In the past few years, Generative Adversarial Network (GAN), which proposed in 2014, has achieved great success. There have been increasing research achievements based on GAN field of computer vision and natural language processing. Image steganography is an information security technique aiming at hiding secret messages common digital images for covert communication. Recently, image demonstrated potential by introducing other neural network techniques. this paper, we review art with GANs...

This paper proposes a fully homomorphic encryption encapsulated difference expansion (FHEE-DE) scheme for reversible data hiding in encrypted domain (RDH-ED). The circuits and ciphertext operations are elaborated. Key-switching bootstrapping techniques introduced to control the extension decryption failure of encryption. A key-switching based least-significant-bit (KS-LSB) method has been designed realize extraction directly from without private key. In application, user first encrypts...

In this paper, a novel data-driven information hiding scheme called "generative steganography by sampling" (GSS) is proposed. Unlike in traditional modification-based steganography, our method the stego image directly sampled powerful generator: no explicit cover used. Both parties share secret key used for message embedding and extraction. The Jensen-Shannon divergence introduced as new criterion evaluating security of generative steganography. Based on these principles, we propose simple...

Schemes of reversible data hiding in encrypted domain (RDH-ED) based on symmetric or public key encryption are mainly applied the scenarios end-to-end communication. To provide security guarantees for multi-party scenarios, a RDH-ED scheme secret image sharing Chinese remainder theorem (CRT) is presented. In application (<inline-formula> <tex-math notation="LaTeX">$t$ </tex-math></inline-formula>, <inline-formula> notation="LaTeX">$n$ </tex-math></inline-formula>) sharing, an first shared...

Traditional image steganography modifies the content of more or less, it is hard to resist detection steganalysis tools. To address this problem, a novel method named generative coverless information hiding based on adversarial networks proposed in paper. The main idea that class label replaced with secret as driver generate hidden directly, and then extract from through discriminator. It's first time achieved by networks. Compared traditional steganography, does not modify original image....

This paper focuses on several theoretical issues and principles in steganography security, defines four security levels by analyzing the corresponding algorithm instances. In analysis, we discuss differences between watermarking security. The two necessary conditions for are obtained. Under current technology situation, then analyze indistinguishability of cover stego-cover, consider that should rely key secrecy with algorithms open. By specifying role steganography, a secure theory formally...

Existing deep-learning-based image steganalysis networks have problems such as large model sizes, significant runtime memory usage, and extensive computational operations, which hinder their deployment in many practical applications. To address these challenges, we applied compression techniques to designed a called SASRNet, slimming-assisted residual network. We observed that the trainable scale factor of BN (batch normalization) layer network can be used channel scaling for pruning. The...

In this paper, we analyze the video steganography technique, which is used to ensure national security and confidentiality of information governmental agencies enterprises. Videos may be transmit secrets conduct covert communication. As such, present an algorithm based on a secret sharing scheme Error-Correcting Code (ECC), combines Grey Relational Analysis (GRA) with partition mode in compression standard H.264/AVC. First, process by sharing, then use ECC obtained information. Moreover,...

Based on the H.264/AVC Video coding standard, a new video steganography algorithm is proposed and realized in this paper. The designed motion vector component feature to control embedding, also be secret carrier. information embedded will not significantly affect sequence's visual invisibility statistical invisibility. Experiments show that has large embedding capacity with high carrier utilization, can implement fast effectively.

NTRU (Number Theory Research Unit) has the characteristics of resistance to quantum computing attacks, fast encryption and decryption, high security. It is very suitable for wireless confidential data networks authentication systems. Combined with reversible hiding technology, a separable scheme in homomorphic encrypted domain based on proposed. The image owner directly divides cover-image into groups reference pixel T adjacent pixels. Then, grouped by NTRU. Finally, uploaded hider. divided...