The recent progress and development of deep generative models have led to remark- able improvements in research topics in computer vision and machine learning. In this paper, we address the task of cross-domain feature disentanglement. We advance the idea of unsupervised domain adaptation→