{"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/learning-to-predict-crisp-boundaries","title":"Learning to predict crisp boundaries","arxiv_id":"1807.10097","date":"2018-07-26","proceeding":"ECCV 2018 9","authors":["Ruoxi Deng","Chunhua Shen","Shengjun Liu","Huibing Wang","Xinru Liu"],"abstract":"Recent methods for boundary or edge detection built on Deep Convolutional\nNeural Networks (CNNs) typically suffer from the issue of predicted edges being\nthick and need post-processing to obtain crisp boundaries. Highly imbalanced\ncategories of boundary versus background in training data is one of main\nreasons for the above problem. In this work, the aim is to make CNNs produce\nsharp boundaries without post-processing. We introduce a novel loss for\nboundary detection, which is very effective for classifying imbalanced data and\nallows CNNs to produce crisp boundaries. Moreover, we propose an end-to-end\nnetwork which adopts the bottom-up/top-down architecture to tackle the task.\nThe proposed network effectively leverages hierarchical features and produces\npixel-accurate boundary mask, which is critical to reconstruct the edge map.\nOur experiments illustrate that directly making crisp prediction not only\npromotes the visual results of CNNs, but also achieves better results against\nthe state-of-the-art on the BSDS500 dataset (ODS F-score of .815) and the NYU\nDepth dataset (ODS F-score of .762).","url_abs":"http://arxiv.org/abs/1807.10097v1","url_pdf":"http://arxiv.org/pdf/1807.10097v1.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":"learning-to-predict-crisp-boundaries","repo_url":"https://github.com/yunfan1202/Delving-into-Crispness","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"boundary-detection","task_name":"Boundary Detection"},{"task_slug":"edge-detection","task_name":"Edge Detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1807.10097","atlas_url":"https://app.syntology.ai/?focus=1807.10097","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}