{"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/deeply-learned-part-aligned-representations","title":"Deeply-Learned Part-Aligned Representations for Person Re-Identification","arxiv_id":"1707.07256","date":"2017-07-23","proceeding":"ICCV 2017 10","authors":["Liming Zhao","Xi Li","Jingdong Wang","Yueting Zhuang"],"abstract":"In this paper, we address the problem of person re-identification, which\nrefers to associating the persons captured from different cameras. We propose a\nsimple yet effective human part-aligned representation for handling the body\npart misalignment problem. Our approach decomposes the human body into regions\n(parts) which are discriminative for person matching, accordingly computes the\nrepresentations over the regions, and aggregates the similarities computed\nbetween the corresponding regions of a pair of probe and gallery images as the\noverall matching score. Our formulation, inspired by attention models, is a\ndeep neural network modeling the three steps together, which is learnt through\nminimizing the triplet loss function without requiring body part labeling\ninformation. Unlike most existing deep learning algorithms that learn a global\nor spatial partition-based local representation, our approach performs human\nbody partition, and thus is more robust to pose changes and various human\nspatial distributions in the person bounding box. Our approach shows\nstate-of-the-art results over standard datasets, Market-$1501$, CUHK$03$,\nCUHK$01$ and VIPeR.","url_abs":"http://arxiv.org/abs/1707.07256v1","url_pdf":"http://arxiv.org/pdf/1707.07256v1.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":"deeply-learned-part-aligned-representations","repo_url":"https://github.com/Phoebe-star/part_aligned","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"person-re-identification","task_name":"Person Re-Identification"},{"task_slug":null,"task_name":"Triplet"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/person-re-identification-on-market-1501","task":"Person Re-Identification","dataset":"Market-1501","model":"DF","rank_in_archive_order":112,"of":135,"metrics":{"Rank-1":"81.0","mAP":"63.4"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1707.07256","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}