Evasion Attack STeganography: Turning Vulnerability Of Machine Learning ToAdversarial Attacks Into A Real-world Application

Ghamizi, Salah; Cordy, Maxime; Papadakis, Mike; Le Traon, Yves

Reference : Evasion Attack STeganography: Turning Vulnerability Of Machine Learning ToAdversarial...


	Document type :	Scientific journals : Article
	Discipline(s) :	Engineering, computing & technology : Computer science
	Focus Areas :	Security, Reliability and Trust
	To cite this reference:	http://hdl.handle.net/10993/47832

Title :	Evasion Attack STeganography: Turning Vulnerability Of Machine Learning ToAdversarial Attacks Into A Real-world Application
Language :	English
Author, co-author :	Ghamizi, Salah [University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Computer Science (DCS) >]
	Cordy, Maxime [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SerVal >]
	Papadakis, Mike [University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Computer Science (DCS) >]
	Le Traon, Yves [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SerVal >]
Publication date :	11-Oct-2021
Journal title :	Proceedings of International Conference on Computer Vision 2021
Peer reviewed :	Yes
Audience :	International
Keywords :	[en] Adversarial Attacks ; Steganography ; Watermarking
Abstract :	[en] Evasion Attacks have been commonly seen as a weakness of Deep Neural Networks. In this paper, we flip the paradigm and envision this vulnerability as a useful application. We propose EAST, a new steganography and watermarking technique based on multi-label targeted evasion attacks. Our results confirm that our embedding is elusive; it not only passes unnoticed by humans, steganalysis methods, and machine-learning detectors. In addition, our embedding is resilient to soft and aggressive image tampering (87% recovery rate under jpeg compression). EAST outperforms existing deep-learning-based steganography approaches with images that are 70% denser and 73% more robust and supports multiple datasets and architectures.
Target :	Researchers
Permalink :	http://hdl.handle.net/10993/47832

File(s) associated to this reference

Fulltext file(s):

	File	Commentary	Version	Size	Access
Limited access	EAST_ICCV_CameraReady.pdf		Author postprint	1.45 MB	Request a copy

All documents in ORBi^lu are protected by a user license.

University of Luxembourg Library | Feedback | Legal notices