{"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/patch-mix-transformer-for-unsupervised-domain","title":"Patch-Mix Transformer for Unsupervised Domain Adaptation: A Game Perspective","arxiv_id":"2303.13434","date":"2023-03-23","proceeding":"CVPR 2023 1","authors":["Jinjing Zhu","Haotian Bai","Lin Wang"],"abstract":"Endeavors have been recently made to leverage the vision transformer (ViT) for the challenging unsupervised domain adaptation (UDA) task. They typically adopt the cross-attention in ViT for direct domain alignment. However, as the performance of cross-attention highly relies on the quality of pseudo labels for targeted samples, it becomes less effective when the domain gap becomes large. We solve this problem from a game theory's perspective with the proposed model dubbed as PMTrans, which bridges source and target domains with an intermediate domain. Specifically, we propose a novel ViT-based module called PatchMix that effectively builds up the intermediate domain, i.e., probability distribution, by learning to sample patches from both domains based on the game-theoretical models. This way, it learns to mix the patches from the source and target domains to maximize the cross entropy (CE), while exploiting two semi-supervised mixup losses in the feature and label spaces to minimize it. As such, we interpret the process of UDA as a min-max CE game with three players, including the feature extractor, classifier, and PatchMix, to find the Nash Equilibria. Moreover, we leverage attention maps from ViT to re-weight the label of each patch by its importance, making it possible to obtain more domain-discriminative feature representations. We conduct extensive experiments on four benchmark datasets, and the results show that PMTrans significantly surpasses the ViT-based and CNN-based SoTA methods by +3.6% on Office-Home, +1.4% on Office-31, and +17.7% on DomainNet, respectively.","url_abs":"https://arxiv.org/abs/2303.13434v2","url_pdf":"https://arxiv.org/pdf/2303.13434v2.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":[],"tasks":[{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"unsupervised-domain-adaptation","task_name":"Unsupervised Domain Adaptation"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"mixup","method_name":"Mixup"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"vision-transformer","method_name":"Vision Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/domain-adaptation-on-domainnet-1","task":"Domain Adaptation","dataset":"DomainNet","model":"PMTrans","rank_in_archive_order":4,"of":4,"metrics":{"Accuracy":"52.4"},"uses_additional_data":false},{"leaderboard":"/sota/domain-adaptation-on-office-31","task":"Domain Adaptation","dataset":"Office-31","model":"PMTrans","rank_in_archive_order":2,"of":40,"metrics":{"Average Accuracy":"95.3"},"uses_additional_data":false},{"leaderboard":"/sota/domain-adaptation-on-office-home","task":"Domain Adaptation","dataset":"Office-Home","model":"PMTrans","rank_in_archive_order":4,"of":29,"metrics":{"Accuracy":"89.0"},"uses_additional_data":false},{"leaderboard":"/sota/domain-adaptation-on-visda2017","task":"Domain Adaptation","dataset":"VisDA2017","model":"PMtrans","rank_in_archive_order":9,"of":28,"metrics":{"Accuracy":"88.8"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-domain-adaptation-on-domainnet-1","task":"Unsupervised Domain Adaptation","dataset":"DomainNet","model":"PMTrans","rank_in_archive_order":2,"of":4,"metrics":{"Accuracy":"52.4"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-domain-adaptation-on-office-31","task":"Unsupervised Domain Adaptation","dataset":"Office-31","model":"PMTrans","rank_in_archive_order":1,"of":5,"metrics":{"Accuracy":"95.3"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-domain-adaptation-on-office-home","task":"Unsupervised Domain Adaptation","dataset":"Office-Home","model":"PMTrans","rank_in_archive_order":4,"of":20,"metrics":{"Accuracy":"89.0"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-domain-adaptation-on-visda2017","task":"Unsupervised Domain Adaptation","dataset":"VisDA2017","model":"PMTrans","rank_in_archive_order":7,"of":13,"metrics":{"Accuracy":"88.8"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2303.13434","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}