{"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/orientation-driven-bag-of-appearances-for","title":"Orientation Driven Bag of Appearances for Person Re-identification","arxiv_id":"1605.02464","date":"2016-05-09","proceeding":null,"authors":["Liqian Ma","Hong Liu","Liang Hu","Can Wang","Qianru Sun"],"abstract":"Person re-identification (re-id) consists of associating individual across\ncamera network, which is valuable for intelligent video surveillance and has\ndrawn wide attention. Although person re-identification research is making\nprogress, it still faces some challenges such as varying poses, illumination\nand viewpoints. For feature representation in re-identification, existing works\nusually use low-level descriptors which do not take full advantage of body\nstructure information, resulting in low representation ability.\n%discrimination. To solve this problem, this paper proposes the mid-level\nbody-structure based feature representation (BSFR) which introduces body\nstructure pyramid for codebook learning and feature pooling in the vertical\ndirection of human body. Besides, varying viewpoints in the horizontal\ndirection of human body usually causes the data missing problem, $i.e.$, the\nappearances obtained in different orientations of the identical person could\nvary significantly. To address this problem, the orientation driven bag of\nappearances (ODBoA) is proposed to utilize person orientation information\nextracted by orientation estimation technic. To properly evaluate the proposed\napproach, we introduce a new re-identification dataset (Market-1203) based on\nthe Market-1501 dataset and propose a new re-identification dataset (PKU-Reid).\nBoth datasets contain multiple images captured in different body orientations\nfor each person. Experimental results on three public datasets and two proposed\ndatasets demonstrate the superiority of the proposed approach, indicating the\neffectiveness of body structure and orientation information for improving\nre-identification performance.","url_abs":"http://arxiv.org/abs/1605.02464v1","url_pdf":"http://arxiv.org/pdf/1605.02464v1.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":"orientation-driven-bag-of-appearances-for","repo_url":"https://github.com/charliememory/Market1203-Reid-Dataset","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null},{"paper_slug":"orientation-driven-bag-of-appearances-for","repo_url":"https://github.com/charliememory/PKU-Reid-Dataset","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"person-re-identification","task_name":"Person Re-Identification"}],"methods":[],"datasets_introduced":[{"slug":"market1203-reid-dataset","name":"Market1203-Reid-Dataset","full_name":null},{"slug":"pku-reid","name":"PKU-Reid","full_name":null}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1605.02464","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}