Thyroid carcinoma (TC) remains a critical health challenge, where timely and accurate diagnosis is essential for improving patient outcomes. This review provides a comprehensive examination of artificial intelligence (AI) applications — including machine learning (ML), deep learning (DL), and emerging transformer-based approaches — in the detection and classification of TC. We first outline standardized evaluation metrics and analyze publicly available datasets, highlighting their limitations in diversity, annotation quality, and representativeness. Next, we survey AI-driven diagnostic frameworks across three domains: classification, segmentation, and prediction, with emphasis on ultrasound imaging, histopathology, and genomics. A comparative analysis of ML and DL approaches illustrates their respective strengths, such as interpretability in smaller datasets versus automated feature extraction in large-scale imaging tasks. Advanced methods leveraging vision transformers (ViT) and large language models (LLMs) are discussed alongside traditional models, situating them within a broader ecosystem of feature engineering, ensemble learning, and hybrid strategies. We also examine key challenges — imbalanced datasets, computational demands, model generalizability, and ethical concerns — before outlining future research directions, including explainable AI, federated and privacy-preserving learning, reinforcement learning, and integration with the Internet of Medical Things (IoMT). By bridging technical insights with clinical considerations, this review establishes a roadmap for next-generation TC diagnostics and highlights pathways toward robust, patient-centric, and ethically responsible AI deployment in oncology.

This paper presents an innovative approach to mitigating the peak-to-average power ratio (PAPR). The proposed method uses a deep learning model called autoencoders (AEs) to simplify the process and avoid the complex calculations of traditional methods such as selective mapping (SLM). Unlike SLM, our approach does not need side information about the PAPR distribution. Through simulations of coherent optical orthogonal frequency division multiplexing (CO-OFDM) systems, the AEbased model offers substantial enhancements in both PAPR reduction and bit error rate (BER) performance when compared to conventional techniques. An error-free transmission can be acheived with a reduction in PAPR exceeding 10 dB compared to the original signal and a 1 dB advantage over SLM. In particular, the AE model achieves the best BER performance of$2 \times 10^{-6}$at 44 dB OSNR, surpassing traditional methods. Furthermore, the model demonstrates robustness against noise and nonlinear distortions, making it appropriate for optical channels experiencing diverse levels of impairment. This innovative technique has the potential to revolutionize next-generation optical communication systems by enabling efficient and reliable data transmission.

The analysis of complex building time-series for actionable insights and recommendations remains challenging due to the nonlinear and multi-scale characteristics of energy data. To address this, we propose a framework that fine-tunes visual language large models (VLLMs) on 3D graphical representations of the data. The approach converts 1D time-series into 3D representations using continuous wavelet transforms (CWTs) and recurrence plots (RPs), which capture temporal dynamics and localize frequency anomalies. These 3D encodings enable VLLMs to visually interpret energy-consumption patterns, detect anomalies, and provide recommendations for energy efficiency. We demonstrate the framework on real-world building-energy datasets, where fine-tuned VLLMs successfully monitor building states, identify recurring anomalies, and generate optimization recommendations. Quantitatively, the Idefics-7B VLLM achieves validation losses of 0.0952 with CWTs and 0.1064 with RPs on the University of Sharjah energy dataset, outperforming direct fine-tuning on raw time-series data (0.1176) for anomaly detection. This work bridges time-series analysis and visualization, providing a scalable and interpretable framework for energy analytics.

This study presents an Adaptive Transfer Learning and Thresholding-based Deep Learning Model (ATL-TDLM) for automated breathing pattern recognition using thermal imaging. Unlike conventional methods that rely on sound-based respiratory data, our approach leverages hierarchical deep feature extraction and adaptive multi-thresholding (AMT) to enhance feature segmentation. The model integrates knowledge distillation-based fine-tuning (KD-FT) to optimize learning transfer and contrastive representation learning (CRL) to improve inter-class separability between inhalation (INH) and exhalation (EXH) phases. The ATL-TDLM framework achieves an accuracy of 98.8%, significantly outperforming state-of-the-art models while ensuring computational efficiency. This approach has potential applications in respiratory disorder detection, including sleep apnea and asthma monitoring.

This manuscript presents the design and evaluation of a flexible, optically transparent monopole antenna tailored for sub-6 GHz 5G applications. The antenna utilizes a polyimide substrate and a meshed radiating structure to balance electromagnetic performance with high optical transparency. Three variations, non-meshed, circularly-meshed, and square-meshed, were designed and analyzed through CST Microwave Studio simulations. The square-meshed design achieved the best tradeoff, offering a transparency of $79.89 \%$, a peak realized gain of 3.39 dBi, and stable performance under 30° mechanical bending. The antenna demonstrates omnidirectional radiation patterns, wide impedance bandwidth ($2.83-3.8 \mathrm{GHz}$), and strong resilience to deformation, confirming its suitability for integration into wearable, transparent, and conformal devices in next-generation 5G systems.

With the rapid advancements in quantum computing, traditional cryptographic schemes like Rivest-Shamir-Adleman (RSA) and elliptic curve cryptography (ECC) are becoming vulnerable, necessitating the development of quantum-resistant algorithms. The National Institute of Standards and Technology (NIST) has initiated a standardization process for PQC algorithms, and several candidates, including CRYSTALS-Kyber and McEliece, have reached the final stages. This paper first provides a comprehensive analysis of the hardware complexity of post-quantum cryptography (PQC) in embedded systems, categorizing PQC algorithms into families based on their underlying mathematical problems: lattice-based, code-based, hash-based and multivariate / isogeny-based schemes. Each family presents distinct computational, memory, and energy profiles, making them suitable for different use cases. To address these challenges, this paper discusses optimization strategies such as pipelining, parallelization, and high-level synthesis (HLS), which can improve the performance and energy efficiency of PQC implementations. Finally, a detailed complexity analysis of CRYSTALS-Kyber and McEliece, comparing their key generation, encryption, and decryption processes in terms of computational complexity, has been conducted.

As intravehicular communication systems become increasingly complex and high-dimensional within the evolving paradigm of the Internet of Vehicles (loV), the need for resilient and intelligent cybersecurity mechanisms is rendered more critical. In this paper, a hybrid deep neural architecture is proposed, integrating Convolutional Neural Networks (CNNs) with autoencoder frameworks to extract both spatial feature hierarchies and latent structural patterns from multivariate vehicular telemetry. Unlike conventional approaches in the literature that rely on abstracted or non-vehicular datasets, the proposed method is trained and evaluated on the CICIoV2024 dataset-collected from a production-grade 2019 Ford vehicle under five distinct adversarial scenarios, including spoofing and Denial of Service (DoS) attacks. Through extensive experimentation, the effectiveness of the model is validated, achieving a classification accuracy of 98.88%, a mean precision of 93. 00%, recall of 93. 83%, and a Fl score of 92. 26%.

As intravehicular communication systems become increasingly complex and high-dimensional within the evolving paradigm of the Internet of Vehicles (IoV), the need for resilient and intelligent cybersecurity mechanisms is rendered more critical. In this paper, a hybrid deep neural architecture is proposed, integrating Convolutional Neural Networks (CNNs) with autoencoder frameworks to extract both spatial feature hierarchies and latent structural patterns from multivariate vehicular telemetry. Unlike conventional approaches in the literature that rely on abstracted or non-vehicular datasets, the proposed method is trained and evaluated on the CICIoV2024 dataset– collected from a production-grade 2019 Ford vehicle under five distinct adversarial scenarios, including spoofing and Denial of Service (DoS) attacks. Through extensive experimentation, the effectiveness of the model is validated, achieving a classification accuracy of 98.88%, a mean precision of 93. 00%, recall of 93. 83%, and a F1 score of 92. 26%.

Abstract Occupancy detection is crucial for various applications, including smart buildings, security systems, and energy management. This paper introduces a novel convolutional neural network (CNN) architecture based on an image encoding approach for accurate occupancy detection. Our network effectively extracts relevant features from occupancy images by leveraging deep learning and image processing techniques, enabling reliable and real-time detection. We employed an image encoding method that converts environmental time-series data into 2D image representations—either grayscale or RGB-like—depending on the input requirements of the CNN model. This transformation captures spatial and temporal characteristics of the data, allowing the network to learn more expressive occupancy-related patterns from raw 1D input. Additionally, we developed a custom CNN architecture optimized for the encoded images, enabling the network to identify key features and understand complex spatial relationships. We evaluated the performance of our CNN through extensive testing on well-known occupancy datasets. The results highlight the superiority of our approach, outperforming existing techniques in accuracy, precision, recall, and F1-score. Our model achieved impressive accuracies of 98.45%, 99.05%, and 97.32% across the three datasets used in this study.

Large Language Models (LLMs) have been utilized extensively for cancer detection and diagnosis, benefiting from the vast textual data available in the medical field. However, these models often lack specific training on cancer-related data, which can limit their effectiveness in specialized medical contexts. Traditional methods typically deploy LLMs directly for diagnosis without incorporating domain-specific expertise, potentially compromising outcome reliability. This paper presents a comparative study focusing on the application of Low-Rank Adaptation (LoRA) to fine-tune LLMs for cancer-related tasks. LoRA modifies the self-attention and feed-forward layers of transformer architectures with low-rank matrices, allowing for specialized adaptation with fewer parameters. A general LLM was fine-tuned using LoRA on a dataset derived from four annotated books on breast cancer. The performance of this LoRA-enhanced model was compared against several baseline LLMs fine-tuned through traditional methods. It was found that the LoRA-fine-tuned Biomstral-7B demonstrated the best training loss of 0.91025 and validation loss of 0.912722 scores, indicating enhanced integration of domain-specific knowledge. The potential of adaptive fine-tuning techniques like LoRA in specialized applications is highlighted, suggesting further exploration into their effectiveness across various complex domains requiring expert knowledge. Further research is encouraged to assess such approaches’ broader applicability and impact in diverse AI applications.

In Photovoltaic Systems, detecting partial shading is critical for optimizing energy output and ensuring system reliability. This paper presents a novel method for partial shading detection in photovoltaic systems, leveraging transfer learning to improve accuracy and efficiency. By utilizing a pre-trained InceptionV3 model, discriminative features are extracted from time series signals. To align with the architectural requirements of InceptionV3, these time series signals are transformed into 2D pixel-mapped images. The proposed model is rigorously validated using both balanced and unbalanced scenarios on the Grid-connected PV System Faults (GPVS-Faults) dataset, achieving remarkable accuracies of 96.73% and 94.59% in the respective scenarios. This approach represents a significant advancement in accurately identifying partial shading, thus enhancing the performance and reliability of solar energy systems.

Tomato plant disease detection and classification, utilizing leaf images through deep learning, intersects the fields of plant pathology and agriculture. Deep learning has demonstrated significant potential in accurately identifying and classifying various plant diseases from leaf images. In this study, we introduce a hybrid system that combines a potent machine learning algorithm, Exponential Discriminant Analysis (EDA), with a transfer learning process leveraging recent and advanced deep networks, including ResNet50, Darknet53, DenseNet201, and EfficientNetB0. This system was evaluated using two challenging datasets: Taiwan and PlantVillage tomato leaf datasets. The experimental results underscore the high competitiveness of the proposed method, achieving mean accuracies of 98.29% and $\mathbf{9 8. 0 9 \%}$ on these datasets, respectively.

Across the globe, agricultural yield faces numerous challenges, including unpredictable weather patterns, resource constraints, and the ever-present threat of plant diseases. Early and accurate disease detection is crucial for mitigating losses, optimizing resource allocation, and promoting sustainable farming practices. Machine Learning (ML) and Deep Learning (DL) techniques, particularly convolutional neural networks (CNNs), offer immense potential for tackling this challenge. This paper investigates the potential of DL for disease detection in agricultural crops. We delve into data scarcity, a major obstacle, and analyze the suitability of existing datasets like PlantVillage for training robust models. Furthermore, we explore popular CNN architectures, such as LeNet, AlexNet, and VGGNet, along with their strengths and limitations in the context of plant disease detection.

Person re-identification (PRe-ID) is a crucial task in security, surveillance, and retail analysis, which involves identifying an individual across multiple cameras and views. However, it is a challenging task due to changes in illumination, background, and viewpoint. Efficient feature extraction and metric learning algorithms are essential for a successful PRe-ID system. This paper proposes a novel approach for PRe-ID, which combines a Convolutional Neural Network (CNN) based feature extraction method with Cross-view Quadratic Discriminant Analysis (XQDA) for metric learning. Additionally, a matching algorithm that employs Mahalanobis distance and a score normalization process to address inconsistencies between camera scores is implemented. The proposed approach is tested on four challenging datasets, including VIPeR, GRID, CUHK01, and PRID450S. The proposed approach has demonstrated its effectiveness through promising results obtained from the four challenging datasets.

This paper proposes an improved version of the Lifelong Topic Model (LTM) called the HC-LTM. The traditional LTM is known to be biased in the domain selection process and does not fully consider the contextual information of target words when determining similarity. The HC-LTM addresses these issues by combining Word2vec cosine similarity and Hellinger distance between topics to identify similar words and topics, leading to better selection and more effective knowledge acquisition during iterative learning. Additionally, the problem of repetitive calculation of cosine distance is resolved by pre-loading the similarity matrix of word vectors and using Hellinger distance to calculate topic similarity accelerates the convergence of the model. The experimental results on the Amazon product review dataset demonstrate the effectiveness of the HC-LTM model, with a 49% improvement in topic consistency and a 44.57% reduction in time compared to the LTM model.

Machine learning and computer vision techniques have grown rapidly in recent years due to their automation, suitability, and ability to generate astounding results. Hence, in this paper, we survey the key studies that are published between 2014 and 2022, showcasing the different machine learning algorithms researchers have used to segment the liver, hepatic tumors, and hepatic-vasculature structures. We divide the surveyed studies based on the tissue of interest (hepatic-parenchyma, hepatic-tumors, or hepatic-vessels), highlighting the studies that tackle more than one task simultaneously. Additionally, the machine learning algorithms are classified as either supervised or unsupervised, and they are further partitioned if the amount of work that falls under a certain scheme is significant. Moreover, different datasets and challenges found in literature and websites containing masks of the aforementioned tissues are thoroughly discussed, highlighting the organizers’ original contributions and those of other researchers. Also, the metrics used excessively in the literature are mentioned in our review, stressing their relevance to the task at hand. Finally, critical challenges and future directions are emphasized for innovative researchers to tackle, exposing gaps that need addressing, such as the scarcity of many studies on the vessels’ segmentation challenge and why their absence needs to be dealt with sooner than later.

Recently, developing automated video surveillance systems (VSSs) has become crucial to ensure the security and safety of the population, especially during events involving large crowds, such as sporting events. While artificial intelligence (AI) smooths the path of computers to think like humans, machine learning (ML) and deep learning (DL) pave the way more, even by adding training and learning components. DL algorithms require data labeling and high-performance computers to effectively analyze and understand surveillance data recorded from fixed or mobile cameras installed in indoor or outdoor environments. However, they might not perform as expected, take much time in training, or not have enough input data to generalize well. To that end, deep transfer learning (DTL) and deep domain adaptation (DDA) have recently been proposed as promising solutions to alleviate these issues. Typically, they can (i) ease the training process, (ii) improve the generalizability of ML and DL models, and (iii) overcome data scarcity problems by transferring knowledge from one domain to another or from one task to another. Although the increasing number of articles proposed to develop DTL- and DDA-based VSSs, a thorough review that summarizes and criticizes the state-of-the-art is still missing. To that end, this paper introduces, to the best of the authors’ knowledge, the first overview of existing DTL- and DDA-based video surveillance to (i) shed light on their benefits, (ii) discuss their challenges, and (iii) highlight their future perspectives.

Building occupancy information could aid energy preservation while simultaneously maintaining the end-user comfort level. Energy conservation becomes essential since energy resources are scarce and human dependency on appliances is only exponentially increasing. While intrusive sensors (i.e., cameras and microphones) can raise privacy concerns, this paper presents an innovative non-intrusive occupancy detection approach using environmental sensor data (e.g., temperature, humidity, carbon dioxide (CO2), and light sensors). The proposed scheme transforms multivariate time-series data into images for better encoding and extracting relevant features. The utilized image transformation method is based on data normalization and matrix conversion. Precisely, by representing time-series in 2D space, an encoding kernel can move in two directions while it can move only in one direction when applied to a 1D signal. Moreover, machine learning (ML) and deep learning (DL) techniques are utilized to classify occupancy patterns. Several simulations are used to evaluate the approach; mainly, we investigated pre-trained and custom convolutional neural network (CNN) models. The latter attained an accuracy of 99.00%. Additionally, pixel data are extracted from the generated images and subjected to traditional ML methods. Throughout the numerous comparison settings, it was observed that the latter strategy provided the optimal balance of 99.42% accuracy performance and minimal training time across the occupancy datasets.

The building internet of things (BIoT) is quite a promising concept for curtailing energy consumption, reducing costs, and promoting building transformation. Besides, integrating artificial intelligence (AI) into the BIoT is essential for data analysis and intelligent decision-making. Thus, data-driven approaches to infer occupancy patterns usage are gaining growing interest in BIoT applications. Typically, analyzing big occupancy data gathered by BIoT networks helps significantly identify the causes of wasted energy and recommend corrective actions. Within this context, building occupancy data aids in the improvement of the efficacy of energy management systems, allowing the reduction of energy consumption while maintaining occupant comfort. Occupancy data might be collected using a variety of devices. Among those devices are optical/thermal cameras, smart meters, environmental sensors such as carbon dioxide (CO2), and passive infrared (PIR). Even though the latter methods are less precise, they have generated considerable attention owing to their inexpensive cost and low invasive nature. This article provides an in-depth survey of the strategies used to analyze sensor data and determine occupancy. The article’s primary emphasis is on reviewing deep learning (DL), and transfer learning (TL) approaches for occupancy detection. This work investigates occupancy detection methods to develop an efficient system for processing sensor data while providing accurate occupancy information. Moreover, the paper conducted a comparative study of the readily available algorithms for occupancy detection to determine the optimal method in regards to training time and testing accuracy. The main concerns affecting the current occupancy detection system in terms of privacy and precision were thoroughly discussed. For occupancy detection, several directions were provided to avoid or reduce privacy problems by employing forthcoming technologies such as edge devices, Federated learning, and Blockchain-based IoT.

Smart nonintrusive load monitoring (NILM) represents a cost-efficient technology for observing power usage in buildings. It tackles several challenges in transitioning into a more effective, sustainable, and digital energy efficiency environment. This paper presents a comprehensive review of recent trends in the NILM field, in which we propose a multiperspective classification of existing smart NILM techniques. More attention is devoted to describing the contributions of deep learning, feature extraction, computing platforms, and application scenarios for NILM development. Accordingly, NILM technical aspects are first investigated, including data collection devices and public data sets. Next, event-based and non-event-based NILM algorithms are overviewed. Furthermore, potential limitations of existing solutions are identified, highlighting their technical challenges, especially those related to security and privacy preservation, data scarcity, results reproduction, and implementation and business difficulties. Lastly, future directions are explored to overcome the identified limitations.

Nonintrusive load monitoring (NILM) is the de facto technique for extracting device-level power consumption fingerprints at (almost) no cost from only aggregated mains readings. Specifically, there is no need to install an individual meter for each appliance. However, a robust NILM system should incorporate a precise appliance identification module that can effectively discriminate between various devices. In this context, this paper proposes a powerful method to extract accurate power fingerprints for electrical appliance identification. Rather than relying solely on time-domain (TD) analysis, this framework abstracts the phase encoding of the TD description of power signals using a two-dimensional (2D) representation. This allows mapping power trajectories to a novel 2D binary representation space, and then performing a histogramming process after converting binary codes to new decimal representations. This yields the final histogram of 2D phase encoding of power signals, namely, 2D-PEP. An empirical performance evaluation conducted with three realistic power consumption databases collected at distinct resolutions indicates that the proposed 2D-PEP descriptor achieves outperformance for appliance identification in comparison with other recent techniques. Accordingly, high identification accuracies are attained on the GREEND, UK-DALE, and WHITED data sets, where 99.54%, 98.78%, and 100% rates have been achieved, respectively, using the proposed 2D-PEP descriptor.

Anomaly detection in energy consumption is a crucial step towards developing efficient energy saving systems, diminishing overall energy expenditure and reducing carbon emissions. Therefore, implementing powerful techniques to identify anomalous consumption in buildings and providing this information to end-users and managers is of significant importance. Accordingly, two novel schemes are proposed in this paper; the first one is an unsupervised abnormality detection based on one-class support vector machine, namely UAD-OCSVM, in which abnormalities are extracted without the need of annotated data; the second is a supervised abnormality detection based on micromoments (SAD-M2), which is implemented in the following steps: (i) normal and abnormal power consumptions are defined and assigned; (ii) a rule-based algorithm is introduced to extract the micromoments representing the intent-rich moments, in which the end-users make decisions to consume energy; and (iii) an improved K-nearest neighbors model is introduced to automatically classify consumption footprints as normal or abnormal. Empirical evaluation conducted in this framework under three different data sets demonstrates that SAD-M2 achieves both a highest abnormality detection performance and real-time processing capability with considerably lower computational cost in comparison with other machine learning methods. For instance, up to 99.71% accuracy and 99.77% F1 score have been achieved using a real-world data set collected at the Qatar University energy lab.

The recent advances in artificial intelligence namely in machine learning and deep learning, have boosted the performance of intelligent systems in several ways. This gave rise to human expectations, but also created the need for a deeper understanding of how intelligent systems think and decide. The concept of explainability appeared, in the extent of explaining the internal system mechanics in human terms. Recommendation systems are intelligent systems that support human decision making, and as such, they have to be explainable to increase user trust and improve the acceptance of recommendations. In this study, we focus on a context-aware recommendation system for energy efficiency and develop a mechanism for explainable and persuasive recommendations, which are personalized to user preferences and habits. The persuasive facts either emphasize on the economical saving prospects (Econ) or on a positive ecological impact (Eco) and explanations provide the reason for recommending an energy saving action. Based on a study conducted using a Telegram bot, different scenarios have been validated with actual data and human feedback. Current results show a total increase of 19% on the recommendation acceptance ratio when both economical and ecological persuasive facts are employed. This revolutionary approach on recommendation systems, demonstrates how intelligent recommendations can effectively encourage energy saving behavior.

Identifying domestic appliances in the smart grid leads to a better power usage management and further helps in detecting appliance-level abnormalities. An efficient identification can be achieved only if a robust feature extraction scheme is developed with a high ability to discriminate between different appliances on the smart grid. Accordingly, we propose in this paper a novel method to extract electrical power signatures after transforming the power signal to 2D space, which has more encoding possibilities. Following, an improved local binary patterns (LBP) is proposed that relies on improving the discriminative ability of conventional LBP using a post-processing stage. A binarized eigenvalue map (BEVM) is extracted from the 2D power matrix and then used to post-process the generated LBP representation. Next, two histograms are constructed, namely up and down histograms, and are then concatenated to form the global histogram. A comprehensive performance evaluation is performed on two different datasets, namely the GREEND and WITHED, in which power data were collected at 1 Hz and 44000 Hz sampling rates, respectively. The obtained results revealed the superiority of the proposed LBP-BEVM based system in terms of the identification performance versus other 2D descriptors and existing identification frameworks.

Consciousness about power consumption at the appliance level can assist user in promoting energy efficiency in households. In this paper, a superior non-intrusive appliance recognition method that can provide particular consumption footprints of each appliance is proposed. Electrical devices are well recognized by the combination of different descriptors via the following steps: (a) investigating the applicability along with performance comparability of several time-domain (TD) feature extraction schemes; (b) exploring their complementary features; and (c) making use of a new design of the ensemble bagging tree (EBT) classifier. Consequently, a powerful feature extraction technique based on the fusion of TD features is proposed, namely fTDF, aimed at improving the feature discrimination ability and optimizing the recognition task. An extensive experimental performance assessment is performed on two different datasets called the GREEND and WITHED, where power consumption signatures were gathered at 1 Hz and 44000 Hz sampling frequencies, respectively. The obtained results revealed prime efficiency of the proposed fTDF based EBT system in comparison with other TD descriptors and machine learning classifiers.