User: Guest  Login
Title:

Copy and Paste: A Simple But Effective Initialization Method for Black-Box Adversarial Attacks

Document type:
Konferenzbeitrag
Author(s):
Brunner, Thomas; Diehl, Frederik; Knoll, Alois
Abstract:
Many optimization methods for generating black-box adversarial examples have been proposed, but the aspect of initializing said optimizers has not been considered in much detail. We show that the choice of starting points is indeed crucial, and that the performance of state-of-the-art attacks depends on it. First, we discuss desirable properties of starting points for attacking image classifiers, and how they can be chosen to increase query efficiency. Notably, we find that simply copying small...     »
Book / Congress title:
2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
Congress (additional information):
Long Beach, CA, USA
Year:
2019
Month:
Jun
Reviewed:
ja
Notes:
Poster presentation at the Workshop for Adversarial Machine Learning in Real-World Computer Vision Systems.
 BibTeX