Steganographic visual story with mutual-perceived joint attention

Abstract Social media plays an increasingly important role in providing information and social support to users. Due the easy dissemination of content, as well difficulty track on network, we are motivated study way concealing sensitive messages this channel with high confidentiality. In paper, design a steganographic visual stories generation model that enables users automatically post stego status without any direct user intervention use mutual-perceived joint attention (MPJA) maintain imperceptibility text. We demonstrate our approach storytelling (VIST) dataset show it yields high-quality texts. Since proposed work realizes steganography by auto-generating story using deep learning, us move real-world online networks intelligent bots.