{"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/cross-view-asymmetric-metric-learning-for","title":"Cross-view Asymmetric Metric Learning for Unsupervised Person Re-identification","arxiv_id":"1708.08062","date":"2017-08-27","proceeding":"ICCV 2017 10","authors":["Hong-Xing Yu","An-Cong Wu","Wei-Shi Zheng"],"abstract":"While metric learning is important for Person re-identification (RE-ID), a\nsignificant problem in visual surveillance for cross-view pedestrian matching,\nexisting metric models for RE-ID are mostly based on supervised learning that\nrequires quantities of labeled samples in all pairs of camera views for\ntraining. However, this limits their scalabilities to realistic applications,\nin which a large amount of data over multiple disjoint camera views is\navailable but not labelled. To overcome the problem, we propose unsupervised\nasymmetric metric learning for unsupervised RE-ID. Our model aims to learn an\nasymmetric metric, i.e., specific projection for each view, based on asymmetric\nclustering on cross-view person images. Our model finds a shared space where\nview-specific bias is alleviated and thus better matching performance can be\nachieved. Extensive experiments have been conducted on a baseline and five\nlarge-scale RE-ID datasets to demonstrate the effectiveness of the proposed\nmodel. Through the comparison, we show that our model works much more suitable\nfor unsupervised RE-ID compared to classical unsupervised metric learning\nmodels. We also compare with existing unsupervised RE-ID methods, and our model\noutperforms them with notable margins. Specifically, we report the results on\nlarge-scale unlabelled RE-ID dataset, which is important but unfortunately less\nconcerned in literatures.","url_abs":"http://arxiv.org/abs/1708.08062v2","url_pdf":"http://arxiv.org/pdf/1708.08062v2.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":"cross-view-asymmetric-metric-learning-for","repo_url":"https://github.com/KovenYu/CAMEL","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"metric-learning","task_name":"Metric Learning"},{"task_slug":"person-re-identification","task_name":"Person Re-Identification"},{"task_slug":"unsupervised-person-re-identification","task_name":"Unsupervised Person Re-Identification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/person-re-identification-on-market-1501","task":"Person Re-Identification","dataset":"Market-1501","model":"CAMEL*","rank_in_archive_order":121,"of":135,"metrics":{"Rank-1":"54.5","mAP":"26.3"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.08062","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}