Papers › Patch-Mix Transformer for Unsupervised Domain Adaptation: A Game Perspective

Patch-Mix Transformer for Unsupervised Domain Adaptation: A Game Perspective

23 Mar 2023CVPR 2023 1arXiv:2303.13434archive 2025-07-28

Jinjing Zhu, Haotian Bai, Lin Wang

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.

PaperPDFConference PDF

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Domain AdaptationUnsupervised Domain Adaptation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Domain Adaptation DomainNet PMTrans Accuracy 52.4 #4 of 4 Archive leaderboard report
Domain Adaptation Office-31 PMTrans Average Accuracy 95.3 #2 of 40 Archive leaderboard report
Domain Adaptation Office-Home PMTrans Accuracy 89.0 #4 of 29 Archive leaderboard report
Domain Adaptation VisDA2017 PMtrans Accuracy 88.8 #9 of 28 Archive leaderboard report
Unsupervised Domain Adaptation DomainNet PMTrans Accuracy 52.4 #2 of 4 Archive leaderboard report
Unsupervised Domain Adaptation Office-31 PMTrans Accuracy 95.3 #1 of 5 Archive leaderboard report
Unsupervised Domain Adaptation Office-Home PMTrans Accuracy 89.0 #4 of 20 Archive leaderboard report
Unsupervised Domain Adaptation VisDA2017 PMTrans Accuracy 88.8 #7 of 13 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

AttentionDense ConnectionsLayer NormalizationLinear LayerMixupMulti-Head AttentionResidual ConnectionSoftmaxVision Transformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections