Steganalysis of Spread Spectrum Hiding Based on DWT and GMM

Audio steganalysis has attracted more attentions recently. DSSS steganalysis is one of the most challenging research fields. In this paper, a novel algorithm to detect DSSS steganography in audio signal is proposed. Firstly, it takes DWT transform of special segment of audio and takes the detail sub-band coefficients, and then uses GMM to model the coefficients. Secondly, in order to monitor the effect of DSSS hiding, we calculate the GMM PDF (possibility density function) as to measure the difference. Thirdly, considering the two variables composed of wavelet coefficients and GMM PDF, the multivariate skewness and kurtosis were taken as features. Lastly, the SVM classifier is utilized for classification. All of the 800 various audios are trained and tested in our experimental work. With various embedding parameters for training and testing audios, the proposed algorithm can achieve a good classification, and the correct rate of detecting is up to 80%.