{"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/multiregion-bilinear-convolutional-neural","title":"Multiregion Bilinear Convolutional Neural Networks for Person Re-Identification","arxiv_id":"1512.05300","date":"2015-12-16","proceeding":null,"authors":["Evgeniya Ustinova","Yaroslav Ganin","Victor Lempitsky"],"abstract":"In this work we propose a new architecture for person re-identification. As\nthe task of re-identification is inherently associated with embedding learning\nand non-rigid appearance description, our architecture is based on the deep\nbilinear convolutional network (Bilinear-CNN) that has been proposed recently\nfor fine-grained classification of highly non-rigid objects. While the last\nstages of the original Bilinear-CNN architecture completely removes the\ngeometric information from consideration by performing orderless pooling, we\nobserve that a better embedding can be learned by performing bilinear pooling\nin a more local way, where each pooling is confined to a predefined region. Our\narchitecture thus represents a compromise between traditional convolutional\nnetworks and bilinear CNNs and strikes a balance between rigid matching and\ncompletely ignoring spatial information.\n  We perform the experimental validation of the new architecture on the three\npopular benchmark datasets (Market-1501, CUHK01, CUHK03), comparing it to\nbaselines that include Bilinear-CNN as well as prior art. The new architecture\noutperforms the baseline on all three datasets, while performing better than\nstate-of-the-art on two out of three. The code and the pretrained models of the\napproach can be found at https://github.com/madkn/MultiregionBilinearCNN-ReId.","url_abs":"http://arxiv.org/abs/1512.05300v5","url_pdf":"http://arxiv.org/pdf/1512.05300v5.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":"multiregion-bilinear-convolutional-neural","repo_url":"https://github.com/madkn/MultiregionBilinearCNN-ReId","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":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}