{"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/instance-level-meta-normalization","title":"Instance-Level Meta Normalization","arxiv_id":"1904.03516","date":"2019-04-06","proceeding":"CVPR 2019 6","authors":["Songhao Jia","Ding-Jie Chen","Hwann-Tzong Chen"],"abstract":"This paper presents a normalization mechanism called Instance-Level Meta\nNormalization (ILM~Norm) to address a learning-to-normalize problem. ILM~Norm\nlearns to predict the normalization parameters via both the feature\nfeed-forward and the gradient back-propagation paths. ILM~Norm provides a meta\nnormalization mechanism and has several good properties. It can be easily\nplugged into existing instance-level normalization schemes such as Instance\nNormalization, Layer Normalization, or Group Normalization. ILM~Norm normalizes\neach instance individually and therefore maintains high performance even when\nsmall mini-batch is used. The experimental results show that ILM~Norm well\nadapts to different network architectures and tasks, and it consistently\nimproves the performance of the original models. The code is available at\nurl{https://github.com/Gasoonjia/ILM-Norm.","url_abs":"http://arxiv.org/abs/1904.03516v1","url_pdf":"http://arxiv.org/pdf/1904.03516v1.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":"instance-level-meta-normalization","repo_url":"https://github.com/Gasoonjia/ILM-Norm","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[{"method_slug":"group-normalization","method_name":"Group Normalization"},{"method_slug":"instance-normalization","method_name":"Instance Normalization"},{"method_slug":"instance-level-meta-normalization","method_name":"Instance-Level Meta Normalization"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"}],"datasets_introduced":[],"methods_introduced":[{"slug":"instance-level-meta-normalization","name":"Instance-Level Meta Normalization","full_name":"Instance-Level Meta Normalization"}],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.03516","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}