Unsupervised domain adaptation based on the predictive uncertainty of models

Unsupervised domain adaptation (UDA) aims to improve the prediction performance in the target domain under distribution shifts from the source domain. The key principle of UDA is to minimize the divergence between the source and the target domains. To follow this principle, many methods employ a domain discriminator to match the feature distributions. Some recent methods evaluate the discrep-ancy between two predictions on target samples to detect those that deviate from the source distribution. However, their performance is limited because they either match the marginal distributions or measure the divergence conservatively. In this paper, we present a novel UDA method that learns domain -invariant features that minimize the domain divergence. We propose model uncertainty as a measure of the domain divergence. Our UDA method based on model uncertainty (MUDA) adopts a Bayesian framework and provides an efficient way to evaluate model uncertainty by means of Monte Carlo dropout sampling. Experiment results on image recognition tasks show that our method is superior to existing state-of-the-art methods. We also extend MUDA to multi-source domain adaptation problems.(c) 2022 Elsevier B.V. All rights reserved.

Automated summary

Unsupervised domain adaptation (UDA) aims to improve prediction performance in the target domain by minimizing the divergence between the source and target domains. This paper proposes a novel UDA method, called Model Uncertainty-based UDA (MUDA), which learns domain-invariant features to minimize domain divergence. MUDA utilizes a Bayesian framework and Monte Carlo dropout sampling to evaluate model uncertainty. Experimental results on image recognition tasks demonstrate the superiority of MUDA over existing state-of-the-art methods. MUDA is also extended to multi-source domain adaptation problems.