"Unrestricted Adversarial Examples via Semantic Manipulation."

Anand Bhattad et al. (2020)
a service of Schloss Dagstuhl - Leibniz Center for Informatics