Informative feature disentanglement for unsupervised domain adaptation
|Author:||Deng, Wanxia1; Zhao, Lingjun1; Liao, Qing2;|
1State Key Laboratory of Complex Electromagnetic Environment Effects on Electronics, and Information System, College of Electronic Science, National University of Defense technology, Changsha 410073, China
2Department of Computer Science, and Technology, Harbin Institute of Technology, Shenzhen 518055, China
3College of System Engineering, National University of Defense Technology, Changsha 410073, China
4College of Intelligent Science, National University of Defense Technology, Changsha 410073, China
5Center for Machine Vision, and Signal analysis, University of Oulu, 90570 Oulu, Finland
6ollege of System Engineering, National University of Defense Technology, Changsha 410073, China
|Online Access:||PDF Full Text (PDF, 4.9 MB)|
|Persistent link:|| http://urn.fi/urn:nbn:fi-fe2022082956605
Institute of Electrical and Electronics Engineers,
|Publish Date:|| 2022-08-29
Unsupervised Domain Adaptation (UDA) aims at learning a classifier for an unlabeled target domain by transferring knowledge from a labeled source domain with a related but different distribution. The strategy of aligning the two domains in latent feature space via metric discrepancy or adversarial learning has achieved considerable progress. However, these existing approaches mainly focus on adapting the entire image and ignore the bottleneck that occurs when forced adaptation of uninformative domain-specific variations undermines the effectiveness of learned features. To address this problem, we propose a novel component called Informative Feature Disentanglement (IFD), which is equipped with the adversarial network or the metric discrepancy model, respectively. Accordingly, the new network architectures, named IFDAN and IFDMN, enable informative feature refinement before the adaptation. The proposed IFD is designed to disentangle informative features from the uninformative domain-specific variations, which are produced by a Variational Autoencoder (VAE) with lateral connections from the encoder to the decoder. We cooperatively apply the IFD to conduct supervised disentanglement for the source domain and unsupervised disentanglement for the target domain. In this way, informative features are disentangled from the domain-specific details before the adaptation. Extensive experimental results on three gold-standard domain adaptation datasets, e.g., Office31, Office-Home and VisDA-C, demonstrate the effectiveness of the proposed IFDAN and IFDMN models for UDA.
IEEE transactions on multimedia
|Pages:||2407 - 2421|
|Type of Publication:||
A1 Journal article – refereed
|Field of Science:||
113 Computer and information sciences
This work was supported in part by the National Natural Science Foundation of China under Grants 62022091, 61872379, 71701205,61701508, and 62036013, in part by the Academy of Finland under Grant 331883, and in part by Hunan Provincial Natural Science Foundation of China under Grant 2018JJ3613.
|Academy of Finland Grant Number:||
331883 (Academy of Finland Funding decision)
© The Author(s) 2021. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.