Anomaly Detection (AD) is an important area of research because it helps identify outliers in data, enabling early detection of errors, fraud, and potential security breaches. Machine Learning (ML) can be utilized for distinct AD systems, and Generative Adversarial Networks (GANs) have emerged as a promising technique due to their ability to generate new data that closely resembles a given dataset, allowing for the creation of realistic images, videos, audio, text, and other types of synthetic data. This paper explores state-of-the-art approaches in AD using GANs. The paper starts by providing a comprehensive overview of ML techniques for AD, including supervised, unsupervised, and semi-supervised approaches. This survey also explores various AD approaches based on GANs and provides an application-based classification of GANs-based AD approaches in the Internet-of-Things (IoT), Industrial IoT, Digital Healthcare, Energy Management Systems, and Cellular Network domains. Moreover, the paper discusses several datasets used in evaluating the performance of GANs-based AD techniques such as BOT-IoT, TON-IoT, CIC-IoT, CIC-IDS, and NSL-KDD. These datasets serve as valuable resources for researchers and practitioners to develop and test AD systems, particularly in the context of IoT and network security. Furthermore, the paper discusses the challenges and limitations of GANs-based AD techniques and proposes future research directions to address these challenges.

Recent advancements have highlighted the effectiveness of combining deep learn-ing models with attention mechanisms for fingerprint recognition. However, existing approaches often struggle to achieve high accuracy when processing low-quality or latent fingerprint images, as they fail to capture important critical ridge and valley patterns. To address this limitation, we propose a Multi-head Self-Attention based EfficientNetB4 model with advanced preprocessing techniques for enhanced fingerprint matching, leveraging the efficient feature extraction capability of EfficientNet. The proposed approach consists of three key components: advanced preprocessing techniques, feature extraction using EfficientNet, and a Multi-Head Self-Attention (MHSA) mechanism to enhance feature representation. Preprocessing steps include image augmentation to increase data diversity, Contrast Limited Adaptive Histogram Equalization for improving contrast, and Enhanced Super-Resolution Generative Adversarial Network for super-resolution enhancement, ensuring the clarity of fingerprint patterns even in noisy or distorted images. EfficientNet processes these preprocessed images to extract spatial and structural features, while the MHSA mechanism assigns varying importance to different features, enabling the network to focus on key fingerprint details. Hyperparameter tuning was employed to optimize model performance and ensure adaptability across diverse datasets. To validate the performance of ENET-EMHSA, experiments were conducted on benchmark datasets, including FVC 2000, FVC 2002, and FVC 2004. The proposed ENET-EMHSA model demonstrated state-of-the-art accuracy, achieving 99.57%, 99.72%, and 99.86% on the respective datasets, while maintaining a notably lower Equal Error Rate (EER) across all evaluations, thereby outperforming existing fingerprint matching methods. The ENET-EMHSA demonstrates the capability to differentiate fine-grained fingerprint patterns, improving the accuracy and robustness of biometric identification systems.

Most of the existing face photo de-meshing methods have accomplished promising results; there are certain quality problems with these methods like the inpainted regions would appear blurry and unpleasant boundaries becoming visible. Such artifacts cause generated face photos unreal. Therefore, we propose an effective image-to-image translation framework called Face De-meshing Using Generative Adversarial Networks (De-mesh GANs). The De-mesh GANs is a two-stage model: (i) binary mask generating module, is a three convolution layers-based encoder–decoder network architecture that automatically generates a binary mask for the meshed region, and (ii) face photo de-meshing module, is a GANs-based network that eliminates the mesh mask and synthesizes the meshed area. An arrangement of careful losses (reconstruction loss, adversarial loss, and perceptual loss) is used to reassure the better quality of the de-mesh face photos. To facilitate the training of the proposed model, we have designed a dataset of clean/corrupted photo pairs using the CelebA dataset. Qualitative and quantitative evaluations of the De-mesh GANs on real-world corrupted face photo images show better performance than the previously proposed face photo de-meshing models. Furthermore, we also offer the ablation study for performance assessment of the additional network i.e., perceptual network.

Abstract Currently, there is no unified Electronic Health Record (EHR) system connecting major healthcare organizations such as hospitals, medical centers, and specialists. Blockchain technology, with its unique features, provides an ideal platform for developing a large‐scale electronic health record system. In this article, the authors introduce HealthChain, a novel blockchain‐based secure EHR system that integrates advanced encryption techniques, a robust consent management system, cross‐platform interoperability, and enhanced scalability. Unlike existing EHR systems, HealthChain allows patients to have comprehensive control over their health data, ensuring that access is strictly regulated according to their preferences. The experimental results demonstrate several significant improvements over traditional EHR systems. HealthChain reduces data access times by 30%, and its interoperability rate with various healthcare systems is 40% higher than that of other blockchain‐based EHR solutions. Security is greatly enhanced, with HealthChain experiencing 50% fewer data breaches due to its advanced encryption and smart contract‐based access controls. Moreover, patient satisfaction has increased by 35% as a result of better control and access to their health records. These findings highlight HealthChain as not only a feasible and effective solution for managing health records but also a significant advancement over existing systems.

Abstract With the rapid expansion of the Internet of Things (IoT), cloud storage has emerged as one of the cornerstones of data management, facilitating ubiquitous access and seamless sharing of information. However, with the involvement of a third party, traditional cloud‐based storage systems are plagued by security and availability concerns, stemming from centralized control and management architectures. A novel blockchain‐IoT model that leverages blockchain technology and decentralized storage mechanisms to address these challenges is presented. The model combines the Ethereum blockchain, interplanetary file system, and attribute‐based encryption to ensure secure and resilient storage and sharing of IoT data. Through an in‐depth exploration of the system architecture and underlying mechanisms, it is demonstrated how the framework decouples storage functionality from resource‐constrained IoT devices, mitigating security risks associated with on‐device storage. In addition, data owners and users can easily exchange data with one another through the use of Ethereum smart contracts, fostering a collaborative environment and providing incentives for data sharing. Moreover, an incentive mechanism powered by the FileCoin cryptocurrency is introduced, which motivates and ensures data sharing transparency and integrity between stakeholders. Furthermore, in the proposed blockchain‐IoT model, the proof‐of‐authority system consensus algorithm has been replaced by a delegated proof‐of‐capacity system, which reduces transaction costs and energy consumption. Using the Rinkby Ethereum official testing network, the proposed model has been demonstrated to be feasible and economical, emphasizing its potential to redefine IoT data management.

The cloud-fog computing paradigm is a novel hybrid computing model that delivers computational services to Fog nodes situated near data sources. This paradigm features a volatile and dynamic network topology, comprising heterogeneous IoT devices with varying computational capabilities, alongside a large number of diverse end-user requests. These complexities present significant challenges for researchers in establishing a robust, energy-efficient, and reliable communication environment. Efficient and optimal task scheduling is among these challenges, as it involves finding appropriate computing resources for processing tasks. Assigning tasks to fog nodes reduces delay but increases energy consumption, while routing tasks to cloud servers conserves energy but prolongs transmission delay. Therefore, it is essential to develop an optimal task scheduling algorithm for a reliable, delay-efficient, and energy-efficient communication environment. To address this, we propose a Multi-objective Harris Hawks Optimization (HHO)-based task scheduling algorithm (MoHHOTS) for cloud-fog computing networks, aiming to optimize task scheduling with the objectives of minimizing delay and energy consumption. MoHHOTS is implemented in MATLAB and evaluated against state-of-the-art benchmark algorithms, including MOGWO and the cloud-fog cooperation algorithm. Leveraging the high convergence and stochastic operators of the HHO algorithm, alongside a balanced approach to iteration between diversification and intensification, the proposed algorithm provides a set of trade-off solutions via the Pareto-optimal Front. Simulation results demonstrate the efficacy of the proposed solution, achieving improvements of up to 25% over a similar scheduling algorithm in terms of optimizing transmission delay and energy consumption.

Abstract In recent years, the COVID-19 outbreak has affected humanity across the globe. The frequent symptoms of COVID-19 are identical to the normal flu, such as fever and cough. COVID-19 disseminates rapidly, and it has become a prominent cause of mortality. Nowadays, the new wave of COVID-19 has created significant impacts in China. This virus can have detrimental effects on people of all ages, particularly the elderly, due to their weak immune systems. The real-time polymerase chain reaction (RT-PCR) examination is typically performed for the identification of coronavirus. RT-PCR is an expensive and time requiring method, accompanied by a significant rate of false negative detections. Therefore, it is mandatory to develop an inexpensive, fast, and reliable method to detect COVID-19. X-ray images are generally utilized to detect diverse respiratory conditions like pulmonary infections, breathlessness syndrome, lung cancer, air collection in spaces of the lungs, etc. This study has also utilized a chest X-ray dataset to identify COVID-19 and pneumonia. In this research work, we proposed a novel deep learning model CP_DeepNet, which is based on a pre-trained deep learning model such as SqueezeNet, and further added three blocks of convolutional layers to it for assessing the classification efficacy. Furthermore, we employed a data augmentation method for generating more images to overcome the problem of model overfitting. We utilized COVID-19 radiograph dataset for evaluating the performance of the proposed model. To elaborate further, we obtained significant results with accuracy of 99.32%, a precision of 100%, a recall of 99%, a specificity of 99.2%, an area under the curve of 99.78%, and an F1-score of 99.49% on CP_DeepNet for the binary classification of COVID-19 and normal class. We also employed CP_DeepNet for the multiclass classification of COVID-19, pneumonia, and normal person, in which CP_DeepNet achieved accuracy, precision, recall, specificity, area under curve, and F1-score of 99.62%, 99.79%, 99.52%, 99.69, 99.62, and 99.72%, respectively. Comparative analysis of experimental results with different preexisting techniques shows that the proposed model is more dependable as compared to RT-PCR and other prevailing modern techniques for the detection of COVID-19.

Abstract Timely and accurate detection and diagnosis of heart disorders is a significant problem in the medical community since the mortality rate is increasing. Pulsing of cardiac structures and blood turbulence creates heart sounds recorded and detected through Phonococardiogram (PCG). As a non‐invasive technique, PCG signals have a strong ability to be used for designing automatic classification of possible heart disorders. This paper presents an expert system design for the detection and classification of PCG signals for five classes, namely, healthy, aortic stenosis, mitral stenosis, mitral regurgitation, and mitral valve prolapse. In this work, a single‐channel PCG signal is first decomposed using Empirical Mode Decomposition (EMD) into different modes known as intrinsic mode functions (IMFs). Manual signal analysis is applied to identify the relevant IMFs to construct a preprocessed signal. We proposed an automated energy‐based signal reconstruction through IMFs. The proposed algorithms automatically identify the relevant IMFs and added them together to form a preprocessed signal. After preprocessing, the first nine features of Mel Frequency Cepstral Coefficients (MFCC) were computed and passed to several classification methods such as Fine Tree, Quadratic Discriminant, Kernel Naive Bayes, Support Vector Machines (SVM), Fine K‐Nearest Neighbours (Fine‐KNN), Ensemble Bagged Trees and Neural Network. The best performance of 99.3% accuracy was obtained via Fine‐KNN using 10‐fold cross‐validation. The proposed method was evaluated on a publicly available dataset of heart sounds. The proposed method demonstrated improved performance as compared to the existing state‐of‐the‐art methods.

The technology of visual servoing, with the digital twin as its driving force, holds great promise and advantages for enhancing the flexibility and efficiency of smart manufacturing assembly and dispensing applications. The effective deployment of visual servoing is contingent upon the robust and accurate estimation of the vision‐motion correlation. Network‐based methodologies are frequently employed in visual servoing to approximate the mapping between 2D image feature errors and 3D velocities, offering promising avenues for improving the accuracy and reliability of visual servoing systems. These developments have the potential to fully leverage the capabilities of digital twin technology in the realm of smart manufacturing. However, obtaining sufficient training data for these methods is challenging, and thus improving model generalization to reduce data requirements is imperative. To address this issue, we offer a learning‐based approach for estimating Jacobian matrices of visual servoing that organically combines an extreme learning machine (ELM) and a differential evolutionary algorithm (DE). In the first stage, the pseudoinverse of the image Jacobian matrix is approximated using the ELM, which solves the problems associated with traditional visual servoing and is resistant to outside influences such as image noise and mistakes in camera calibration. In the second stage, differential evolution is utilized to select input weights and hidden layer bias and to determine ELM’s output weights. Experimental results conducted on a digital twin operating platform for 4‐DOF robot with an eye‐in‐hand configuration demonstrate better performance than classical visual servoing and traditional ELM‐based visual servoing in various cases.

In recent years, ozone (O3) has gradually become the primary pollutant plaguing urban air quality. Accurate and efficient ozone prediction is of great significance to the prevention and control of ozone pollution. The air quality monitoring network provides multisource pollutant concentration monitoring data for ozone prediction, but ozone prediction based on multisource monitoring data still faces the challenges of each station’s series of data. Aiming at the problems of low prediction accuracy and low computational efficiency in traditional atmospheric ozone concentration prediction, ozone concentration prediction using dual series decomposition was proposed by variational mode decomposition (VMD), ensemble empirical mode decomposition (EEMD), and long short‐term memory (LSTM). First, the historical data series of Nanjing air quality monitoring stations is decomposed by VMD, and then the EEMD algorithm is applied to the residual of VMD to obtain several characteristic intrinsic mode function (IMF) components; each characteristic IMF component is trained by LSTM to obtain the prediction result of each component, and then the final result can be obtained by linear superposition. The proposed method achieved the best results with R2 = 99%, MSE = 5.38, MAE = 4.54, and MAPE = 3.12. Because LSTM has strong adaptive learning ability and good memory function, it has the learning advantage of long‐term memory for long‐term data, and the prediction results are more accurate. According to the data, the proposed method is superior to the baseline models in terms of statistical metrics. As a result, the proposed hybrid method can serve as a reliable model for ozone forecasting.