Yedrouj-Net: An efficient CNN for spatial steganalysis Results Conclusions • An efficient approach based on deep learning (CNN) for steganalysis. • Our method outperforms the state-of-the-art and others CNN-based models with and without taking extra measu (original) (raw)
Related papers
Yedroudj-Net: An Efficient CNN for Spatial Steganalysis
2018
For about 10 years, detecting the presence of a secret message hidden in an image was performed with an Ensemble Classifier trained with Rich features. In recent years, studies such as Xu et al. have indicated that well-designed convolutional Neural Networks (CNN) can achieve comparable performance to the two-step machine learning approaches. In this paper, we propose a CNN that outperforms the state-ofthe-art in terms of error probability. The proposition is in the continuity of what has been recently proposed and it is a clever fusion of important bricks used in various papers. Among the essential parts of the CNN, one can cite the use of a pre-processing filterbank and a Truncation activation function, five convolutional layers with a Batch Normalization associated with a Scale Layer, as well as the use of a sufficiently sized fully connected section. An augmented database has also been used to improve the training of the CNN. Our CNN was experimentally evaluated against S-UNIWARD and WOW embedding algorithms and its performances were compared with those of three other methods: an Ensemble Classifier plus a Rich Model, and two other CNN steganalyzers.
Convolutional Neural Network-based Steganalysis on Spatial Domain
Steganalysis has been studied to detect the existence of hidden messages by steganography. However, human intervention is required to determine the flaws of the steganography and it is time consuming. This paper presents a steganalysis method using a convolutional neural network for spatial domain steganography whereby there is no human intervention is required. We have designed a convolutional neural network-based steganalysis model to have 5 convolutional layers and 2 full connected layers. Especially, binarized differential filter and high pass filter are applied to extract hidden messages. After the model is trained with cover images and LSB-based stego-images, unknown images are tested to determine if secret messages have been embedded. Experiments are performed using BOSS and SIPI database and the presented models show over 99% and 96% accuracy for stego-images with the same key and different keys.
PeerJ Computer Science, 2021
In recent years, Deep Learning techniques applied to steganalysis have surpassed the traditional two-stage approach by unifying feature extraction and classification in a single model, the Convolutional Neural Network (CNN). Several CNN architectures have been proposed to solve this task, improving steganographic images’ detection accuracy, but it is unclear which computational elements are relevant. Here we present a strategy to improve accuracy, convergence, and stability during training. The strategy involves a preprocessing stage with Spatial Rich Models filters, Spatial Dropout, Absolute Value layer, and Batch Normalization. Using the strategy improves the performance of three steganalysis CNNs and two image classification CNNs by enhancing the accuracy from 2% up to 10% while reducing the training time to less than 6 h and improving the networks’ stability.
GBRAS-Net: A Convolutional Neural Network Architecture for Spatial Image Steganalysis
IEEE Access, 2021
Advances in Deep Learning (DL) have provided alternative approaches to various complex problems, including the domain of spatial image steganalysis using Convolutional Neural Networks (CNN). Several CNN architectures have been developed in recent years, which have improved the detection accuracy of steganographic images. This work presents a novel CNN architecture which involves a preprocessing stage using filter banks to enhance steganographic noise, a feature extraction stage using depthwise and separable convolutional layers, and skip connections. Performance was evaluated using the BOSSbase 1.01 and BOWS 2 datasets with different experimental setups, including adaptive steganographic algorithms, namely WOW, S-UNIWARD, MiPOD, HILL and HUGO. Our results outperformed works published in the last few years in every experimental setting. This work improves classification accuracies on all algorithms and bits per pixel (bpp), reaching 80.3% on WOW with 0.2 bpp and 89.8% on WOW with 0.4 bpp, 73.6% and 87.1% on S-UNIWARD (0.2 and 0.4 bpp respectively), 68.3% and 81.4% on MiPOD (0.2 and 0.4 bpp), 68.5% and 81.9% on HILL (0.2 and 0.4 bpp), 74.6% and 84.5% on HUGO (0.2 and 0.4 bpp), using BOSSbase 1.01 test data.
IFIP Advances in Information and Communication Technology, 2017
Conventional state-of-the-art image steganalysis approaches usually consist of a classifier trained with features provided by rich image models. As both features extraction and classification steps are perfectly embodied in the deep learning architecture called Convolutional Neural Network (CNN), different studies have tried to design a CNN-based steganalyzer. The network designed by Xu et al. is the first competitive CNN with the combination Spatial Rich Models (SRM) and Ensemble Classifier (EC) providing detection performances of the same order. In this work we propose a criterion to choose either the CNN or the SRM+EC method for a given input image. Our approach is studied with three different steganographic spatial domain algorithms: S-UNIWARD, MiPOD, and HILL, using the Tensorflow computing platform, and exhibits detection capabilities better than each method alone. Furthermore, as SRM+EC and the CNN are both only trained with a single embedding algorithm, namely MiPOD, the proposed method can be seen as an approach for blind steganalysis. In blind detection, error rates are respectively of 16% for S-UNIWARD, 16% for MiPOD, and 17% for HILL on the BOSSBase with a payload of 0.4 bpp. For 0.1 bpp, the respective corresponding error rates are of 39%, 38%, and 41%, and are always better than the ones provided by SRM+EC.
Sensitivity of deep learning applied to spatial image steganalysis
2021
In recent years, the traditional approach to spatial image steganalysis has shifted to deep learning (DL) techniques, which have improved the detection accuracy while combining feature extraction and classification in a single model, usually a convolutional neural network (CNN). The main contribution from researchers in this area is new architectures that further improve detection accuracy. Nevertheless, the preprocessing and partition of the database influence the overall performance of the CNN. This paper presents the results achieved by novel steganalysis networks (Xu-Net, Ye-Net, Yedroudj-Net, SR-Net, Zhu-Net, and GBRAS-Net) using different combinations of image and filter normalization ranges, various database splits, different activation functions for the preprocessing stage, as well as an analysis on the activation maps and how to report accuracy. These results demonstrate how sensible steganalysis systems are to changes in any stage of the process, and how important it is fo...
A Review on Deep Learning Solutions for Steganalysis
International Journal of Computing and Digital Systems
Steganalysis methods have developed to attack steganography, a technique used to hide secret information in a digital media. The traditional way of steganalysis is performed as feature extraction followed by classification. With the popularity of Deep Learning (DL) in the field of computer vision, researchers started applying deep learning for steganalysis problems also. Soon they found promising results with DL as it automates the feature extraction step and classification results can be used to better learn the features. Thus, the tedious task of manual extraction of features with a separate classification step is unified in deep learning giving optimistic results. This work provides a better insight into steganalysis evolution using deep learning and provides a broad review on how researchers have successfully applied Convolutional Neural Network (CNN) by using steganalysis specific activation functions, different convolutional layers and others. Researchers have compared their results with each other as well as state-of-the-art before deep learning (Rich Models + Ensemble Classifier). Initially, CNNs were created from scratch in the field of steganalysis but later researchers moved to highly efficient pretrained networks such as SRNet, ResNet and EfficientNet and found significant improvement in results on more challenging datasets such as ALASKA-I and ALASKA-II. The reason for such improvement is that pretrained networks are already trained on a very large dataset of images for some classification tasks and thus can be finetuned easily to other classification tasks with improved results.
Deep learning in steganography and steganalysis
Digital Media Steganography, 2020
For almost 10 years, the detection of a hidden message in an image has been mainly carried out by the computation of Rich Models (RM), followed by classification using an Ensemble Classifier (EC). In 2015, the first study using a convolutional neural network (CNN) obtained the first results of steganalysis by Deep Learning approaching the performances of the two-step approach (EC + RM). Between 2015-2018, numerous publications have shown that it is possible to obtain improved performances, notably in spatial steganalysis, JPEG steganalysis, Selection-Channel-Aware steganalysis, and in quantitative steganalysis. This chapter deals with deep learning in steganalysis from the point of view of current methods, by presenting different neural networks from the period 2015-2018, that have been evaluated with a methodology specific to the discipline of steganalysis. The chapter is not intended to repeat the basic concepts of machine learning or deep learning. So, we will present the structure of a deep neural network, in a generic way and present the networks proposed in existing literature for the different scenarios of steganalysis, and finally, we will discuss steganography by deep learning.
Histogram Layer, Moving Convolutional Neural Networks Towards Feature-Based Steganalysis
Feature-based steganalysis has been an integral tool for detecting the presence of steganography in communication channels for a long time. In this paper, we explore the possibility to utilize powerful optimization algorithms available in convolutional neural network packages to optimize the design of rich features. To this end, we implemented a new layer that simulates the formation of histograms from truncated and quantized noise residuals computed by convolution. Our goal is to show the potential to compactify and further optimize existing features, such as the projection spatial rich model (PSRM).
IET Image Processing, 2020
Steganography, a branch of data hiding techniques aims to hide confidential information within any digital media by obscuring the existence of hidden information. On the contrary, steganalysis aims to detect steganography. With the advent of powerful steganographic algorithms, the process of cracking them became very challenging. Traditional steganalysis following machine learning principle employs a two-step process, with first process extracting highly sophisticated features capable of discriminating hidden message from original data and second process classifying the input as innocent or guilty. In recent years, deep learning which has its roots in artificial neural networks emerged as a brilliant alternative for many computer vision tasks. A review of recent research works in deep learning based digital image steganalysis is presented here. The paradigm shift from machine learning approaches to employing more promising deep learning architectures, observed with the current research community and hence in literature has been presented here in chronological order. Deep learning can unify the two-step process into a single process by giving the ability for machine to learn end-to-end by itself. The use of Convolutional Neural Networks to perform steganalysis in spatial or transform or combination of both domains has effectively lowered the detection error rates. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.