{"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/horizontal-pyramid-matching-for-person-re","title":"Horizontal Pyramid Matching for Person Re-identification","arxiv_id":"1804.05275","date":"2018-04-14","proceeding":null,"authors":["Yang Fu","Yunchao Wei","Yuqian Zhou","Honghui Shi","Gao Huang","Xinchao Wang","Zhiqiang Yao","Thomas Huang"],"abstract":"Despite the remarkable recent progress, person re-identification (Re-ID)\napproaches are still suffering from the failure cases where the discriminative\nbody parts are missing. To mitigate such cases, we propose a simple yet\neffective Horizontal Pyramid Matching (HPM) approach to fully exploit various\npartial information of a given person, so that correct person candidates can be\nstill identified even even some key parts are missing. Within the HPM, we make\nthe following contributions to produce a more robust feature representation for\nthe Re-ID task: 1) we learn to classify using partial feature representations\nat different horizontal pyramid scales, which successfully enhance the\ndiscriminative capabilities of various person parts; 2) we exploit average and\nmax pooling strategies to account for person-specific discriminative\ninformation in a global-local manner. To validate the effectiveness of the\nproposed HPM, extensive experiments are conducted on three popular benchmarks,\nincluding Market-1501, DukeMTMC-ReID and CUHK03. In particular, we achieve mAP\nscores of 83.1%, 74.5% and 59.7% on these benchmarks, which are the new\nstate-of-the-arts. Our code is available on Github","url_abs":"http://arxiv.org/abs/1804.05275v4","url_pdf":"http://arxiv.org/pdf/1804.05275v4.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":"horizontal-pyramid-matching-for-person-re","repo_url":"https://github.com/OasisYang/HPM","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"person-re-identification","task_name":"Person Re-Identification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/person-re-identification-on-dukemtmc-reid","task":"Person Re-Identification","dataset":"DukeMTMC-reID","model":"HPM","rank_in_archive_order":59,"of":94,"metrics":{"Rank-1":"86.6","mAP":"74.3"},"uses_additional_data":false},{"leaderboard":"/sota/person-re-identification-on-market-1501","task":"Person Re-Identification","dataset":"Market-1501","model":"HPM","rank_in_archive_order":81,"of":135,"metrics":{"Rank-1":"94.2","mAP":"82.7"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1804.05275","atlas_url":"https://app.syntology.ai/?focus=1804.05275","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}