Provably Secure Steganography on Generative Media

In this paper, we propose provably secure steganography on generative media. Firstly, we discuss the essence of the steganographic security, which is identical to behavioral security. The behavioral security implies that the generative media are suitable for information hiding as well. Based on the duality of source coding and generating discrete distribution from fair coins and the explicit probability distribution yielded by generative model, perfectly secure steganography on generative media is proposed. Instead of random sampling from the probability distribution as ordinary generative models do, we combine the source decoding into the process of generation, which can implement the sampling according to the probability distribution as well as embed the encrypted message. Adaptive Arithmetic Coding is selected as the source coding method, and it is proved theoretically that the proposed generative steganography framework using adaptive Arithmetic Coding is asymptotically perfect secure. Taking text-to-speech system based on WaveNet as an instance, we describe the process of embedding and extracting message in detail, and the experimental results show that the proposed method is nearly perfectly secure when resists state-of-the-art steganalysis.