Poster: Characterizing adversarial subspaces by mutual information

Chia Yi Hsu, Pin Yu Chen, Chia Mu Yu

研究成果: Conference contribution同行評審

摘要

Deep learning is well-known for its great performances on images classification, object detection, and natural language processing. However, the recent research has demonstrated that visually indistinguishable images called adversarial examples can successfully fool neural networks by carefully crafting. In this paper, we design a detector named MID, calculating mutual information to characterize adversarial subspaces. Meanwhile, we use the defense framework called MagNet and mount the detector MID on it. Experimental results show that projected gradient descent (PGD), basic iterative method (BIM), Carlini and Wanger's attack (C&W attack) and elastic-net attack to deep neural network (elastic-net and L1 rules) can be effectively defended by our method.

原文English
主出版物標題AsiaCCS 2019 - Proceedings of the 2019 ACM Asia Conference on Computer and Communications Security
發行者Association for Computing Machinery, Inc
頁面667-669
頁數3
ISBN(電子)9781450367523
DOIs
出版狀態Published - 2 7月 2019
事件2019 ACM Asia Conference on Computer and Communications Security, AsiaCCS 2019 - Auckland, 新西蘭
持續時間: 9 7月 201912 7月 2019

出版系列

名字AsiaCCS 2019 - Proceedings of the 2019 ACM Asia Conference on Computer and Communications Security

Conference

Conference2019 ACM Asia Conference on Computer and Communications Security, AsiaCCS 2019
國家/地區新西蘭
城市Auckland
期間9/07/1912/07/19

指紋

深入研究「Poster: Characterizing adversarial subspaces by mutual information」主題。共同形成了獨特的指紋。

引用此