{"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/deep-learning-prototype-domains-for-person-re","title":"Deep Learning Prototype Domains for Person Re-Identification","arxiv_id":"1610.05047","date":"2016-10-17","proceeding":null,"authors":["Arne Schumann","Shaogang Gong","Tobias Schuchert"],"abstract":"Person re-identification (re-id) is the task of matching multiple occurrences\nof the same person from different cameras, poses, lighting conditions, and a\nmultitude of other factors which alter the visual appearance. Typically, this\nis achieved by learning either optimal features or matching metrics which are\nadapted to specific pairs of camera views dictated by the pairwise labelled\ntraining datasets. In this work, we formulate a deep learning based novel\napproach to automatic prototype-domain discovery for domain perceptive\n(adaptive) person re-id (rather than camera pair specific learning) for any\ncamera views scalable to new unseen scenes without training data. We learn a\nseparate re-id model for each of the discovered prototype-domains and during\nmodel deployment, use the person probe image to select automatically the model\nof the closest prototype domain. Our approach requires neither supervised nor\nunsupervised domain adaptation learning, i.e. no data available from the target\ndomains. We evaluate extensively our model under realistic re-id conditions\nusing automatically detected bounding boxes with low-resolution and partial\nocclusion. We show that our approach outperforms most of the state-of-the-art\nsupervised and unsupervised methods on the latest CUHK-SYSU and PRW benchmarks.","url_abs":"http://arxiv.org/abs/1610.05047v2","url_pdf":"http://arxiv.org/pdf/1610.05047v2.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":[],"tasks":[{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"person-re-identification","task_name":"Person Re-Identification"},{"task_slug":"unsupervised-domain-adaptation","task_name":"Unsupervised Domain Adaptation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/person-re-identification-on-cuhk-sysu","task":"Person Re-Identification","dataset":"CUHK-SYSU","model":"Deep *","rank_in_archive_order":3,"of":3,"metrics":{"MAP":"74.0","Rank-1":"76.7"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}