{"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/pose-driven-deep-models-for-person-re","title":"Pose-Driven Deep Models for Person Re-Identification","arxiv_id":"1803.08709","date":"2018-03-23","proceeding":null,"authors":["Andreas Eberle"],"abstract":"Person re-identification (re-id) is the task of recognizing and matching\npersons at different locations recorded by cameras with non-overlapping views.\nOne of the main challenges of re-id is the large variance in person poses and\ncamera angles since neither of them can be influenced by the re-id system. In\nthis work, an effective approach to integrate coarse camera view information as\nwell as fine-grained pose information into a convolutional neural network (CNN)\nmodel for learning discriminative re-id embeddings is introduced. In most\nrecent work pose information is either explicitly modeled within the re-id\nsystem or explicitly used for pre-processing, for example by pose-normalizing\nperson images. In contrast, the proposed approach shows that a direct use of\ncamera view as well as the detected body joint locations into a standard CNN\ncan be used to significantly improve the robustness of learned re-id\nembeddings. On four challenging surveillance and video re-id datasets\nsignificant improvements over the current state of the art have been achieved.\nFurthermore, a novel reordering of the MARS dataset, called X-MARS is\nintroduced to allow cross-validation of models trained for single-image re-id\non tracklet data.","url_abs":"http://arxiv.org/abs/1803.08709v1","url_pdf":"http://arxiv.org/pdf/1803.08709v1.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":"pose-driven-deep-models-for-person-re","repo_url":"https://github.com/andreas-eberle/x-mars","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"person-re-identification","task_name":"Person Re-Identification"}],"methods":[],"datasets_introduced":[{"slug":"x-mars","name":"X-MARS","full_name":null}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}