{"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-spread-out-local-feature-descriptors","title":"Learning Spread-out Local Feature Descriptors","arxiv_id":"1708.06320","date":"2017-08-21","proceeding":"ICCV 2017 10","authors":["Xu Zhang","Felix X. Yu","Sanjiv Kumar","Shih-Fu Chang"],"abstract":"We propose a simple, yet powerful regularization technique that can be used\nto significantly improve both the pairwise and triplet losses in learning local\nfeature descriptors. The idea is that in order to fully utilize the expressive\npower of the descriptor space, good local feature descriptors should be\nsufficiently \"spread-out\" over the space. In this work, we propose a\nregularization term to maximize the spread in feature descriptor inspired by\nthe property of uniform distribution. We show that the proposed regularization\nwith triplet loss outperforms existing Euclidean distance based descriptor\nlearning techniques by a large margin. As an extension, the proposed\nregularization technique can also be used to improve image-level deep feature\nembedding.","url_abs":"http://arxiv.org/abs/1708.06320v1","url_pdf":"http://arxiv.org/pdf/1708.06320v1.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-spread-out-local-feature-descriptors","repo_url":"https://github.com/ColumbiaDVMM/Spread-out_Local_Feature_Descriptor","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"learning-spread-out-local-feature-descriptors","repo_url":"https://github.com/ColumbiaDVMM/AutoGAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":null,"task_name":"Triplet"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.06320","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}