{"id":1211,"date":"2026-01-25T19:31:04","date_gmt":"2026-01-25T19:31:04","guid":{"rendered":"https:\/\/www.cs.ubbcluj.ro\/~meco\/railway-switch-classification-using-deep-neural-networks-2023\/"},"modified":"2026-02-01T12:08:02","modified_gmt":"2026-02-01T12:08:02","slug":"railway-switch-classification-using-deep-neural-networks-2023","status":"publish","type":"post","link":"https:\/\/www.cs.ubbcluj.ro\/~meco\/railway-switch-classification-using-deep-neural-networks-2023\/","title":{"rendered":"Railway Switch Classification Using Deep Neural Networks (2023)"},"content":{"rendered":"<div class=\"entry-content\">\n<p>VISIGRAPP<\/p>\n<h2>Authors<\/h2>\n<p>Andrei-Robert Alexandrescu, Alexandru Manole, L. Dio\u015fan<\/p>\n<h2>Abstract<\/h2>\n<p>Railway switches represent the mechanism which slightly adjusts the rail blades at the intersection of two rail<br \/>\n tracks in order to allow trains to exchange their routes. Ensuring that the switches are correctly set represents<br \/>\n a critical task. If switches are not correctly set, they may cause delays in train schedules or even loss of lives.<br \/>\n In this paper we propose an approach for classifying switches using various deep learning architectures with<br \/>\n a small number of parameters. We exploit various input modalities including: grayscale images, black and<br \/>\n white binary masks and a concatenated representation consisting of both. The experiments are conducted on<br \/>\n RailSem19, the most comprehensive dataset for the task of switch classification, using both fine-tuned models<br \/>\n and modelstrained from scratch. The switch bounding boxes from the dataset are pre-processed by introducing<br \/>\n three hyper-parameters over the boxes, improving the models performance. We manage to achieve an overall<br \/>\n accuracy of up to 96% in a ternary multi-class classification setting where our model is able to distinguish<br \/>\n between images containing left, right or no switches at all. The results for the left and right switch classes are<br \/>\n compared with two other existing approaches from the literature. We obtain competitive results using deep<br \/>\n neural networks with considerably fewer learnable parameters than the ones from the literature.<\/p>\n<h2>Citation<\/h2>\n<pre class=\"wp-block-preformatted\">@Inproceedings{Alexandrescu2023RailwaySC,\n author = {Andrei-Robert Alexandrescu and Alexandru Manole and L. Dio\u015fan},\n booktitle = {VISIGRAPP},\n title = {Railway Switch Classification Using Deep Neural Networks},\n year = {2023}\n}<\/pre>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Railway switches represent the mechanism which slightly adjusts the rail blades at the intersection of two rail tracks in order to allow trains to exchange their routes. Ensuring that the switches are correctly set represents a critical task. If switches are not correctly set, they may cause delays in train schedules or even loss of lives. In this paper we propose an approach for classifying switches using various deep learning architectures with a small number of parameters. We exploit various input modalities including: grayscale images, black and white binary masks and a concatenated representation consisting of both. The experiments are conducted on RailSem19, the most comprehensive dataset for the task of switch classification, using both fine-tuned models and modelstrained from scratch. The switch bounding boxes from the dataset are pre-processed by introducing three hyper-parameters over the boxes, improving the models performance. We manage to achieve an overall accuracy of up to 96% in a ternary multi-class classification setting where our model is able to distinguish between images containing left, right or no switches at all. The results for the left and right switch classes are compared with two other existing approaches from the literature. We obtain competitive results using deep neural networks with considerably fewer learnable parameters than the ones from the literature.<\/p>\n","protected":false},"author":6,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":[],"categories":[4],"tags":[11,66],"_links":{"self":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts\/1211"}],"collection":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/comments?post=1211"}],"version-history":[{"count":1,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts\/1211\/revisions"}],"predecessor-version":[{"id":1468,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts\/1211\/revisions\/1468"}],"wp:attachment":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/media?parent=1211"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/categories?post=1211"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/tags?post=1211"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}