The Second International Conference on Computer Science's Complex Systems and their Applications (Oum El Bouaghi, Algeria, May 25-26, 2021)
Permanent URI for this collection
Browse
Recent Submissions
Item Using Association Rules for Ontology Enrichment(University of Oum El Bouaghi, 2021-05-25) Benali, KhaledAn ontology is a formal description of knowledge as a set of concepts within a domain and the relationships that hold between them. At the same time, data mining techniques are used to discover hidden structures in large databases. In particular, Association Rules are used to discover implicative trends among items in a transactional database. In this context, we propose to develop a method to enrich existing ontologies with the identification of new semantic relationships between concepts in order to have a better coverage of domain knowledge. The enrichment process is realized by association rules discovered by applying the Apriori algorithm. We demonstrate the applicability of this method using an existing ontology.Item Transfer Learning Using VGG Based on Deep Convolutional Neural Network For Finger-Knuckle-Print Recognition(University of Oum El Bouaghi, 2021-05-25) Hamidi, Amira; Khemgani, Salma; Bensid, KhledTransfer learning is an example of Convolutional Neural Network (CNN) method. It based to reusing a pretrained model knowledge for another task. which used for image classification, feature extraction, and clustering problems. In this paper, we used two types of the pre-trained models VGG–16 and VGG-19 with deep convolutional neural network to extract the features of Finger-Knuckle-Print FKP images in order to develop an efficient multimodal identification system. The results obtained in this work show an excellent performance for unimodal and multimodal identification systems.Item Training Cellular Automata with Extended Neighborhood for Edge Detection(University of Oum El Bouaghi, 2021-05-25) Safia, DjemameEdge detection refers to the process of identifying and locating sharp discontinuities in an image. Since edge detection is in the forefront of image processing for object detection, it has attracted much attention from scientific research. More accurate results and less time consuming are there still the main issues when extracting edges from images. To cope with this challenge, we propose a complex system: Cellular Automata (CA) that has proven high performances in image processing domain. Unlike previous works, which used in majority Von Neumann or Moore neighborhood, We use a particular kind of CA, with extended Moore neighborhood. This allows a large exploration of the search space. We trained a QPSO algorithm for extracting the adequante subset of rules. Experiments were carried on several images from Mathworks and Berkeley dataset. Visual and numerical results show that our CA provides excellent performances, and edges with high accuracy.Item Towards emotion recognition in immersive virtual environments: A method for Facial emotion recognition(University of Oum El Bouaghi, 2021-05-25) Amara, Kahina; Ramzan, Naeem; Zenati, Nadia; Djekoune, Oualid; Larbes, Cherif; Guerroudji, Mohamed Amine; Aouam, DjamelVirtual Reality (VR) is, thus, proposed as a powerful tool to simulate complex, real situations and environments, offering researchers unprecedented opportunities to investigate human behaviour in closely controlled designs in controlled laboratory conditions. Facial emotion recognition has attracted a great deal of interest for interaction in virtual reality, healthcare system: therapeutic applications, surveillance video application etc. In this paper, we propose a method for facial emotion recognition for immersive virtual environment based on 2D and 3D geometrical features. We used our collected dataset of 17 subjects’ performance of six basic facial emotions (anger, fear, happiness, surprise, sadness, and neutral) using three devices: Kinect (v1), Kinect (v2), and RGB HD camera. In addition, we present the performance results of the RGB data for facial emotion recognition using Bagged Trees algorithm. To assess the performance of the proposed system, we used leave-oneout- subject cross-validation. We compared the 2D and 3D data performance for facial expression recognition. The obtained results show the superior performance of the RGB-D features provided by Kinect (v2). Our findings highlight that the 2D images are not robust enough for facial emotion recognition. The built facial emotion models will animate virtual characters that can express emotions via facial expressions. This could be deployed for Chatting, Learning and Therapeutic Intervention.Item SIRATa : a Real-Time Indexing Arabic Text Editor Based on the Extraction of Keywords(University of Oum El Bouaghi, 2021-05-25) Dilekh, Tahar; Benharzallah, Saber; Mokeddem, AyoubIndexing stage in information retrieval process has a great importance as an essential tool for the performance of recall and precision. Despite the many studies that have been done on the indexing conducted in the last few decades, to our knowledge, no study has investigated whether indexing realtime based on keywords extraction is efficient to perform of recall and precision. Moreover, relatively fewer Arabic text indexing studies are currently available despite the enormous efforts put together to satisfy the needs of the growing number of Arabic internet users. This paper suggests a method for Arabic text indexing based on keywords extraction. The proposed method consists of two stages. The first stage conducts a real-time indexing. The second stage is a keywords extraction and updating of initial index taking into account the output of keywords extraction process. We illustrate application and the performance of this method of indexing using an Arabic text editor (SIRAT) developed and designed for this aim. We also illustrate the process of building a new form of Arabic corpus appropriate to conduct the necessary experiments. Our findings show that SIRAT successfully identifies the keywords most relevant to the document. Finally, the main contribution of this experiment is to demonstrate the effectiveness of this method compared to other methods. In addition, the paper proposes a solution to issues and deficiencies Arabic language processing suffers from, especially regarding corpora building and keywords extraction evaluation systems.Item Robust characteristics for texture classification(University of Oum El Bouaghi, 2021-05-25) Maarouf, Abderrazak Ayoub; Hachouf, FellaIn this paper, an exhaustive search for relevant characteristics for automatic texture classification has been carried out. These features have been extracted from different cooperative methods dealing with texture characterization. An optimal features vector has been constructed using genetic algorithms (GA) to avoid characteristics redundancy . Then texture classification has been performed using multi-class SVM, k-nearest neighbors, and random forest classifier algorithms. Obtained results on three texture databases are very satisfying against those produced by existing methods.Item On the Drivers’ Behavior Evaluation using Vehicular Networks(University of Oum El Bouaghi, 2021-05-25) Bendouma, Tahar; Tahari, Abdou El Karim; Kerrache, Chaker Abdelaziz; Boukhelkhaly, Mama Chima; Bendoumay, Rekaia; Lagraa, NasreddineWith the emergence of connected and intelligent vehicles, various research projects aiming at reducing traffic accidents by detecting driver behavior have also emerged. These vehicles are generally equipped with cameras and sensors that can be used to detect driver’s fatigue, drowsiness, and distraction using different technologies and a multitude of classification techniques. In this work, we propose a new real-time driver behavior-detection technique based on vehicle-to-vehicle communication (V2V) and by exploiting the information carried by the periodically exchanged messages known as Cooperative Awareness Message (CAM) that are a part of the European ETSI-ITS standard (or Basic safety message BSM in the US standard). These information include the vehicle’s current speed, the average speed, the position, the acceleration, to name a few. In our proposal, each vehicle can classify its neighbors (normal, aggressive) according to its driver’s driving style. An audio or video message can be then generated to warn the driver of any vehicle presenting a danger. Simulations conduct in both rural and urban environments depict that our proposal called ”Vehicular Ad-Hoc Network Exchange Message (VanetExM)” can determine the state of the driver with a relatively high success rate and low overhead.Item NECS-based Cache Management in the Named Data Networking(University of Oum El Bouaghi, 2021-05-25) Fathallah, Nour Al Huda; Bouziane, Hafiza; Sharafiya, AbdullahThe Information-Centric Networking ICN architectures proposed to overcome the problems of the actual internet architecture. One of the main straight points of the ICN architectures is in-network caching. The effectiveness of the adopted caching strategy, which manages and decides where to store them, influences the performance of the ICN. However, the major issue that faces the caching strategies in the ICN architectures is the strategic selection of the cache routers to store the data through its delivery path. This will reduce congestion, optimize the distance between the consumers and the required data furthermore improve latency and alleviate the viral load on the servers. In this paper, we propose a new efficient caching strategy for the Named Data Networking architecture NDN named NECS, which is the most promising architecture between all the ICN architectures. The proposed strategy reduces the traffic redundancy, eliminates the useless replication of contents, and improves the replay time for users due to the strategic position of cache routers. Besides, we evaluate the performance of this proposed strategy and compare it with three other NDN caching strategies, using the simulator network environment NdnSIM. Based on the simulations carried out, we obtained interesting and convincing results.Item Navigation of a Differential Drive Mobile Robot Using Nonlinear Model Predictive Control(University of Oum El Bouaghi, 2021-05-25) Benchouche, Welid; Mellah, Rabah; Bennouna, Mohammed SalahIn this paper, an implementation of a very fast nonlinear model-based predictive controller using a newly developed open-source toolkit (CasADi) was used to attain the two control goals of differential drive mobile robots, point stabilization (regulation) and trajectory following (time-varying reference). The controller’s stability was assured by the addition of final state equality constraints, which in general require a long optimization horizon for feasibility. In the work presented here, we performed a full-scale simulation proving the applicability of the terminal stabilization equality constraint have been performed. The obstacle avoidance problem has been solved by adding the obstacle position as a constraint in the main optimal control problem.Item Modularity maximization to find community structure in complex networks(University of Oum El Bouaghi, 2021-05-25) Saoud, BilalComplex networks have in generally communities. These communities are very important. Network’s communities represent sets of nodes, which are very connected. In this research, we developed a new method to find the community structure in networks. Our method is based on flower pollination algorithm (FPA) witch is used in the splitting process. The splitting of networks in our method maximizes a function of quality called modularity. We provide a general framework for implementing our new method to find community structure in networks. We present the effectiveness of our method by comparison with some known methods on computer-generated and real-world networks.Item Lips Recognition for Biometric Identification Systems(University of Oum El Bouaghi, 2021-05-25) Boucetta, Aldjia; Boussaad, LeilaIn recent years, researches in biometric methods have gained much attention and they have advanced to a wide scope in security concepts. Therefore, many biometric technologies have been developed and enhanced with many of the most successful security applications. Lately, lip-based biometric identification becomes one of the most relevant emerging tools, which comes from criminal and forensic real-life applications. The main purpose of this paper is to prove the benefit of lips as a biometric modality, by using both handcraft and deeplearning based feature extraction methods. So, we consider three different techniques, Histogram of Oriented Gradients(HOG), Local Binary Pattern(LBP) and pretrained Deep-CNN. All results are confirmed by a ten-fold cross-validation method using two datasets, NITRLipV1 and database1. The mean accuracy is found to be very high in all the experiments carried out. Also the feature extraction using the Inceptionv3 model always achieve highest mean accuracy.Item Image denoising algorithms using norm minimization techniques(University of Oum El Bouaghi, 2021-05-25) Diffellah, Nacira; Bekkouche, Tewfik; Hamdini, RabahImage denoising is one of the fundamental image processing problems. Noise removal is an important step in the image restoration process. In this paper, firstly we develop and implement two different image denoising algorithms based on norm minimization, namely `1 and `2-regularization applied to images contaminated by gaussian noise. Then, after their discretization and implementation, we perform a comparison between the two methods using several test images. Through this study, the algorithm which minimizes `2-norm of gradient of image has a unique solution and it’s easy to implement, but it doesn’t accept contour discontinuities, causing the obtained solution to be smooth. The `2-norm will blur the edges of the image. In order to preserve sharp edges, `1-norm is introduced. There are different methods to solve the problem of energy minimization. In this work, we have chosen the discretization finite difference method before applying the gradient descent algorithm to optimize the signal (2D grayscale images) denoising functionality. Experiments results, show that `1 regularization encourages image smoothness while allowing for presence of jumps and discontinuities, a key feature for image processing because of the importance of edges in human vision.Item Evaluation of ANN, ICA-ANN and PSO-ANN predicting ability in the prediction of CO2 emissions during the calcination of cement raw material(University of Oum El Bouaghi, 2021-05-25) Boukhari, YakoubCement industry releases large amounts of carbon dioxide CO2 as by-product to the atmosphere during the calcination of cement raw material. In fact, the calcination is a complex process and not completely understood. The amount of CO2 emitted varies with the grain size, chemical composition, burning temperature and time to pass through the kiln during calcination process. However, due to interaction of several parameters, it is not easy to establish accurate mathematic model to calculate the real amount of CO2 emission. Moreover, using the laboratory experiments to determine the amount of CO2 emissions are not usually easy, time-consuming, expensive and require good quality of reagents and equipments. To overcome the above problems, artificial neural network (ANN), ANN optimised by imperialist competitive algorithm (ICAANN), ANN optimised by particle swarm optimization (PSOANN) are applied to predict amount of CO2 emissions. A comparative accuracy of these tools is evaluated based on the coefficient of determination R2, R2 adjusted, mean absolute percentage error (MAPE) and scatter index (SI). The results obtained are promising and demonstrate that all proposed tools represent a good alternative for the prediction of CO2 emission with adequate accuracy. PSO and ICA are capable to improve the predicting accuracy of ANN. In addition, PSO-ANN can predict slightly better than ICA-ANN. Based on testing data, the results obtained show that 98.61%, 98.18% and 97.5% of experimental data are explained by PSO-ANN, ICAANN and ANN, respectively with average relative error less than 1.41%and SI less than 0.1.Item Evaluation and comparison study of video streaming routing protocols in vehicular ad-hoc networks(University of Oum El Bouaghi, 2021-05-25) Zaidi, Sofiane; Ogab, Mostafa; Khamer, LazharVideo streaming is a challenging issue in Vehicular Ad-Hoc Networks (VANETs), due to the strict video streaming Quality of Service (QoS) requirements, such as throughput, delivery ratio, and transmission delay. Moreover, video streaming is influenced by VANET characteristics, such as the high dynamic topology, fluctuation of vehicle density, and environmental obstacles. In VANET, video streaming can be achieved through different VANET communication types, such as Vehicle to Vehicle (V2V), Vehicle to Infrastructure (V2I), and Vehicle to Broadband cloud (V2B). Based on these communications, the vehicles can exchange between them the video stream over single or multi-hop link. When the video content is delivered over a multi-hop link, the vehicles have to use a routing protocol to disseminate the video stream through a path (s) between the sender (s) end the receiver (s) vehicles. In this paper, we have presented an overview of popular existing routing protocols for video streaming in VANET, such as AODV, AOMDV, DSR, and DSDV. Furthermore, we have evaluated and compared these protocols in terms of some QoS evaluation metrics, such as throughput, packet delivery ratio, and end-to-end delay in function with vehicles density in order to judge which one is outperforming for video streaming in VANET. The simulation results show that the reactive routing protocols (AODV, AOMDV, DSR) provide higher throughput and packet delivery ratio than DSDV proactive routing protocol. However, DSDV achieves lower end-to-end delay than AODV, AOMDV, DSR routing protocolsItem Distributed Secure Services Based on IoT and Blockchain for e-Health remote care(University of Oum El Bouaghi, 2021-05-25) Reffad, Hamza; Djenaoui, Abdelatif; Alti, AdelNowadays, the Internet of Thing (IoT) is a potentially powerful solution for health applications. It is a smart technology that provides remote care in real time and requires low latency health data processing and transmission. The large number of connected objects to Cloud can be a problem for low-latency workloads, which is the case of several health mobile applications. To this end, Fog Computing, has emerged, where Cloud computing is extended to the edge of the network to reduce latency and network congestion. It provides a highly virtualized platform that provides health data storage on remote public Cloud servers to which users cannot be fully trusted, especially when we are dealing with sensitive data like health data. In fact, it becomes necessary to rethink a new more robust secure technique. To provide such technique, we proposed a new secure solution called IoToDChain for e-Health mobile application, based on cryptographic techniques especially Elliptic Curve Diffie Hellman-RSA and the Blockchain paradigm. They exchange of a secret key in confidential and robust manner and protect patients’ privacy in a mobile-Fog-Cloud environment. The experiments achieved promising results for good data protection against the most known attacks in healthcare systems.Item Deep Neural Transformer Model for Mono and Multi Lingual Machine Translation(University of Oum El Bouaghi, 2021-05-25) Khaber, Mohamed Islam; Frahta, Nabila; Moussaoui, Abdelouahab; Saidi, MohamedIn recent years, the Transformers have emerged as the most relevant deep architecture, especially machine translation. These models, which are based on attention mechanisms, outperformed previous neural machine translation architectures in several tasks. This paper proposes a new architecture based on the transformer model for the monolingual and multilingual translation system. The tests were carried out on the IWSLT 2015 and 2016 dataset. The Transformers attention mechanism increases the accuracy to more than 92% that we can quantify by more than 4 BLEU points (a performance metric used in machine translation systems).Item Compressed VGG16 Auto-Encoder for Road Segmentation from Aerial Images with Few Data Training(University of Oum El Bouaghi, 2021-05-25) Abdeldjalil, Kebir; Taibi, Mahmoud; Serradilla, Francisco; Spain, MadridDeep Learning methods have found many applications such as segmentation, recognition and classification. However, almost all of these methods require large data-set for the training step and a long training time. Indeed, in surveillance video domain, as for many real-world applications, samples are only accessible in limited amounts owing to acquisition and experiments complexity. In this work, we introduce compressed VGG Auto-Encoder system for road image segmentation in highresolution aerial imagery. The objective of our experiments is to improve the methodology of distinguishing the road network when only few Data is available. We propose an approach based on compressed Auto-encoder; focus on avoiding the over-fitting effect by generating new data augmentation, based on basic filter transformation to increase and enhance the quality of data training, in the aim of learn an appropriate and simplified representation of data from the original data set in order to obtain a deeper insight from large data-set, and to achieve a quick segmentation training time. Our model achieve a good result and is considered as the best network for fast and accurate segmentation of road images, compared to other models. Furthermore, we provide an explanation of these techniques and some recommendation for their use in the field of deep learning.Item Biometric Image Encryption Scheme based on Modified Double Random Phase Encoding System(University of Oum El Bouaghi, 2021-05-25) Yahi, Amina; Bekkouche, Tewfik; Diffellah, Nacira; Daachi, Mohamed El HossineIn this paper, an opto-digital encryption scheme based on a modified Double Random Phase Encoding (DRPE) system is proposed. Two biometric modalities are used in this work which is the face and the corresponding finger print of the same person. Firstly the face biometric image is encrypted chaotically using the permutation-diffusion architecture. Then obtained encrypted face is multiplied element by element by a constructed mask formed by injecting the finger print image within the phase of this mask. The obtained result will be transformed into a frequency domain by the two-dimensional. Fourier transform or any of its derivatives, resulting complex image is exactly the encrypted biometric image. Experiment computer simulations confirm the efficiency of this work in terms of histogram analysis, loss data and sensitivity test when compared with existing works.Item Behavioural verification of limited resources systems under true concurrency semantics(University of Oum El Bouaghi, 2021-05-25) Bouneb; Messaouda; Saidouni; Djamel EddineIn this paper we propose a true concurrency semantics for limited resources systems using K-bounded Petri net as modeling formalism and maximality labeled transition system (MLTS) as semantics model. Indeed the model of MLTS expresses clearly the semantics of true parallelism of concurrent systems. The proposed operational maximality semantics for Kbounded Petri nets makes it possible to interpret any K-bounded Petri net in terms of MLTS. Through an example we show the interest of the proposed semantics in comparison with the interleaving semantics and the ST semantics. The comparison concerns the preservation of true concurrency and the reduction of the size of the semantics model. Furthermore, we will show that expected CTL properties may be verified on the corresponding maximality labeled transition system of a modeled system using our developed tool.Item Bandwidth Provision through Disjoint Multipath RPL in the IoMT(University of Oum El Bouaghi, 2021-05-25) Kettouche, Souhila; Maimour, Moufida; Derdouri,LakhdarInternet of Multimedia Things (IoMT) is one extensively current topic of the Internet of Things (IoT) due to the immersive growth of multimedia applications in several fields. In LowPower and Lossy Networks (LLNs) where sensor nodes are a key component, providing a satisfactory quality of service (QoS) as well as a user quality of experience (QoE) for such applications is a challenging task. In fact, high bandwidth and substantial computation resources are required. To provide sufficient bandwidth to handle these high data rate applications, we propose to extend RPL to enable for simultaneous use of disjoint multiple paths. This is done on top of the already maintained DODAG structure with the least induced overhead. Furthermore, we suggest applying a low-complexity encoding method on the captured images. Based on both QoS and QoE metrics, we evaluate the performance of our disjoint multipath RPL (DM-RPL) for real video clip transmission using the IoTLAB testbed. Our results show that multipath provides more bandwidth as the PDR is increased. Video quality is further improved thanks to the adopted data reduction at the source. All of this translates into less energy being consumed.