Detecting information-hiding in WAV audios

In this article, we propose a steganalysis method for detecting the presence of information-hiding behavior in wav audios. We extract the neighboring joint distribution features and the Markov features of the second order derivative, and combine these features with the error response by randomly modifying the least significant bit, then apply learning machines to the features for distinguishing the stegoaudios from cover videos. Experimental results show that our method performs well in steganalysis of the audio stegograms that are produced by using Hide4PGP, Invisible Secrets and S-tools4.

• Type: Conference
• Year: 2008
• Where: ICPR
• Authors: Qingzhong Liu, Andrew H. Sung, Mengyu Qiao
• Keywords: Computer Vision, ICPR 2008, Joint Distribution Features, Steganalysis Method, Wav Audios