{"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/deep-imbalanced-learning-for-face-recognition","title":"Deep Imbalanced Learning for Face Recognition and Attribute Prediction","arxiv_id":"1806.00194","date":"2018-06-01","proceeding":null,"authors":["Chen Huang","Yining Li","Chen Change Loy","Xiaoou Tang"],"abstract":"Data for face analysis often exhibit highly-skewed class distribution, i.e.,\nmost data belong to a few majority classes, while the minority classes only\ncontain a scarce amount of instances. To mitigate this issue, contemporary deep\nlearning methods typically follow classic strategies such as class re-sampling\nor cost-sensitive training. In this paper, we conduct extensive and systematic\nexperiments to validate the effectiveness of these classic schemes for\nrepresentation learning on class-imbalanced data. We further demonstrate that\nmore discriminative deep representation can be learned by enforcing a deep\nnetwork to maintain inter-cluster margins both within and between classes. This\ntight constraint effectively reduces the class imbalance inherent in the local\ndata neighborhood, thus carving much more balanced class boundaries locally. We\nshow that it is easy to deploy angular margins between the cluster\ndistributions on a hypersphere manifold. Such learned Cluster-based Large\nMargin Local Embedding (CLMLE), when combined with a simple k-nearest cluster\nalgorithm, shows significant improvements in accuracy over existing methods on\nboth face recognition and face attribute prediction tasks that exhibit\nimbalanced class distribution.","url_abs":"http://arxiv.org/abs/1806.00194v2","url_pdf":"http://arxiv.org/pdf/1806.00194v2.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":"deep-imbalanced-learning-for-face-recognition","repo_url":"https://github.com/JoyLuo/face-attribute-recognition-paper-list","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"attribute","task_name":"Attribute"},{"task_slug":"face-recognition","task_name":"Face Recognition"},{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.00194","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}