Intelligent Communication Technologies and Virtual Mobile Networks

A super wideband compact antenna for Microwave Imaging (MI) is proposed. The antenna radiator is a truncated patch with slots and has a defected ground structure (DGS). The slots in the patch and truncation allowed the radiator to couple well with the partial ground plane loaded with DGS. This led to a super wide bandwidth of 25 GHz ranging from 5 to 30 GHz with a compact size of 0.27λ0 × 0.33λ0. Initially, with a truncated circular patch, a triple wideband antenna is achieved. To increase the impedance matching of the antenna and realise the super wideband, a circular ring slot has been etched in patch and ground is also truncated. To further enhance the bandwidth, slots in triangular shape are etched in the patch and a rectangular slot is etched in ground. The proposed antenna is a compact space-saving one. The results reveal a super wideband performance of 167% (5–30 GHz) with a consistent radiation pattern and peak gain of 10.2 dB in a compact area.

The performance of mathematical analyses for the learning of ad hoc wireless setups yielded limited results for the transformation of movement and traffic patterns, the theory of dynamic games, and the connection irregularity that separates these networks. Inspired by various studies, this work would theoretically influence all other researchers, addressing game theory, and generally, it would be preferred to calculate the probability of networks beforehand. This article describes several connections in advanced wireless networks that are often designed to be readily available. It allows the investigation of code of conduct and management plans for the existing resources, as well as the creation of mechanisms that promote balance and encourage individual users to work generally in a fruitful manner. The modern literature is reviewed on situational setting research game theory, proving its compatibility with power control and waveform change, mediocre input control and beat, and knot engrossment.

In the past decades, there have been traditional network architectures that were responsible for routing data between various nodes using different types of IP addresses through a router using different types of routable protocols such as TCP/IP. But these networks presented a problem with respect to the management of large volumes of information. Faced with this situation, the Telco industry proposes a network structure called Information Centered Networks (ICN), which is a part of the Content Centered Network paradigm. The ICN help distribute and retrieve content on the network once they are requested by the user, which allows independence in the location and storage of this cached content in its various ICN nodes. In turn, the ICN architectures use protocols that allow the unique assignment and sending of data objects when making use of the respective network. Hence, the goal of the ICN topology is to deliver these data objects to users located in the network, in contrast to networks that use TCP/IP protocols and in which communications is carried out between the various nodes of the network. The main objective of this article is to familiarize users by comparing different proposed ICN architectures, as well as network caching and ICN transformation.

The proposed accident prevention system decreases the likelihood of accidents on roads and also send its location and notify people who can take immediate action. The system combines detection of alcohol, driver’s fatigue, fuel level, overheating of fuel tank and if the appropriate conditions are not satisfied, the vehicle stops. Additionally, the vehicle’s speed is tracked and in case of over speeding, the vehicle stops automatically. In case of an accident, the Global Positioning System locates the geographic coordinates and using Global System for Mobile (GSM), it will send a message to the registered mobile number. A microprocessor and an ultrasonic sensor are deployed to determine the distance between two driving vehicles in the same lane and notify the driver coming behind the vehicle using RF technology.

The most complicated problems in the field of self-driving aircraft are the design of a strong real-time obstacle discovery and evasion framework. The issue is becoming a little complex due to the size and shape of the vehicle, and it becomes a little complex to accomplish the mission. As a result, we will be using a Flight Time Sensor (TOF sensor—time of flight) which is a lightweight sensor, and it is known as a MicroLiDAR sensor. The man’s behavior of detecting the collision status of the obstacles to the approach using the time-of-flight sensor is proposed here. The control board has related to a time-of-flight sensor and Artificial Intelligence (AI) in order to recognize the obstruction and prevent a collision. During the motion of the aerial vehicle (UAV), the detection set of rules estimates modifications inside the length of the upcoming obstruction zone. The strategy primarily recognizes the characteristic focuses of the deterrents and after that the impediments which are likely to approach the UAV. Another, by comparing the obstacle surface ratio, the UAV's position helps determine whether an obstacle can cause a collision. The algorithm was tested by doing actual flights, and the results show that it is accurate. With a focus on unmanned aerial vehicles, the study offers a thorough investigation of collision evasion methods for unmanned vehicles (UAVs); it could be a careful examination of a few collision avoidance strategies that are categorically characterized, besides a comparison of the ways taken into thought in connection to different circumstances and specialized contemplations. Also covered are how various sensor types are used in the context of UAVs to avoid collisions. It is highly accurate and collision avoidance ultimately happens.

The goal of this study is to analyze the dynamics of the collaborative communication model for the implementation of the Non-Cash Food Assistance (Bantuan Pangan Non-Tunai, BPNT) Program in Takalar Regency, Indonesia, to fight poverty toward community resilience. In order to go more into the subject issue, this study employs a qualitative exploratory methodology. Based on the results of the study, it shows that the implementation of the Non-Cash Food Assistance Program is intended to increase community resilience in the midst of a global crisis in Indonesia. Because outreach to the community is lacking, the implementation of the Non-Cash Food Assistance (Bantuan Pangan Non-Tunai, BPNT) aid program in Takalar Regency, Indonesia, has not been sufficiently successful. This is evident from a number of program targets that were missed. The distribution of the non-cash food Assistance (Bantuan Pangan Non-Tunai, BPNT) aid program must promote collaborative communication with all Takalar Regency stakeholders in order to resolve this issue. This will have an impact on the extent to which distribution policies and management will be implemented in supporting the success of the program.

In this article, a review on the use of unmanned aerial vehicles (UAVs) for retransmission has been conducted. The description of the basic network architecture and the main characteristics of communication channels are given. Various types of data transmission using UAVs, an approach to the organization of communication using unmanned aerial vehicles—short-range repeaters, which allows for information exchange in emergency situations, are considered. The general principles of UAV operation when used for weather monitoring, forest fire detection, traffic jam tracking, cargo transportation, use in rescue operations, etc., are analyzed. The advantages and disadvantages of such use of UAVs, the general architecture of communication systems and the place of the UAV in it, as well as the general characteristics of communication channels and types of data transmission are described. The technology was implemented using the example of establishing communication between separate units during training in a mountainous area using a low-flying UAV of the DJI Mavic Pro 2 type with suspended equipment for retransmission of radio signals SURECOM SR-112.

Reliability (information management) and lifetime (energy management) are the two most important factors for the real-time application of WSNs. The reliability and lifetime must be managed properly as there are on-demand individual networks, subnets, extended networks, and other real-time applications of WSNs. The overall reliability of a wireless sensor network (WSN) depends on its lifetime, which in turn depends on battery-powered nodes, node-sink distances, and traffic near the sink. The referred literature proved that energy efficiency and energy balancing depend upon optimized node-sink distances that are obtained by a proactive heuristic ACO (ant colony optimization) routing algorithm. This algorithm also finds the minimum value of average nodal energy consumption. The efficient and balanced energy paths lead to good packet delivery. The majority of the referred literature uses a programmed random node deployment strategy and a clustered WSN model. This work uses a programmed manual node deployment strategy with uniform and random procedures and a simple sectored network model. The performance comparison results are placed with respect to different strategies such as node deployment planning, different transmission planning, different routing methods, different network sizes, different energy levels, and iterations (rounds). This comparison paves the way for futuristic applications of WSN. The applications include subnets, extended networks, real-time patient monitoring systems, and driverless cars.

The improvement of mobile application has improved over the years with the number of libraries and APIs that are shared for the developers. Resources such as APIs and libraries will brings more items that developers can use and utilized in their software projects mainly mobile applications. Right up to this moment, finding a complete documentation of integrating a 3D object model to a mobile application is very difficult. This brings up the question on to “why developers are not using 3D models to be used in their mobile applications?” It is a valid question because developers will always try and use the newest and latest thing on the market to build their mobile applications, for a 3D object model not included in that category is surprising. For this paper, a deep dive into 3D visualizations API called SceneView will be conducted to see if the 3D model technologies for mobile application is relevant for the current style of mobile application design and understand on what a 3D model can do for a mobile android application when used correctly.

The demand for machine learning (ML) algorithms in wireless communication have increased over the last decade. Anyhow to enhance the prediction quality in complex problems, a significant amount of training is needed. ML requires centralized training which consumes more processing power. İn the case of massive networks, consequently, it is necessary to train and test the algorithms in a distributed manner, which results in distributed learning. Recent development and methodologies used for distributed learning in network optimization of wireless networks are presented in a comprehensive manner in this paper.

One of the biggest and most severe risks on the Internet today is malicious software, generally known as malware. Attackers are producing malware that has the ability to change its source code as it spreads and is polymorphic and metamorphic. Furthermore, the variety and quantity of their variants seriously compromise the effectiveness of current defences, which frequently rely on signature-based techniques and are unable to identify malicious executables that have not yet been detected. Variants from different malware families have behavioural traits that are indicative of their function and place in society. Utilizing the behavioural patterns obtained either statically or dynamically, deep learning techniques can be utilized to discover and classify novel viruses into their recognized families. In this digital age, security failures brought on by malware attacks are on the rise and pose a serious security concern. Malware detection is still a strongly contested academic topic because of the significant implications that malware attacks have on businesses, governments, and computer users. For the real-time identification of unknown malware, the efficacy of current malware detection techniques, which entail the static and dynamic analysis of malware signatures and behaviour patterns, has not been shown. For classifying malware, we mostly utilize CNN and ELM deep learning algorithms.

The technical problems during the application of the optical satellite communication network (FSON) in free space were analyzed, with the aim of creating new optical photon network switches for the construction of the optical communication network between nanosatellites and terminals in ground stations (GS) of the FSON nanosatellite-nano-satellite and nanosatellite-ground terminals, whereas a new generation 3D optical photon switch (OS) is proposed for switching optical channels between transmitters, and the principle of its operation at different wavelengths is explained appropriately. Based on the working principle of the new generation 3D optical switch, which is intended to be installed on nanosatellites and optical terminals, the reliability indicators of its elements were calculated and the sustained operation period was investigated.

Different reasons make data packet to be lost in Mobile Ad-hoc Networks (MANETs) over wired networks, in which packet loss is typically brought on by network congestion. Packet losses in MANETs can either be attributed to the dynamic character of such networks or to the wireless communication situations (fading, interference, multi-path routing, link failures, network partitioning). This latter issue might be brought on by the node's mobility or by its battery running out. Terrain dimensions, speed, and density of nodes are the important features that set apart MANETs from wired networks. The major objective is to reduce packet loss and delay and improve packet delivery ratio and throughput. In this paper, we have implemented MANET by using AODV, DSR, and DSDV and simulated on network simulator. The significance of terrain areas and node density is analyzed over reactive and proactive routing protocols. The expansion of terrain dimensions and node density is analyzed through simulation experiments by using NS2 and BonnMotion tools. The comparative outcomes are analyzed with a pictorial evaluation of five QoS performance metrics through node density under various terrain dimensions.

The wireless sensor network (WSN) is gaining huge importance these days. It forms the backbone of IoT systems, a cutting-edge technology. As the applications are rapidly increasing, it is important to address the security concerns around this network. The fault-tolerant algorithms form an integral part of the wireless sensor distributed network. These algorithms make the wireless sensor network act as fault-tolerant, thus making the sensor readings more reliable. Even today, many sensor nodes are transmitting the data within the network in a wireless mode, with inappropriate security mechanisms in place. To overcome this gap, elliptic curve cryptography (ECC) and its variants are integrated with fault-tolerant sensor communications recently. In this study, we focus on presenting the challenges involved in implementing ECC-based secure sensor communication in the existing systems and the suitability of ECC-based secure communication for many real-time applications.

A quick access recorder (QAR) that is carried on the aircraft captures hundreds of parameter data generated during flight. However, these information are not increasingly being applied well. The use of QAR data for fault prediction is becoming more and more essential. First, an illustration of the technological path for civil aviation defect prediction is shown. 2, 4 commercial airplane defect prediction procedure rely on QAR information is presented. These methodologies include advanced methodology based on Enhanced Grey Prototype, performance analysis method based on Template Matching, Time-series data Extrapolation, and Non-parametric Correlation and assessment on predicated efficiency regarding the trend prediction Technique. Finally, a thorough description of the forecasting system’s architecture is provided. As a result, it can keep an eye on how the aircraft’s systems and components are functioning in order to promptly identify any defect symptoms, devise an appropriate routine maintenance, and guarantee flying security. Finally, using the approach outlined in this study, the temperatures’ characteristics of the air conditioning systems of Boeing are anticipated. The outcome predictions validate the approach’s efficacy.

The rise in fatal and permanently disabling road accidents is a severe public health issue. Human lives are frequently lost in road accidents when medical assistance arrives too slowly. Therefore, deaths from road accidents are more common. Several accident prevention systems can, to some extent, avoid accidents. Still, they lack in giving collision warning while driving and the ability to contact emergency services in the event of an accident. A collision warning is one of the great difficulties regarding active safety for automobiles on the road. In this paper, the proposed system is to design and implement collision warning and accident detection for automobiles. In the proposed work, the collision warning is done using object detection, and accident detection is implemented by Raspberry Pi, a vibrator sensor, GPS, and a Telegram bot. The detection of the vehicles is based on the Single-Shot Detection (SSD) MobileNetV1 algorithm. Suppose the vehicles did not maintain a minimum distance between each other. In that case, it will send a collision warning message to the user using a Telegram bot via a Wi-Fi network. If the user made accident, then the accident occurred message and location are sent to emergency services based on the vibration sensor and Global Positioning System module.

The pipe robots are autonomous machines that operate within pipe networks and perform several tasks including surveillance, monitoring, cleaning, rescue, etc., therein. Various elements, such as rust and fractures, could harm pipelines. Consequently, a reliable monitoring system is necessitated to guarantee the security of these pipes. It is difficult for an individual to examine each component of pipes and repair the problems. Pipe robots have been created as a result of such issues. Moreover, the demand for studies on pipe robots has increased in latest days. A variety of pipe robots are usable which include wheel type, screw type, wall-press type, walking type, inchworm type, etc. Relying on its specifications and assessment goal, every model has a unique set of benefits and drawbacks. This paper presents an overall review on the advancement of pipe robots from past few years. Authors believe that ongoing advancements in pipe robot technology are anticipated to be significantly influenced through this paper.

As the population is increasing, the environmental resources consumed also started to increase, due to this pollution which becomes a major concern, and it is a day-to-day problem, which is faced by all biological species. Various solutions have been proposed by researchers as countermeasures to control the pollution in past few decades. In this paper, IoT-based pollution monitoring system is presented which encompasses various sensors, which is capable of measuring air pollutants (ammonia, carbon monoxide, smoke, LPG, and methane), temperature, humidity, and sound levels. The microcontroller used is MSP430. The real-time data can be monitored virtually across the world by an application. The application which is used is ThingSpeak, where the level of harmful pollutants can be monitored in the form of tables and graphs.

Entrepreneurship and business require Internet of Things support to be able to develop to meet market and consumer needs. This study aims to map the current state of global research and the future trends of IoT development in entrepreneurship studies. Based on the Scopus database, this study uses a bibliometric intellectual structure perspective by analyzing 1,075 scientific publications for half a century. The findings show that the United States Environmental Protection Agency and the United States were the most prolific countries and research institutions studying IoT in entrepreneurship. The findings of this study propose the ISEEEII research theme concept for IoT in entrepreneurship research.

In this paper, the groundwork on commuter-based travel system is studied by considering the several existing techniques in this problem. In the present scenario, IoT plays an inevitable role in all emerging technologies because of the increasing usage of the internet. It aims at smart usage of human time in real-time applications. Transportation has become the essential factor of regular activities. Hence, this paper reviews all the system that can automatically measure the distance travelled by passengers on the bus and collect the distance travelled from the accounts of passengers. Several new systems could be implemented in the comfort zone of passengers only after a thorough review on literature. A few ideas of passenger’s convenience like a method of charging tickets are by distance travelled is one of the most innovative ways. The usage of most secure smart card as the prepaid travel card, storing the amount in its internal memory, allowing users to easily board any bus in the area. The idea of incorporating the GPS receiver to calculate position and driving distance avoids relying on the vehicle's built-in distance of metres. To develop an effective system for IoT-based APP for commuter travel tracking system, this paper provides an exhaustive survey for real-time applications and compares the existing systems in various aspects guiding the future researchers.

Due to the difficulty and inefficiency of the reserved data hiding approach, several encryption domain-based data (reversible) concealing algorithms have drawbacks such as a low embedding rate and poor resolution of the immediately retrieved data. This article will give complete details about different reserved model approaches for data hiding and proposes a different algorithm for data hiding. By taking advantage of the pixel redundancy inside each block, the technique can incorporate confidential data into 2X2 picture blocks. Our proposed mode extends this idea to the encrypted realm (ABPEE-RDHEI). In order to maintain spatial redundancy for data embedding, to further boost the security level, a stream cipher is then applied to the block-permutated image. The proposed ABPEE-RDHEI can produce marked decrypted images with a high embedding rate and good resolution due to the advanced data combining techniques using iteration and data encryption. ABPEE-RDHEI outperforms a number of state-of-the-art approaches, according to experimental findings and analyses. The ability to forecast an image accurately and insert a message into the image with minimal distortion are two essential components in the field of RDHEI. However, it is difficult to increase forecast accuracy because linear regression prediction is susceptible to outliers. To overcome this issue, this work suggests an RDHEI approach based on an adaptive prediction-error label map.

The electronic Know Your Customer (e-KYC) is basically a system for all the banking enterprises or for the identity providers to establish and provide a customer identity such as data verification process between agreeing parties. Nowadays, most of the banks are implementing e-KYC on cloud due to its efficient resource consumption and high degree of accessibility and availability of cloud computing. KYC is also a procedure by which banks gather information about their customers' identities and addresses. It is a regulated practice of completing due diligence on clients to verify their identity. This procedure aids in ensuring that clients are genuine and the bank's services are not being abused. The banks are in charge of completing the KYC process. In this paper, we have discussed the demerits of the process for establishing accounts and KYC may be a time-consuming, manual process that is duplicated across institutions. Financial institutions would be able to use blockchain to share KYC information, to improve compliance outcomes, efficiency, and customer service, and to avoid redundancy by gaining experience.

Today, with an ever-increasing number of computer users, the number of cyberattacks to steal data and invade privacy is of utmost importance. A group of applications uses the Advanced Encryption Standard (AES) to encrypt data for security reasons. This mainly concerns enterprises and businesses, which ultimately handle user data. But many implementations of the AES algorithm consume large amounts of CPU horsepower and are not up to the mark in terms of throughput. To tackle this problem, the proposed system makes use of GPUs, which are targeted for parallel applications. These enable parallel operations to be performed much faster than the CPU, ultimately increasing throughput and reducing resource consumption to some extent. The vital aspect of this approach is the speedup that is achieved due to massive parallelism. This research aims to implement AES encryption and decryption using CUDA and benchmark it on various compute devices.

Routing protocols in the Internet of Things (IoT) are quite vulnerable to attacks by design. There are various types of the routing attacks like the blackhole attack, selective forwarding, sinkhole attack, wormhole attack, and decreased rank attack. Other attacks such as version number modification, hello flooding attack, and Sybil attack are considered resource depletion attacks. These attacks can be designed in a way to corrupt information, reduce bandwidth, and threaten the integrity of the network which is why it is critical to identify the attacks and avoid as much damage as possible to the network. In this paper, we will be focusing on some of the attacks mentioned above. We have proposed a methodology to detect and predict IoT attacks. We propose different machine learning algorithms and conclude which algorithm is the most accurate in detecting every attack. The paper primarily focuses on dataset creation, followed by implementing various machine learning algorithms and identifying if a network is under attack and then what attack is taking place in that scenario.

Online Ticketing systems face the issue of eliminating ticket counterfeit and scalping while ensuring privacy protection and information openness. Another concern is ticketing fraud, when duplicate tickets allow unauthorised entry and cost hosts money. An Ethereum-based ticketing DApp would fix all the issues. Ticket holders may easily sell their tickets using the DApp. Ticket numbers are set, and each ticket is owned by a concert goer. When paying using an Ethereum wallet, the consumer gains ownership of the tickets. The ownership of tickets cannot be modified once changed, thereby preventing ticket fraud. When tickets are purchased, they become the purchasers’ property. This paper proposes a hybrid online ticketing model based on the blockchain to address these problems. It makes need for blockchain technology to make ticketing data transparent and utilises encryption techniques to keep user information private. It also employs the use of digital signature advanced technologies to assure ticket validity and includes a revolutionary ticket approval process to avoid digital piracy. An assessment of the system concludes with a description and analysis of the tests carried out during the installation of the system.

Consumers can now access services using a wide range of handheld devices due to the recent explosive expansion of various mobile network payment gateways. The method used to assess the security of mobile network payment gateways covers complex, complex situations such as malware detection, two-factor authentication, and fraudulent payment system detection. To help with mobile transaction fraud detection, virus detection, and authentication issues, this work recommends using the Secure Mobile Electronic Payment Framework Enhanced by Machine Learning.

The demand for healthcare services is increasing with the growing population. Integration of the Internet of Things (IoT) is a key move to cater to this demand. Routing protocols form a key aspect of a network. The three principal elements of routing protocols deployed in the healthcare sector include compliance with the constrained nature of devices, capacity to manage heterogeneous traffic, and data security. In this study, the RPL routing protocol is being explored. This paper compares the usage of Multiple RPL Instances with Multi Sink against Single RPL Instance with Single Sink over Latency, Energy consumption, Control traffic overhead, and Packet Delivery Ratio in Cooja Simulator. The results obtained illustrate that the Multiple Instance and Multi Sink approach performs better, thereby indicating its ability to handle heterogeneity. Further, cryptographic mechanisms are added to ensure confidentiality. Additionally, data from the sinks are transferred to a backend server hosting a website.

The background of our prototype—‘Safar’, is to develop a web-based application based on the effectiveness of QR codes for authenticating an E-ticket. The aim is to provide a simple and sophisticated system to generate an E-ticket, it will be able to increase ticket sales. The proposed system will allow the user to book a ticket for a particular monument on a particular day. After booking a ticket, the user is directed to the payment portal where the money transaction is carried out using the RazorPay testing API. If the payment is successful, an E-ticket containing the QR is generated which is sent to the email of the tourist. The tourist then scans this QR code at the monument to gain entry into the monument. The proposed website also contains an admin portal where the live count of visitors with their credentials and the history of tourists for that monument is stored. All the user data will be encrypted using a novel cryptographic approach before it is saved to the database, to provide added security. As tourism accounts for 7% of India’s GDP, the proposed system aims to attract tourists by giving them an all-round, complete experience. It not only reduces the time to wait for a long duration in queues but also makes the process of visiting a monument seamless and hassle free. The proposed system is developed using React, HTML, CSS, JavaScript, and Django. It will give a boost to the heritage and culture of our country and also ensure a greener and cleaner economy.

In modern era, smart home automation system provides sophistication and comfort in day-to-day life. Secure, scalable, efficient authenticated protocols are designed for indoor localization, real-time tracking, and flexibility of applications in home automation. In this paper, a novel, secure, and cost effective IoT framework is implemented using Raspberry Pi 3 B+ controller and sensor network. The proposed model is an advanced access control system encompassing face recognition, biometric authentication that provides decentralization of information in the home environment including protection of the PET animals. The proposed model is analysed for three different scenarios and an exhaustive theoretical and extensive experimental analysis is carried to prove the significance of the proposed model.

A recent development in industry and a cutting-edge means of generating funds is blockchain-based crowdfunding. While it has similarities with conventional crowdfunding, it has some unique characteristics of its own. In the light of this, the success criteria that affect how traditional crowdfunding turns out might have a different impact on crowdfunding enabled by blockchain. It is currently unknown what distinguishes the success criteria for blockchain-based fundraising efforts from those for conventional crowdfunding, given the fact that the amount of these projects has grown significantly over the past several years. Such information is necessary for organisation to properly organise their blockchain-based fundraising operations and to help potential investors find the important characteristics and motivators of outstanding initiatives. Regulators and market players would also benefit from understanding how the present regulatory framework applies to blockchain-based crowdfunding. Due to the specific properties of blockchain-based crowdfunding, legal structures may need to be interpreted in order for law to be successfully enforced. In order to fill this information vacuum, a variety of significant literature on the success determinants for both conventional and blockchain-based crowdfunding was looked upon. The results of this literature evaluation offer recommendations for the direction of future research and development. The study has helped to clarify the differences and parallels between traditional crowdfunding and blockchain-based crowdfunding.

Increased Internet use has brought about a greater focus on security in recent years. The daily rate of data exchange is increasing as Internet usage grows. Hackers might potentially compromise the daily flow of information. Videos feature a wide variety of sceneries and objects, some of which are static, some of which are constantly moving, some of which are sparsely populated, and some of which undergo complex, non-repetitive motions. The Dynamic Texture (DT) in particular presents a number of difficult challenges due to its constantly shifting appearance and motion. With cloud computing, users have unprecedented flexibility in terms of video data storage, sharing, and accessibility. It has also contributed to the explosive growth of digital data. In order to keep up with this speedy expansion, video deduplication has become a major strategy for Cloud Storage Providers (CSPs) by enabling them to efficiently delete duplicate data from their storage facilities. This research proposes a new compression plot that can safely deduplicate videos stored in the cloud. To efficiently transmit and store digital video files over a network and on computer discs, video compression techniques are used to solve the problem of reducing and deleting superfluous video data. Video steganography, in which information is concealed within video frames, is used in this research. The video compression and deduplication techniques and feature extraction techniques for video steganography for integrity preservation are analysed in this research.

In real-world circumstances, like as logistics, it may be highly valuable to consider all of the features of the products and the container, such as value, weight, and form, for better optimum solutions. Two well-known optimization problems involving placing products in a container are the packing problem and the Knapsack Problem. Knapsack Problem is an optimization problem in the sense that it finds the best solution with maximum profits among a group of objects with varying weights and values. Packing problems are a sort of optimization problem in which objects are packed into containers. The combination of these two issues (called the Joint Problem) defines the set of things that must be placed in a bin such that the greatest amount of space in the bin is used by obtaining the maximum profits. The geometric forms associated with the objects and the container in the packing problem were combined with the value and weight conceptions of the Knapsack Problem. The Joint Problem's purpose is to maximize the total value sum of all things in the container while remaining within the container's maximum weight constraint and avoiding geometric crossings. In this article, a detailed study has been performed on various techniques of two-dimensional bin packing problem. The main categorization of mathematical models or algorithms, as well as the purpose of usage, are also studied and listed so that it will be useful for the researchers to have an insight on the working of each technique.

Assessing cognitive abilities in children is one of the contemporary challenges. The classical psychometric approach for assessing is based on a questionnaire of behavioral markers. It does not offer great prospects for developing a possible computer-aided assessment. Recent studies proposed a new approach that provides a computerized diagnosis based on evaluating shared intentionality in mother–child dyads. This new approach does not employ behavior markers. However, the proposed high-tech method is limited by the applied mathematical model described in probabilistic terms. This theoretical study develops the computer-aided approach, shaping the concept design for translational research. The new high-tech method also emulates the mother-newborn communication model by employing human–computer interaction to detect shared intentionality in mother–child dyads. Its novelty lies in introducing the mean baseline value into computations that forms the new mathematical model and algorithm, reducing the shortcomings of the former computerized method.

If all aspects of vehicle transportation management are handled manually, it is a time-consuming operation that produces significant errors. To address the issues raised above, it is important to design an automatic license plate recognition system that will automatically identify numbers from an image of the vehicle's front side. Recognition of license plates (LPR), commonly known because ANPR, has emerged as one of the most reliable techniques for vehicle surveillance in recent years. It can be utilized to achieve a range of objectives in numerous public spaces, including traffic safety regulation, automatic toll text collection, parking systems, and automatic car park systems. To automatically recognize license plates, different image processing, and techniques must be employed in a single application. Text localization, extraction, enhancement, segmentation, and identification algorithms are utilized to locate the number plate number in each frame of an image or video. The previous studies only partially covered the complete process of a typical LPR device, from picture acquisition to verification. This project developed an entire real-time, restriction-based license plate identification system. They put this mechanism in place to find lost automobiles and find out how vehicle insurance is doing.

In this computerized world, any individual can create deep fake videos, images, and audios, and it’s becoming a challenge to differentiate between the real and the fake ones. Such fake videos and images of a person that spread like wildfire on social media ultimately destroys the person's life by leaving a deep mark that is impossible to remove. Hence, it’s important to discern the fake and stop the circulation of these types of videos and images. Different technologies with different algorithms have been used to identify fake images. In this research, to distinguish fake and real images, long short-term memory network methods in deep learning technology are used to recognize faces from real-time dataset for better result because of its hidden layers and some inbuilt functions of this algorithm. Haar Cascade is used for detecting the images from real time; even though it’s an oldest one, it still has its own place in detecting face by providing more accuracy than others. Backpropagation is then used to feed both forward and backward, so that the values of bias can be reduced according to the predicted output. Sequence transformer makes the images get into cell state with continuous step-up time, so that the image will not collide while capturing. By using all these conditions, the output will be predicted with actual value which gives the desired outcome. Detecting the swapped and manipulated images from the real one is very difficult even though different algorithms are used. Therefore, long short-term memory networks are used in this research, because it's easy to identify fake images due to the continuous networks in the layer.

An avant-garde and futuristic Modified Quasi-Admittance Source Inverter (MQYSI) is put forward in this paper. Two administering modes govern how an admittance source inverter works. The redesigned q-YSI has a stronger boost voltage inversion and a lower duty ratio value to enhance output voltage quality compared to the previous YSIs. When compared to inverters of a more traditional design, this topology minimises the stress from voltage athwart the capacitors, solves the issue of a larger early inrush current, and makes decentralised generating systems possible. The inverter working with space vector-based pulse width modulation (SVPWM) is designed and analysed in MATLAB/Simulink environment and further contrasted with sine PWM technique in order to justify the better control technique.

The purpose of this study is to examine the integration of information communication and technology and the development of eTourism as a field in tourism industry, as well as the implications of ICT usage on the Tourism industry in Indonesia. Using various types of literature such as scientific articles, articles, and books, this study emphasizes and examines the importance of ICT and how these acts contribute to the tourism and hospitality sector. Tourism sector businesses should be praised for their use of ICT and use of current technology such as social networks for client engagement. The use of ICT has been carefully monitored and deployed at a low cost, needing little technical knowledge from the workforce. The purpose of this article is to raise awareness among scholars, educators, legislators, leisure entrepreneurs, and public officials on the usefulness of ICT applications in the tourism and hospitality industries in Indonesia. The work is limited to Indonesia, and the conclusion is based on previous research. The report is totally prepared using Indonesian culture. The ICT application is only used as an independent variable in the study. As a result, there may be some other factors influencing the tourist and hospitality industries in Indonesia.

In the recent few decades, due to the rapid and extensive growth of renewable sources of generation, the use of solar photovoltaics is at a large extent. Many individuals prefer a grid integrating system instead of a stand-alone system. However, this large usage of energy from solar photovoltaic with grid integration brings some severe challenges along with an increase in PV penetration. Generally, there are two major categories of these challenges: feeder level (distribution) and grid level. In this paper, the feeder-level challenges of solar photovoltaic integrated systems are discussed, and remedies are suggested to overcome these problems. In addition, analysis of PV penetration on several loads after the application of the capacitor and the impact of the on-load tap changer are presented.

User sentiment analysis from the online social media has become an interest, and importantly, Twitter opens a valuable source for user emotions, opinions, attitudes, confessions, and so on. However, analyzing the massive reviews and posts dumped in the Twitter media is a tedious and computationally ineffective scenario, raising the demand for implementing the automatic sentiment analysis models. Accordingly, Megaptera whale optimization-based adaptive long short-term memory (MWO-based adaptive LSTM) classifier is proposed for sentiment analysis, where the significance of the research relies in the proposed adaptive Megaptera whale hunt optimization that optimizes the classifier parameters toward the effective classification performance. The analysis of the model based on the accuracy reveals that the proposed model acquires the maximal performance with 98.48% of accuracy.

Electric vehicle (EV) is seen as a feasible approach to lowering pollution. Because the number of EVs is increasing, it is critical to include charging stations (CS). Conversely, many unscheduled EVs survive owing to a lack of charging options or adequate energy. This paper proposes an optimization-aware method for EVs charge scheduling. In this work, the EVs simulation is the initial step. The revelation of changing requests from EVs and available charging stations is carried out. Afterward, the charge scheduling algorithm is called for scheduling the EV. The charge scheduler algorithm termed as adaptive circle-inspired optimization algorithm (Adaptive CIOA) is proposed here. In this case, a new multi-objective fitness function is created with parameters such as distance parameter, remaining power, user preference, and charging cost. The EVs are assigned to the CS based on the scheme. Moreover, the characteristics of the EV charging scheduling systems are then revised to reveal the method's efficiency. The proposed adaptive CIOA tends to outperform with the lowest charging cost of 16.22, lowest fitness of 0.0091, maximum user convenience of 0.826, and maximum power of 10.06 J.

Artificial intelligence has a remarkable effect in every field. Various tools of artificial intelligence have proved themselves in different sectors. Whether it is the banking sector, health industry, or any other sector, it is. One such sector is cloud computing. Cloud computing is Internet-enabled technology and has proved to be a boon in information technology. The aim of this research is to study the two tools of AI in the field of cloud computing. In this paper, we are targeting fuzzy logic and ANN as a tool. A comparative investigation has been made by applying these tools in cloud computing-based resource scheduling. Fuzzy logic works very well in cloud computing because it handles uncertainties efficiently. ANN, on the other hand, provides a trained expert machine for predicting future behavior. After comparing the results in the case of both tools, we get that fuzzy logic outperforms artificial neural networks in the field of resource scheduling. This research is very helpful for researchers researching scheduling cloud-based jobs by setting up their priorities based on chosen attributes.

Artificial intelligence (AI) is used to personalise learning experiences for students, adapt to their individual needs and abilities, and provide real-time feedback on their progress, whereas virtual and augmented reality (VR/AR) are used to create immersive learning experiences that allow students to explore and interact with virtual environments and simulations. Online learning platforms provide students with access to educational resources and courses from anywhere in the world and allow for greater flexibility in terms of when and where they learn. Adaptive learning is a form of technology-enabled learning that adjusts to the student’s learning style, pace, and progress. This is done using algorithms that analyse student data, such as their performance on assessments, and adjust the content or pedagogy accordingly. There is also a wide-ranging emphasis on gamification for teaching–learning. Incorporating game-like elements into the learning process makes it more engaging and interactive for students. AI has completely revolutionised the formal education space. Thus, in this article, the researcher investigates how the most advanced technologies are currently being developed and used to enhance the way we educate students, and how it is integrated into the curriculum and classroom.

In the past years, understanding human behavior and classification of human actions and intentions by researching human activity recognition (HAR) using traditional pattern recognition has made great progress. In this paper, a novel approach is proposed based on convolutional neural network (CNN) and attention-based long short-term memory (attentıon LSTM) architecture. Human activity recognition (HAR) or recognizing human behavior is one of the challenging tasks due to human tendencies as the activities are not only complex but also multitasking. This deep learning-based long short-term memory network using convolutional neural networks (CNN-attentıon LSTM) architecture predicts the activities performed by humans and improves the accuracy by reducing the complexity of raw data and also by removing unnecessarily complex data. The convolutional layers act as a feature extractor, where they learn hierarchical representations of the image by applying multiple filters to the input image and passing the resulting feature maps through multiple activation functions. These learned features are then used as input to another classifier, such as a attention LSTM networks, to make the final prediction. On the internal UCF50 dataset, the proposed model achieves an 84.43% accuracy. The outcomes demonstrate that the suggested model is more robust and capable of activity detection than some of the results that have been reported.

The influence of globalization and the rapid development of technology is certainly very helpful for our daily lives. Every day people use technology for their daily needs, such as seeking knowledge. Developed technology makes it easier for us to live our daily lives, especially during a pandemic. These technological advances also help the community in dealing with problems that occur during a pandemic. During this pandemic, we are required to stay at home. Therefore, the government implemented large-scale social restrictions (PSBB). Stay at home if you do not have an important need, so that the virus does not spread. The government must find a way so that daily activities can be carried out even at home. One of the easiest ways to do this is by using an online meeting application. This online meeting application is an application that can be used to communicate with one or many people indirectly with the help of the Internet. This study aims to determine the impact of using online meeting applications during the pandemic and to see the impacts that occur in society. The use of easy-to-use online meeting applications is expected to increase consumer comfort in daily activities such as studying or working online during a pandemic.

Applying an artificial neurological network algorithm to a Web-based system, this study hopes to discover neural diseases. Critical to this study is identifying neurological states based on symptom data and analyzing them with an artificial neural network algorithm. The artificial brain network method is implemented; the artificial neural network algorithm is utilized to conduct an analysis that begins with the gathering of data in the form of symptoms and kinds of neural illnesses. The results of this study are based on an analysis of the identification system for neural disorders using an artificial neural network algorithm with an accuracy of 92%.

An important aspect of sustainable drug development is drug-target interaction. In cancer cell lines, the drug response target ratio is critical. It is important to estimate the drug reaction in a cancer cell line. In prior research, we employed ensemble algorithms with voting methods to predict medication response and achieved 97.5% accuracy. A hybrid ensemble algorithm for the revised drug response (HBEA) method is developed to improve drug-target strategy in cell lines. Rather than generating several homogeneous weak learners to generate a single model in the ensemble, this enhanced algorithm uses a diverse collection of weak learners such as random forest, Naive Bayes, and decision tree to create a strong meta-classifier. Cross-validation of hard and soft data would be used to accomplish this. The concentrations of various drugs are used as inputs, and the cell line predicts the relevant drug response. The goal of this enhanced ensemble algorithm is to suggest a new medicine based on a single licensed drug or a combination of drugs. This approach increased the drug responsiveness from 97.5 to 100%, according to our findings. The proposed method is applied in an open-source and freely available at https://decrease.fimm.fi .

The network load balancer acts as a reverse proxy and reroutes the traffic across a cluster of instances of the application service. The proposed integrated algorithm with a novelty of load sharing approach disperses the oncoming requests such that an optimal server ends up fulfilling the request. Along with that, the load balancer has a number of other utility functions that enable it to work similarly in a dynamic Web application. This proposed algorithm is evaluated with the horizontally-scaled architecture on the basis of latency and throughput as recommended in the RFC 2544 for evaluation on a LAN network. Through this research work, it was found that the ability of this integrated load balancing algorithm was successful in reducing the load for a set of four particular use-cases based upon reducing the response time while simultaneously maintaining all the characteristics of a distributed system.

Deep learning is a powerful computing strategy that has changed the landscape of computer vision. It has been used to tackle complicated cognitive tasks such as detecting abnormalities in videos. Anomalies in the video are events or objects in the footage that don’t fit the typical, learned patterns. Using deep learning, it is possible to automatically and in real-time identify unusual actions and objects like fights, riots, traffic rule violations, abrupt rushes, and the presence of weapons in restricted areas or abandoned luggage. Despite the challenges posed by video anomaly detection, this review offers a comprehensive assessment of published deep learning algorithms for the task. Future research can build on this work by understanding the existing methods to create more effective solutions. First, the challenges of video anomaly identification are discussed as the benefits of deep learning in anomaly detection. Furthermore, several types of abnormalities were explored, followed by diverse methodologies for anomaly identification. Furthermore, significant aspects of anomaly detection using deep learning, including learning approaches, were presented. Finally, numerous datasets used in anomaly detection were examined, followed by a discussion of deep learning-based algorithms for spotting video anomalies.

The ubiquity of Wi-Fi over the last decade has led to increased popularity of intelligent indoor positioning systems (IPS). In particular, machine learning has been recently utilized to develop intelligent IPS. Most of the existing research focus on developing intelligent IPS using balanced data. In this paper, we investigate a hitherto unexamined issue of imbalanced data in the context of machine learning-based IPS. We consider several traditional machine learning algorithms to determine the optimal method for training IPS on imbalanced data. We also analyze the effect of imbalance ratio on the performance of the IPS. The results show that the k-nearest neighbors algorithm provides the best approach to developing intelligent IPS for imbalanced data.

The mechanism for digital transformation of transportation systems is described in this article. In particular, cloud technologies, distributed ledger technology, blockchain technologies, big data technology, the Internet of Things, augmented and virtual reality technologies, technologies of artificial intelligence, and additive technologies were considered for their use in an intelligent transport system.

The metaverse is a global and immersive virtual shared environment formed by the fusion of virtual and physical reality, made possible through the use of augmented reality and virtual reality headsets. The vision of an immersive Internet as a massive, determined, united, and communal realm is vital to metaverse technology. In the context of digital reality, the metaverse was conceptualized as Web 3.0 or 3D Internet. A new open and decentralized virtual reality will emerge as a result of Web 3.0. The metaverse, which has entirely opened up in numerous fields, is made up of three key technical advancements: artificial intelligence, augmented reality, and virtual reality. The traditional tourism industry is having a serious impact because of the COVID-19 pandemic, which has a higher influence on lowering the country's GDP. As a result, it is critical to address this issue through digital technology with enhanced user experiences. This paper focuses on the fundamental concepts, components, and key challenges of metaverse technology in the travel and tourism sectors. This technology is still being developed and will require extensive research before it can be realized to its full potential.

Student motivation to learn becomes the encouragement that comes from himself. This encouragement makes students able to perform the learning activities to increase their knowledge. However, during this pandemic, students are studied at home using video conferencing facilities which could affect students’ learning motivation. This is the motivation of this paper to analyze the learning motivation of junior high school students in Information Systems during Covid-19. This study aims to determine the student’s learning outcomes, one of which is the learning motivation. The method used in this research is a survey method, wherein the data is collected from a sample or respondents of a population to represent the entire population using a questionnaire or interview. After the data needed is collected, in the form of a questionnaire, the analysis is discussed to determine the learning motivation. Likert scale is used to assist in analyzing the data. Based on the analysis of the data that has been collected, it is observed that junior high school students majoring in Information Systems during Covid-19 have relatively strong learning motivation. They have the willingness to succeed, have a driving force and need to learn, have hopes and dreams for the future, have an appreciation for learning, have learning activities that are interesting, and have a conductive learning environment, which allow the students to perform well.

Insight from product reviews from customers is important messages to the business. It provides feedback to the business and enhances the customers’ experiences and leading to sustainable competitive advantage for the business. Sentiment analysis can be performed from the customers’ product reviews regarding to particular product or service. Capturing sentiment analysis can be done automatically using machine learning techniques. This research aims to explore several machine learning algorithms to model automatic sentiment analysis on Indonesian product reviews. The product reviews data was gathered from an e-commerce platform in Indonesia. All of the product reviews are in the local (i.e. Indonesian) language. The research contributes to a state of the art of automatic sentiment analysis model on Indonesian product reviews. Eight algorithms and 4507 settings are explored to find the best setting and algorithm to model sentiment analysis in Indonesian product reviews. The results demonstrate that the best model was achieved by the one trained with five layers of artificial neural network with 99.1% for testing model accuracy, precision, recall, and F1-score. The best AUC of the model was 99.6%. The model requires 34.68 minutes to be trained.

Technology allows users to store a variety of datasets. Among these datasets, retrieving the correct source from the vast respiratory system is one of the most critical tasks. This extraction process is used in various applications in the engineering, medicine, and science domains. Extracting the image attributes from the stored respiratory is one of the more complex processes. Due to the intricacy of the input set, many scholars find these techniques among the most difficult undertakings. The process is the consolidation of various other attribute sets. From this combination, extracting the specific domain or specific patterns is more difficult for the investigators. The work is made considerably more complicated inputs like video sets. The proposed work finds the best solution for extracting image datasets using image attributes as an input. Investigational results prove that the suggested system will work with various types of inputs.

Essay writing is a common component of educational assessments used to evaluate student learning. Automated Text Scoring (ATS) is a collection of statistical and natural language processing techniques used to grade a text automatically on a scoring scale. Grading essays manually is time-consuming, expensive, and prone to inconsistencies. With the current student-to-teacher ratio disparity in the education system, there is an urgent need to develop a grading system. The goal of this project is to automate the essay evaluation process while providing feedback on the quality of work and detecting plagiarism. Students can use it to evaluate themselves. It can be used to evaluate or grade the essay-style responses to a specific question, for which sample answers are provided. The OCR detection and plagiarism checks can also be utilized independently.

Recently, intense research and development have been conducted in the research area of Artificial Intelligence, to find various techniques to implement this and create more advanced versions of the technology. Emotion Artificial Intelligence or opinion mining happens to be one such technique that provides advancement to Artificial Intelligence. By definition, opinion mining refers to the collection of different opinions or reviews of people with respect to a particular subject. Classification and categorization of these reviews in a consolidated manner of broad categories are known as summarization of these opinions. The representation of this information thus obtained in the forms of models, charts, and graphs is known as visualization of this summary. In the past couple of years, online platforms have increased rapidly. These platforms advanced themselves to not just providing items that can be purchased but also other activities that can be done with the help of the internet such as reservations, financial transactions. The services also include purchasing/booking various kinds of tickets, for traveling or for entertainment, or ordering food using these online portals. In the following work, the various restaurants which deliver food via online portals and their services have been discussed with respect to the feedbacks or opinions received by customers. In this work, the various reviews left by the customers have been summarized and presented and categorized manner classifying them as positive, negative. These have also been visualized for a better understanding and help in comparison to two or more restaurants. The main goal of this work is to ease the difficulty faced by people to decide which place to go and will also increase the efficiency as the time consumed will be reduced as compared to visiting each website and reading the reviews.

This paper provides a comprehensive review of the performance comparison of object detection-based deep learning techniques. Object detection is a crucial task in computer vision that involves identifying and localizing objects within an image, signal, or video. Deep learning has emerged as a promising approach for object detection due to its ability to learn and extract features automatically. In this paper, we review the recent literature on object detection-based deep learning techniques, including region-based, edge-based, texture-based, etc. This paper also analyzes the performance metrics used to evaluate these techniques. Through this review, we aim to provide a better view of the strengths and weaknesses of various object detection techniques and their applicability to different applications. Our findings suggest that while some techniques perform better in specific scenarios, there is no single best-performing technique for all scenarios. Therefore, the choice of object detection technique should be based on the specific application and performance requirements.

Information criteria are used in many applications including statistical model selection and intelligent systems. The traditional information criteria such as the Akaike information criterion (AIC) do not always provide an adequate penalty on the number of model covariates. To address this issue, we propose a novel method for evaluating statistical models based on information criterion. The proposed method, called regularized information criterion (RIL), modifies the penalty term in AIC to reduce model overfitting. The results of numerical experiments show that RIL provides a better reflection of model predictive error than AIC. Thus, RIL can be a useful tool in model selection.

Software development using agile techniques is gaining popularity nowadays just because of its benefits over traditional development. Many software practitioners and researchers are taking interest in agile development, but they need to be aware about all agile techniques currently available and about all the challenges they will face while migrating to the agile technologies. It has been seen that only big companies are getting benefits of agile, but not small or startup companies, while agile can be beneficial more for such type of organization. More awareness for agile development should be spread. This paper presents agile goals, principles, team ethics, and moral values with different agile methodologies which will help to create a truly agile team for developing software.

One of the biggest issues in metropolitan areas is traffic congestion, despite having well-planned road systems and adequate infrastructure. The main cause of this issue is the 40% annual increase in the number of cars on the road. Most current traffic control systems are fixed cycle types, which always cycle through red, yellow, and green. The deployment of these pilots is accompanied by the deployment of traffic police officers to maintain order in the streets. Unlike human traffic cops, these inflexible systems cannot adjust to changing circumstances on the fly. Intelligent traffic management systems are needed immediately. In order to measure traffic volume, our proposed system will use AI and image processing to analyse live feeds from cameras placed at intersections. The amount of vehicles passing through the intersection is predicted to increase by around 32% based on simulation results, which is a substantial gain over the status quo. More training and calibration of the model with actual CCTV data can bring about significant improvements in the system's performance.

This paper presents a strategy for charging the combined energy storage (CES) system that contains supercapacitor and battery. When battery suddenly charges or discharges very quickly within a few seconds, reduction in the life of the battery occurs. To enhance the battery life, supercapacitor is used along with the battery. Whenever the hybrid storage system is connected to the load, it should give power to the load only if it has enough energy. When the battery-state of charge is below a defined level it has to be charged. Likewise, whenever the voltage across the supercapacitor is below the certain level, it has to be charged in order to supply the load. So, here charging strategy is based on battery-state of charge (SOC) and voltage across the supercapacitor.

In this work, using the flocking algorithm based on reaction directly controls agents in the swarm on a virtual structure as a single control target, achieve obstacle avoidance and flying in a swarm. The human operator controls the self-organized swarm with a higher level of autonomy, flying in fixed formations or switching between predefined formations, thus allowing each quadcopter follows the desired flight trajectory throughout the maneuver processing and completes actions to maintain, move, or switch formations.

In making strategic decisions in the military, military aircraft detection has become increasingly crucial. The identification of military aircraft continues to be a problematic issue. In operations and wars, military aircraft detection is crucial for detecting unknown aircraft. The difficulty is always in accurately recognizing the unfamiliar aircraft, regardless of class and orientation. Military aircraft, such as stealth aircraft, are still difficult to detect because stealth aircraft are more challenging to detect or track using conventional radar. Still, these aircraft can be detected using object detection. In this article, we proposed identifying five types of airplanes independent of class or direction using object detection. The You Only Look Once version 5 (YOLOv5) method and the PyTorch military aircraft dataset were used to identify various aircraft. The identification of different aircraft was discovered using the YOLOv5 algorithm and the PyTorch military aircraft dataset. Bounding boxes for the dataset, data pre-processing, and data augmentation are made using Roboflow. The goal is to employ computer vision and object identification to identify whether a particular aircraft is a military aircraft. This military aircraft detection may be used in the border area, air force, and marine force.

In this scientific paper, the researcher develops a course on “Safety, Fairness, Privacy, and Ethics of Artificial Intelligence” (SFPE-AI) designed for university students. The course aims to provide students with a comprehensive understanding of the technical and ethical issues associated with the development and deployment of AI systems. The course is designed to be interdisciplinary, drawing on concepts and techniques from computer science, philosophy, and law. The curriculum is divided into four modules: safety, fairness, privacy, and ethics. To facilitate student learning, the course employs a variety of pedagogical tools, such as interactive lectures, case studies, group discussions, and hands-on projects. The case studies used in the course include real-world examples of AI applications and their associated ethical and societal implication, thus providing students with a diverse perspective on the challenges and opportunities associated with AI. After the completion of this course, students are expected to understand the technical and ethical issues associated with AI, design and develop AI systems that are safe, fair, private, and ethical, and critically evaluate the societal implications of AI. The SFPE-AI course is expected to prepare the next generation of AI professionals to build responsible and trustworthy AI systems. The course will also serve as a model for other universities and educational institutions looking to integrate the discussion of AI safety, fairness, privacy, and ethics into their curriculum.

The COVID-19 outbreak is the greatest global health catastrophe and the largest threat facing humanity ever since the Second World War. It started in the Asia region in late 2019 and since then it has spread to almost every country on the planet, causing millions of people to be sick. The deadly disease has now reached the tragic milestone of one million deaths. It has surely impacted the world in most factors, especially in the education factor. The government has introduced social-distancing protocols that affect the teaching and learning process to decrease the infection rate. To cope with the situation, schools and universities implement long-distance teaching with video conference methods. This long-distance learning and teaching with video conference methods has several impacts on undergraduate students because, they are not accustomed to learning using this method, and that some courses require direct practices in the field. Therefore, this study is to gain a conclusion if long-distance teaching with video conference methods is effective, if it brings positive effects to the students in their learning process and if it motivates to study. This study’s sampling method is submitted through Google form wherein the questionnaires are aimed at undergraduate students to obtain their outcomes about the video conference method used in their learning process.

One abbreviation could have multiple meanings, depending on the area in which it is used. Of course, within a single domain, its meaning can be ambiguous. This is also the case with medical data, where the challenge for artificial intelligence techniques is to identify them correctly and efficiently, as any mistake can cost human lives. This paper disambiguates medical abbreviations from Medical Dataset for Abbreviation Disambiguation (MeDAL). It compares the DistilBERT Transformer network with bag-of-words encoding-based classification and typical machine learning algorithms such as XGBoost, Random Forest, and Decision Tree. The study shows that DistilBERT Transformer achieved worse accuracy and F1 than the best baseline model when each tested shortcut was examined. The study shows that DistilBERT Transformer achieved worse accuracy and F1 than the best baseline model when each tested shortcut was examined. The basic model was better in the categories of shortcuts that have less than 500 examples and 500–5000 examples. The results in the group with 5000–20,000 examples are relatively equal for both approaches.

With the increase of automobiles on roadways, a vehicle system for recognizing numbers is necessary which can be implemented in the most simple way possible for effective traffic control, surveillance, smart parking management systems, and tollbooth record management. The system should be able to recognize number plates from live video feeds so that it can be used for automated survelliance and other practical purposes. The system takes a live video/image and localizes the license plate. License plate localization is performed using a trained YOLOv4 or YOLOv7 object detection models. Real-ESRGAN, a super-resolution technique is applied on the obtained license plate to enhance the image quality. Next, an OCR such as Tessaract and Easy OCR are used to perform image-to-text conversion extracting the required characters present on the number plate. The resultant output is validated as per the limitations of the Indian vehicle registration number format to improve accuracy.