Intelligent Deep Convolutional Neural Network Based Object Detection Model for Visually Challenged People

Artificial Intelligence (AI) and Computer Vision (CV) advancements have led to many useful methodologies in recent years, particularly to help visually-challenged people. Object detection includes a variety of challenges, for example, handling multiple class images, images that get augmented when captured by a camera and so on. The test images include all these variants as well. These detection models alert them about their surroundings when they want to walk independently. This study compares four CNN-based pre-trained models: Residual Network (ResNet-50), Inception v3, Dense Convolutional Network (DenseNet-121), and SqueezeNet, predominantly used in image recognition applications. Based on the analysis performed on these test images, the study infers that Inception V3 outperformed other pre-trained models in terms of accuracy and speed. To further improve the performance of the Inception v3 model, the thermal exchange optimization (TEO) algorithm is applied to tune the hyperparameters (number of epochs, batch size, and learning rate) showing the novelty of the work. Better accuracy was achieved owing to the inclusion of an auxiliary classifier as a regularizer, hyperparameter optimizer, and factorization approach. Additionally, Inception V3 can handle images of different sizes. This makes Inception V3 the optimum model for assisting visually challenged people in real-world communication when integrated with Internet of Things (IoT)-based devices.     

Expectation Maximization Method for Effective Detection and Tracking of Object Using Machine Learning Technique for Secure Wireless Communication

Wireless Personal Communications - Object detection is one of the most important computer vision tasks that is used synonymous to object recognition which comprises the mission of identifying the...     

Stability limits and NO emissions of premixed swirl ammonia-air flames enriched with hydrogen or methane at elevated pressures

This study reports measurements of stability limits and exhaust NO mole fractions of technically-premixed swirl ammonia-air flames enriched with either methane or hydrogen. Experiments were conducted at different pressures from atmospheric to 5 bar, representative of commercial micro gas turbines. The full range of ammonia fractions in the fuel blend, x_NH3, was considered, from 0 (pure methane or hydrogen) to 1 (pure ammonia), covering very lean (φ = 0.25) to rich (φ = 1.60) equivalence ratios. Results show that increasing pressure widens the range of stable equivalence ratios for pure ammonia-air flames. Regardless of pressure, there is a critical ammonia fraction above which the range of stable equivalence ratios suddenly widens. This is because flashback does not occur anymore when the equivalence ratio is progressively increased towards stoichiometric and rich blowout occurs instead. This critical ammonia fraction increases with pressure and is larger for ammonia-hydrogen than for ammonia-methane. Provided that enough hydrogen is blended with ammonia (x_NH3 < 0.9), flames with very lean equivalence ratios (φ < 0.7) can be stabilized and these yield competitively low NO emissions (<200 ppm), regardless of pressure. For this reason, very lean swirl ammonia-hydrogen-air flames are promising candidates for micro gas turbines. However, N₂O emissions have the potential to be unacceptably large for these operating conditions if heat loss is too large or residence time is too short. As a consequence, the post flame region must be considered carefully. Due to the lower reactivity of methane compared to that of hydrogen, very lean swirl ammonia-methane-air flames could not be stabilized and good NO performance is limited to rich equivalence ratios for ammonia-methane fuel blends. The equivalence ratio above which good NO performance depends on pressure and bulk velocity.     

Internet of things for healthcare monitoring applications based on RFID clustering scheme

Wireless Networks - COVID-19 surprised the whole world by its quick and sudden spread. Coronavirus pushes all community sectors: government, industry, academia, and nonprofit organizations to take...     

Performance computation and implementation of distributed controllers for reliable software-defined networks

The Journal of Supercomputing - Software-defined networks (SDNs) are designed to cover the dynamic operations of network factors and the complex role of controlling components to achieve...     

A Bioinformatics Systems Biology Analysis of the Current Oral Proteomic Biomarkers and Implications for Diagnosis and Treatment of External Root Resorption

External root resorption (ERR) is a silent destructive phenomenon detrimental to dental health. ERR may have multiple etiologies such as infection, inflammation, traumatic injuries, pressure, mechanical stimulations, neoplastic conditions, systemic disorders, or idiopathic causes. Often, if undiagnosed and untreated, ERR can lead to the loss of the tooth or multiple teeth. Traditionally, clinicians have relied on radiographs and cone beam computed tomography (CBCT) images for the diagnosis of ERR; however, these techniques are not often precise or definitive and may require exposure of patients to more ionizing radiation than necessary. To overcome these shortcomings, there is an immense need to develop non-invasive approaches such as biomarker screening methods for rapid and precise diagnosis for ERR. In this review, we performed a literature survey for potential salivary or gingival crevicular fluid (GCF) proteomic biomarkers associated with ERR and analyzed the potential pathways leading to ERR. To the best of our knowledge, this is the first proteomics biomarker survey that connects ERR to body biofluids which represents a novel approach to diagnose and even monitor treatment progress for ERR.     

Enhanced Metaheuristics with Trust Aware Route Selection for Wireless Sensor Networks

Recently, a trust system was introduced to enhance security and cooperation between nodes in wireless sensor networks (WSN). In routing, the trust system includes or avoids nodes related to the estimated trust values in the routing function. This article introduces Enhanced Metaheuristics with Trust Aware Secure Route Selection Protocol (EMTA-SRSP) for WSN. The presented EMTA-SRSP technique majorly involves the optimal selection of routes in WSN. To accomplish this, the EMTA-SRSP technique involves the design of an oppositional Aquila optimization algorithm to choose safe routes for data communication. For the clustering process, the nodes with maximum residual energy will be considered cluster heads (CHs). In addition, the OAOA technique gets executed to choose optimal routes based on objective functions with multiple parameters such as energy, distance, and trust degree. The experimental validation of the EMTA-SRSP technique is tested, and the results exhibited a better performance of the EMTA-SRSP technique over other approaches.     

FSO-LSTM IDS: hybrid optimized and ensembled deep-learning network-based intrusion detection system for smart networks

The Internet of Things (IoT) has achieved exponential growth worldwide. Although the IoT is used by millions of users, these networks are handicapped by attacks such as denial of service, man in the middle, and spoofing. These attacks threaten the entire IoT ecosystem and affect the integrity and security of the user. Hence, the prediction and identification of novel network attacks in an IoT network remains a challenge for researchers. Recently, machine learning and deep learning have played a pivotal role in predicting and classifying different attacks in an IoT network. However, these algorithms suffer from computational complexity as the number of attacks increases. Hence, a novel hybrid optimized long short-term memory (LSTM) approach is proposed. Whereas a convolutional neural network is used to extract the temporal and spatial correlated features of the IoT network, the optimized LSTM is used to predict the different attacks in the network. Furthermore, firefly swarm optimization is integrated with LSTM to reduce the computational overhead, which in turn increases the prediction accuracy. Nearly 19,00,503 real-time normal and attack data were collected from the experimental simulation setup based on the OMNET++–Python–IoT framework. Extensive experimentation was carried out to evaluate the proposed algorithm, and various metrics, such as accuracy, sensitivity, specificity, and F1-score, were calculated and compared with state-of-the-art learning-based network intrusion detection systems. Furthermore, other benchmarks, such as the CIDCC-15, UNSW-NB15, and NSL-KDD datasets, were used to evaluate the performance of the different deep-learning-based intrusion detection systems. The results demonstrate that the proposed deep-learning method outperforms other classical learning models with low complexity and high prediction performance.

     

Optimization and sensitivity study of biodiesel synthesis from Jojoba oil using mixed-integer programming

Jojoba oil-based biodiesel is promising alternative fuel due to its versatile properties. Renewable transportation fuels are considered as promising alternatives to conventional fuels. The physical and chemical properties of these fuels enabled them to be used in modern internal combustion engines; this makes them attractive for use as direct replacements or as additives of fossil fuels. Jojoba oil is extracted from Jojoba seeds, and it is an excellent feedstock for biodiesel after the transesterification process. The plant is highly adaptable to harsh weather including salty water, desert, and hot temperatures; thus, it can be grown in Saudi Arabia. This research work comprises a detailed optimization study of biodiesel production from Jojoba oil using mixed-integer programming. golden section search method was used for the optimization and sensitivity study was conducted for reaction time and temperature. The result shows that 54.1 minutes and 47.5 °C are the optimized reaction time and temperature to produce biodiesel which is considerably low as compared to previous studies.