{"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/correlation-alignment-for-unsupervised-domain","title":"Correlation Alignment for Unsupervised Domain Adaptation","arxiv_id":"1612.01939","date":"2016-12-06","proceeding":null,"authors":["Baochen Sun","Jiashi Feng","Kate Saenko"],"abstract":"In this chapter, we present CORrelation ALignment (CORAL), a simple yet\neffective method for unsupervised domain adaptation. CORAL minimizes domain\nshift by aligning the second-order statistics of source and target\ndistributions, without requiring any target labels. In contrast to subspace\nmanifold methods, it aligns the original feature distributions of the source\nand target domains, rather than the bases of lower-dimensional subspaces. It is\nalso much simpler than other distribution matching methods. CORAL performs\nremarkably well in extensive evaluations on standard benchmark datasets. We\nfirst describe a solution that applies a linear transformation to source\nfeatures to align them with target features before classifier training. For\nlinear classifiers, we propose to equivalently apply CORAL to the classifier\nweights, leading to added efficiency when the number of classifiers is small\nbut the number and dimensionality of target examples are very high. The\nresulting CORAL Linear Discriminant Analysis (CORAL-LDA) outperforms LDA by a\nlarge margin on standard domain adaptation benchmarks. Finally, we extend CORAL\nto learn a nonlinear transformation that aligns correlations of layer\nactivations in deep neural networks (DNNs). The resulting Deep CORAL approach\nworks seamlessly with DNNs and achieves state-of-the-art performance on\nstandard benchmark datasets. Our code is available\nat:~\\url{https://github.com/VisionLearningGroup/CORAL}","url_abs":"http://arxiv.org/abs/1612.01939v1","url_pdf":"http://arxiv.org/pdf/1612.01939v1.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":"correlation-alignment-for-unsupervised-domain","repo_url":"https://github.com/VisionLearningGroup/CORAL","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"correlation-alignment-for-unsupervised-domain","repo_url":"https://github.com/eridgd/WCT-TF","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"correlation-alignment-for-unsupervised-domain","repo_url":"https://github.com/adapt-python/adapt","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"correlation-alignment-for-unsupervised-domain","repo_url":"https://github.com/domainadaptation/salad","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MPL-2.0"}}],"tasks":[{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"unsupervised-domain-adaptation","task_name":"Unsupervised Domain Adaptation"}],"methods":[{"method_slug":"coral","method_name":"CORAL"},{"method_slug":"lda","method_name":"LDA"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/domain-adaptation-on-office-caltech","task":"Domain Adaptation","dataset":"Office-Caltech","model":"CORAL[[Sun, Feng, and Saenko2017]]","rank_in_archive_order":8,"of":8,"metrics":{"Average Accuracy":"84.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1612.01939","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}