DeepVision: Enhanced Drone Detection and Recognition in Visible Imagery through Deep Learning Networks.

Drones are increasingly capturing the world's attention, transcending mere hobbies to revolutionize areas such as engineering, disaster aid, logistics, and airport protection, among myriad other fascinating applications. However, there is growing concern about the risks that they pose to physical infrastructure, particularly at airports, due to potential misuse. In recent times, numerous incidents involving unauthorized drones at airports disrupting flights have been reported. To solve this issue, this article introduces an innovative deep learning method proposed to effectively distinguish between drones and birds. Evaluating the suggested approach with a carefully assembled image dataset demonstrates exceptional performance, surpassing established detection systems previously proposed in the literature. Since drones can appear extremely small compared to other aerial objects, we developed a robust image-tiling technique with overlaps, which showed improved performance in the presence of very small drones. Moreover, drones are frequently mistaken for birds due to their resemblances in appearance and movement patterns. Among the various models tested, including SqueezeNet, MobileNetV2, ResNet18, and ResNet50, the SqueezeNet model exhibited superior performance for medium area ratios, achieving higher average precision (AP) of 0.770. In addition, SqueezeNet's superior AP scores, faster detection times, and more stable precision-recall dynamics make it more suitable for real-time, accurate drone detection than the other existing CNN methods. The proposed approach has the ability to not only detect the presence or absence of drones in a particular area but also to accurately identify and differentiate between drones and birds. The dataset utilized in this research was obtained from a real-world dataset made available by a group of universities and research institutions as part of the 2020 Drone vs. Bird Detection Challenge. We have also tested the performance of the proposed model on an unseen dataset, further validating its better performance.

Nature-inspired spider web shaped UHF RFID reader antenna for IoT and healthcare applications.

This paper proposes a nature-inspired spider web-shaped ultra-high frequency (UHF) radio frequency identification (RFID) reader antenna and battery-free sensor-based system for healthcare applications. This antenna design consists of eight concentric decagons of various sizes and five straight microstrip lines.These lines are connected to the ground using 50 [Formula: see text] resistors from both ends, except for one microstrip line that is reserved for connecting a feeding port. The reader antenna design features fairly strong and uniform electric and magnetic field characteristics. It also exhibits wideband characteristics, covering whole UHF RFID band (860-960 MHz) and providing a tag reading volume of 200 [Formula: see text] 200 [Formula: see text] 20 mm[Formula: see text]. Additionally, it has low gain characteristics, which are necessary for the majority of nearfield applications to prevent the misreading of other tags. Moreover, the current distribution in this design is symmetric throughout the structure, effectively resolving orientation sensitivity issues commonly encountered in low-cost linearly polarized tag antennas. The measurement results show that the reader antenna can read medicine pills tagged using low-cost passive/battery-free RFID tags, tagged expensive jewelry, intervenes solution, and blood bags positioned in various orientations. As a result, the proposed reader antenna-based system is a strong contender for near-field RFID, healthcare, and IoT applications.

A Two-Stage Multi-Agent EV Charging Coordination Scheme for Maximizing Grid Performance and Customer Satisfaction.

Advancements in technology and awareness of energy conservation and environmental protection have increased the adoption rate of electric vehicles (EVs). The rapidly increasing adoption of EVs may affect grid operation adversely. However, the increased integration of EVs, if managed appropriately, can positively impact the performance of the electrical network in terms of power losses, voltage deviations and transformer overloads. This paper presents a two-stage multi-agent-based scheme for the coordinated charging scheduling of EVs. The first stage uses particle swarm optimization (PSO) at the distribution network operator (DNO) level to determine the optimal power allocation among the participating EV aggregator agents to minimize power losses and voltage deviations, whereas the second stage at the EV aggregator agents level employs a genetic algorithm (GA) to align the charging activities to achieve customers' charging satisfaction in terms of minimum charging cost and waiting time. The proposed method is implemented on the IEEE-33 bus network connected with low-voltage nodes. The coordinated charging plan is executed with the time of use (ToU) and real-time pricing (RTP) schemes, considering EVs' random arrival and departure with two penetration levels. The simulations show promising results in terms of network performance and overall customer charging satisfaction.

Dental caries detection using a semi-supervised learning approach.

Early diagnosis of dental caries progression can prevent invasive treatment and enable preventive treatment. In this regard, dental radiography is a widely used tool to capture dental visuals that are used for the detection and diagnosis of caries. Different deep learning (DL) techniques have been used to automatically analyse dental images for caries detection. However, most of these techniques require large-scale annotated data to train DL models. On the other hand, in clinical settings, such medical images are scarcely available and annotations are costly and time-consuming. To this end, we present an efficient self-training-based method for caries detection and segmentation that leverages a small set of labelled images for training the teacher model and a large collection of unlabelled images for training the student model. We also propose to use centroid cropped images of the caries region and different augmentation techniques for the training of self-supervised models that provide computational and performance gains as compared to fully supervised learning and standard self-supervised learning methods. We present a fully labelled dental radiographic dataset of 141 images that are used for the evaluation of baseline and proposed models. Our proposed self-supervised learning strategy has provided performance improvement of approximately 6% and 3% in terms of average pixel accuracy and mean intersection over union, respectively as compared to standard self-supervised learning. Data and code will be made available to facilitate future research.

Personalized wearable electrodermal sensing-based human skin hydration level detection for sports, health and wellbeing.

Personalized hydration level monitoring play vital role in sports, health, wellbeing and safety of a person while performing particular set of activities. Clinical staff must be mindful of numerous physiological symptoms that identify the optimum hydration specific to the person, event and environment. Hence, it becomes extremely critical to monitor the hydration levels in a human body to avoid potential complications and fatalities. Hydration tracking solutions available in the literature are either inefficient and invasive or require clinical trials. An efficient hydration monitoring system is very required, which can regularly track the hydration level, non-invasively. To this aim, this paper proposes a machine learning (ML) and deep learning (DL) enabled hydration tracking system, which can accurately estimate the hydration level in human skin using galvanic skin response (GSR) of human body. For this study, data is collected, in three different hydration states, namely hydrated, mild dehydration (8 hours of dehydration) and extreme mild dehydration (16 hours of dehydration), and three different body postures, such as sitting, standing and walking. Eight different ML algorithms and four different DL algorithms are trained on the collected GSR data. Their accuracies are compared and a hybrid (ML+DL) model is proposed to increase the estimation accuracy. It can be reported that hybrid Bi-LSTM algorithm can achieve an accuracy of 97.83%.

Economic load dispatch using memetic sine cosine algorithm.

In this paper, the economic load dispatch (ELD) problem which is an important problem in electrical engineering is tackled using a hybrid sine cosine algorithm (SCA) in a form of memetic technique. ELD is tackled by assigning a set of generation units with a minimum fuel costs to generate predefined load demand with accordance to a set of equality and inequality constraints. SCA is a recent population based optimizer turned towards the optimal solution using a mathematical-based model based on sine and cosine trigonometric functions. As other optimization methods, SCA has main shortcoming in exploitation process when a non-linear constraints problem like ELD is tackled. Therefore, ß -hill climbing optimizer, a recent local search algorithm, is hybridized as a new operator in SCA to empower its exploitation capability to tackle ELD. The proposed hybrid algorithm is abbreviated as SCA- ß HC which is evaluated using two sets of real-world generation cases: (i) 3-units, two versions of 13-units, and 40-units, with neglected Ramp Rate Limits and Prohibited Operating Zones constraints. (ii) 6-units and 15-units with Ramp Rate Limits and Prohibited Operating Zones constraints. The sensitivity analysis of the control parameters for SCA- ß HC is initially studied. The results show that the performance of the SCA- ß HC algorithm is increased by tuning its parameters in proper value. The comparative evaluation against several state-of-the-art methods show that the proposed method is able to produce new best results for some tested cases as well as the second-best for others. In a nutshell, hybridizing ß HC optimizer as a new operator for SCA is very powerful algorithm for tackling ELD problems.

An Intensive and Comprehensive Overview of JAYA Algorithm, its Versions and Applications.

In this review paper, JAYA algorithm, which is a recent population-based algorithm is intensively overviewed. The JAYA algorithm combines the survival of the fittest principle from evolutionary algorithms as well as the global optimal solution attractions of Swarm Intelligence methods. Initially, the optimization model and convergence characteristics of JAYA algorithm are carefully analyzed. Thereafter, the proposed versions of JAYA algorithm have been surveyed such as modified, binary, hybridized, parallel, chaotic, multi-objective and others. The various applications tackled using relevant versions of JAYA algorithm are also discussed and summarized based on several problem domains. Furthermore, the open sources code of JAYA algorithm are identified to provide enrich resources for JAYA research communities. The critical analysis of JAYA algorithm reveals its advantages and limitations in dealing with optimization problems. Finally, the paper ends up with conclusion and possible future enhancements suggested to improve the performance of JAYA algorithm. The reader of this overview will determine the best domains and applications used by JAYA algorithm and can justify their JAYA-related contributions.

Making assembly line in supply chain robust and secure using UHF RFID.

This paper presents a block-chain enabled inkjet-printed ultrahigh frequency radiofrequency identification (UHF RFID) system for the supply chain management, traceability and authentication of hard to tag bottled consumer products containing fluids such as water, oil, juice, and wine. In this context, we propose a novel low-cost, compact inkjet-printed UHF RFID tag antenna design for liquid bottles, with 2.5 m read range improvement over existing designs along with robust performance on different liquid bottle products. The tag antenna is based on a nested slot-based configuration that achieves good impedance matching around high permittivity surfaces. The tag was designed and optimized using the characteristic mode analysis. Moreover, the proposed RFID tag was commercially tested for tagging and billing of liquid bottle products in a conveyer belt and smart refrigerator for automatic billing applications. With the help of block-chain based product tracking and a mobile application, we demonstrate a real-time, secure and smart supply chain process in which items can be monitored using the proposed RFID technology. We believe the standalone system presented in this paper can be deployed to create smart contracts that benefit both the suppliers and consumers through the development of trust. Furthermore, the proposed system will paves the way towards authentic and contact-less delivery of food, drinks and medicine in recent Corona virus pandemic.

Kinetics and Thermodynamics of Acoustic Release of Doxorubicin from Non-stabilized polymeric Micelles.

This paper studies the thermodynamic characteristics of ultrasound-activated release of Doxorubicin (Dox) from micelles. The release and re-encapsulation of Dox into Pluronic® P105 micelles was measured by recording the fluorescence of a solution of 10 µg/ml Dox and 10% wt P105 polymer in phosphate-buffered saline, during and after insonation by ultrasound at three temperatures, (25 °C, 37 °C and 56 °C). The experimental data were modeled using a previously-published model of the kinetics of the system. The model was simplified by the experimental measurement of one of the parameters, the maximum amount of Dox that can be loaded into the poly(propyleneoxide) cores of the micelles, which was found to be 89 mg/ml PPO and 150 mg Dox/ml PPO at 25 °C and 37 °C, respectively. From the kinetic constants and drug distribution parameters, we deduced the thermodynamic activation energy for micelle re-assembly and the residual activation energies for micelle destruction. Our model showed that the residual activation energy for destruction decreased with increasing acoustic intensity. In addition, higher temperatures were found to encourage micelle destruction and hinder micelle re-assembly.

Spatio-temporal feature-extraction techniques for isolated gesture recognition in Arabic sign language.

This paper presents various spatio-temporal feature-extraction techniques with applications to online and offline recognitions of isolated Arabic Sign Language gestures. The temporal features of a video-based gesture are extracted through forward, backward, and bidirectional predictions. The prediction errors are thresholded and accumulated into one image that represents the motion of the sequence. The motion representation is then followed by spatial-domain feature extractions. As such, the temporal dependencies are eliminated and the whole video sequence is represented by a few coefficients. The linear separability of the extracted features is assessed, and its suitability for both parametric and nonparametric classification techniques is elaborated upon. The proposed feature-extraction scheme was complemented by simple classification techniques, namely, K nearest neighbor (KNN) and Bayesian, i.e., likelihood ratio, classifiers. Experimental results showed classification performance ranging from 97% to 100% recognition rates. To validate our proposed technique, we have conducted a series of experiments using the classical way of classifying data with temporal dependencies, namely, hidden Markov models (HMMs). Experimental results revealed that the proposed feature-extraction scheme combined with simple KNN or Bayesian classification yields comparable results to the classical HMM-based scheme. Moreover, since the proposed scheme compresses the motion information of an image sequence into a single image, it allows for using simple classification techniques where the temporal dimension is eliminated. This is actually advantageous for both computational and storage requirements of the classifier.

Extraction of fetal electrocardiogram using adaptive neuro-fuzzy inference systems.

In this paper, we investigate the use of adaptive neuro-fuzzy inference systems (ANFIS) for fetal electrocardiogram (FECG) extraction from two ECG signals recorded at the thoracic and abdominal areas of the mother's skin. The thoracic ECG is assumed to be almost completely maternal (MECG) while the abdominal ECG is considered to be composite as it contains both the mother's and the fetus' ECG signals. The maternal component in the abdominal ECG signal is a nonlinearly transformed version of the MECG. We use an ANFIS network to identify this nonlinear relationship, and to align the MECG signal with the maternal component in the abdominal ECG signal. Thus, we extract the FECG component by subtracting the aligned version of the MECG signal from the abdominal ECG signal. We validate our technique on both real and synthetic ECG signals. Our results demonstrate the effectiveness of the proposed technique in extracting the FECG component from abdominal signals of very low maternal to fetal signal-to-noise ratios. The results also show that the technique is capable of extracting the FECG even when it is totally embedded within the maternal QRS complex.

A novel technique for the extraction of fetal ECG using polynomial networks.

In this paper, we propose a novel technique for extracting fetal electrocardiogram (FECG) from a thoracic ECG recording and an abdominal ECG recording of a pregnant woman. The polynomial networks technique is used to nonlinearly map the thoracic ECG signal to the abdominal ECG signal. The FECG is then extracted by subtracting the mapped thoracic ECG from the abdominal ECG signal. Visual test results obtained from real ECG signals show that the proposed algorithm is capable of reliably extracting the FECG from two leads only. The visual quality of the FECG extracted by the proposed technique is found to meet or exceed that of published results using other techniques such as the independent component analysis.

Spectral subtraction and cepstral distance for enhancing EEG entropy.

Electroencephalographic (EEG) signals are normally acquired in the presence of background noise which causes inaccurate or false entropy measurement throughout the signal. In this paper, spectral subtraction is used to pre-process EEG signals to improve the accuracy of computing the subband wavelet entropy (SWE). The silent period in the EEG signal is identified via cepstral distance which allows its entropy to be set to zero. The EEG signal presented in this paper represents various stages of brain recovery obtained from a rodent following global cerebral ischemia. The various subband entropies are calculated using wavelet decomposition in EEG subbands, namely Delta, Theta, Alpha, Beta and Gamma. The utilization of spectral subtraction improved the accuracy of the SWE as compared to energy thresholding.