User: Guest  Login
Document type:
Konferenzbeitrag 
Contribution type:
Textbeitrag / Aufsatz 
Author(s):
Brunner, Thomas; Diehl, Frederik; Truong Le, Michael; Knoll, Alois 
Title:
Guessing Smart: Biased Sampling for Efficient Black-Box Adversarial Attacks 
Abstract:
We consider adversarial examples for image classification in the black-box decision-based setting. Here, an attacker cannot access confidence scores, but only the final label. Most attacks for this scenario are either unreliable or inefficient. Focusing on the latter, we show that a specific class of attacks, Boundary Attacks, can be reinterpreted as a biased sampling framework that gains efficiency from domain knowledge. We identify three such biases, image frequency, regional masks and surroga...    »
 
Book / Congress title:
The IEEE International Conference on Computer Vision (ICCV) 
Year:
2019