Reducing the Covariate Shift by Mirror Samples in Cross Domain Alignment

by   Yin Zhao, et al.

Eliminating the covariate shift cross domains is one of the common methods to deal with the issue of domain shift in visual unsupervised domain adaptation. However, current alignment methods, especially the prototype based or sample-level based methods neglect the structural properties of the underlying distribution and even break the condition of covariate shift. To relieve the limitations and conflicts, we introduce a novel concept named (virtual) mirror, which represents the equivalent sample in another domain. The equivalent sample pairs, named mirror pairs reflect the natural correspondence of the empirical distributions. Then a mirror loss, which aligns the mirror pairs cross domains, is constructed to enhance the alignment of the domains. The proposed method does not distort the internal structure of the underlying distribution. We also provide theoretical proof that the mirror samples and mirror loss have better asymptotic properties in reducing the domain shift. By applying the virtual mirror and mirror loss to the generic unsupervised domain adaptation model, we achieved consistent superior performance on several mainstream benchmarks.


page 1

page 2

page 3

page 4


Generalized Domain Adaptation with Covariate and Label Shift CO-ALignment

Unsupervised knowledge transfer has a great potential to improve the gen...

Mitigating Both Covariate and Conditional Shift for Domain Generalization

Domain generalization (DG) aims to learn a model on several source domai...

Joint Geometrical and Statistical Alignment for Visual Domain Adaptation

This paper presents a novel unsupervised domain adaptation method for cr...

Implicit Class-Conditioned Domain Alignment for Unsupervised Domain Adaptation

We present an approach for unsupervised domain adaptation—with a strong ...

Searching for Optimal Subword Tokenization in Cross-domain NER

Input distribution shift is one of the vital problems in unsupervised do...

Please sign up or login with your details

Forgot password? Click here to reset