{"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-multiple-visual-domains-with","title":"Learning multiple visual domains with residual adapters","arxiv_id":"1705.08045","date":"2017-05-22","proceeding":"NeurIPS 2017 12","authors":["Sylvestre-Alvise Rebuffi","Hakan Bilen","Andrea Vedaldi"],"abstract":"There is a growing interest in learning data representations that work well\nfor many different types of problems and data. In this paper, we look in\nparticular at the task of learning a single visual representation that can be\nsuccessfully utilized in the analysis of very different types of images, from\ndog breeds to stop signs and digits. Inspired by recent work on learning\nnetworks that predict the parameters of another, we develop a tunable deep\nnetwork architecture that, by means of adapter residual modules, can be steered\non the fly to diverse visual domains. Our method achieves a high degree of\nparameter sharing while maintaining or even improving the accuracy of\ndomain-specific representations. We also introduce the Visual Decathlon\nChallenge, a benchmark that evaluates the ability of representations to capture\nsimultaneously ten very different visual domains and measures their ability to\nrecognize well uniformly.","url_abs":"http://arxiv.org/abs/1705.08045v5","url_pdf":"http://arxiv.org/pdf/1705.08045v5.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-multiple-visual-domains-with","repo_url":"https://github.com/srebuffi/residual_adapters","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"learning-multiple-visual-domains-with","repo_url":"https://github.com/YuWang24/MultiTune","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"continual-learning","task_name":"Continual Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/continual-learning-on-visual-domain-decathlon","task":"Continual Learning","dataset":"visual domain decathlon (10 tasks)","model":"Res. adapt. (large)","rank_in_archive_order":6,"of":14,"metrics":{"decathlon discipline (Score)":"3131"},"uses_additional_data":false},{"leaderboard":"/sota/continual-learning-on-visual-domain-decathlon","task":"Continual Learning","dataset":"visual domain decathlon (10 tasks)","model":"Res. adapt. finetune all","rank_in_archive_order":9,"of":14,"metrics":{"decathlon discipline (Score)":"2643"},"uses_additional_data":false},{"leaderboard":"/sota/continual-learning-on-visual-domain-decathlon","task":"Continual Learning","dataset":"visual domain decathlon (10 tasks)","model":"Res. adapt. decay","rank_in_archive_order":10,"of":14,"metrics":{"decathlon discipline (Score)":"2621"},"uses_additional_data":false},{"leaderboard":"/sota/continual-learning-on-visual-domain-decathlon","task":"Continual Learning","dataset":"visual domain decathlon (10 tasks)","model":"Res. adapt. dom-pred","rank_in_archive_order":12,"of":14,"metrics":{"decathlon discipline (Score)":"2503"},"uses_additional_data":false},{"leaderboard":"/sota/continual-learning-on-visual-domain-decathlon","task":"Continual Learning","dataset":"visual domain decathlon (10 tasks)","model":"Res. adapt.","rank_in_archive_order":13,"of":14,"metrics":{"decathlon discipline (Score)":"2118"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1705.08045","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}