Abstract Unsupervised domain adaptation transfers empirical knowledge from a label-rich source domain to a fully unlabeled target domain with a different distribution. A core idea of many existing approaches is to reduce the distribution divergence between domains. However. they focused only on part of the discrimination. https://www.wholisticpet.shop/product-category/gear-pants/