1 paper
Alexander Matyasko, Xin Lou, Indriyati Atmosukarto +1
Attacking semantic segmentation models is significantly harder than image classification models because an attacker must flip thousands of pixel predictions simultaneously. Standar…