{"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/eanet-enhancing-alignment-for-cross-domain","title":"EANet: Enhancing Alignment for Cross-Domain Person Re-identification","arxiv_id":"1812.11369","date":"2018-12-29","proceeding":null,"authors":["Houjing Huang","Wenjie Yang","Xiaotang Chen","Xin Zhao","Kaiqi Huang","Jinbin Lin","Guan Huang","Dalong Du"],"abstract":"Person re-identification (ReID) has achieved significant improvement under\nthe single-domain setting. However, directly exploiting a model to new domains\nis always faced with huge performance drop, and adapting the model to new\ndomains without target-domain identity labels is still challenging. In this\npaper, we address cross-domain ReID and make contributions for both model\ngeneralization and adaptation. First, we propose Part Aligned Pooling (PAP)\nthat brings significant improvement for cross-domain testing. Second, we design\na Part Segmentation (PS) constraint over ReID feature to enhance alignment and\nimprove model generalization. Finally, we show that applying our PS constraint\nto unlabeled target domain images serves as effective domain adaptation. We\nconduct extensive experiments between three large datasets, Market1501, CUHK03\nand DukeMTMC-reID. Our model achieves state-of-the-art performance under both\nsource-domain and cross-domain settings. For completeness, we also demonstrate\nthe complementarity of our model to existing domain adaptation methods. The\ncode is available at https://github.com/huanghoujing/EANet.","url_abs":"http://arxiv.org/abs/1812.11369v1","url_pdf":"http://arxiv.org/pdf/1812.11369v1.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":"eanet-enhancing-alignment-for-cross-domain","repo_url":"https://github.com/huanghoujing/EANet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"eanet-enhancing-alignment-for-cross-domain","repo_url":"https://github.com/ithuanhuan/EANet-master","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"eanet-enhancing-alignment-for-cross-domain","repo_url":"https://github.com/whu-fanghan/EANet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"person-re-identification","task_name":"Person Re-Identification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1812.11369","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}