{"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/learning-more-universal-representations-for","title":"Learning More Universal Representations for Transfer-Learning","arxiv_id":"1712.09708","date":"2017-12-27","proceeding":null,"authors":["Youssef Tamaazousti","Hervé Le Borgne","Céline Hudelot","Mohamed El Amine Seddik","Mohamed Tamaazousti"],"abstract":"A representation is supposed universal if it encodes any element of the\nvisual world (e.g., objects, scenes) in any configuration (e.g., scale,\ncontext). While not expecting pure universal representations, the goal in the\nliterature is to improve the universality level, starting from a representation\nwith a certain level. To do so, the state-of-the-art consists in learning\nCNN-based representations on a diversified training problem (e.g., ImageNet\nmodified by adding annotated data). While it effectively increases\nuniversality, such approach still requires a large amount of efforts to satisfy\nthe needs in annotated data. In this work, we propose two methods to improve\nuniversality, but pay special attention to limit the need of annotated data. We\nalso propose a unified framework of the methods based on the diversifying of\nthe training problem. Finally, to better match Atkinson's cognitive study about\nuniversal human representations, we proposed to rely on the transfer-learning\nscheme as well as a new metric to evaluate universality. This latter, aims us\nto demonstrates the interest of our methods on 10 target-problems, relating to\nthe classification task and a variety of visual domains.","url_abs":"http://arxiv.org/abs/1712.09708v5","url_pdf":"http://arxiv.org/pdf/1712.09708v5.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":"learning-more-universal-representations-for","repo_url":"https://github.com/youssefTamaazousti/MulDiPNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1712.09708","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}