{"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/3d-segmentation-with-exponential-logarithmic","title":"3D Segmentation with Exponential Logarithmic Loss for Highly Unbalanced Object Sizes","arxiv_id":"1809.00076","date":"2018-08-31","proceeding":null,"authors":["Ken C. L. Wong","Mehdi Moradi","Hui Tang","Tanveer Syeda-Mahmood"],"abstract":"With the introduction of fully convolutional neural networks, deep learning\nhas raised the benchmark for medical image segmentation on both speed and\naccuracy, and different networks have been proposed for 2D and 3D segmentation\nwith promising results. Nevertheless, most networks only handle relatively\nsmall numbers of labels (<10), and there are very limited works on handling\nhighly unbalanced object sizes especially in 3D segmentation. In this paper, we\npropose a network architecture and the corresponding loss function which\nimprove segmentation of very small structures. By combining skip connections\nand deep supervision with respect to the computational feasibility of 3D\nsegmentation, we propose a fast converging and computationally efficient\nnetwork architecture for accurate segmentation. Furthermore, inspired by the\nconcept of focal loss, we propose an exponential logarithmic loss which\nbalances the labels not only by their relative sizes but also by their\nsegmentation difficulties. We achieve an average Dice coefficient of 82% on\nbrain segmentation with 20 labels, with the ratio of the smallest to largest\nobject sizes as 0.14%. Less than 100 epochs are required to reach such\naccuracy, and segmenting a 128x128x128 volume only takes around 0.4 s.","url_abs":"http://arxiv.org/abs/1809.00076v2","url_pdf":"http://arxiv.org/pdf/1809.00076v2.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":"3d-segmentation-with-exponential-logarithmic","repo_url":"https://github.com/ibm/multimodal-3d-image-segmentation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"brain-segmentation","task_name":"Brain Segmentation"},{"task_slug":"image-segmentation","task_name":"Image Segmentation"},{"task_slug":"medical-image-segmentation","task_name":"Medical Image Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1809.00076","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}