The instruments of advertising on the Internet are increasing. Advertisements may be a source of inconvenience and threats. This paper presents a comparison among three different tools for blocking advertisements, namely parental control tools, AdBlock Plus, and Pi-hole. This comparison aims to evaluate the effectiveness of these tools in blocking advertisements. The result showed that the Pi-hole is the most effective tool to block advertisements on the entire home network, while the other tools vary in the level of blocking advertisements on the network.
Participatory Budget (PB) is a process that distributes part of the city budgets among projects submitted and selected by the dwellers. The key challenge for IT-supported e-PB is the projects’ comparison and ranking. This paper focuses on an empirical test of a hybrid method for comparing BP projects. In this study, we investigate two difficult to measure dimensions: beneficiaries and categories. We use ontology to describe and map distances between concept and then generate ranking based on the fuzzy TOPSIS method. The method is validated through experiments with annotators and working methods. The results surpass semantic measure, but also show the space for further development.
In this paper, we have focused on the implementation of a mechanism for detecting malicious nodes in the MP-OLSR routing protocol through the integration of trust reasoning. The study carried out has enabled the emergence of trust rules between nodes based on local information and messages exchanged during the neighborhood discovery process and network topology to make the MP-OLSR protocol more trusted. The correlation between the received control messages allows nodes to validate the neighborhood at one hop and two hops. When the information received is consistent with the MP-OLSR protocol specifications, this reasoning allows the creation of a trust relationship. Otherwise, it allows nodes to be wary of the sources of its inconsistencies. When an attack is detected, the countermeasure proposed allows the attacking node to be isolated from the network and to guarantee its non-participation in the routing mechanism by simply integrating this node in the HELLO message with a modified Link Code, so that it is declared by all the nodes of the network as an attacker.
In cryptography, chaotic systems are commonly used for image encryption due to the complex properties of chaos such as sensitivity to initial conditions, unpredictability, and ergodicity. An image encryption method using one-dimensional chaotic maps and Josephus permutation sequence is depicted in this paper. The keystreams required for the encryption process are generated from one-dimensional chaotic maps. Double scrambling operation is proposed in the pixel permutation phase to enhance security. Josephus permutation sequence is used in the first level scrambling. For the second level of scrambling, a novel method is implemented using the keystream generated from a chaotic tent map. Also, the keystream generated from the logistic map is used in the substitution phase. Experimental results and security analysis shows that the proposed encryption scheme can resist various attacks and ensures high security.
In an educational environment, classifying the cognitive aspect of students is critical. It is because an accurate classification is needed by a lecturer to take the right decision for enhancing a better educational environment. To the best of our knowledge, there is no previous research that focuses on this classification process. In this paper, we propose discretization and feature selection methods before the classification. For this purpose, we adopt the equal frequency for the discretization whose result is evaluated by using logistic regression with two regularizations: lasso and ridge. The experimental result shows that four-intervals on the ridge achieve the highest accuracy. It is to be the base to determine the level of the student’s performance: excellent, good, fair, and poor. Next, we remove unnecessary features, by using the Gain Ratio and Gini Index. Also, we build classifiers to evaluate our proposed methods by using k-Nearest Neighbors (k-NN), Neural Network (NN), and CN2 Rule Induction. The experimental result indicates that both discretization and feature selection can enhance the performance of the classification process. Concerning the accuracy level, there is an increase of about 35%, 2.14%, and 3.8% on average of k-NN, NN, and CN2 Rule Induction respectively, from those with original features.
Computed tomography (CT) has gained extensive application in nondestructive testing and medical diagnosis. Most of the practical application of CT imaging applications demands the development of more efficient reconstruction techniques for low-dose and limited-angle scanning to avoid the risk of ionizing radiation associated with CT. The widely used reconstruction algorithms, filtered back-projection (FBP), fails to produce satisfactory results because of lacking the projection data. Statistical techniques have been recently introduced for low-dose CT but the success of these algorithms rely on the assumption of noise characteristics and makes clinical application difficult. Aiming to improve the quality of CT reconstruction, this work investigates the incorporation of an alternate filter for FBP reconstruction. To this end, a multi-resolution 1-D Gabor filter is employed due to its band-pass capability and texture segmentation nature. The proposed filter was developed as reusable software components using the object oriented and platform independent features of Java language. The feasibility and validity of Gabor filterfor FBP was evaluated for its potential on projection simulated from Shepp-Logan phantom image. The results demonstrate that the Gabor filter outperformed its counter parts in artifacts reduction, noise suppression, and structure preservation.
Conventional monitoring devices are usually kept at fixed locations which yields a fixed surveillance coverage. Unmanned aerial vehicles (UAVs) are receiving much attention from researchers in traffic monitoring due to their low cost, high flexibility, and wide view. Unlike stationary surveillance, the camera platform of UAVs is in constant motion and makes it difficult to process for data extraction. The inaccuracy in detection rates of vehicles from UAV videos becomes the motivation for combining optical flow methods with supervised learning algorithms. The proposed method incorporates steps that make use of the Kanade–Lucas optical flow method for moving object detection, connected graphs theory and CNN-SVM for further classification. Optical flow generated contains some background objects detected as vehicle when the camera platforms are moving. The classifier rules out the presence of any other moving objects to be detected as vehicles. The proposed method is tested on few stationary and moving aerial videos. The system is found to be 100% accurate in case of stationary aerial videos and 98% accurate in moving videos.
Human actions recognition (HAR) and understanding become very popular topics in the field of computer vision and signal processing. The purpose of human activities recognition is to automatically examine and characterize actions from a video sequence. The main goal of a HAR system is to identify simple actions of everyday life (like walking, running, jumping ...) from videos. Each of these actions, performed by one person or many persons within a specific period of time, must be represented by a simple movement model. In recent years, a large number of applications of HAR have been proposed in literature such as video surveillance, human-computer interaction and video indexing. In this line, we present our method for HAR using Mask Region Based CNN, MRCNN. This technique will help us to make the accent on the body of individual and this step facilitates the recognition of the current action. With the mask RCNN we used key frame extraction and background extraction. This framework was tested on two datasets KTH and WEIZMANN datasets and experimental results showed the performance of the proposed technique.
Online learning concerns analyzing a continuous stream of transient data and progressively updating the knowledge model without revisiting previously encountered examples. This paper proposes a new incremental fuzzy learning approach for online classification of data streams. It enables an existing fuzzy rule set to be efficiently updated based on a new training example without learning from scratch. The proposed algorithm can not only incrementally construct new fuzzy classification rules but also update the content with old rules to assimilate information from new data. The efficacy of the proposed incremental fuzzy learning method has been demonstrated in a set of simulation tests where the benchmark data sets were treated as data streams for learning.
Renewable sources of energy and chemicals are a viable solution in current situation where world is facing extreme energy crisis. Achieving carbon neutral options for energy supply and high productive options for bulk chemicals by renewable source. Microalgal Biomass is a promising option and has the potential to provide renewable energy and products for future grown in minimum inputs and gives higher output. In this review paper three cultivation techniques are discussed such as autotrophic, heterotrophic, and mixotrophic. Mixotrophic cultivation involves combination of both auto and heterotrophic modes where both light reaction and dark reaction are combined for maximum results. Cost effective options of bioenergy and valuable co products are efficiently produced from microalgae in mixotrophic cultivation conditions. Bioprospection of microalgal options for bioenergy and cultivating those strains in mixotrophic conditions combining with wastewater treatment and CO2 fixation from atmosphere to produce biofuel and valuable co products in the main theme of this review. Concept of biorefinery is prescribed for future options.
The segmentation of magnetic resonance images of the fetal brain has been emerging as a clinical tool to detect abnormalities during the development of the foetus. Since the brain is still in development, a mixture of regions of white matter, grey matter and transition structures that are related to brain growth are still associated with it. In this work, two versions of the K-nearest neighbour algorithm are proposed as the core method for the recognition of different regions of an image; the first one is a refinement of the standard algorithm since some statistics are associated with the pixel, and the second a reinforcing iterative version of the same method. Both versions are used to identify 3 apriori selected regions—the brain, the intracranial region and the remaining part of the foetus. The effectiveness of the method has been demonstrated in a magnetic resonance image segmentation that was first pre-processed with digital filters for feature extraction. The present study illustrates the capabilities of this type of method to support obstetricians and general practitioners to assess the foetus and, in particular, its brain.
Many learner devices with multimedia capacities such as computers, laptops and telephone cells derive from the nowadays technology evolution. The wide range of these electronic devices brings a real opportunity to build an ad hoc network in the DOUNG [1, 2] classroom with real time option of following courses. In addition of previous works that define significant Quality of Service (QoS) parameters for following courses alive in the DOUNG model, this paper aims to contribute by using the learner devices to extend the DOUNG network architecture. The idea is to use an ad hoc network in the classroom after a Wi-Fi antenna instead of duplicating this antenna when the signal is low. Thus, with an access mode that brings the network resource consumption to reach a critical threshold, we measure some QoS parameters by simulation in this new architecture using SCTP (Stream Control Transport Protocol) at the transport layer and DSR (Dynamic Source Routing) at the network layer. Curves are produced and the results are discussed.
The process of democratization is developing and promoting due to electronic voting. However, the increase of frauds and the increasing amount of attacks launched by hackers, gave birth to privacy and authentication problems. Cryptography offers multiple solutions to overcome the sensitive data protection issues in e-voting. In this paper, we study the application of elliptic curve cryptography, and use the homomorphic encryption properties to present a new electronic voting system. Our new scheme is based on the homomorphic cryptosystem EC-ELGAMAL [7], and zero knowledge algorithm of Schnorr algorithm for identification and authentication.
Medical ultrasonography is widely used because it is a safe, non-invasive and non-ionizing diagnosis method. For these reasons, medical ultrasound is recommended for monitoring parts of body. This paper proposes a new contour methodology for ultrasound images, based on the Dijkstra algorithm. It will is used to detect structures or boundaries regions. It belongs to Live wire segmentation techniques group, for extracting quickly and accurately regions of interests in images. This can be done through a set of seed points, generated by an automatic process or selected manually by the user. Experimental results show that this method is robust and efficient in segmentation of fetal ultrasound images. It is less sensible to noise and, shadows and to imprecise boundary regions.
Churn studies have been used for years to achieve profitability and to establish a sustainable customer-company relationship. Deep learning is one of the contemporary methods used in churn analysis due to its ability to process huge amounts of customer data. In this study, a deep learning model is proposed to predict whether customers in the retail industry will churn in the future. The model was compared with logistic regression and artificial neural network models, which are also frequently used in the churn prediction studies. The results of the models were compared with accuracy classification tools, which are precision, recall and AUC. The results showed that the deep learning model achieved better classification and prediction success than other compared models.
The greatest challenge of this century is the protection of stored and transmitted data over the network. This paper provides a new hybrid algorithm designed based on combination algorithms, in the proposed algorithm we combine the Hill and the Advanced Encryption Standard Algorithms, to increase the efficiency of color image encryption and increase the sensitivity of the key to protect the RGB image from Keyes attackers. The proposed algorithm has proven its efficiency in encryption of color images with high security and countering attacks. The strength and efficiency of combination the Hill Chipper and Advanced Encryption Standard Algorithm is tested by statical analysis for RGB images histogram and correlation of RGB images before and after encryption using hill cipher and proposed algorithm and also analysis of the secret key and key space to protect the RGB image from brute force attack. The result of combining Hill and Advanced Encryption Standard Algorithm achieve strength in encrypting images rather than encrypting images by using the single Hill algorithm.
Copy-move is one type of attack to forge a digital image where the attacker duplicates several areas of the image and paste them in different places to conceal a particular object on the original image. After the forgery, advanced methods such as noise addition and blurring, are often performed in the forged image to make it more challenging to recognize the attack. Therefore, it is required to do a preprocessing before conducting the detection. The preprocessing can be eliminated using a copy-move detection that is more resistant to noise addition and blurring. This paper proposes a new, flexible, and robust method that perform forensic analysis of both regular and advanced copy-move using modification and addition from two methods. The first method is designed to identify a regular copy-move attack, while the second one is effective for an advanced attack. The proposed method combines these two methods, can adapt to the forged image condition, and no preprocessing is required.
Use of biometrics in digital society has raised the questions of biometric template protection and secure authentication. The biometric template protection mechanisms known so far hardly maintain a trade-off between security of template database and recognition performance. This paper proposes a hybrid technique of template protection for a multibiometric system that provides better matching performance and infallible from fraudulent attacks. The multimodal system is prepared from face and ECG biometrics. The ECG as a biometrics not only supplements the face biometrics in a multimodal system but also ensures security for robust recognition. The pre-trained deep learning models are used to process both biometrics and prepare multimodal templates. The templates are mapped to their corresponding classes represented by randomly generated unique binary codes. These binary codes are further encrypted using cryptographic hash for non-invertiblity and hide information of fused templates. Finally, the matching is performed using hash codes for ensuring an additional layer of defense against adversarial attacks.
