Rethinking Distributional Matching Based Domain Adaptation

Abstract

Domain adaptation (DA) is a technique that transfers predictive models trained on a labeled source domain to an unlabeled target domain, with the core difficulty of resolving distributional shift between domains. Currently, most popular DA algorithms are based on distributional matching (DM). However in practice, realistic domain shifts (RDS) may violate their basic assumptions and as a result these methods will fail. In this paper, in order to devise robust DA algorithms, we first systematically analyze the limitations of DM based methods, and then build new benchmarks with more realistic domain shifts to evaluate the well-accepted DM methods. We further propose InstaPBM, a novel Instance-based Predictive Behavior Matching method for robust DA. Extensive experiments on both conventional and RDS benchmarks demonstrate both the limitations of DM methods and the efficacy of InstaPBM: Compared with the best baselines, InstaPBM improves the classification accuracy respectively by $4.5\%$, $3.9\%$ on Digits5, VisDA2017, and $2.2\%$, $2.9\%$, $3.6\%$ on DomainNet-LDS, DomainNet-ILDS, ID-TwO. We hope our intuitive yet effective method will serve as a useful new direction and increase the robustness of DA in real scenarios. Code will be available at anonymous link: https://github.com/pikachusocute/InstaPBM-RobustDA.

Preprint version

References in corpus (18)

• mixup: Beyond Empirical Risk Minimization
• Learning Transferable Features with Deep Adaptation Networks
• Unsupervised Data Augmentation for Consistency Training
• Conditional Adversarial Domain Adaptation
• VisDA: The Visual Domain Adaptation Challenge
• A DIRT-T Approach to Unsupervised Domain Adaptation
• Wasserstein Distance Guided Representation Learning for Domain Adaptation
• Domain-Adversarial Neural Networks
• Unsupervised Domain Adaptation through Self-Supervision
• On Learning Invariant Representation for Domain Adaptation
• Multi-source Domain Adaptation for Semantic Segmentation
• Domain Adaptation with Conditional Distribution Matching and Generalized Label Shift
• Self-ensembling for visual domain adaptation
• Domain Adaptation with Asymmetrically-Relaxed Distribution Alignment
• Adversarial Dropout Regularization
• Towards Discriminability and Diversity: Batch Nuclear-norm Maximization under Label Insufficient Situations
• Cross-domain Object Detection through Coarse-to-Fine Feature Adaptation
• HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation

Cited by in corpus (5)

• Cycle Self-Training for Domain Adaptation
• Partial Domain Adaptation without Domain Alignment
• A Theory of Label Propagation for Subpopulation Shift
• Cross-Domain Sentiment Classification with In-Domain Contrastive Learning
• Cross-Domain Sentiment Classification with Contrastive Learning and Mutual Information Maximization