IBM Systems Journal
Chaotic-Type Features for Speech Steganalysis
IEEE Transactions on Information Forensics and Security
Steganalysis using image quality metrics
IEEE Transactions on Image Processing
Optimal Signature Design for Spread-Spectrum Steganography
IEEE Transactions on Image Processing
A novel echo-hiding scheme with backward and forward kernels
IEEE Transactions on Circuits and Systems for Video Technology
Hi-index | 0.00 |
In this paper, we describe an effective steganalysis for four kinds of echo hiding based on combinatorial high order statistics. We analyze effect on power cepstrum, due to statistic characteristic changing with echo in stego speech. High order statistics of each segment of power cepstrum firstly are calculated, including coefficients of skewness and kurtosis. Then we compute high order statistics of segmented skewness and kurtosis in combination. The combinatorial high order statistics of original speech and echoed speech are trained with SVM. By obtaining combinatorial high order statistics thresholds, we can used them for blind steganalysis of echo hiding. Experimental results are compared and show that our proposed steganalysis with three combinatorial high order statistics performs well in detecting Ker.1, Ker.3 and Ker.4 echo hiding, and skewness-kurtosis achieves high detecting correct rate in four kinds of echo hiding.