{"id":1185252,"date":"2021-04-07T09:52:33","date_gmt":"2021-04-07T16:52:33","guid":{"rendered":"https:\/\/www.esri.com\/arcgis-blog\/?post_type=blog&#038;p=1185252"},"modified":"2021-04-13T11:59:43","modified_gmt":"2021-04-13T18:59:43","slug":"devsummit-2021-pre-trained-models-in-arcgis","status":"publish","type":"blog","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis","title":{"rendered":"Dev Summit 2021: Pre-Trained Models in ArcGIS"},"author":7461,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_acf_changed":false,"_searchwp_excluded":""},"categories":[23341],"tags":[186132,759812,759612],"industry":[],"product":[36841,36581,36561],"class_list":["post-1185252","blog","type-blog","status-publish","format-standard","hentry","category-analytics","tag-deep-learning","tag-dev-summit-2021-demo","tag-model-builder","product-api-python","product-arcgis-living-atlas","product-arcgis-pro"],"acf":{"short_description":"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point cloud data.","flexible_content":[{"acf_fc_layout":"content","content":"<p>I<span data-contrast=\"auto\">n this demonstration, Saranya M from Esri India <\/span><span data-contrast=\"auto\">reveals how they used AI to extract railway assets<\/span><span data-contrast=\"auto\"> (e.g., <\/span><span data-contrast=\"auto\">powerlines, poles, railway track and cantilever)<\/span><span data-contrast=\"auto\"> and their dimensions <\/span><span data-contrast=\"auto\">from point cloud for their client, L&amp;T NXT. L&amp;T was working on one of India\u2019s largest railway project<\/span><span data-contrast=\"auto\">s where they observed that during field inspections<\/span><span data-contrast=\"auto\"> manual<\/span><span data-contrast=\"auto\">ly measuring t<\/span><span data-contrast=\"auto\">he dimensions of railway assets was both time consuming and susceptible to human error. <\/span><span data-contrast=\"auto\">They approached Esri India to design an end-to-end workflow to classify these point clouds and extract the dimensions, for the trains to run safely.\u00a0<\/span><span data-ccp-props=\"{&quot;201341983&quot;:0,&quot;335559739&quot;:160,&quot;335559740&quot;:259}\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">They decided to use the <\/span><b><span data-contrast=\"auto\">\u201cPowerline Classification Model\u201d<\/span><\/b><span data-contrast=\"auto\"> which is a pre-trained model available on <\/span><span data-contrast=\"auto\">ArcGIS Living Atlas<\/span><b><span data-contrast=\"auto\">. <\/span><\/b><span data-contrast=\"auto\">Even though this model was trained on electric wires and poles, it gave very good results for <\/span><span data-contrast=\"auto\">the <\/span><span data-contrast=\"auto\">classif<\/span><span data-contrast=\"auto\">ication of railway <\/span><span data-contrast=\"auto\">pole<\/span><span data-contrast=\"auto\">s<\/span><span data-contrast=\"auto\"> and powerline<\/span><span data-contrast=\"auto\">s<\/span> <span data-contrast=\"auto\">too<\/span><span data-contrast=\"auto\">.<\/span><\/p>\n"},{"acf_fc_layout":"image","image":{"ID":1185262,"id":1185262,"title":"1-predefined-models","filename":"1-predefined-models.png","filesize":357986,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\/1-predefined-models","alt":"pre-trained-models1","author":"7461","description":"","caption":"L-R: Input point cloud data, features classified using pre-trained model, features classified with transfer learning","name":"1-predefined-models","status":"inherit","uploaded_to":1185252,"date":"2021-04-07 03:47:46","modified":"2021-04-07 03:49:53","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":1920,"height":966,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models.png","medium-width":464,"medium-height":233,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models.png","medium_large-width":768,"medium_large-height":386,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models.png","large-width":1920,"large-height":966,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models-1536x773.png","1536x1536-width":1536,"1536x1536-height":773,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models.png","2048x2048-width":1920,"2048x2048-height":966,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models-826x416.png","card_image-width":826,"card_image-height":416,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/1-predefined-models.png","wide_image-width":1920,"wide_image-height":966}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"youtube","start_time":"0","end_time":"","youtube_video_url":"<iframe title=\"AI User Story: Extracting Railway Assets from LiDAR from L&amp;T\" width=\"640\" height=\"360\" src=\"https:\/\/www.youtube.com\/embed\/HWVmyyEKn8Q?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>"},{"acf_fc_layout":"content","content":"<p>&nbsp;<\/p>\n<p><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\">But since the goal was to obtain extremely precise dime<\/span><\/span><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\">nsions in the interest of safety, Saranya explains how they achieved it<\/span><\/span><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\"> with transfer learning<\/span><\/span><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\">, which is a field of machine learning that applies the knowledge gained from solving one problem to another similar problem. <\/span><\/span>Fine-tuning the pre-trained model helped save time, cost and computational resources. <span class=\"TrackChangeTextInsertion TrackedChange BCX2 SCXW240670125\"><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\">As verified by their team, t<\/span><\/span><\/span><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun CommentStart ContextualSpellingAndGrammarErrorV2 BCX2 SCXW240670125\">he<\/span><span class=\"NormalTextRun BCX2 SCXW240670125\"> process also resulted in improved precision and recall by <\/span><\/span><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\">a substantial margin<\/span><\/span> <span class=\"TrackChangeTextInsertion TrackedChange BCX2 SCXW240670125\"><span class=\"TextRun BCX2 SCXW240670125\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW240670125\">as compared to what would have been achieved if the model was trained from scratch. <\/span><\/span><\/span>Fine tuning the pre-trained models not only helps in saving time &amp; resources but also helps in achieving better results than what would have been achieved had the models been trained from scratch.<\/p>\n<p><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">She <\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">then demonstrates through <\/span><\/span><strong><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun SpellingErrorV2 BCX2 SCXW174130810\">Jupyter<\/span><\/span> <span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">notebooks <\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">in ArcGIS Pro <\/span><\/span><\/strong><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun CommentStart BCX2 SCXW174130810\">how <\/span><\/span><span class=\"TrackChangeTextInsertion TrackedChange BCX2 SCXW174130810\"><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">a small subset <\/span><\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">manually labelled data<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\"> was used to fine-tune <\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">their pre-trained model. They also trained a new model from scratch to classify the objects that were<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\"> not supported by the Powerline Classification Model, <\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">i<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">.e., the railway tracks and the cantilevers. <\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">As we see below<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun CommentStart BCX2 SCXW174130810\">,<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\"> t<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">he outputs from these two models were combined using<\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\"> the functionalities of <\/span><\/span><strong><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">selective classification<\/span><\/span><\/strong><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\"> and <\/span><\/span><strong><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">class preservation<\/span><\/span><\/strong><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">, which are available in the <\/span><\/span><span class=\"TextRun BCX2 SCXW174130810\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"auto\"><span class=\"NormalTextRun BCX2 SCXW174130810\">ArcGIS API for Python<\/span><\/span><\/p>\n"},{"acf_fc_layout":"image","image":{"ID":1185302,"id":1185302,"title":"predefined-models-4","filename":"predefined-models-4.png","filesize":305167,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\/predefined-models-4","alt":"pre-trained-4","author":"7461","description":"","caption":"L-R: output from the fine-tuned model, output from the new model, combined output created using new functionalities in the API","name":"predefined-models-4","status":"inherit","uploaded_to":1185252,"date":"2021-04-07 04:04:18","modified":"2021-04-07 04:06:12","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":1920,"height":968,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4.png","medium-width":464,"medium-height":234,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4.png","medium_large-width":768,"medium_large-height":387,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4.png","large-width":1920,"large-height":968,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4-1536x774.png","1536x1536-width":1536,"1536x1536-height":774,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4.png","2048x2048-width":1920,"2048x2048-height":968,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4-826x416.png","card_image-width":826,"card_image-height":416,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-4.png","wide_image-width":1920,"wide_image-height":968}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"<p><span data-contrast=\"auto\">After successfully classifying all their data, they used the <\/span><b><span data-contrast=\"auto\">ModelBuilder<\/span><\/b> <span data-contrast=\"auto\">framework<\/span><span data-contrast=\"auto\"> in ArcGIS Pro<\/span><span data-contrast=\"auto\"> to<\/span><span data-contrast=\"auto\"> automate the dimension extraction process<\/span> <span data-contrast=\"auto\">by bringing in <\/span><span data-contrast=\"auto\">several geoprocessing tools such as <\/span><b><span data-contrast=\"auto\">density based<\/span><\/b><b><span data-contrast=\"auto\"> clustering, minimum bounding volume, zonal statistics<\/span><\/b><b><span data-contrast=\"auto\">. <\/span><\/b><span data-contrast=\"auto\">This provided <\/span><span data-contrast=\"auto\">dimensions o<\/span><span data-contrast=\"auto\">f railway assets such as <\/span><span data-contrast=\"auto\">pole heights, centerline geometry of the railway track, catenary height and contact height<\/span><span data-contrast=\"auto\">,<\/span><span data-contrast=\"auto\"> and even complex measurements such as top fitting and bottom fitting. And<\/span><span data-contrast=\"auto\"> the best part about it? A<\/span><span data-contrast=\"auto\">ll their ex<\/span><span data-contrast=\"auto\">tracted<\/span><span data-contrast=\"auto\"> measurements matched <\/span><span data-contrast=\"auto\">those of the railway assets.<\/span><span data-ccp-props=\"{&quot;201341983&quot;:0,&quot;335551550&quot;:1,&quot;335551620&quot;:1,&quot;335559739&quot;:160,&quot;335559740&quot;:259}\">\u00a0<\/span><\/p>\n"},{"acf_fc_layout":"image","image":{"ID":1185292,"id":1185292,"title":"predefined-models-3","filename":"predefined-models-3.png","filesize":139531,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\/predefined-models-3","alt":"model3","author":"7461","description":"","caption":"The model builder framework to automate dimension extraction","name":"predefined-models-3","status":"inherit","uploaded_to":1185252,"date":"2021-04-07 03:56:19","modified":"2021-04-07 03:57:59","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":1920,"height":962,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3.png","medium-width":464,"medium-height":232,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3.png","medium_large-width":768,"medium_large-height":385,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3.png","large-width":1920,"large-height":962,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3-1536x770.png","1536x1536-width":1536,"1536x1536-height":770,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3.png","2048x2048-width":1920,"2048x2048-height":962,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3-826x414.png","card_image-width":826,"card_image-height":414,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/predefined-models-3.png","wide_image-width":1920,"wide_image-height":962}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"<p><span data-contrast=\"auto\">Additional Resources:<\/span><span data-ccp-props=\"{&quot;201341983&quot;:0,&quot;335551550&quot;:1,&quot;335551620&quot;:1,&quot;335559739&quot;:160,&quot;335559740&quot;:259}\">\u00a0<\/span><\/p>\n<ul>\n<li data-leveltext=\"\uf0b7\" data-font=\"Symbol\" data-listid=\"1\" data-aria-posinset=\"1\" data-aria-level=\"1\"><a href=\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis\/imagery\/introducing-ready-to-use-deep-learning-models\/\"><span data-contrast=\"none\">Introducing pre-trained geospatial deep learning models<\/span><\/a><span data-ccp-props=\"{&quot;134233279&quot;:true,&quot;201341983&quot;:0,&quot;335551550&quot;:1,&quot;335551620&quot;:1,&quot;335559739&quot;:160,&quot;335559740&quot;:259}\">\u00a0<\/span><\/li>\n<li data-leveltext=\"\uf0b7\" data-font=\"Symbol\" data-listid=\"1\" data-aria-posinset=\"2\" data-aria-level=\"1\"><a href=\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis\/announcements\/announcing-new-pretrained-models-at-fedgis\/\"><span data-contrast=\"none\">Pre-trained deep learning models update (February 2021)<\/span><\/a><span data-ccp-props=\"{&quot;134233279&quot;:true,&quot;201341983&quot;:0,&quot;335551550&quot;:1,&quot;335551620&quot;:1,&quot;335559739&quot;:160,&quot;335559740&quot;:259}\">\u00a0<\/span><\/li>\n<li data-leveltext=\"\uf0b7\" data-font=\"Symbol\" data-listid=\"1\" data-aria-posinset=\"3\" data-aria-level=\"1\"><a href=\"https:\/\/livingatlas.arcgis.com\/en\/browse\/#d=2&amp;q=deep%20learning&amp;type=tool\"><span data-contrast=\"none\">List of pre-trained models in Living Atlas<\/span><\/a><span data-ccp-props=\"{&quot;134233279&quot;:true,&quot;201341983&quot;:0,&quot;335551550&quot;:1,&quot;335551620&quot;:1,&quot;335559739&quot;:160,&quot;335559740&quot;:259}\">\u00a0<\/span><\/li>\n<\/ul>\n"}],"authors":[{"ID":7461,"user_firstname":"Manushi","user_lastname":"Majumdar","nickname":"Manushi Majumdar","user_nicename":"mmajumdar_dcdev","display_name":"Manushi Majumdar","user_email":"MMajumdar@esri.com","user_url":"","user_registered":"2018-03-21 18:21:20","user_description":"Product Engineer - Applied Data Science with ArcGIS API for Python. Or in other words, a (Data, Maps, Analyses, Python, Books) Nerd.","user_avatar":"<img data-del=\"avatar\" src='https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/03\/me_cropped-213x200.jpg' class='avatar pp-user-avatar avatar-96 photo ' height='96' width='96'\/>"}],"related_articles":"","card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2021\/04\/pretrained-banner.jpg","wide_image":false},"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v25.9 (Yoast SEO v25.9) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Dev Summit 2021: Pre-Trained Models in ArcGIS<\/title>\n<meta name=\"description\" content=\"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point clouds.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Dev Summit 2021: Pre-Trained Models in ArcGIS\" \/>\n<meta property=\"og:description\" content=\"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point clouds.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\" \/>\n<meta property=\"og:site_name\" content=\"ArcGIS Blog\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/esrigis\/\" \/>\n<meta property=\"article:modified_time\" content=\"2021-04-13T18:59:43+00:00\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:site\" content=\"@ESRI\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":[\"Article\",\"BlogPosting\"],\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#article\",\"isPartOf\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\"},\"author\":{\"name\":\"Manushi Majumdar\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/person\/a4b2b6288c6864279ac548b1846ec901\"},\"headline\":\"Dev Summit 2021: Pre-Trained Models in ArcGIS\",\"datePublished\":\"2021-04-07T16:52:33+00:00\",\"dateModified\":\"2021-04-13T18:59:43+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\"},\"wordCount\":6,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#organization\"},\"keywords\":[\"deep learning\",\"Dev Summit 2021 demo\",\"model builder\"],\"articleSection\":[\"Analytics\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\",\"url\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\",\"name\":\"Dev Summit 2021: Pre-Trained Models in ArcGIS\",\"isPartOf\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#website\"},\"datePublished\":\"2021-04-07T16:52:33+00:00\",\"dateModified\":\"2021-04-13T18:59:43+00:00\",\"description\":\"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point clouds.\",\"breadcrumb\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/www.esri.com\/arcgis-blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Dev Summit 2021: Pre-Trained Models in ArcGIS\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#website\",\"url\":\"https:\/\/www.esri.com\/arcgis-blog\/\",\"name\":\"ArcGIS Blog\",\"description\":\"Get insider info from Esri product teams\",\"publisher\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.esri.com\/arcgis-blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#organization\",\"name\":\"Esri\",\"url\":\"https:\/\/www.esri.com\/arcgis-blog\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2018\/04\/Esri.png\",\"contentUrl\":\"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2018\/04\/Esri.png\",\"width\":400,\"height\":400,\"caption\":\"Esri\"},\"image\":{\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.facebook.com\/esrigis\/\",\"https:\/\/x.com\/ESRI\",\"https:\/\/www.linkedin.com\/company\/5311\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/person\/a4b2b6288c6864279ac548b1846ec901\",\"name\":\"Manushi Majumdar\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/03\/me_cropped-213x200.jpg\",\"contentUrl\":\"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/03\/me_cropped-213x200.jpg\",\"caption\":\"Manushi Majumdar\"},\"description\":\"Product Engineer - Applied Data Science with ArcGIS API for Python. Or in other words, a (Data, Maps, Analyses, Python, Books) Nerd.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/manushimajumdar\/\",\"https:\/\/x.com\/Manushi_M\"],\"url\":\"https:\/\/www.esri.com\/arcgis-blog\/author\/mmajumdar_dcdev\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Dev Summit 2021: Pre-Trained Models in ArcGIS","description":"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point clouds.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis","og_locale":"en_US","og_type":"article","og_title":"Dev Summit 2021: Pre-Trained Models in ArcGIS","og_description":"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point clouds.","og_url":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis","og_site_name":"ArcGIS Blog","article_publisher":"https:\/\/www.facebook.com\/esrigis\/","article_modified_time":"2021-04-13T18:59:43+00:00","twitter_card":"summary_large_image","twitter_site":"@ESRI","schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":["Article","BlogPosting"],"@id":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#article","isPartOf":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis"},"author":{"name":"Manushi Majumdar","@id":"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/person\/a4b2b6288c6864279ac548b1846ec901"},"headline":"Dev Summit 2021: Pre-Trained Models in ArcGIS","datePublished":"2021-04-07T16:52:33+00:00","dateModified":"2021-04-13T18:59:43+00:00","mainEntityOfPage":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis"},"wordCount":6,"commentCount":0,"publisher":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/#organization"},"keywords":["deep learning","Dev Summit 2021 demo","model builder"],"articleSection":["Analytics"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis","url":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis","name":"Dev Summit 2021: Pre-Trained Models in ArcGIS","isPartOf":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/#website"},"datePublished":"2021-04-07T16:52:33+00:00","dateModified":"2021-04-13T18:59:43+00:00","description":"Saranya M from Esri India reveals how they used AI and pre-trained models to extract railway assets and their dimensions from point clouds.","breadcrumb":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/analytics\/devsummit-2021-pre-trained-models-in-arcgis#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.esri.com\/arcgis-blog\/"},{"@type":"ListItem","position":2,"name":"Dev Summit 2021: Pre-Trained Models in ArcGIS"}]},{"@type":"WebSite","@id":"https:\/\/www.esri.com\/arcgis-blog\/#website","url":"https:\/\/www.esri.com\/arcgis-blog\/","name":"ArcGIS Blog","description":"Get insider info from Esri product teams","publisher":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.esri.com\/arcgis-blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.esri.com\/arcgis-blog\/#organization","name":"Esri","url":"https:\/\/www.esri.com\/arcgis-blog\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/logo\/image\/","url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2018\/04\/Esri.png","contentUrl":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2018\/04\/Esri.png","width":400,"height":400,"caption":"Esri"},"image":{"@id":"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/esrigis\/","https:\/\/x.com\/ESRI","https:\/\/www.linkedin.com\/company\/5311\/"]},{"@type":"Person","@id":"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/person\/a4b2b6288c6864279ac548b1846ec901","name":"Manushi Majumdar","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.esri.com\/arcgis-blog\/#\/schema\/person\/image\/","url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/03\/me_cropped-213x200.jpg","contentUrl":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/03\/me_cropped-213x200.jpg","caption":"Manushi Majumdar"},"description":"Product Engineer - Applied Data Science with ArcGIS API for Python. Or in other words, a (Data, Maps, Analyses, Python, Books) Nerd.","sameAs":["https:\/\/www.linkedin.com\/in\/manushimajumdar\/","https:\/\/x.com\/Manushi_M"],"url":"https:\/\/www.esri.com\/arcgis-blog\/author\/mmajumdar_dcdev"}]}},"text_date":"April 7, 2021","author_name":"Manushi Majumdar","author_page":"https:\/\/www.esri.com\/arcgis-blog\/author\/mmajumdar_dcdev","custom_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2025\/08\/Newsroom-Keyart-Wide-1920-x-1080.jpg","primary_product":"ArcGIS Pro","tag_data":[{"term_id":186132,"name":"deep learning","slug":"deep-learning","term_group":0,"term_taxonomy_id":186132,"taxonomy":"post_tag","description":"","parent":0,"count":65,"filter":"raw"},{"term_id":759812,"name":"Dev Summit 2021 demo","slug":"dev-summit-2021-demo","term_group":0,"term_taxonomy_id":759812,"taxonomy":"post_tag","description":"","parent":0,"count":17,"filter":"raw"},{"term_id":759612,"name":"model builder","slug":"model-builder","term_group":0,"term_taxonomy_id":759612,"taxonomy":"post_tag","description":"","parent":0,"count":4,"filter":"raw"}],"category_data":[{"term_id":23341,"name":"Analytics","slug":"analytics","term_group":0,"term_taxonomy_id":23341,"taxonomy":"category","description":"","parent":0,"count":1329,"filter":"raw"}],"product_data":[{"term_id":36841,"name":"ArcGIS API for Python","slug":"api-python","term_group":0,"term_taxonomy_id":36841,"taxonomy":"product","description":"","parent":36601,"count":151,"filter":"raw"},{"term_id":36581,"name":"ArcGIS Living Atlas","slug":"arcgis-living-atlas","term_group":0,"term_taxonomy_id":36581,"taxonomy":"product","description":"","parent":0,"count":1171,"filter":"raw"},{"term_id":36561,"name":"ArcGIS Pro","slug":"arcgis-pro","term_group":0,"term_taxonomy_id":36561,"taxonomy":"product","description":"","parent":0,"count":2037,"filter":"raw"}],"primary_product_link":"https:\/\/www.esri.com\/arcgis-blog\/?s=#&products=arcgis-pro","_links":{"self":[{"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/blog\/1185252","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/blog"}],"about":[{"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/types\/blog"}],"author":[{"embeddable":true,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/users\/7461"}],"replies":[{"embeddable":true,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/comments?post=1185252"}],"version-history":[{"count":0,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/blog\/1185252\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/media?parent=1185252"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/categories?post=1185252"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/tags?post=1185252"},{"taxonomy":"industry","embeddable":true,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/industry?post=1185252"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/www.esri.com\/arcgis-blog\/wp-json\/wp\/v2\/product?post=1185252"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}