Advscod: bayesian-based out-of-distribution detection via curvature sketching and adversarial sample enrichment

HIGHLIGHTS

• who: Jiacheng Qiao et al. from the Cyberspace Institute Advanced Technology, Guangzhou University, Guangzhou, China have published the research: AdvSCOD: Bayesian-Based Out-of-Distribution Detection via Curvature Sketching and Adversarial Sample Enrichment, in the Journal: Mathematics 2023, 11, 692. of /2023/
• what: The authors propose a novel framework that enriches the input with a small set its neighborhoods generated by applying perturbation which the authors believe can better reflect the influence on model predictions and then the authors average their uncertainties measured by SCOD. To resolve the above issue, the authors propose a novel . . .
  

   

  

  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.