Unsupervised Domain Adaptation With Hierarchical Gradient Synchronization

Hu, Lanqing; Kan, Meina; Shan, Shiguang; Chen, Xilin

doi:10.1109/cvpr42600.2020.00410

Cited by 82 publications

(24 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(k) GVB-GD [9], which proposes a gradually vanishing bridge mechanism for adversarial-based domain adaptation. (l) GSDA [16], which aims to learn domain invariant representations by hierarchical domain alignment. (m) STAR [25], which tries to employ more classifiers by sampling from Gaussian distribution without more parameters.…”

Section: Datasets and Compared Approachesmentioning

confidence: 99%

Contrastive Adaptation Network for Unsupervised Domain Adaptation

Kang¹,

Jiang

Yang³

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

781

457

View full text Add to dashboard Cite

Unsupervised Domain Adaptation (UDA) makes predictions for the target domain data while manual annotations are only available in the source domain. Previous methods minimize the domain discrepancy neglecting the class information, which may lead to misalignment and poor generalization performance. To address this issue, this paper proposes Contrastive Adaptation Network (CAN) optimizing a new metric which explicitly models the intra-class domain discrepancy and the inter-class domain discrepancy. We design an alternating update strategy for training CAN in an end-to-end manner. Experiments on two real-world benchmarks Office-31 and VisDA-2017 demonstrate that CAN performs favorably against the state-of-the-art methods and produces more discriminative features.

show abstract

Section: Datasets and Compared Approachesmentioning

confidence: 99%

Contrastive Adaptation Network for Unsupervised Domain Adaptation

Kang¹,

Jiang

Yang³

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

781

457

View full text Add to dashboard Cite

show abstract

“…The Structurally Regularized Deep Clustering (SRDC) [49] defines the target discrimination using the clustering of intermediate network features [33], [34]. Although the two methods proposed in [50], [48] have achieved much better performance, they ignore the fact that the exploration of discriminative features cannot guarantee the alignment of the corresponding categorical features of two domains; accordingly, there is a need to adapt the condition distribution.…”

Section: A Mainstream Approaches In Udamentioning

confidence: 99%

“…MADA [54] trains multiple class-wise domain discriminators to capture multi-mode structures, thereby enabling the fine-grained alignment of different data distributions. Similarly, GSDA [50] implements hierarchical domain alignments including class-wise, groupwise and global alignment. Although this type of method significantly boosts performance by exploring the class-wise adaptation, there is also a drawback that many class-level alignment methods incorporate target pseudo labels, most of which depend only on the source classifier.…”

Section: A Mainstream Approaches In Udamentioning

confidence: 99%

Joint Clustering and Discriminative Feature Alignment for Unsupervised Domain Adaptation

Deng

Liao

Zhao

et al. 2021

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Unsupervised Domain Adaptation (UDA) aims to learn a classifier for the unlabeled target domain by leveraging knowledge from a labeled source domain with a different but related distribution. Many existing approaches typically learn a domain-invariant representation space by directly matching the marginal distributions of the two domains. However, they ignore exploring the underlying discriminative features of the target data and align the cross-domain discriminative features, which may lead to suboptimal performance. To tackle these two issues simultaneously, this paper presents a Joint Clustering and Discriminative Feature Alignment (JCDFA) approach for UDA, which is capable of naturally unifying the mining of discriminative features and the alignment of class-discriminative features into one single framework. Specifically, in order to mine the intrinsic discriminative information of the unlabeled target data, JCDFA jointly learns a shared encoding representation for two tasks: supervised classification of labeled source data, and discriminative clustering of unlabeled target data, where the classification of the source domain can guide the clustering learning of the target domain to locate the object category. We then conduct the cross-domain discriminative feature alignment by separately optimizing two new metrics: 1) an extended supervised contrastive learning, i.e., semi-supervised contrastive learning 2) an extended Maximum Mean Discrepancy (MMD), i.e., conditional MMD, explicitly minimizing the intra-class dispersion and maximizing the inter-class compactness. When these two procedures, i.e., discriminative features mining and alignment are integrated into one framework, they tend to benefit from each other to enhance the final performance from a cooperative learning perspective. Experiments are conducted on four real-world benchmarks (e.g., Office-31, ImageCLEF-DA, Office-Home and VisDA-C). All the results demonstrate that our JCDFA can obtain remarkable margins over stateof-the-art domain adaptation methods. Comprehensive ablation studies also verify the importance of each key component of our proposed algorithm and the effectiveness of combining two learning strategies into a framework.

show abstract

“…MADA [39] implements class-wise alignment with multi-discriminators. GSDA [23] performs class-, group-and domain-wise alignments simultaneously, where the three types of alignment are enforced to be consistent in their gradients for more precise alignment. HDA [11] leverages domain-specific representations as heuristics to obtain domain-invariant representations from a heuristic search perspective.…”

Section: Related Workmentioning

confidence: 99%

ToAlign: Task-oriented Alignment for Unsupervised Domain Adaptation

Wei¹,

Lan²,

Zeng³

et al. 2021

Preprint

View full text Add to dashboard Cite

Unsupervised domain adaptive classification intends to improve the classification performance on unlabeled target domain. To alleviate the adverse effect of domain shift, many approaches align the source and target domains in the feature space. However, a feature is usually taken as a whole for alignment without explicitly making domain alignment proactively serve the classification task, leading to sub-optimal solution. What sub-feature should be aligned for better adaptation is under-explored. In this paper, we propose an effective Task-oriented Alignment (ToAlign) for unsupervised domain adaptation (UDA). We study what features should be aligned across domains and propose to make the domain alignment proactively serve classification by performing feature decomposition and alignment under the guidance of the prior knowledge induced from the classification task itself. Particularly, we explicitly decompose a feature in the source domain into a task-related/discriminative feature that should be aligned, and a task-irrelevant feature that should be avoided/ignored, based on the classification meta-knowledge. Extensive experimental results on various benchmarks (e.g., Office-Home, Visda-2017, and DomainNet) under different domain adaptation settings demonstrate the effectiveness of ToAlign which helps achieve the state-of-the-art performance.

show abstract

Unsupervised Domain Adaptation With Hierarchical Gradient Synchronization

Cited by 82 publications

References 24 publications

Contrastive Adaptation Network for Unsupervised Domain Adaptation

Contrastive Adaptation Network for Unsupervised Domain Adaptation

Joint Clustering and Discriminative Feature Alignment for Unsupervised Domain Adaptation

ToAlign: Task-oriented Alignment for Unsupervised Domain Adaptation

Contact Info

Product

Resources

About