Unsupervised domain adaptation via importance sampling

Publication Type

Journal Article

Publication Date

12-2020

Abstract

Unsupervised domain adaptation aims to generalize a model from the label-rich source domain to the unlabeled target domain. Existing works mainly focus on aligning the global distribution statistics between source and target domains. However, they neglect distractions from the unexpected noisy samples in domain distribution estimation, leading to domain misalignment or even negative transfer. In this paper, we present an importance sampling method for domain adaptation (ISDA), to measure sample contributions according to their "informative" levels. In particular, informative samples, as well as outliers, can be effectively modeled using feature-norm and prediction entropy of the network. The importance of information is further formulated as the importance sampling losses in features and label spaces. In this way, the proposed model mitigates the noisy outliers while enhancing the important samples during domain alignment. In addition, our model is easy to implement yet effective, and it does not introduce any extra parameters. Extensive experiments on several benchmark datasets show that our method outperforms state-of-the-art methods under both the standard and partial domain adaptation settings.

Keywords

Feature extraction, Entropy, Tuning, Estimation, Monte Carlo methods, Adaptation models, Noise measurement, Domain adaptation, deep learning, distribution sampling

Discipline

Information Security

Research Areas

Information Systems and Management

Publication

IEEE Transactions on Circuits and Systems for Video Technology

Volume

30

Issue

12

First Page

4688

Last Page

4699

ISSN

1051-8215

Identifier

10.1109/TCSVT.2019.2963318

Publisher

Institute of Electrical and Electronics Engineers

Additional URL

https://doi.org/10.1109/TCSVT.2019.2963318

This document is currently not available here.

Share

COinS