{"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/contextdesc-local-descriptor-augmentation","title":"ContextDesc: Local Descriptor Augmentation with Cross-Modality Context","arxiv_id":"1904.04084","date":"2019-04-08","proceeding":"CVPR 2019 6","authors":["Zixin Luo","Tianwei Shen","Lei Zhou","Jiahui Zhang","Yao Yao","Shiwei Li","Tian Fang","Long Quan"],"abstract":"Most existing studies on learning local features focus on the patch-based\ndescriptions of individual keypoints, whereas neglecting the spatial relations\nestablished from their keypoint locations. In this paper, we go beyond the\nlocal detail representation by introducing context awareness to augment\noff-the-shelf local feature descriptors. Specifically, we propose a unified\nlearning framework that leverages and aggregates the cross-modality contextual\ninformation, including (i) visual context from high-level image representation,\nand (ii) geometric context from 2D keypoint distribution. Moreover, we propose\nan effective N-pair loss that eschews the empirical hyper-parameter search and\nimproves the convergence. The proposed augmentation scheme is lightweight\ncompared with the raw local feature description, meanwhile improves remarkably\non several large-scale benchmarks with diversified scenes, which demonstrates\nboth strong practicality and generalization ability in geometric matching\napplications.","url_abs":"http://arxiv.org/abs/1904.04084v1","url_pdf":"http://arxiv.org/pdf/1904.04084v1.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":"contextdesc-local-descriptor-augmentation","repo_url":"https://github.com/lzx551402/contextdesc","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"geometric-matching","task_name":"Geometric Matching"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1904.04084","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}