Advanced Network Research Articles

TCSRNet: a lightweight tobacco leaf curing stage recognition network model

Due to the constraints of the tobacco leaf curing environment and computational resources, current image classification models struggle to balance recognition accuracy and computational efficiency, making practical deployment challenging. To address this issue, this study proposes the development of a lightweight classification network model for recognizing tobacco leaf curing stages (TCSRNet). Firstly, the model utilizes an Inception structure with parallel convolutional branches to capture features at different receptive fields, thereby better adapting to the appearance variations of tobacco leaves at different curing stages. Secondly, the incorporation of Ghost modules significantly reduces the model’s computational complexity and parameter count through parameter sharing, enabling efficient recognition of tobacco leaf curing stages. Lastly, the design of the Multi-scale Adaptive Attention Module (MAAM) enhances the model’s perception of key visual information in images, emphasizing distinctive features such as leaf texture and color, which further improves the model’s accuracy and robustness. On the constructed tobacco leaf curing stage dataset (with color images sized 224×224 pixels), TCSRNet achieves a classification accuracy of 90.35% with 158.136 MFLOPs and 1.749M parameters. Compared to models such as ResNet34, GhostNet, ShuffleNetV2×1.5, EfficientNet-b0, MobileViT-xs, MobileNetV2, MobileNetV3-large, and MobileNetV3-small, TCSRNet demonstrates superior performance in terms of accuracy, FLOPs, and parameter count. Furthermore, when evaluated on the public V2 Plant Seedlings dataset, TCSRNet maintains an impressive accuracy of 97.15% compared to other advanced network models. This research advances the development of lightweight models for recognizing tobacco leaf curing stages, providing theoretical support for smart tobacco curing technologies and injecting new momentum into the digital transformation of the tobacco industry.

AeroNet: Efficient YOLOv7 for Tiny-Object Detection in UAV Imagery

The detection of multiple tiny objects from diverse perspectives using unmanned aerial vehicles (UAVs) and onboard edge devices presents a significant challenge in computer vision. To address that, this study proposes AeroNet, a lightweight and efficient detection algorithm based on YOLOv7 (You Only Look Once version7).This algorithm features the LHGNet (Lightweight High-Performance GhostNet) backbone, an advanced feature extraction network that integrates depth-wise separable convolution and channel shuffle modules.These modules enable deeper exploration of network features, promoting the fusion of local detail information and channel characteristics. Additionally, this research introduces the LGS(Lightweight Gradient-Sensitive) bottleneck and LGSCSP(Lightweight Gradient-Sensitive Cross Stage Partial Network) fusion module in the neck to reduce computational complexity while maintaining accuracy. Structural modifications and adjusted feature map sizes further enhance detection accuracy. Evaluated on the SkyFusion dataset,this method demonstrated a 25.0% reduction in parameter count and a 12.8% increase in mAP (0.5) compared to YOLOv7. These results underscore the effectiveness of this proposed approach in improving detection accuracy and model efficiency through the proposed enhancements.

Garbage Classification: A Deep Learning Perspective

Garbage Classification using deep learning focuses on techniques to automate and improve the sorting of waste materials. The objective is to enhance recycling processes and promote environmental sustainability by accurately categorizing waste into six types: glass, paper, cloth, trash, cardboard, and plastic. The study implements advanced convolutional neural networks (CNNs) to analyze and classify images of garbage, automating a task that is traditionally manual and labor-intensive.To achieve this, several deep learning models were used, including MobileNet, NASNet, LeNet, Inception, and DenseNet. These models were trained on a carefully curated dataset to ensure balanced representation across all waste categories, allowing them to extract complex features from the images and make precise classifications. Each model was evaluated based on its performance, with NASNet delivering the highest accuracy, making it the most suitable for real-world applications where resources might be limited, such as mobile or edge devices.The results demonstrate that NASNet is the most effective algorithm for garbage classification, outperforming the other models in terms of accuracy. By automating the classification process, this research offers a practical solution to improve recycling efficiency, reduce the need for manual sorting, and contribute to sustainable waste management. The study highlights the significant role that deep learning can play in transforming waste management systems for a cleaner and more sustainable environment.

Predictive Abilities of Machine Learning and Deep Learning Approaches for Exchange Rate Prediction

This study evaluates the efficacy of forecasting models in predicting USD/TRY exchange rate fluctuations. We assess Support Vector Machine (SVM), XGBoost, Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) models with 96 and 21 feature sets. Data from 01.01.2010 to 30.04.2024 were sourced from Bloomberg, CBRT, and BDDK. Findings indicate that LSTM and GRU models outperform traditional models, with GRU showing the highest predictive accuracy. SVM performs poorly with highdimensional data, while XGBoost offers moderate predictive power but lacks in capturing intricate patterns. This study highlights the importance of model and feature selection in financial time series forecasting and underscores the advantages of advanced neural networks. The results provide valuable insights for analysts and policymakers in developing robust economic forecasting models.

GLCANet: Global–Local Context Aggregation Network for Cropland Segmentation from Multi-Source Remote Sensing Images

Cropland is a fundamental basis for agricultural development and a prerequisite for ensuring food security. The segmentation and extraction of croplands using remote sensing images are important measures and prerequisites for detecting and protecting farmland. This study addresses the challenges of diverse image sources, multi-scale representations of cropland, and the confusion of features between croplands and other land types in large-area remote sensing image information extraction. To this end, a multi-source self-annotated dataset was developed using satellite images from GaoFen-2, GaoFen-7, and WorldView, which was integrated with public datasets GID and LoveDA to create the CRMS dataset. A novel semantic segmentation network, the Global–Local Context Aggregation Network (GLCANet), was proposed. This method integrates the Bilateral Feature Encoder (BFE) of CNNs and Transformers with a global–local information mining module (GLM) to enhance global context extraction and improve cropland separability. It also employs a multi-scale progressive upsampling structure (MPUS) to refine the accuracy of diverse arable land representations from multi-source imagery. To tackle the issue of inconsistent features within the cropland class, a loss function based on hard sample mining and multi-scale features was constructed. The experimental results demonstrate that GLCANet improves OA and mIoU by 3.2% and 2.6%, respectively, compared to the existing advanced networks on the CRMS dataset. Additionally, the proposed method also demonstrated high precision and practicality in segmenting large-area croplands in Chongzhou City, Sichuan Province, China.

Role of Computer-Aided Design and Development in Modern Pharmaceuticals

The integration of molecular biology and computational technologies has transformed modern drug discovery and development processes. Advanced computational methodologies, particularly artificial intelligence (AI) and machine learning (ML), have reshaped traditional drug development approaches through unprecedented access to ligand property data, target binding interactions, three-dimensional protein structures, and virtual libraries containing billions of drug-like molecules. AI and deep learning (DL) implementations have enhanced multiple stages of drug discovery, from target identification to lead optimization. These computational advances, backed by improved hardware capabilities and sophisticated algorithms, now enable targeting of previously "undruggable" proteins. This review presents modern computational approaches in pharmaceutical development, including strategies for challenging protein targets through covalent regulation, allosteric inhibition, protein-protein interaction modulation, and targeted protein degradation. AI-driven methods have accelerated drug discovery pipelines, reduced development costs, and improved clinical trial success rates. The transition from traditional broad-spectrum approaches to precision medicine, supported by computational tools, has enabled personalized therapeutic strategies. Current limitations in computer-aided drug design persist, yet the combination of computational predictions with experimental validation continues to advance therapeutic development. Recent developments in quantum computing and advanced neural networks promise to further enhance drug discovery efficiency and success rates in the coming decades.

A multi-frame fusion video deraining neural network based on depth and luminance features

Rainy weather is a common natural occurrence, but capturing clear and accurate images in rainy conditions can be challenging. Consequently, addressing the effects of rain on video imagery is essential for enhancing visual quality and the accuracy and reliability of computer vision applications. In this work, we introduce a novel rain model accompanied by a multi-frame fusion video rain removal algorithm that exploits depth and luminance features. We present an innovative three-stage video deraining methodology that synergizes multi-frame fusion with advanced neural network design to achieve highly realistic and clear rain removal. Our Option-Flow-Depth-Luminance (OFDL) Derain algorithm propels the research forward by offering a new perspective on rain removal. Incorporating depth and luminance features into image processing, we devise a novel rain model that allows the algorithm to handle complex environments and a variety of rain patterns effectively. Additionally, our model takes into account degradation factors such as rain streaks, accumulation, runoff, and occlusion, leading to the generation of more authentic rain-affected images that improve the training and performance evaluation of the model. Experimental tests conducted on the RainSynLight25, RainSynComplex25, and NTURain datasets demonstrate that our method outperforms current state-of-the-art techniques, achieving increases of 4.3 dB, 2.2 dB, and 2.2 dB in PSNR, and enhancements of 0.062, 0.109, and 0.008 in Structural Similarity Index Measure (SSIM), respectively. Moreover, our approach exhibits superior processing speed, further underscoring its practical advantages.

Ecological response mechanism of alkaline wastewater periphytic biofilms to pH alteration

The environmental impact of indiscriminate alkaline wastewater discharge is increasingly concerning. Studying the ecological response mechanisms of periphytic biofilms to such wastewater is crucial for deepening our understanding of its broader environmental implications. In this study, periphytic biofilms collected from the drainage outlet of a tunnel construction project were developed for 56 days at initial pH values of 10 (pH0 10) and 12 (pH0 12). These stabilized biofilms were then transferred to a neutral environment (pH 7) for another 35 days. High-throughput 16S rRNA sequencing revealed changes in the growth characteristics, microbial interactions, and microbial community composition, function, and assembly patterns of the biofilms. Specifically, results revealed decreased microbial activity, live/dead bacteria ratios, and polysaccharide content in extracellular polymeric substances when the pH was reduced from alkaline to neutral, with incomplete recovery observed after 35 days of incubation in a neutral environment. During neutral pH incubation, proteobacterial abundance increased by 58.39 % and 33.08 % in biofilms cultured at pH0 10 and pH0 12, respectively, becoming the dominant flora, with chemoheterotrophic functions increasing by 8.11 % and 38.23 %, respectively. Network analysis indicated that biofilms in the pH0 12 culture group had a more advanced organizational co-occurrence network, with more complex and stable interactions, than those in the pH0 10 culture group. Microbial community α diversity, key differential species count, the proportion of positively correlated network edges, and null model data showed that microbial communities in biofilms cultured at pH0 10 were more resilient, whereas those cultured at pH0 12 were more resistant. Overall, this study reveals that alkaline wastewater periphytic biofilms adapt and develop high tolerance and metabolic activity but struggle to recover their original state in a short-term neutral environment, with biofilms cultured at pH0 12 exhibiting more complex interactions and higher resistance and those cultured at pH0 10 showing greater resilience.

Real-time instance segmentation of recyclables from highly cluttered construction and demolition waste streams.

Efficient recycling strategies are crucial for mitigating the adverse environmental impacts of escalating construction and demolition waste (CDW). While automated identification via deep-learning is a promising direction, localizing CDW recyclables is uniquely challenging due to significant clutter and compositional complexity. Recognizing that accurate and fast localization is a strong prerequisite for swift robotic action, this study provides a comprehensive assessment of state-of-the-art (s.o.t.a) real-time instance segmentation of recyclables from complex CDW streams. Unlike previous studies that simplify the real-world conditions, this study curates and employs a high-quality CDW instance segmentation dataset tailored to capture often-ignored domain intricacies such as deformation, contamination, intricate class distinctions, scale variations, and high levels of clutter, ensuring practical applicability. The results show that even advanced networks struggle with complexity of real CDW streams due to high clutter, with segmentation accuracy not surpassing 50%. To address this, the study proposes integrating patch-based inferencing techniques, which help the model focus on cluttered regions more effectively, boosting overall performance by a notable 12.9%. Additionally, to enhance the zero-shot capabilities of s.o.t.a prompt-based real-time segmentation for identifying CDW recyclables, a simple yet effective domain-transfer framework is proposed, substantially increasing number of correct mask predictions by 68%. This study serves as a practical reference for applying deep-learning tools in automated environmental management tasks such as waste sorting and suggests the best architectural framework for practical use.

The JCMT Legacy Release: SCUBA-2 850 μm Coadds and Catalogs

We present the James Clerk Maxwell Telescope (JCMT) 850 μm Legacy Release, containing uniformly reduced, coadded tiles, and catalogs of detected emission, for the 850 μm data from all Submillimetre Common User Bolometer Array 2 (SCUBA-2) observations taken between 2011 February 2 and 2020 August 1. This release provides the fastest and easiest way to identify uniformly determined 850 μm detections and calibrated fluxes for any position observed by the JCMT. The coadded observations include 11,722 hr of observing time and cover 1516 square degrees of the sky, with detections of contiguous areas of emission at better than 5σ covering 2.46 square degrees of area. Within these regions 21,059 individual local maxima were detected. Fifteen tiles contain regions with a noise level better than 0.0025 mJy arcsec−2. The data are gridded onto HEALPix tiles of ≈1° a side, using the HEALPix projection with pixels of size ≈3.″22. The coadds have been calibrated into units of mJy arcsec−2 using the standard date-varying SCUBA-2 flux-conversion factor (FCF) from Mairs et al. (2021). We then examined the accuracy of this calibration for our data set by calculating self-derived FCF values for our two most-used standard sources, finding a result within the expected errors but with a larger standard deviation of 9%. The coadds and catalogs can be searched and retrieved via the Canadian Astronomy Data Centre (project code: JCMT-LR). Combined catalogs of all detected regions and of all local maxima can be downloaded from the Canadian Advanced Network for Astronomy Research Data Publication Service (doi:10.11570/23.0013), along with masks showing the full area observed in this release and the full area of detected emission.

Severity Classification of Parkinson’s Disease via Synthesis of Energy Skeleton Images from Videos Produced in Uncontrolled Environments

Background/Objectives: Parkinson’s Disease is a prevalent neurodegenerative disorder affecting millions worldwide, primarily marked by motor and non-motor symptoms due to the degeneration of dopamine-producing neurons. Despite the absence of a cure, current treatments focus on symptom management, often relying on pharmacotherapy and surgical interventions. Early diagnosis remains a critical challenge, particularly in underserved areas, as existing diagnostic protocols lack standardization and accessibility. This paper proposes a novel framework for the diagnosis and severity classification of PD using video data captured in uncontrolled environments. Methods: Leveraging deep learning techniques, our approach synthesizes Skeleton Energy Images (SEIs) from gait sequences and employs three advanced models—a Convolutional Neural Network (CNN), a Residual Network (ResNet), and a Vision Transformer (ViT)—to analyze these images. Our methodology allows for the accurate detection of PD and differentiation of its severity without requiring specialized equipment or professional oversight. The dataset used consists of labeled videos capturing the early stages of the disease, facilitating the potential for timely intervention. Results: The four models performed very accurately during the training phase. In fact, an accuracy higher than 99% was achieved by the ViT and ResNet models. Moreover, a lesser accuracy of 90% was achieved by the CNN five-layer model. During the test phase, only the best-performing models from the training experiments were tested. The ResNet-18 model has achieved a 100% accuracy. However, the ViT and the CNN five-layer models have achieved, respectively, 99.96% and 96.40% test accuracy. Conclusions: The results demonstrate high accuracy, highlighting the framework’s capabilities, and in particular the effectiveness of the workflow used for generating the SEI images. Given the nature of the dataset used, the proposed framework stands to function as a cost-effective and accessible tool for early PD detection in various healthcare settings. This study contributes to the advancement of mobile health technologies, aiming to enhance early diagnosis and monitoring of Parkinson’s Disease.

Advancing Healthcare: Intelligent Speech Technology for Transcription, Disease Diagnosis, and Interactive Control of Medical Equipment in Smart Hospitals

Intelligent Speech Technology (IST) is revolutionizing healthcare by enhancing transcription accuracy, disease diagnosis, and medical equipment control in smart hospital environments. This study introduces an innovative approach employing federated learning with Multi-Layer Perceptron (MLP) and Gated Recurrent Unit (GRU) neural networks to improve IST performance. Leveraging the “Medical Speech, Transcription, and Intent” dataset from Kaggle, comprising a variety of speech recordings and corresponding medical symptom labels, noise reduction was applied using a Wiener filter to improve audio quality. Feature extraction through MLP and sequence classification with GRU highlighted the model’s robustness and capacity for detailed medical understanding. The federated learning framework enabled collaborative model training across multiple hospital sites, preserving patient privacy by avoiding raw data exchange. This distributed approach allowed the model to learn from diverse, real-world data while ensuring compliance with strict data protection standards. Through rigorous five-fold cross-validation, the proposed Fed MLP-GRU model demonstrated an accuracy of 98.6%, with consistently high sensitivity and specificity, highlighting its reliable generalization across multiple test conditions. In real-time applications, the model effectively performed medical transcription, provided symptom-based diagnostic insights, and facilitated hands-free control of healthcare equipment, reducing contamination risks and enhancing workflow efficiency. These findings indicate that IST, powered by federated neural networks, can significantly improve healthcare delivery, accuracy in patient diagnosis, and operational efficiency in clinical settings. This research underscores the transformative potential of federated learning and advanced neural networks for addressing pressing challenges in modern healthcare and setting the stage for future innovations in intelligent medical technology.

Advancing Brain MRI Image Classification: Integrating VGG16 and ResNet50 with a Multi-Verse Optimization Method

Background/Objectives: The accurate categorization of brain MRI images into tumor and non-tumor categories is essential for a prompt and effective diagnosis. This paper presents a novel methodology utilizing advanced Convolutional Neural Network (CNN) designs to tackle the complexity and unpredictability present in brain MRI data. Methods: The methodology commences with an extensive preparation phase that includes image resizing, grayscale conversion, Gaussian blurring, and the delineation of the brain region for preparing the MRI images for analysis. The Multi-verse Optimizer (MVO) is utilized to optimize data augmentation parameters and refine the configuration of trainable layers in VGG16 and ResNet50. The model’s generalization capabilities are significantly improved by the MVO’s ability to effectively balance computational cost and performance. Results: The amalgamation of VGG16 and ResNet50, further refined by the MVO, exhibits substantial enhancements in classification metrics. The MVO-optimized hybrid model demonstrates enhanced performance, exhibiting a well-calibrated balance between precision and recall, rendering it exceptionally trustworthy for medical diagnostic applications. Conclusions: The results highlight the effectiveness of MVO-optimized CNN models for classifying brain tumors in MRI data. Future investigations may examine the model’s applicability to multiclass issues and its validation in practical clinical environments.

Mathematical modelling-based blockchain with attention deep learning model for cybersecurity in IoT-consumer electronics

Security in the Internet of Things (IoT)-consumer electronics is decisive in safeguarding connected devices from possible vulnerabilities and threats. Strong security measures are required to protect against unauthorized access, data breaches, and cyberattacks as these smart devices gather, transfer, and save sensitive information. Employing regular software updates, secure authentication protocols, and strong encryption are indispensable approaches to guarantee the integrity and privacy of user details. Drones offer users a bird's-eye view that can be started and implemented anywhere and anytime. But, the malicious use of drones has developed among criminals and cyber-criminals. The possibility and frequency of these attacks are maximum, and their effect is highly unsafe and devastating. Thus, the desire for protective, preventive counter-measures and detective are needed. Intrusion detection utilizing deep learning (DL) drones control advanced neural network (NN) structures to improve security surveillance in dynamic outdoor environments. Prepared with sophisticated sensors and onboard processing abilities, these drones autonomously examine aerial imagery to identify and classify possible risks like suspicious activities, unauthorized personnel, or vehicles. DL approaches allow drones to learn complex patterns and anomalies in real-time, enabling quick response and proactive security procedures. This study introduces an enhanced Mathematical Modeling-based Blockchain with Mountain Gazelle Optimization and Attention to Deep Learning for Cybersecurity (MGOADL-CS) technique in the drone's platform. The MGOADL-CS method aims to improve cybersecurity using BC technology in the drone's environment by detecting attacks using optimal DL models. In the initial stage, the MGOADL-CS technique uses a linear scaling normalization (LSN) approach to normalize the input data. The MGOADL-CS technique uses an improved tunicate swarm algorithm (ITSA) based feature selection approach for dimensionality reduction. Besides, the attention long short-term memory neural network (ALSTM-NN) model is employed to detect and classify cyberattacks. Finally, the MGO-based hyperparameter tuning process is performed to adjust the hyperparameter values of the ALSTM-NN model. To highlight the enhanced attack detection results of the MGOADL-CS technique, a detailed simulation set is accomplished under the NSL dataset. The performance validation of the MGOADL-CS method portrayed a superior accuracy value of 99.71 % over existing approaches.

Data-specific activation function learning for constructive neural networks

Activation functions play a crucial role in learning and expressive capabilities of advanced neural networks due to their non-linear or non-saturated properties. However, how to determine the appropriate activation function from various candidates is a challenging yet not well-addressed topic. To address the issue, a novel self-learning approach, called as data-specific activation function learning (DSAFL) algorithm, is proposed to establish constructive neural network on one-time by adaptively selecting appropriate activation function based on the specific data characteristics. To assess the space dimension mapping abilities of different activation functions, the configuration probabilities are used to guide the generation of various candidate activation functions and corresponding candidate hidden node. In the learning stage, an exploration-exploitation mechanism composed of the random algorithm and the greedy strategy is developed to obtain the influence of different candidate activation functions, thereby avoiding configuration probabilities falling into local optimum. A reward-penalty mechanism is built to update the configuration probabilities and enhance the robustness of network by integrating the simulated annealing strategy. In final, the activation function with the highest configuration probability, as the best one, is used to reconstruct the neural network. Experimental results on both regression and classification tasks demonstrate the efficiency and effectiveness of DSAFL in the activation function selection problems of a class of constructive neural networks.

Optimizing Fire Scene Analysis: Hybrid Convolutional Neural Network Model Leveraging Multiscale Feature and Attention Mechanisms

The rapid and accurate detection of fire scenes in various environments is crucial for effective disaster management and mitigation. Fire scene classification is a critical aspect of modern fire detection systems that directly affects public safety and property preservation. This research introduced a novel hybrid deep learning model designed to enhance the accuracy and efficiency of fire scene classification across diverse environments. The proposed model integrates advanced convolutional neural networks with multiscale feature extraction, attention mechanisms, and ensemble learning to achieve superior performance in real-time fire detection. By leveraging the strengths of pre-trained networks such as ResNet50, VGG16, and EfficientNet-B3, the model captures detailed features at multiple scales, ensuring robust detection capabilities. Including spatial and channel attention mechanisms further refines the focus on critical areas within the input images, reducing false positives and improving detection precision. Extensive experiments on a comprehensive dataset encompassing wildfires, building fires, vehicle fires, and non-fire scenes demonstrate that the proposed framework outperforms existing cutting-edge techniques. The model also exhibited reduced computational complexity and enhanced inference speed, making it suitable for deployment in real-time applications on various hardware platforms. This study sets a new benchmark for fire detection and offers a powerful tool for early warning systems and emergency response initiatives.

Complex quadrupled-power-law nonlinearity form of Radhakrishnan-Kundu-Lakshmanan equation: novel optical soliton analysis

Abstract In this manuscript, the novel optical wave solutions to the newly developed Radhakrishnan-Kundu-Lakshmanan (NRKL) model are studied. Two analytical methods namely, the new extended generalized Kudryashov and the extended modified auxiliary equation mapping methods are used to obtain these novel solutions. Furthermore, the dynamics of the obtained solutions are analyzed thoroughly with the help of different graphical forms such as three-dimensional (3-D), contour, and two-dimensional (2-D) figures. To get the graphical view of the derived solutions, numerical values for the unknown parameters are assigned while balancing the nonlinearity with dispersion. Moreover, the characteristics analysis of the obtained solutions depict anti-kink, periodic, bright, dark-shaped periodic, anti-bell shape, bright-shaped periodic, and dark waveforms Furthermore, the practical implications of these waveforms in optical fiber transmission are profound, with the potential to influence nonlinear signal processing and advanced optical network design.

Developing the BRANCH Initiative-Building Relationship science for Advanced Networks in Communication and Health: Year One S

Developing the BRANCH Initiative-Building Relationship science for Advanced Networks in Communication and Health: Year One S

Facets of security and safety problems and paradigms for smart aerial mobility and intelligent logistics

AbstractThe use of unmanned aerial vehicles (UAVs) for smart and speedy logistics is still relatively nascent compared to traditional delivery methods. However, it is witnessing sporadic and steady growth due to booming demands, technological advancement, and regulatory support. The intelligence and integrity of UAV systems depend largely on the underlying cognitive and cybersecurity models, which serve as both eyes and brains to perceive and respond to the myriad of scenarios around them. Smart mobility and intelligent logistic ecosystems (SMiLE) are complex and advanced technological networks which are exposed to several issues. The incorporation of UAVs for priority logistics, thereby extending the coverage and capacity of SMiLE, further heightens these vulnerabilities and questions its security, safety, and sustainability. This review scrutinizes the significant security disruptions, smartness dynamics, and sundry developments for the sustainable deployment of UAVs as an aerial logistics‐based vehicle. Using the PRISMA‐SPIDER methodology, 157 articles were selected for quantitative analysis and 20 review articles for qualitative evaluation. Security and safety issues in UAVs cut across all the layers of logistics operations: components, communication, network architecture, navigation, supply chain etc. Expanding the capacity of SMiLE using UAV demands an intentional and incremental convergence‐based integration of an agile explainable artificial framework for reliable and safety‐conscious smart mobility, a scalable and tamperproof blockchain for multi‐factor authentication, and a zero trust cybersecurity paradigm for inclusive enterprise‐based authorization.

Enhancing human activity recognition for the elderly and individuals with disabilities through optimized Internet-of-Things and artificial intelligence integration with advanced neural networks.

Elderly and individuals with disabilities can greatly benefit from human activity recognition (HAR) systems, which have recently advanced significantly due to the integration of the Internet of Things (IoT) and artificial intelligence (AI). The blending of IoT and AI methodologies into HAR systems has the potential to enable these populations to lead more autonomous and comfortable lives. HAR systems are equipped with various sensors, including motion capture sensors, microcontrollers, and transceivers, which supply data to assorted AI and machine learning (ML) algorithms for subsequent analyses. Despite the substantial advantages of this integration, current frameworks encounter significant challenges related to computational overhead, which arises from the complexity of AI and ML algorithms. This article introduces a novel ensemble of gated recurrent networks (GRN) and deep extreme feedforward neural networks (DEFNN), with hyperparameters optimized through the artificial water drop optimization (AWDO) algorithm. This framework leverages GRN for effective feature extraction, subsequently utilized by DEFNN for accurately classifying HAR data. Additionally, AWDO is employed within DEFNN to adjust hyperparameters, thereby mitigating computational overhead and enhancing detection efficiency. Extensive experiments were conducted to verify the proposed methodology using real-time datasets gathered from IoT testbeds, which employ NodeMCU units interfaced with Wi-Fi transceivers. The framework's efficiency was assessed using several metrics: accuracy at 99.5%, precision at 98%, recall at 97%, specificity at 98%, and F1-score of 98.2%. These results then were benchmarked against other contemporary deep learning (DL)-based HAR systems. The experimental outcomes indicate that our model achieves near-perfect accuracy, surpassing alternative learning-based HAR systems. Moreover, our model demonstrates reduced computational demands compared to preceding algorithms, suggesting that the proposed framework may offer superior efficacy and compatibility for deployment in HAR systems designed for elderly or individuals with disabilities.