Deep Unsupervised Domain Adaptation for Image Classification via Low Rank Representation Learning

Domain adaptation is a powerful technique given a wide amount of labeled data from similar attributes in different domains. In real-world applications, there is a huge number of data but almost more of them are unlabeled. It is effective in image classification where it is expensive and time-consuming to obtain adequate label data. We propose a novel method named DALRRL, which consists of deep architecture with domain-general and domain-specific representations across domains for deep unsupervised domain adaptation. Also, we apply low-rank representation learning to reduce source and target domains discrepancy. The low-rank constraint can uncover more related information between domains and it can transfer more relevant knowledge from the source domain to the target domain. The DALRRL guarantees to minimize marginal and conditional distributions difference between the source and target domains. The experimental results conducted on two benchmark domain adaptation datasets demonstrate the effectiveness of our method in image classification tasks.