Abstract:Adversarial learning has proven to be an effective method for capturing transferable features for unsupervised domain adaptation. However, some existing conditional adversarial domain adaptation methods assign equal importance to different samples, ignoring the fact that hard-to-transfer samples might damage the conditional adversarial adaptation procedure. Meanwhile, some methods can only roughly align marginal distributions across domains, but cannot ensure category distributions alignment, causing classifiers to make uncertain or even wrong predictions for some target data. Furthermore, we find that the feature norms of real images usually follow a complex distribution, so directly matching the mean feature norms of two domains cannot effectively reduce the statistical discrepancy of feature norms and may potentially induce feature degradation. In this paper, we develop a Trust-aware Conditional Adversarial Domain Adaptation (TCADA) method for solving the aforementioned issues. To quantify data transferability, we suggest utilizing posterior probability modeled by a Gaussian-uniform mixture, which effectively facilitates conditional domain alignment. Based on this posterior probability, a confidence-guided alignment strategy is presented to promote precise alignment of category distributions and accelerate the learning of shared features. Moreover, a novel optimal transport-based strategy is introduced to align the feature norms and facilitate shared features becoming more informative. To encourage classifiers to make more accurate predictions for target data, we also design a mixed information-guided entropy regularization term to promote deep features being away from the decision boundaries. Extensive experiments show that our method greatly improves transfer performance on various tasks.

Multi-objective dynamic distribution adaptation with instance reweighting for transfer feature learning

HOMDA: High-Order Moment-Based Domain Alignment for Unsupervised Domain Adaptation

Trust-aware Conditional Adversarial Domain Adaptation with Feature Norm Alignment.

Transfer Learning with Dynamic Distribution Adaptation

When to transfer: a dynamic domain adaptation method for effective knowledge transfer

Heterogeneous Domain Adaptation Through Progressive Alignment.

Multisource Deep Transfer Learning Based on Balanced Distribution Adaptation

A Multi-objective Bayesian Optimization Algorithm with Online Knowledge Transfer

Hierarchical Knowledge Amalgamation with Dual Discriminative Feature Alignment

Transfer Feature Learning with Joint Distribution Adaptation

A Knowledge guided Transfer Strategy for Evolutionary Dynamic Multiobjective Optimization

Transfer Learning based Dynamic Multiobjective Optimization Algorithms

Dynamic Distribution Adaptation Based Transfer Network for Cross Domain Bearing Fault Diagnosis

Reliable hybrid knowledge distillation for multi-source domain adaptive object detection

Transfer Learning-Based Dynamic Multiobjective Optimization Algorithms

Dual-Level Adaptive and Discriminative Knowledge Transfer for Cross-Domain Recognition

Active Dynamic Weighting for multi-domain adaptation

A novel joint distinct subspace learning and dynamic distribution adaptation method for fault transfer diagnosis

Rethinking Distributional Matching Based Domain Adaptation

Multisource Heterogeneous Domain Adaptation With Conditional Weighting Adversarial Network