{"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/backbone-can-not-be-trained-at-once-rolling","title":"Backbone Can Not be Trained at Once: Rolling Back to Pre-trained Network for Person Re-Identification","arxiv_id":"1901.06140","date":"2019-01-18","proceeding":null,"authors":["Youngmin Ro","Jongwon Choi","Dae Ung Jo","Byeongho Heo","Jongin Lim","Jin Young Choi"],"abstract":"In person re-identification (ReID) task, because of its shortage of trainable\ndataset, it is common to utilize fine-tuning method using a classification\nnetwork pre-trained on a large dataset. However, it is relatively difficult to\nsufficiently fine-tune the low-level layers of the network due to the gradient\nvanishing problem. In this work, we propose a novel fine-tuning strategy that\nallows low-level layers to be sufficiently trained by rolling back the weights\nof high-level layers to their initial pre-trained weights. Our strategy\nalleviates the problem of gradient vanishing in low-level layers and robustly\ntrains the low-level layers to fit the ReID dataset, thereby increasing the\nperformance of ReID tasks. The improved performance of the proposed strategy is\nvalidated via several experiments. Furthermore, without any add-ons such as\npose estimation or segmentation, our strategy exhibits state-of-the-art\nperformance using only vanilla deep convolutional neural network architecture.","url_abs":"http://arxiv.org/abs/1901.06140v1","url_pdf":"http://arxiv.org/pdf/1901.06140v1.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":"backbone-can-not-be-trained-at-once-rolling","repo_url":"https://github.com/youngminPIL/rollback","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"person-re-identification","task_name":"Person Re-Identification"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"}],"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}