Multi-target Unsupervised Domain Adaptation without Exactly Shared Categories

Yu, Huanhuan; Hu, Menglei; Chen, Songcan

doi:10.48550/arxiv.1809.00852

Cited by 16 publications

(19 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another group of methods utilize a model trained on source data to generate pseudo label on the target data in a self-training (ST) manner. Instead of assuming that the target data comes from a single domain, Yu et al [42] and Gholami et al [10] address a challenging multi-target UDA problem. However, these methods are not directly applicable to the FMTDA setting since they assume centralized data on a server.…”

Section: Unsupervised Domain Adaptationmentioning

confidence: 99%

“…The problem of domain mismatch among multiple clients, or target domains, is challenging even in a centralized training framework. Although some methods have been proposed for centralized multi-target DA [42,10], the best way to adapt to diverse domains in real-world scenarios remains unclear. To disentangle the challenges of multi-target DA and distributed training, we also look into the problem in a centralized setting.…”

Section: Centralized Multi-target Domain Adaptationmentioning

confidence: 99%

See 1 more Smart Citation

Federated Multi-Target Domain Adaptation

Yao¹,

Gong²,

Cui³

et al. 2021

Preprint

View full text Add to dashboard Cite

Federated learning methods enable us to train machine learning models on distributed user data while preserving its privacy. However, it is not always feasible to obtain high-quality supervisory signals from users, especially for vision tasks. Unlike typical federated settings with labeled client data, we consider a more practical scenario where the distributed client data is unlabeled, and a centralized labeled dataset is available on the server. We further take the server-client and inter-client domain shifts into account and pose a domain adaptation problem with one source (centralized server data) and multiple targets (distributed client data). Within this new Federated Multi-Target Domain Adaptation (FMTDA) task, we analyze the model performance of exiting domain adaptation methods and propose an effective DualAdapt method to address the new challenges. Extensive experimental results on image classification and semantic segmentation tasks demonstrate that our method achieves high accuracy, incurs minimal communication cost, and requires low computational resources on client devices.

show abstract

Section: Unsupervised Domain Adaptationmentioning

confidence: 99%

Section: Centralized Multi-target Domain Adaptationmentioning

confidence: 99%

Federated Multi-Target Domain Adaptation

Yao¹,

Gong²,

Cui³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Multi-Source/Target UDA. Multi-source domain adaptation (MSDA) [21,36,45,26] and multi-target domain adaptation (MTDA) [44,12] consider UDA where either the source or target domain consists of multiple subdomains. While these studies assume that the subdomain labels of all samples are available, some others assumed that they are unavailable in the source [20] or target domain (Blendingtarget domain adaptation: BTDA) [7,27].…”

Section: Related Workmentioning

confidence: 99%

Generalized Domain Adaptation

Mitsuzumi¹,

Irie²,

Ikami³

et al. 2021

Preprint

View full text Add to dashboard Cite

Many variants of unsupervised domain adaptation (UDA) problems have been proposed and solved individually. Its side effect is that a method that works for one variant is often ineffective for or not even applicable to another, which has prevented practical applications. In this paper, we give a general representation of UDA problems, named Generalized Domain Adaptation (GDA). GDA covers the major variants as special cases, which allows us to organize them in a comprehensive framework. Moreover, this generalization leads to a new challenging setting where existing methods fail, such as when domain labels are unknown, and class labels are only partially given to each domain. We propose a novel approach to the new setting. The key to our approach is self-supervised class-destructive learning, which enables the learning of class-invariant representations and domain-adversarial classifiers without using any domain labels. Extensive experiments using three benchmark datasets demonstrate that our method outperforms the state-of-the-art UDA methods in the new setting and that it is competitive in existing UDA variations as well.

show abstract

“…According to the numbers of source domain and target domain, we can group UDA methods into three scenarios: single source domain with single target domain (1S1T) [23], [24], [25], [26], single source domain with multi-target domains (1SmT) [27] and multi-source domains with single target domain (mS1T), as shown in Fig. 1.…”

Section: Classifier-level Adaptationmentioning

confidence: 99%

“…To this end, the 1SmT modeling strategy has been proposed. One of the representative is the PA-1SmT [27], which achieved UDA by transferring knowledge from the source domain to each of the target domains. However, it does not consider the domain shift in distributions when aligning the target domains to the source.…”

Section: Classifier-level Adaptationmentioning

confidence: 99%

Unsupervised Domain Adaptation Through Transferring both the Source-Knowledge and Target-Relatedness Simultaneously

Qiu¹,

Ma²,

Cao³

et al. 2020

Preprint

View full text Add to dashboard Cite

Unsupervised domain adaptation (UDA) is an emerging research topic in the field of machine learning and pattern recognition, which aims to help the learning of unlabeled target domain by transferring knowledge from the source domain. To perform UDA, a variety of methods have been proposed, most of which concentrate on the scenario of single source and single target domain (1S1T). However, in real applications, usually single source domain with multiple target domains are involved (1SmT), which cannot be handled directly by those 1S1T models. Unfortunately, although a few related works on 1SmT UDA have been proposed, nearly none of them model the source domain knowledge and leverage the target-relatedness jointly. To overcome these shortcomings, we herein propose a more general 1SmT UDA model through transferring both the Source-Knowledge and Target-Relatedness, UDA-SKTR for short. In this way, not only the supervision knowledge from the source domain, but also the potential relatedness among the target domains are simultaneously modeled for exploitation in the process of 1SmT UDA. In addition, we construct an alternating optimization algorithm to solve the variables of the proposed model with convergence guarantee. Finally, through extensive experiments on both benchmark and real datasets, we validate the effectiveness and superiority of the proposed method.

show abstract

Multi-target Unsupervised Domain Adaptation without Exactly Shared Categories

Cited by 16 publications

References 17 publications

Federated Multi-Target Domain Adaptation

Federated Multi-Target Domain Adaptation

Generalized Domain Adaptation

Unsupervised Domain Adaptation Through Transferring both the Source-Knowledge and Target-Relatedness Simultaneously

Contact Info

Product

Resources

About