PDF) of the scores with an Exponential distribution (a special case of the\nGeneralized Pareto distribution), and using the Gamma conjugate prior learned\nfrom the training data, it is possible to reduce the variability in false alarm\nrate and improve the overall performance. This method has shown an increase in\nthe defect detection rate of rail fasteners in the presence of clutter (at PFA\n0.1%) from 95.40% to 99.26% on the 85-mile Northeast Corridor (NEC) 2012-2013\nconcrete tie dataset.","url_abs":"http://arxiv.org/abs/1510.05822v1","url_pdf":"http://arxiv.org/pdf/1510.05822v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"sequential-score-adaptation-with-extreme","repo_url":"https://github.com/xavigibert/EvtTrack","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"defect-detection","task_name":"Defect Detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}