{"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/efficient-smoothing-of-dilated-convolutions","title":"Efficient Smoothing of Dilated Convolutions for Image Segmentation","arxiv_id":"1903.07992","date":"2019-03-19","proceeding":null,"authors":["Thomas Ziegler","Manuel Fritsche","Lorenz Kuhn","Konstantin Donhauser"],"abstract":"Dilated Convolutions have been shown to be highly useful for the task of\nimage segmentation. By introducing gaps into convolutional filters, they enable\nthe use of larger receptive fields without increasing the original kernel size.\nEven though this allows for the inexpensive capturing of features at different\nscales, the structure of the dilated convolutional filter leads to a loss of\ninformation. We hypothesise that inexpensive modifications to Dilated\nConvolutional Neural Networks, such as additional averaging layers, could\novercome this limitation. In this project we test this hypothesis by evaluating\nthe effect of these modifications for a state-of-the art image segmentation\nsystem and compare them to existing approaches with the same objective. Our\nexperiments show that our proposed methods improve the performance of dilated\nconvolutions for image segmentation. Crucially, our modifications achieve these\nresults at a much lower computational cost than previous smoothing approaches.","url_abs":"http://arxiv.org/abs/1903.07992v1","url_pdf":"http://arxiv.org/pdf/1903.07992v1.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":"efficient-smoothing-of-dilated-convolutions","repo_url":"https://github.com/ThomasZiegler/Efficient-Smoothing-of-Dilated-Convolutions","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"image-segmentation","task_name":"Image Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}