BASN—Learning Steganography with a Binary Attention Mechanism

Abstract

Secret information sharing through image carriers has aroused much research attention in recent years with images’ growing domination on the Internet and mobile applications. The technique of embedding secret information in images without being detected is called image steganography. With the booming trend of convolutional neural networks (CNN), neural-network-automated tasks have been embedded more deeply in our daily lives. However, a series of wrong labeling or bad captioning on the embedded images has left a trace of skepticism and finally leads to a self-confession like exposure. To improve the security of image steganography and minimize task result distortion, models must maintain the feature maps generated by task-specific networks being irrelative to any hidden information embedded in the carrier. This paper introduces a binary attention mechanism into image steganography to help alleviate the security issue, and, in the meantime, increase embedding payload capacity. The experimental results show that our method has the advantage of high payload capacity with little feature map distortion and still resist detection by state-of-the-art image steganalysis algorithms.

A Comprehensive study of digital Image Steganography Techniques

“…They built a novel encoding and decoding algorithm in Matlab that outperforms the old approach in terms of performance evaluation [21]. Wu et al (2020) proposed an improved security of image steganography with the least amount of distortion; these models must keep the feature maps created by task-specific networks independent of any concealed information buried in the carrier. To help alleviate the security issue, this paper incorporates a binary attention method into picture steganography to aid enhance embedding payload capacity, and in the meantime [22].…”

“…Wu et al (2020) proposed an improved security of image steganography with the least amount of distortion; these models must keep the feature maps created by task-specific networks independent of any concealed information buried in the carrier. To help alleviate the security issue, this paper incorporates a binary attention method into picture steganography to aid enhance embedding payload capacity, and in the meantime [22]. Kumar et al (2020) suggested a steganalytic method based on artificial intelligence that uses deep features extracted from stego pictures using deep CNN techniques.…”

Steganography in color images with random order of pixel selection and encrypted text message embedding

“…It is one of the most effective approaches to protecting important information by hiding it (data hiding). High-tech steganography summarizes the areas for hiding messages using communication and computer technology, nanotechnology and modern advances in sciences such as biology, chemistry and others (Wu et al, 2020;Koptyra & Ogiela, 2020;Abd-El-Atty et al, 2020).…”

High-Capacity Image Steganography Based on Improved Xception

“…Chen et al [ 11 ] introduced advanced ideas, such as skip connection and feature fusion, and designed a novel two-channel deep hidden network (TDHN). Wu et al [ 12 ] introduced an attention mechanism in steganography to alleviate security issues and increase the load capacity.…”