Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

Defending DNN Adversarial Attacks with Pruning and Logits Augmentation

Abstract: 

Deep neural networks (DNNs) have been shown to be powerful models and perform extremely well on many complicated artificial intelligent tasks. However, recent research found that these powerful models are vulnerable to adversarial attacks, i.e., intentionally added imperceptible perturbations to DNN inputs can easily mislead the DNNs with extremely high confidence. In this work, we enhance the robustness ofDNNs under adversarial attacks by using pruning method and logits augmentation, we achieve both effective defense against adversarial examples and DNN model compression. We have observed defense against adversarial attacks under the white box attack assumption. Our defense mechanisms work even better under the grey box attack assumption.

up
0 users have voted:

Paper Details

Authors:
Siyue Wang, Xiao Wang, Shaokai Ye, Pu Zhao, Xue Lin
Submitted On:
28 November 2018 - 8:40pm
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Pu Zhao
Paper Code:
1228
Document Year:
2018
Cite

Document Files

GlobalSip_Final.pptx

(9 downloads)

Subscribe

[1] Siyue Wang, Xiao Wang, Shaokai Ye, Pu Zhao, Xue Lin, "Defending DNN Adversarial Attacks with Pruning and Logits Augmentation", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3828. Accessed: Dec. 16, 2018.
@article{3828-18,
url = {http://sigport.org/3828},
author = {Siyue Wang; Xiao Wang; Shaokai Ye; Pu Zhao; Xue Lin },
publisher = {IEEE SigPort},
title = {Defending DNN Adversarial Attacks with Pruning and Logits Augmentation},
year = {2018} }
TY - EJOUR
T1 - Defending DNN Adversarial Attacks with Pruning and Logits Augmentation
AU - Siyue Wang; Xiao Wang; Shaokai Ye; Pu Zhao; Xue Lin
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3828
ER -
Siyue Wang, Xiao Wang, Shaokai Ye, Pu Zhao, Xue Lin. (2018). Defending DNN Adversarial Attacks with Pruning and Logits Augmentation. IEEE SigPort. http://sigport.org/3828
Siyue Wang, Xiao Wang, Shaokai Ye, Pu Zhao, Xue Lin, 2018. Defending DNN Adversarial Attacks with Pruning and Logits Augmentation. Available at: http://sigport.org/3828.
Siyue Wang, Xiao Wang, Shaokai Ye, Pu Zhao, Xue Lin. (2018). "Defending DNN Adversarial Attacks with Pruning and Logits Augmentation." Web.
1. Siyue Wang, Xiao Wang, Shaokai Ye, Pu Zhao, Xue Lin. Defending DNN Adversarial Attacks with Pruning and Logits Augmentation [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3828