Y. Z. Yang, X. W. Yi, K. Yang, X. F. Zhao
The existing non-adaptive MP3 steganography methods generally have the problems of weak security. Therefore, an adaptive MP3 steganography method is proposed in this paper. In order to avoid the influence of frame offset in MP3 encoder, the algorithm embeds hidden information by modifying symbol bits of entropy coding, so as to ensure the length of the encoded stream unchanged.The algorithm using psychoacoustic model II to establish a mathematical model of the masking effect, and proposes a method to construct distortion cost function which is based on the auditory sensitivity coefficient and the modified magnitude. Finally, we use STC to adaptively embed secret information into the coefficients sign bits to minimize the distortion caused by steganography. The results of experiment show that the proposed algorithm has better performance in terms of imperceptibility and secure payload. With the steganalysis accuracy is less than 70%, the maximum embedding capacity can reach 13kbps.
Cite this paper as:
Y. Z. Yang, X. W. Yi, K. Yang, X. F. Zhao. Adaptive MP3 Steganography Based on Sign Bit Modification. In Proc. 14th China Information Hiding Workshop (CIHW 2018), Mar.31-Apr.1, 2018, Guangzhou, China.