Simultaneously improve transferability and discriminability for adversarial domain adaptation

HIGHLIGHTS

• who: Ting Xiao et al. from the School of Computer Science, Harbin Institute of Technology, Harbin, China have published the research: Simultaneously Improve Transferability and Discriminability for Adversarial Domain Adaptation, in the Journal: Entropy
• what: The authors propose a matrix rank embedding (MRE) method to enhance feature and MRE restores a low-rank structure for data in the same class and enforces a maximum separation structure for data in different classes. The authors propose aligning the class-conditional label distribution P(Yc ) across two domains to enhance the transferability further. The authors propose an approach . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

  If you don't have an account, you can create one here.