Concealment Charm ( ConcealGAN ): Automatic Generation of Steganographic Text Using Generative Models to Bypass Censorship
Nurpeiis Baimukan, Quanyan Zhu · 2021
The ever-growing demand for messaging services has raised concerns over users' privacy when exchanging texts with each other. We show how those concerns are justified by summarizing the research in digital censorship. Further, we propose our linguistic steganographic system called ConcelGAN that aims to reduce the detectability of cover text and increase the difficulty of extraction by employing generative models, such as LSTM and LeakGAN, and by proposing double layer of encoding. Besides proposing our system, this chapter aims to promote research in linguistic steganography by showing the practical benefits of this field in cybersecurity.