The challenges of rich features in universal steganalysis

Tomáš Pevný; Andrew D. Ker

doi:10.1117/12.2006790

22 March 2013 The challenges of rich features in universal steganalysis

Tomáš Pevný, Andrew D. Ker

Author Affiliations +

Proceedings Volume 8665, Media Watermarking, Security, and Forensics 2013; 86650M (2013) https://doi.org/10.1117/12.2006790
Event: IS&T/SPIE Electronic Imaging, 2013, Burlingame, California, United States

Abstract

Contemporary steganalysis is driven by new steganographic rich feature sets, which consist of large numbers of weak features. Although extremely powerful when applied to supervised classification problems, they are not compatible with unsupervised universal steganalysis, because the unsupervised method cannot separate the signal (evidence of steganographic embedding) from the noise (cover content). This work tries to alleviate the problem, by means of feature extraction algorithms. We focus on linear projections informed by embedding methods, and propose a new method which we call calibrated least squares with the specific aim of making the projections sensitive to stego content yet insensitive to cover variation. Different projections are evaluated by their application to the anomaly detector from Ref. 1, and we are able to retain both the universality and the robustness of the method, while increasing its performance substantially.

Citation Download Citation

Tomáš Pevný, Tomáš Pevný, Andrew D. Ker, Andrew D. Ker, } "The challenges of rich features in universal steganalysis", Proc. SPIE 8665, Media Watermarking, Security, and Forensics 2013, 86650M (22 March 2013); doi: 10.1117/12.2006790; https://doi.org/10.1117/12.2006790

ACCESS THE FULL ARTICLE

PERSONAL SIGN IN
Full access may be available with your subscription

Email or Username Forgot your username?

Password Forgot your password?

Show

Remember Email/Username on this computer

Remember Password

No SPIE account? Create an account
or Institutional Sign In via Shibboleth

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $14.40

Non-members: $18.00 ADD TO CART

Show All Keywords

Keywords/Phrases

Search In:

Publication Years