02/02/2021

Composite Adversarial Attacks

Xiaofeng Mao, Yuefeng Chen, Shuhui Wang, Hang Su, Yuan He, Hui Xue

Keywords:

Abstract Paper Similar Papers

Abstract: Adversarial attack is a technique for deceiving Machine Learning (ML) models, which provides a way to evaluate the adversarial robustness. In practice, attack algorithms are artificially selected and tuned by human experts to break a ML system. However, manual selection of attackers tends to be sub-optimal, leading to a mistakenly assessment of model security. In this paper, a new procedure called Composite Adversarial Attack (CAA) is proposed for automatically searching the best combination of attack algorithms and their hyper-parameters from a candidate pool of 32 base attackers. We design a search space where attack policy is represented as an attacking sequence, i.e., the output of the previous attacker is used as the initialization input for successors. Multi-objective NSGA-II genetic algorithm is adopted for finding the strongest attack policy with minimum complexity. The experimental result shows CAA beats 10 top attackers on 11 diverse defenses with less elapsed time (6 × faster than AutoAttack), and achieves the new state-of-the-art on linf, l2 and unrestricted adversarial attacks.

The video of this talk cannot be embedded. You can watch it here:
https://slideslive.com/38948313
(Link will open in new window)
 0
 0
 0
 0
  Share    
This is an embedded video. Talk and the respective paper are published at AAAI 2021 virtual conference. If you are one of the authors of the paper and want to manage your upload, see the question "My papertalk has been externally embedded..." in the FAQ section.

Comments

Post Comment
no comments yet
code of conduct: tbd Characters remaining: 140

Similar Papers

 0
 0
 0
 0
 9:34 
 0
 0
 0
 0
 14:42 
 0
 0
 0
 0
 5:00 
 0
 0
 0
 0
 1:01 
 0
 0
 0
 0
 3:22 
 0
 0
 0
 0
 15:11 
 0
 0
 0
 0
 11:33 
 0
 0
 0
 0
 4:56 
 0
 0
 0
 0
 12:50 
 0
 0
 0
 0
 5:12 
 0
 0
 0
 0
 15:04 
 0
 0
 0
 0
 15:02 
 1
 0
 0
 0
 12:20 
 0
 0
 0
 0
 1:00 
 0
 0
 0
 0
 6:27 
 0
 0
 0
 0
 3:15 
 0
 0
 0
 0
 5:21 
 0
 0
 0
 0
 3:59 
 0
 0
 0
 0
 15:40 
 0
 0
 0
 0
 1:01 
 0
 0
 0
 1
 14:57 
 0
 0
 0
 0
 5:14 
 0
 0
 0
 0
 14:46 
 0
 0
 0
 0
 14:15 
 1
 0
 0
 1
 3:22 
 0
 0
 0
 0
 5:05 
 0
 0
 0
 0
 9:09 
 0
 0
 0
 0
 12:06 
 0
 0
 0
 0
 10:57 
 0
 0
 0
 0
 14:01 
 0
 0
 0
 0
 14:24 
 0
 0
 0
 0
 17:51 
 0
 0
 0
 0
 4:57 
 0
 0
 0
 0
 13:39 
 0
 0
 0
 0
 14:03 
 0
 0
 0
 0
 14:19 
 0
 0
 0
 0
 4:57 
 0
 0
 0
 0
 14:52 
 0
 1
 0
 0
 5:01 
 0
 0
 0
 0
 3:59