Correlation-aware adversarial domain adaptation and generalization. (April 2020)

• Record Type: Journal Article
• Authors: Rahman, Mohammad Mahfujur; Fookes, Clinton; Baktashmotlagh, Mahsa; Sridharan, Sridha
• Is Part Of: Pattern recognition. Volume 100(2020:Apr.)
• Publication Date: 2020-04
• Subjects: Domain adaptation -- Domain generalization -- Correlation-alignment -- Adversarial learning; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2019.107124
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We propose a new deep domain adaptation (DA) framework. We further extend the proposed DA framework on deep domain generalization (DG) scenarios. The proposed method exceeds state-of-the-art performance on both DA and DG scenarios. Cross-domain testing confirms the suitability for real-world applications. Abstract: Domain adaptation (DA) and domain generalization (DG) have emerged as a solution to the domain shift problem where the distribution of the source and target data is different. The task of DG is more challenging than DA as the target data is totally unseen during the training phase in DG scenarios. The current state-of-the-art employs adversarial techniques, however, these are rarely considered for the DG problem. Furthermore, these approaches do not consider correlation alignment which has been proven highly beneficial for minimizing domain discrepancy. In this paper, we propose a correlation-aware adversarial DA and DG framework where the features of the source and target data are minimized using correlation alignment along with adversarial learning. Incorporating the correlation alignment module along with adversarial learning helps to achieve a more domain agnostic model due to the improved ability to reduce domain discrepancy with unlabeled target data more effectively. Experiments on benchmark datasets serve as evidence that our proposed method yields improved state-of-the-art performance.