Image data hiding with multi-scale autoencoder network

Image steganography is the process of hiding information which can be text, image, or video inside a cover image. The advantage of steganography over cryptography is that the intended secret message does not attract attention and is thus more suitable for secret communication in a highly-surveillant environment such as civil disobedience movements. Internet memes in social media and messaging apps have become a popular culture worldwide, so this folk custom is a good application scenario for image steganography. We try to explore and adopt the steganography techniques on the Internet memes in this work. We implement and improve the HiDDeN model [22] by changing the Conv-BN-ReLU blocks convolution layer with a multiscale autoencoder network so that the neural network learns to embed message bits in higherlevel feature space. Compared to methods that convolve feature filters on the row-pixel domain, our proposed MS-Hidden network learns to hide secrets in both low-level and high-level image features. As a result, the proposed model significantly reduces the bit-error rate to empirically 0% and the required network parameters are much less than the HiDDeN model. Introduction When two parties want to communicate with each other and preserve their privacy securely, the most commonly used strategy is data encryption. The data is converted into ciphertext using cryptography algorithms. The original message is not readable after encryption, but the ciphertext is visible to human eyes, leading to suspicion and further scrutiny. The advantage of steganography over cryptography is that the intended secret message does not attract attention. The visible encrypted messages, no matter how unbreakable they are, arouse interest and may in themselves be criminal in some countries [3]. Even worse, the two parties may be captured by an untrusted leagal authority in a civil disobedience movement [1] and being tortured with Rubber hose cryptanalysis [2]. Messsaging Apps In the Mobile computing and Mobile Internet age, people with a highly cost-effective mobile phone can easily communicate with friends and families using instant messaging apps (such as Facebook Messenger and LINE). The modern messaging apps connect users to social groups and enable users to send messages instantly in all kinds of multimedia formats. The richness of images or photos significantly improves the user experience during communication, and thus, sending interesting photos becomes a dominant user behavior in messaging activities. Figure 1 shows two examples of Internet memes shared among western and Asian societies. This kind of social phenomenon has become a folk custom, and the spread of Internet memes in messaging apps tends to be ignored by the ill-intentioned authority, which is a perfect application scenario for steganography.