Raphaël Couturier

Stéphane Cuenat, Raphaël Couturier

In Digital Holography (DH), it is crucial to extract the object distance from a hologram in order to reconstruct its amplitude and phase. This step is called auto-focusing and it is conventionally solved by first reconstructing a stack of images and then by sharpening each reconstructed image using a focus metric such as entropy or variance. The distance corresponding to the sharpest image is considered the focal position. This approach, while effective, is computationally demanding and time-consuming. In this paper, the determination of the distance is performed by Deep Learning (DL). Two deep learning (DL) architectures are compared: Convolutional Neural Network (CNN) and Vision Transformer (ViT). ViT and CNN are used to cope with the problem of auto-focusing as a classification problem. Compared to a first attempt [11] in which the distance between two consecutive classes was 100$μ$m, our proposal allows us to drastically reduce this distance to 1$μ$m. Moreover, ViT reaches similar accuracy and is more robust than CNN.

Alain Lalande, Zhihao Chen, Thibaut Pommier et al.

A key factor for assessing the state of the heart after myocardial infarction (MI) is to measure whether the myocardium segment is viable after reperfusion or revascularization therapy. Delayed enhancement-MRI or DE-MRI, which is performed several minutes after injection of the contrast agent, provides high contrast between viable and nonviable myocardium and is therefore a method of choice to evaluate the extent of MI. To automatically assess myocardial status, the results of the EMIDEC challenge that focused on this task are presented in this paper. The challenge's main objectives were twofold. First, to evaluate if deep learning methods can distinguish between normal and pathological cases. Second, to automatically calculate the extent of myocardial infarction. The publicly available database consists of 150 exams divided into 50 cases with normal MRI after injection of a contrast agent and 100 cases with myocardial infarction (and then with a hyperenhanced area on DE-MRI), whatever their inclusion in the cardiac emergency department. Along with MRI, clinical characteristics are also provided. The obtained results issued from several works show that the automatic classification of an exam is a reachable task (the best method providing an accuracy of 0.92), and the automatic segmentation of the myocardium is possible. However, the segmentation of the diseased area needs to be improved, mainly due to the small size of these areas and the lack of contrast with the surrounding structures.

Raphaël Couturier, Hassan N. Noura, Ola Salman et al.

Clustering is an unsupervised machine learning method grouping data samples into clusters of similar objects. In practice, clustering has been used in numerous applications such as banking customers profiling, document retrieval, image segmentation, and e-commerce recommendation engines. However, the existing clustering techniques present significant limitations, from which is the dependability of their stability on the initialization parameters (e.g. number of clusters, centroids). Different solutions were presented in the literature to overcome this limitation (i.e. internal and external validation metrics). However, these solutions require high computational complexity and memory consumption, especially when dealing with big data. In this paper, we apply the recent object detection Deep Learning (DL) model, named YOLO-v5, to detect the initial clustering parameters such as the number of clusters with their sizes and centroids. Mainly, the proposed solution consists of adding a DL-based initialization phase making the clustering algorithms free of initialization. Two model solutions are provided in this work, one for isolated clusters and the other one for overlapping clusters. The features of the incoming dataset determine which model to use. Moreover, The results show that the proposed solution can provide near-optimal clusters initialization parameters with low computational and resources overhead compared to existing solutions.

Hassan N. Noura, Ola Salman, Raphaël Couturier

Given the voluminous nature of the multimedia sensed data, the Multimedia Internet of Things (MIoT) devices and networks will present several limitations in terms of power and communication overhead. One traditional solution to cope with the large-size data challenge is to use lossy compression. However, current lossy compression schemes require low compression rate to guarantee acceptable perceived image quality, which results in a low reduction of the communicated data size and consequently a low reduction in the energy and bandwidth consumption. Thus, an efficient compression solution is required for striking a good balance between data size (and consequently communication overhead) and visual degradation. In this paper, a Deep-Learning (DL) super-resolution model is applied to recuperate high quality images (at the application server side) given as input degraded images with a high compression ratio (at the sender side). The experimental analysis shows the effectiveness of the proposed solution in enhancing the visual quality of the compressed and down-scaled images. Consequently, the proposed solution reduces the overall communication overhead and power consumption of limited MIoT devices.

Hassan Noura, Lama Sleem, Raphaël Couturier

Lately, multimedia encryption has been the focus of attention in many researches. Recently, a large number of encryption algorithms has been presented to protect image contents.The main objective of modern image encryption schemes is to reduce the computation complexity in order to respond to the real time multimedia and/or limited resources requirements without degrading the high level of security. In fact, most of the recent solutions are based on the chaotic theory. However, the majority of chaotic systems suffers from different limitations and their implementation is difficult at the hardware level because of the non integer operations that are employed requiring huge resources and latency. In this paper, we analyze the new chaos-based image encryption system presented in~\cite{el2016new}. It uses a static binary diffusion layer, followed by a key dependent bit-permutation layer that only iterates for one round. Based on their results in this paper, we claim that the uniformity and avalanche effect can be reached from the first round. However, we tried to verify the results but our conclusion was that these results were wrong because it was shown that at least 6 iterations are necessary to ensure the required cryptographic performance such as the plain-sensitivity property. Therefore, the required execution time must be multiplied by 6 and consequently this will increase the latency. In addition to all aforementioned problems, we find that ensuring the avalanche effect in the whole image introduces a high error propagation. In order to solve this problem, we recommend to ensure the avalanche effect in the level of blocks instead of the whole image.

Jean-Francois Couchot, Raphaël Couturier, Michel Salomon

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.

Jean-François Couchot, Raphaël Couturier, Yousra Ahmed Fadil et al.

Steganography schemes are designed with the objective of minimizing a defined distortion function. In most existing state of the art approaches, this distortion function is based on image feature preservation. Since smooth regions or clean edges define image core, even a small modification in these areas largely modifies image features and is thus easily detectable. On the contrary, textures, noisy or chaotic regions are so difficult to model that the features having been modified inside these areas are similar to the initial ones. These regions are characterized by disturbed level curves. This work presents a new distortion function for steganography that is based on second order derivatives, which are mathematical tools that usually evaluate level curves. Two methods are explained to compute these partial derivatives and have been completely implemented. The first experiments show that these approaches are promising.

Yousra A. Fadil, Jean-François Couchot, Raphaël Couturier et al.

Steganography and steganalysis are two important branches of the information hiding field of research. Steganography methods consist in hiding information in such a way that the secret message is undetectable for the uninitiated. Steganalyzis encompasses all the techniques that attempt to detect the presence of such hidden information. This latter is usually designed by making classifiers able to separate innocent images from steganographied ones according to their differences on well-selected features. We wonder, in this article whether it is possible to construct a kind of universal steganalyzer without any knowledge regarding the steganographier side. The effects on the classification score of a modification of either parameters or methods between the learning and testing stages are then evaluated, while the possibility to improve the separation score by merging many methods during learning stage is deeper investigated.

Jean-François Couchot, Raphaël Couturier, Christophe Guyeux et al.

For the past few years, in the race between image steganography and steganalysis, deep learning has emerged as a very promising alternative to steganalyzer approaches based on rich image models combined with ensemble classifiers. A key knowledge of image steganalyzer, which combines relevant image features and innovative classification procedures, can be deduced by a deep learning approach called Convolutional Neural Networks (CNN). These kind of deep learning networks is so well-suited for classification tasks based on the detection of variations in 2D shapes that it is the state-of-the-art in many image recognition problems. In this article, we design a CNN-based steganalyzer for images obtained by applying steganography with a unique embedding key. This one is quite different from the previous study of {\em Qian et al.} and its successor, namely {\em Pibre et al.} The proposed architecture embeds less convolutions, with much larger filters in the final convolutional layer, and is more general: it is able to deal with larger images and lower payloads. For the "same embedding key" scenario, our proposal outperforms all other steganalyzers, in particular the existing CNN-based ones, and defeats many state-of-the-art image steganography schemes.