Unsupervised Domain Adaptation via Structured Prediction Based Selective Pseudo-Labeling

Wang, Q.; Breckon, T.P.

Abstract

Unsupervised domain adaptation aims to address the problem of classifying unlabeled samples from the target domain whilst labeled samples are only available from the source domain and the data distributions are different in these two domains. As a result, classifiers trained from labeled samples in the source domain suffer from significant performance drop when directly applied to the samples from the target domain. To address this issue, different approaches have been proposed to learn domain-invariant features or domain-specific classifiers. In either case, the lack of labeled samples in the target domain can be an issue which is usually overcome by pseudo-labeling. Inaccurate pseudo-labeling, however, could result in catastrophic error accumulation during learning. In this paper, we propose a novel selective pseudo-labeling strategy based on structured prediction. The idea of structured prediction is inspired by the fact that samples in the target domain are well clustered within the deep feature space so that unsupervised clustering analysis can be used to facilitate accurate pseudo-labeling. Experimental results on four datasets (i.e. Office-Caltech, Office31, ImageCLEF-DA and Office-Home) validate our approach outperforms contemporary state-of-the-art methods.

Citation

Wang, Q., & Breckon, T. (2020, February). Unsupervised Domain Adaptation via Structured Prediction Based Selective Pseudo-Labeling. Presented at Thirty Fourth AAAI Conference on Artificial Intelligence, New York, USA

• Presentation Conference Type: Conference Paper (published)
• Conference Name: Thirty Fourth AAAI Conference on Artificial Intelligence
• Start Date: Feb 7, 2020
• End Date: Feb 12, 2020
• Acceptance Date: Nov 11, 2019
• Online Publication Date: Apr 3, 2020
• Publication Date: 2020-04
• Deposit Date: Dec 20, 2019
• Publicly Available Date: Mar 18, 2020
• Volume: 34
• Pages: 6243-6250
• Series Number: 4
• Series ISSN: 2159-5399,2374-3468
• Book Title: AAAI-20 / IAAI-20 / EAAI-20 proceedings
• DOI: https://doi.org/10.1609/aaai.v34i04.6091
• Public URL: https://durham-repository.worktribe.com/output/1140803