Electronics

25 pages, 6401 KB

Open AccessArticle

Spiking Neural Network-Based Bidirectional Associative Learning Circuit for Efficient Multibit Pattern Recall in Neuromorphic Systems

by Min Jee Kim, Hyung-Min Lee, YeonJoo Jeong and Joon Young Kwak

Electronics 2025, 14(19), 3971; https://doi.org/10.3390/electronics14193971 - 9 Oct 2025

Viewed by 484

Abstract

Associative learning is a fundamental neural mechanism in human memory and cognition. It has attracted considerable attention in neuromorphic system design owing to its multimodal integration, fault tolerance, and energy efficiency. However, prior studies mostly focused on single inputs, with limited attention to [...] Read more.

Associative learning is a fundamental neural mechanism in human memory and cognition. It has attracted considerable attention in neuromorphic system design owing to its multimodal integration, fault tolerance, and energy efficiency. However, prior studies mostly focused on single inputs, with limited attention to multibit pairs or recall under non-orthogonal input patterns. To address these issues, this study proposes a bidirectional associative learning system using paired multibit inputs. It employs a synapse–neuron structure based on spiking neural networks (SNNs) that emulate biological learning, with simple circuits supporting synaptic operations and pattern evaluation. Importantly, the update and read functions were designed by drawing inspiration from the operational characteristics of emerging synaptic devices, thereby ensuring future compatibility with device-level implementations. The proposed system was verified through Cadence-based simulations using CMOS neurons and Verilog-A synapses. The results show that all patterns are reliably recalled under intact synaptic conditions, and most patterns are still robustly recalled under biologically plausible conditions such as partial synapse loss or noisy initial synaptic weight states. Moreover, by avoiding massive data converters and relying only on basic digital gates, the proposed design achieves associative learning with a simple structure. This provides an advantage for future extension to large-scale arrays. Full article

► Show Figures

Figure 1

17 pages, 3259 KB

Open AccessArticle

A Multivector Direct Model Predictive Control Scheme with Harmonic Suppression for DTP-PMSMs

by Baoyun Qi, Rui Yang, Yu Lu, Zhen Zhang, Bingchen Liang, Bin Deng, Jiancheng Liu, Liwei Yu and Hongyun Wu

Electronics 2025, 14(19), 3970; https://doi.org/10.3390/electronics14193970 - 9 Oct 2025

Viewed by 342

Abstract

A multivector direct model predictive control (DMPC) scheme is proposed for the dual three-phase permanent magnet synchronous machine (DTP-PMSM) drive system to achieve closed-loop control for both fundamental current tracking and harmonic current minimization. The proposed multivector DMPC scheme employs four active voltage [...] Read more.

A multivector direct model predictive control (DMPC) scheme is proposed for the dual three-phase permanent magnet synchronous machine (DTP-PMSM) drive system to achieve closed-loop control for both fundamental current tracking and harmonic current minimization. The proposed multivector DMPC scheme employs four active voltage vectors, including two large vectors and two basic vectors for implicit modulation. Moreover, the control optimization problem is formulated as a four-dimensional quadratic programming problem, which is suitable for real-time implementation. The proposed multivector DMPC scheme enables fast and accurate tracking of the fundamental current as well as effective suppression of harmonic currents in both the fundamental and harmonic subspaces. In addition, a Kalman filter observer is incorporated to enhance robustness against model uncertainties and disturbances. Experimental results on a DTP-PMSM test bench verify that the proposed multivector DMPC scheme effectively reduces torque ripple, improves current quality, and enhances both steady-state and transient performance of the system. Full article

(This article belongs to the Special Issue Emerging Technologies in Wireless Power and Energy Transfer Systems)

► Show Figures

Figure 1

15 pages, 2538 KB

Open AccessArticle

Active Damped PI Speed Loop Design for Motor Direct-Drive Operating Mechanism for High-Voltage Circuit Breakers

by Xiao Wang, Xusheng Wu and Xi Xiao

Electronics 2025, 14(19), 3969; https://doi.org/10.3390/electronics14193969 - 9 Oct 2025

Viewed by 412

Abstract

To address the prevalent issues of oscillation and overshoot in high-voltage circuit breaker motor direct-drive mechanisms under classical PI control, this paper proposes an optimized PI speed loop with active damping characteristics. By first establishing a detailed kinematic and dynamic model of the [...] Read more.

To address the prevalent issues of oscillation and overshoot in high-voltage circuit breaker motor direct-drive mechanisms under classical PI control, this paper proposes an optimized PI speed loop with active damping characteristics. By first establishing a detailed kinematic and dynamic model of the mechanism, we reveal the inherent coupling between tracking performance, disturbance immunity, and the damping ratio within the classical PI speed loop. Our novel method introduces a speed feedback channel at the output of the PI controller to synthesize equivalent viscous damping, thereby enhancing system stability without compromising responsiveness. Through rigorous simulation and experimental validation, the proposed controller’s effectiveness is demonstrated. Compared with the traditional PI controller, the ADPI method reduces the velocity overshoot to only 5.76% in the startup phase, and the maximum velocity tracking error of the velocity is only 18.62% and the cumulative position tracking error is only 0.632 rad under the actual working condition, which is a reduction of 42.7% in the positional error relative to the traditional PI method. The controller also exhibits low sensitivity to changes in the system’s equivalent rotational inertia. This work provides a low-complexity and easy-to-implement speed loop performance enhancement scheme, ideally suited for the short-duration, high-dynamic-load conditions of high-voltage circuit breaker applications. Full article

(This article belongs to the Section Electronic Materials, Devices and Applications)

► Show Figures

Figure 1

16 pages, 7184 KB

Open AccessArticle

Towards Robust Scene Text Recognition: A Dual Correction Mechanism with Deformable Alignment

by Yajiao Feng and Changlu Li

Electronics 2025, 14(19), 3968; https://doi.org/10.3390/electronics14193968 - 9 Oct 2025

Viewed by 448

Abstract

Scene Text Recognition (STR) faces significant challenges under complex degradation conditions, such as distortion, occlusion, and semantic ambiguity. Most existing methods rely heavily on language priors for correction, but effectively constructing language rules remains a complex problem. This paper addresses two key challenges: [...] Read more.

Scene Text Recognition (STR) faces significant challenges under complex degradation conditions, such as distortion, occlusion, and semantic ambiguity. Most existing methods rely heavily on language priors for correction, but effectively constructing language rules remains a complex problem. This paper addresses two key challenges: (1) The over-correction behavior of language models, particularly on semantically deficient input, can result in both recognition errors and loss of critical information. (2) Character misalignment in visual features, which affects recognition accuracy. To address these problems, we propose a Deformable-Alignment-based Dual Correction Mechanism (DADCM) for STR. Our method includes the following key components: (1) We propose a visually guided and language-assisted correction strategy. A dynamic confidence threshold is used to control the degree of language model intervention. (2) We designed a visual backbone network called SCRTNet. The net enhances key text regions through a channel attention module (SENet) and applies deformable convolution (DCNv4) in deep layers to better model distorted or curved text. (3) We propose a deformable alignment module (DAM). The module combines Gumbel-Softmax-based anchor sampling and geometry-aware self-attention to improve character alignment. Experiments on multiple benchmark datasets demonstrate the superiority of our approach. Especially on the Union14M-Benchmark, where the recognition accuracy surpasses previous methods by 1.1%, 1.6%, 3.0%, and 1.3% on the Curved, Multi-Oriented, Contextless, and General subsets, respectively. Full article

► Show Figures

Figure 1

23 pages, 1058 KB

Open AccessArticle

SM-GCG: Spatial Momentum Greedy Coordinate Gradient for Robust Jailbreak Attacks on Large Language Models

by Landi Gu, Xu Ji, Zichao Zhang, Junjie Ma, Xiaoxia Jia and Wei Jiang

Electronics 2025, 14(19), 3967; https://doi.org/10.3390/electronics14193967 - 9 Oct 2025

Viewed by 800

Abstract

Recent advancements in large language models (LLMs) have increased the necessity of alignment and safety mechanisms. Despite these efforts, jailbreak attacks remain a significant threat, exploiting vulnerabilities to elicit harmful responses. While white-box attacks, such as the Greedy Coordinate Gradient (GCG) method, have [...] Read more.

Recent advancements in large language models (LLMs) have increased the necessity of alignment and safety mechanisms. Despite these efforts, jailbreak attacks remain a significant threat, exploiting vulnerabilities to elicit harmful responses. While white-box attacks, such as the Greedy Coordinate Gradient (GCG) method, have demonstrated promise, their efficacy is often limited by non-smooth optimization landscapes and a tendency to converge to local minima. To mitigate these issues, we propose Spatial Momentum GCG (SM-GCG), a novel method that incorporates spatial momentum. This technique aggregates gradient information across multiple transformation spaces—including text, token, one-hot, and embedding spaces—to stabilize the optimization process and enhance the estimation of update directions, thereby more effectively exploiting model vulnerabilities to elicit harmful responses. Experimental results on models including Vicuna-7B, Guanaco-7B, and Llama2-7B-Chat demonstrate that SM-GCG significantly enhances the attack success rate in white-box settings. The method achieves a 10–15% improvement in attack success rate over baseline methods against robust models such as Llama2, while also exhibiting enhanced transferability to black-box models. These findings indicate that spatial momentum effectively mitigates the problem of local optima in discrete prompt optimization, thereby offering a more powerful and generalizable approach for red-team assessments of LLM safety. Warning: This paper contains potentially offensive and harmful text. Full article

► Show Figures

Figure 1

30 pages, 37101 KB

Open AccessArticle

FPGA Accelerated Large-Scale State-Space Equations for Multi-Converter Systems

by Jiyuan Liu, Mingwang Xu, Hangyu Yang, Zhiqiang Que, Wei Gu, Yongming Tang, Baoping Wang and He Li

Electronics 2025, 14(19), 3966; https://doi.org/10.3390/electronics14193966 - 9 Oct 2025

Viewed by 408

Abstract

The increasing integration of high-frequency power electronic converters in renewable energy-grid systems has escalated reliability concerns, necessitating FPGA-accelerated large-scale real-time electromagnetic transient (EMT) computation to prevent failures. However, most existing studies prioritize computational performance and struggle to achieve large-scale EMT computation. To enhance [...] Read more.

The increasing integration of high-frequency power electronic converters in renewable energy-grid systems has escalated reliability concerns, necessitating FPGA-accelerated large-scale real-time electromagnetic transient (EMT) computation to prevent failures. However, most existing studies prioritize computational performance and struggle to achieve large-scale EMT computation. To enhance the computational scale, we propose a scalable hardware architecture comprising domain-specific components and data-centric processing element (PE) arrays. This architecture is further enhanced by a graph-based matrix mapping methodology and matrix-aware fixed-point quantization for hardware-efficient computation. We demonstrate our principles with FPGA implementations of large-scale multi-converter systems. The experimental results show that we set a new record of supporting 1200 switches with a computation latency of 373 ns and an accuracy of 99.83% on FPGA implementations. Compared to the state-of-the-art large-scale EMT computation on FPGAs, our design on U55C FPGA achieves an up-to 200.00× increase in the switch scale, without I/O resource limitations, and demonstrates up-to 71.70% reduction in computation error and 51.43% reduction in DSP consumption, respectively. Full article

► Show Figures

Figure 1

12 pages, 2224 KB

Open AccessArticle

A Memory-Efficient Compensation Algorithm for Vertical Crosstalk in 8K LCD Panels

by Yongwoo Lee, Kiwon Choi, Hyeryoung Park, Yong Ju Kim, Kookhyun Choi, Jae-Hong Jeon and Min Jae Ko

Electronics 2025, 14(19), 3965; https://doi.org/10.3390/electronics14193965 - 9 Oct 2025

Viewed by 372

Abstract

As ultra-high resolution liquid crystal displays (LCDs) advance, crosstalk has become a critical challenge due to the reduced spacing of electronic circuits and increased signal frequencies. In particular, vertical crosstalk (V-CT) in vertical-alignment LCDs arises mainly from fringing electric fields generated by data [...] Read more.

As ultra-high resolution liquid crystal displays (LCDs) advance, crosstalk has become a critical challenge due to the reduced spacing of electronic circuits and increased signal frequencies. In particular, vertical crosstalk (V-CT) in vertical-alignment LCDs arises mainly from fringing electric fields generated by data lines, along with secondary contributions from data line–pixel coupling effect, thin-film transistor leakage, and other factors. To resolve V-CT, we propose a memory-efficient compensation algorithm implemented on a field-programmable gate array as a customized timing controller. The proposed algorithm achieves compensation accuracy within 2% while significantly reducing memory requirements. A conventional 7680 × 4320 pixel LCD panel requires approximately 796 MB of memory for compensation data, whereas our method reduces this to only 0.37 MB—a nearly 2000-fold reduction—by referencing only preceding pixel information. This approach enables cost-effective implementation, faster processing, and enhanced image quality. Overall, the proposed method provides a practical and scalable solution for resolving V-CT in 8K LCD panels, establishing a new benchmark for high-resolution display technologies. Full article

(This article belongs to the Section Electronic Materials, Devices and Applications)

► Show Figures

Figure 1

17 pages, 2289 KB

Open AccessArticle

Aging-Aware Character Recognition with E-Textile Inputs

by Juncong Lin, Yujun Rong, Yao Cheng and Chenkang He

Electronics 2025, 14(19), 3964; https://doi.org/10.3390/electronics14193964 - 9 Oct 2025

Viewed by 305

Abstract

E-textiles, a type of textile integrated with conductive sensors, allows users to freely utilize any area of the body in a convenient and comfortable manner. Thus, interactions with e-textiles are attracting more and more attention, especially for text input. However, the functional aging [...] Read more.

E-textiles, a type of textile integrated with conductive sensors, allows users to freely utilize any area of the body in a convenient and comfortable manner. Thus, interactions with e-textiles are attracting more and more attention, especially for text input. However, the functional aging of e-textiles affects the characteristics and even the quality of the captured signal, presenting serious challenges for character recognition. This paper focuses on studying the behavior of e-textile functional aging and alleviating its impact on text input with an unsupervised domain adaptation technique, named A²TEXT (aging-aware e-textile-based text input). We first designed a deep kernel-based two-sample test method to validate the impact of functional aging on handwriting with an e-textile input. Based on that, we introduced a so-called Gabor domain adaptation technique, which adopts a novel Gabor orientation filter in feature extraction under an adversarial domain adaptation framework. We demonstrated superior performance compared to traditional models in four different transfer tasks, validating the effectiveness of our work. Full article

(This article belongs to the Special Issue End User Applications for Virtual, Augmented, and Mixed Reality)

► Show Figures

Figure 1

26 pages, 5816 KB

Open AccessArticle

Disturbance-Free Switching Control Strategy for Grid-Following/Grid-Forming Modes of Energy Storage Converters

by Geling Jiang, Siyu Kan, Yuhang Li and Xiaorong Zhu

Electronics 2025, 14(19), 3963; https://doi.org/10.3390/electronics14193963 - 9 Oct 2025

Viewed by 437

Abstract

To address the problem of transient disturbance arising during the grid-following (GFL) and grid-forming (GFM) mode switching of energy storage converters, this paper proposes a dual-mode seamless switching control strategy. First, we conduct an in-depth analysis of the mechanism behind switching transients, identifying [...] Read more.

To address the problem of transient disturbance arising during the grid-following (GFL) and grid-forming (GFM) mode switching of energy storage converters, this paper proposes a dual-mode seamless switching control strategy. First, we conduct an in-depth analysis of the mechanism behind switching transients, identifying that sudden changes in current commands and angle-control misalignment are the key factors triggering oscillations in system power and voltage frequency. To overcome this, we design a virtual synchronous generator (VSG) control angle-tracking technique based on the construction of triangular functions, which effectively eliminates the influence of periodic phase-angle jumps on tracking accuracy and achieves precise pre-synchronization of the microgrid phase in GFM mode. Additionally, we employ a current-command seamless switching technique involving real-time latching and synchronization of the inner-loop current references between the two modes, ensuring continuity of control commands at the switching instant. The simulation and hardware-in-the-loop (HIL) experimental results show that the proposed strategy does not require retuning of the parameters after switching, greatly suppresses voltage and frequency fluctuations during mode transition, and achieves smooth, rapid, seamless switching between the GFL and GFM modes of the energy storage converter, thereby improving the stability of microgrid operation. Full article

► Show Figures

Figure 1

2 pages, 132 KB

Open AccessRetraction

RETRACTED: Srinivasan et al. Detection and Grade Classification of Diabetic Retinopathy and Adult Vitelliform Macular Dystrophy Based on Ophthalmoscopy Images. Electronics 2023, 12, 862

by Saravanan Srinivasan, Rajalakshmi Nagarnaidu Rajaperumal, Sandeep Kumar Mathivanan, Prabhu Jayagopal, Sujatha Krishnamoorthy and Seifedine Kardy

Electronics 2025, 14(19), 3962; https://doi.org/10.3390/electronics14193962 - 9 Oct 2025

Viewed by 263

Abstract

The journal retracts the article “Detection and Grade Classification of Diabetic Retinopathy and Adult Vitelliform Macular Dystrophy Based on Ophthalmoscopy Images” [...] Full article

16 pages, 5456 KB

Open AccessArticle

The New Precise Positioning System of the Heavy Hadron Calorimeter FPSD in the NA61/SHINE Experiment Based on the Siemens 1200 Controller Connected with the EPICS Software

by Marcin Bielewicz, Piotr Mazerewicz, Jarosław Szewiński, Krystian Grodzicki, Ian Crotty, Michał Kiecana, Łukasz Świderski, Tomasz Szczęśniak, Piotr Podlaski, Tomasz Kowalski and Konrad Chmielewski

Electronics 2025, 14(19), 3961; https://doi.org/10.3390/electronics14193961 - 9 Oct 2025

Cited by 1 | Viewed by 286

Abstract

The NA61/SHINE collaboration conducts research using the SPS CERN accelerator, focusing primarily on the strong interaction program. In this type of research, it is necessary to use a hadronic calorimeter called PSD to determine the centrality value of nuclear collisions. The detector consists [...] Read more.

The NA61/SHINE collaboration conducts research using the SPS CERN accelerator, focusing primarily on the strong interaction program. In this type of research, it is necessary to use a hadronic calorimeter called PSD to determine the centrality value of nuclear collisions. The detector consists of two separate parts, the MPSD and the FPSD. The FPSD, which is a new detector added to the NA61 SHINE experiment from 2022, has not yet had a functional system for remotely changing and measuring the detector position. Such a remote system is necessary for faster detector calibration, more precise positioning of the detector in the accelerator beam path, and improved safety. For these reasons, in 2023, a group of specialists from the NCBJ laboratory at Poland, prepared a project and built a remote position change system for the FPSD detector. In this work, we describe the main design assumptions and main features of the finished system. We also describe its control system based on the Siemens 1200 PLC controller and the way we supervise its operation through an external DCS system based on the EPICS software (ver.3.16). The introduced changes improved the safety and comfort of work, reduced the radiation risk, and, above all, significantly shortened the time required to change the position of the FPSD detector. Full article

(This article belongs to the Section Systems & Control Engineering)

► Show Figures

Figure 1

21 pages, 3712 KB

Open AccessArticle

CISC-YOLO: A Lightweight Network for Micron-Level Defect Detection on Wafers via Efficient Cross-Scale Feature Fusion

by Yulun Chi, Xingyu Gong, Bing Zhao and Lei Yao

Electronics 2025, 14(19), 3960; https://doi.org/10.3390/electronics14193960 - 9 Oct 2025

Viewed by 502

Abstract

With the development of the semiconductor manufacturing process towards miniaturization and high integration, the detection of microscopic defects on wafer surfaces faces the challenge of balancing precision and efficiency. Therefore, this study proposes a lightweight inspection model based on the YOLOv8 framework, aiming [...] Read more.

With the development of the semiconductor manufacturing process towards miniaturization and high integration, the detection of microscopic defects on wafer surfaces faces the challenge of balancing precision and efficiency. Therefore, this study proposes a lightweight inspection model based on the YOLOv8 framework, aiming to achieve an optimal balance between inspection accuracy, model complexity, and inference speed. First, we design a novel lightweight module called IRB-GhostConv-C2f (IGC) to replace the C2f module in the backbone, thereby significantly minimizing redundant feature computations. Second, a CNN-based cross-scale feature fusion neck network, the CCFF-ISC neck, is proposed to reduce the redundant computation of low-level features and enhance the expression of multi-scale semantic information. Meanwhile, the novel IRB-SCSA-C2f (ISC) module replaces the C2f in the neck to further improve the efficiency of feature fusion. In addition, a novel dynamic head network, DyHeadv3, is integrated into the head structure, aiming to improve the small-scale target detection performance by dynamically adjusting the feature interaction mechanism. Finally, so as to comprehensively assess the proposed algorithm’s performance, an industrial dataset of wafer defects, WSDD, is constructed, which covers “broken edges”, “scratches”, “oil pollution”, and “minor defects”. The experimental results demonstrate that the CISC-YOLO model attains an mAP50 of 93.7%, and the parameter amount is reduced to 1.92 M, outperforming other mainstream leading algorithms in the field. The proposed approach provides a high-precision and low-latency real-time defect detection solution for semiconductor industry scenarios. Full article

(This article belongs to the Special Issue AI-Based Image Processing Detection and Classification Analysis for Multidisciplinary Approaches)

► Show Figures

Figure 1

23 pages, 769 KB

Open AccessArticle

Hierarchical and Clustering-Based Timely Information Announcement Mechanism in the Computing Networks

by Ranran Wei and Rui Han

Electronics 2025, 14(19), 3959; https://doi.org/10.3390/electronics14193959 - 8 Oct 2025

Viewed by 360

Abstract

Information announcement is the process of propagating and synchronizing the information of Computing Resource Nodes (CRNs) within the system of the Computing Networks. Accurate and timely acquisition of information is crucial to ensuring the efficiency and quality of subsequent task scheduling. However, existing [...] Read more.

Information announcement is the process of propagating and synchronizing the information of Computing Resource Nodes (CRNs) within the system of the Computing Networks. Accurate and timely acquisition of information is crucial to ensuring the efficiency and quality of subsequent task scheduling. However, existing announcement mechanisms primarily focus on reducing communication overhead, often neglecting the direct impact of information freshness on scheduling accuracy and service quality. To address this issue, this paper proposes a hierarchical and clustering-based announcement mechanism for the wide-area Computing Networks. The mechanism first categorizes the Computing Network Nodes (CNNs) into different layers based on the type of CRNs they interconnect to, and a top-down cross-layer announcement strategy is introduced during this process; within each layer, CNNs are further divided into several domains according to the round-trip time (RTT) to each other; and in each domain, inspired by the “Six Degrees of Separation” concept from social propagation, a RTT-aware fast clustering algorithm canopy is employed to partition CNNs into multiple overlap clusters. Intra-cluster announcements are modeled as a Traveling Salesman Problem (TSP) and optimized to accelerate updates, while inter-cluster propagation leverages overlapping nodes for global dissemination. Experimental results demonstrate that, by exploiting shortest path optimization within clusters and overlapping-node-based inter-cluster transmission, the mechanism is significantly superior to the comparison scheme in key indicators such as convergence time, Age of Information (AoI), and communication data volume per hop. The mechanism exhibits strong scalability and adaptability in large-scale network environments, providing robust support for efficient and rapid information synchronization in the Computing Networks. Full article

(This article belongs to the Section Networks)

► Show Figures

Figure 1

23 pages, 3069 KB

Open AccessArticle

Fast Discrete Krawtchouk Transform Algorithms for Short-Length Input Sequences

by Marina Polyakova, Aleksandr Cariow and Janusz P. Papliński

Electronics 2025, 14(19), 3958; https://doi.org/10.3390/electronics14193958 - 8 Oct 2025

Viewed by 367

Abstract

This paper presents new fast discrete Krawtchouk transform (DKT) algorithms for input sequences of length 3 to 8. Small-sized DKT algorithms can be utilized in image processing applications to extract local image features formed by a sliding spatial window, and they can also [...] Read more.

This paper presents new fast discrete Krawtchouk transform (DKT) algorithms for input sequences of length 3 to 8. Small-sized DKT algorithms can be utilized in image processing applications to extract local image features formed by a sliding spatial window, and they can also serve as building blocks for developing larger-sized algorithms. Existing strategies to reduce the computational complexity of DKT mainly focus on modifying the recurrence relations for Krawtchouk polynomials, dividing the input signals into blocks or layers, or using different methods to approximate the coefficient values. Algorithms developed using the first two strategies are computationally intensive, which introduces a significant time delay in the computation process. Algorithms based on the approximation of polynomial coefficient values reduce computation time but at the expense of reduced accuracy. We use a different approach based on reducing the block structure of the matrix to one of the previously developed block-structural patterns, which allows us to factorize the resulting matrix in such a way that it leads to a reduction in the computational complexity of the synthesized algorithm. We describe the algorithmic solutions we have obtained through data flow graphs. The proposed DKT algorithms reduce the number of multiplications, additions, and shifts by an average of 58%, 27%, and 68%, respectively, compared to the direct computation of DKT via matrix-vector product. These characteristics were averaged across the considered input sizes (from 3 to 8). Full article

(This article belongs to the Section Circuit and Signal Processing)

► Show Figures

Figure 1

24 pages, 3386 KB

Open AccessArticle

Characterization of Students’ Thinking States Active Based on Improved Bloom Classification Algorithm and Cognitive Diagnostic Model

by Yipeng Liu, Hua Yuan, Zhaoyu Shou, Chenchen Lu and Jianwen Mo

Electronics 2025, 14(19), 3957; https://doi.org/10.3390/electronics14193957 - 8 Oct 2025

Viewed by 326

Abstract

A student’s active thinking state directly affects their learning experience in the classroom. To help teachers understand students’ active thinking states in real-time, this study aims to construct a model which characterizes their active thinking states. The main research objectives are as follows: [...] Read more.

A student’s active thinking state directly affects their learning experience in the classroom. To help teachers understand students’ active thinking states in real-time, this study aims to construct a model which characterizes their active thinking states. The main research objectives are as follows: (1) to achieve accurate classification of the cognitive levels of in-class exercises; (2) to effectively quantify the active thinking state of students through analyzing the correlation between student cognitive levels and exercise cognitive levels. The research methods used in this study to achieve these objectives are as follows: First, LSTM and Chinese-RoBERTa-wwm models are integrated to extract sequential and semantic information from plain text while TBCC is used to extract the semantic features of code text, allowing for comprehensive determination of the cognitive level of exercises. Second, a cognitive diagnosis model—namely, the QRCDM—is adopted to evaluate students’ real-time cognitive levels with respect to knowledge points. Finally, the cognitive levels of exercises and students are input into a self-attention mechanism network, their correlation is analyzed, and the thinking activity state is generated as a state representation. The proposed text classification model outperforms baseline models regarding ACC, micro-F1, and macro-F1 scores on two sets of exercise datasets in Chinese containing mixed code texts, with the highest ACC, micro-F1, and macro-F1 values reaching 0.7004, 0.6941, and 0.6912, respectively. This proves the proposed model’s effectiveness in classifying the cognitive level of exercises. The accuracy of the thinking activity state characterization model reaches 61.54%. In particular, this is higher than the random baseline, thus verifying the model’s feasibility. Full article

(This article belongs to the Special Issue Emerging Theory and Applications in Natural Language Processing, 2nd Edition)

► Show Figures

Figure 1

16 pages, 798 KB

Open AccessArticle

Smart Spectrum Recommendation Approach with Edge Learning for 5G and Beyond Radio Planning

by Ahmet Yazar, Abdulkadir Sönmezışık, Metehan Doğan, Emre Kart and Ayşe Ayhan

Electronics 2025, 14(19), 3956; https://doi.org/10.3390/electronics14193956 - 8 Oct 2025

Viewed by 363

Abstract

Radio spectrum planning has become increasingly important, since the radio spectrum is a scarce resource. Moreover, the utilization of millimeter wave (mmWave) frequencies with fifth-generation (5G) standards has made radio planning more compelling. Considering their different strengths and weaknesses, it is essential to [...] Read more.

Radio spectrum planning has become increasingly important, since the radio spectrum is a scarce resource. Moreover, the utilization of millimeter wave (mmWave) frequencies with fifth-generation (5G) standards has made radio planning more compelling. Considering their different strengths and weaknesses, it is essential to know when mmWave frequencies should be selected in radio planning. In this paper, an approach with edge learning is developed to provide smart spectrum recommendations on which frequency bands should be used for a region. Using the proposed approach, radio spectrum planning can be carried out more efficiently, especially for the frequency ranges of mmWave communications. The proposed approach is designed with a distributed structure, based on awareness of the environment and ambient intelligence. This approach can be performed for each transmission point considering the environment information of the related coverage area. As a result, radio spectrum planning can be conducted for an entire region with the proposed system. The results show that this study both enhances overall user satisfaction and provides reasonable recommendations to operators in the transition to mmWave usage. Thus, the developed approach can be utilized for 5G and beyond communications. Specifically, this methodology is based on applying supervised ML algorithms to a synthetically generated dataset, and the best model achieves around 80% classification accuracy, demonstrating the feasibility of the approach. These quantitative results confirm its practicality and provide a concrete baseline for future studies. Full article

(This article belongs to the Special Issue RF and Millimeter-Wave Technologies for Next-Generation Wireless Communications)

► Show Figures

Figure 1

13 pages, 11609 KB

Open AccessArticle

A 12-bit 100 MSPS Full-Swing Current-Steering Digital-to-Analog Converter with Half-Power Supply Calibration Technique

by Kwangjin Park, Seung Gu Choi, Jintae Kim, Myungsik Kim, Hyunjin Song, Minkyu Song and Soo Youn Kim

Electronics 2025, 14(19), 3955; https://doi.org/10.3390/electronics14193955 - 8 Oct 2025

Viewed by 399

Abstract

We present a digital-to-analog converter (DAC) with full-swing DAC output and a proposed half-power supply calibration technique. To generate a full-swing DAC output, symmetric thermometer decoders and an output selector are implemented to select the appropriate current cell according to the output voltage [...] Read more.

We present a digital-to-analog converter (DAC) with full-swing DAC output and a proposed half-power supply calibration technique. To generate a full-swing DAC output, symmetric thermometer decoders and an output selector are implemented to select the appropriate current cell according to the output voltage range. Furthermore, to improve the linearity, we propose a half-power supply calibration circuit consisting of comparators and calibration counters to control the current of the current cells at the half-power supply voltage point, where the voltage mismatch typically occurs. The DAC was fabricated in a 28 nm CMOS process, with a full chip area of 0.95 mm × 0.93 mm. The measurement results demonstrate a maximum voltage mismatch improvement of 95% when using the proposed half-power supply calibration technique, with DNL and INL values of 0.39 and 1.15 LSB. The total power consumption was 73.8 mW at 100 MSPS, with analog and digital supply voltages of 1.8 and 1.0 V, respectively. Full article

(This article belongs to the Special Issue Low-Voltage Mixed-Signal CMOS Integrated Circuits for Emerging Applications)

► Show Figures

Figure 1

25 pages, 876 KB

Open AccessArticle

Blockchain-Based Self-Sovereign Identity Management Mechanism in AIoT Environments

by Jingjing Ren, Jie Zhang, Yongjun Ren and Jiang Xu

Electronics 2025, 14(19), 3954; https://doi.org/10.3390/electronics14193954 - 8 Oct 2025

Viewed by 654

Abstract

With the rapid growth of Artificial Intelligence of Things (AIoT), identity management and trusted communication have become critical for system security and reliability. Continuous AI learning and large-scale device connectivity introduce challenges such as permission drift, cross-domain access, and fine-grained API calls. Traditional [...] Read more.

With the rapid growth of Artificial Intelligence of Things (AIoT), identity management and trusted communication have become critical for system security and reliability. Continuous AI learning and large-scale device connectivity introduce challenges such as permission drift, cross-domain access, and fine-grained API calls. Traditional identity management often fails to balance privacy protection with efficiency, leading to risks of data leakage and misuse. To address these issues, this paper proposes a blockchain-based self-sovereign identity (SSI) management mechanism for AIoT. By integrating SSI with a zero-trust framework, it achieves decentralized identity storage and continuous verification, effectively preventing unauthorized access and misuse of identity data. The mechanism employs selective disclosure (SD) technology, allowing users to submit only necessary attributes, thereby ensuring user control over self-sovereign identity information and guaranteeing the privacy and integrity of undisclosed attributes. This significantly reduces verification overhead. Additionally, this paper designs a context-aware dynamic permission management that generates minimal permission sets in real time based on device requirements and environmental changes. Combined with the zero-trust principles of continuous verification and least privilege, it enhances secure interactions while maintaining flexibility. Performance experiments demonstrate that, compared with conventional approaches, the proposed zero-trust architecture-based SSI management mechanism better mitigates the risk of sensitive attribute leakage, improves identity verification efficiency under SD, and enhances the responsiveness of dynamic permission management, providing robust support for secure and efficient AIoT operations. Full article

(This article belongs to the Topic Recent Advances in Security, Privacy, and Trust)

► Show Figures

Figure 1

20 pages, 17566 KB

Open AccessArticle

An Isolated AC-DC LED Electronic Lighting Driver Circuit with Power Factor Correction

by Chun-An Cheng, Hung-Liang Cheng, En-Chih Chang and Man-Tang Chang

Electronics 2025, 14(19), 3953; https://doi.org/10.3390/electronics14193953 - 7 Oct 2025

Viewed by 436

Abstract

Light-emitting diodes (LEDs) have gained widespread adoption as solid-state lighting sources due to their compact size, long operational lifetime, high brightness, and mechanical robustness. This paper presents the development and implementation of an isolated AC-DC LED electronic lighting driver circuit that integrates a [...] Read more.

Light-emitting diodes (LEDs) have gained widespread adoption as solid-state lighting sources due to their compact size, long operational lifetime, high brightness, and mechanical robustness. This paper presents the development and implementation of an isolated AC-DC LED electronic lighting driver circuit that integrates a modified flyback converter with a lossless snubber circuit, along with inherent power factor correction (PFC). The proposed design operates the transformer’s magnetizing inductor in the discontinuous conduction mode (DCM), thereby naturally achieving PFC without the need for complex control circuitry. Furthermore, the circuit is capable of recycling the energy stored in the transformer’s leakage inductance, improving overall efficiency. The input current harmonics are shown to comply with the IEC 61000-3-2 Class C standard. A 72 W (36 V/2 A) prototype has been constructed and tested under a 110 V AC input. Experimental results confirm the effectiveness of the proposed design, achieving a power factor of 0.9816, a total harmonic distortion (THD) of 12.094%, an output voltage ripple factor of 9.7%, and an output current ripple factor of 11.22%. These results validate the performance and practical viability of the proposed LED driver architecture. Full article

(This article belongs to the Special Issue Advancements in Electric Motors, Drives, Power Converters and Related Systems, 2nd Edition)

► Show Figures

Figure 1

32 pages, 3383 KB

Open AccessArticle

DLG–IDS: Dynamic Graph and LLM–Semantic Enhanced Spatiotemporal GNN for Lightweight Intrusion Detection in Industrial Control Systems

by Junyi Liu, Jiarong Wang, Tian Yan, Fazhi Qi and Gang Chen

Electronics 2025, 14(19), 3952; https://doi.org/10.3390/electronics14193952 - 7 Oct 2025

Viewed by 436

Abstract

Industrial control systems (ICSs) face escalating security challenges due to evolving cyber threats and the inherent limitations of traditional intrusion detection methods, which fail to adequately model spatiotemporal dependencies or interpret complex protocol semantics. To address these gaps, this paper proposes DLG–IDS—a lightweight [...] Read more.

Industrial control systems (ICSs) face escalating security challenges due to evolving cyber threats and the inherent limitations of traditional intrusion detection methods, which fail to adequately model spatiotemporal dependencies or interpret complex protocol semantics. To address these gaps, this paper proposes DLG–IDS—a lightweight intrusion detection framework that innovatively integrates dynamic graph construction for capturing real–time device interactions and logical control relationships from traffic, LLM–driven semantic enhancement to extract fine–grained embeddings from graphs, and a spatio–temporal graph neural network (STGNN) optimized via sparse attention and local window Transformers to minimize computational overhead. Evaluations on SWaT and SBFF datasets demonstrate the framework’s superiority, achieving a state–of–the–art accuracy of 0.986 while reducing latency by 53.2% compared to baseline models. Ablation studies further validate the critical contributions of semantic fusion, sparse topology modeling, and localized temporal attention. The proposed solution establishes a robust, real–time detection mechanism tailored for resource–constrained industrial environments, effectively balancing high accuracy with operational efficiency. Full article

(This article belongs to the Topic AI and Data-Driven Advancements in Industry 4.0, 2nd Edition)

► Show Figures

Figure 1

18 pages, 864 KB

Open AccessArticle

Enhanced Semantic BERT for Named Entity Recognition in Education

by Ping Huang, Huijuan Zhu, Ying Wang, Lili Dai and Lei Zheng

Electronics 2025, 14(19), 3951; https://doi.org/10.3390/electronics14193951 - 7 Oct 2025

Viewed by 402

Abstract

To address the technical challenges in the educational domain named entity recognition (NER), such as ambiguous entity boundaries and difficulties with nested entity identification, this study proposes an enhanced semantic BERT model (ES-BERT). The model innovatively adopts an education domain, vocabulary-assisted semantic enhancement [...] Read more.

To address the technical challenges in the educational domain named entity recognition (NER), such as ambiguous entity boundaries and difficulties with nested entity identification, this study proposes an enhanced semantic BERT model (ES-BERT). The model innovatively adopts an education domain, vocabulary-assisted semantic enhancement strategy that (1) applies the term frequency–inverse document frequency (TF-IDF) algorithm to weight domain-specific terms, and (2) fuses the weighted lexical information with character-level features, enabling BERT to generate enriched, domain-aware, character–word hybrid representations. A complete bidirectional long short-term memory-conditional random field (BiLSTM-CRF) recognition framework was established, and a novel focal loss-based joint training method was introduced to optimize the process. The experimental design employed a three-phase validation protocol, as follows: (1) In a comparative evaluation using 5-fold cross-validation on our proprietary computer-education dataset, the proposed ES-BERT model yielded a precision of 90.38%, which is higher than that of the baseline models; (2) Ablation studies confirmed the contribution of domain-vocabulary enhancement to performance improvement; (3) Cross-domain experiments on the 2016 knowledge base question answering datasets and resume benchmark datasets demonstrated outstanding precision of 98.41% and 96.75%, respectively, verifying the model’s transfer-learning capability. These comprehensive experimental results substantiate that ES-BERT not only effectively resolves domain-specific NER challenges in education but also exhibits remarkable cross-domain adaptability. Full article

(This article belongs to the Special Issue Advances in Intelligent Data Analysis and Its Applications, 3rd Edition)

► Show Figures

Figure 1

54 pages, 7106 KB

Open AccessReview

Modeling, Control and Monitoring of Automotive Electric Drives

by Pierpaolo Dini, Sergio Saponara, Sajib Chakraborty and Omar Hegazy

Electronics 2025, 14(19), 3950; https://doi.org/10.3390/electronics14193950 - 7 Oct 2025

Viewed by 717

Abstract

The electrification of automotive powertrains has accelerated research efforts in the modeling, control, and monitoring of electric drive systems, where reliability, safety, and efficiency are key enablers for mass adoption. Despite a large corpus of literature addressing individual aspects of electric drives, current [...] Read more.

The electrification of automotive powertrains has accelerated research efforts in the modeling, control, and monitoring of electric drive systems, where reliability, safety, and efficiency are key enablers for mass adoption. Despite a large corpus of literature addressing individual aspects of electric drives, current surveys remain fragmented, typically focusing on either multiphysics modeling of machines and converters, or advanced control algorithms, or diagnostic and prognostic frameworks. This review provides a comprehensive perspective that systematically integrates these domains, establishing direct connections between high-fidelity models, control design, and monitoring architectures. Starting from the fundamental components of the automotive power drive system, the paper reviews state-of-the-art strategies for synchronous motor modeling, inverter and DC/DC converter design, and advanced control schemes, before presenting monitoring techniques that span model-based residual generation, AI-driven fault classification, and hybrid approaches. Particular emphasis is given to the interplay between functional safety (ISO 26262), computational feasibility on embedded platforms, and the need for explainable and certifiable monitoring frameworks. By aligning modeling, control, and monitoring perspectives within a unified narrative, this review identifies the methodological gaps that hinder cross-domain integration and outlines pathways toward digital-twin-enabled prognostics and health management of automotive electric drives. Full article

(This article belongs to the Special Issue Control and Optimization of Power Converters and Drives, 2nd Edition)

► Show Figures

Figure 1

23 pages, 6928 KB

Open AccessArticle

Sustainable Floating PV–Storage Hybrid System for Coastal Energy Resilience

by Yong-Dong Chang, Gwo-Ruey Yu, Ching-Chih Chang and Jun-Hao Chen

Electronics 2025, 14(19), 3949; https://doi.org/10.3390/electronics14193949 - 7 Oct 2025

Viewed by 513

Abstract

Floating photovoltaic (FPV) systems are promising for coastal aquaculture where reliable electricity is essential for pumping, oxygenation, sensing, and control. A sustainable FPV–storage hybrid tailored to monsoon-prone sites is developed, with emphasis on energy efficiency and structural resilience. The prototype combines dual-axis solar [...] Read more.

Floating photovoltaic (FPV) systems are promising for coastal aquaculture where reliable electricity is essential for pumping, oxygenation, sensing, and control. A sustainable FPV–storage hybrid tailored to monsoon-prone sites is developed, with emphasis on energy efficiency and structural resilience. The prototype combines dual-axis solar tracking with a spray-cooling and cleaning subsystem and an active wind-protection strategy that automatically flattens the array when wind speed exceeds 8.0 m/s. Temperature, wind speed, and irradiance sensors are coordinated by an Arduino-based supervisor to optimize tracking, thermal management, and tilt control. A 10 W floating module and a fixed-tilt reference were fabricated and tested outdoors in Penghu, Taiwan. The FPV achieved a 25.17% energy gain on a sunny day and a 40.29% gain under overcast and windy conditions, while module temperature remained below 45 °C through on-demand spraying, reducing thermal losses. In addition, a hybrid energy storage system (HESS), integrating a 12 V/10 Ah lithium-ion battery and a 12 V/24 Ah lead-acid battery, was validated using a priority charging strategy. During testing, the lithium-ion unit was first charged to stabilize the control circuits, after which excess solar energy was redirected to the lead-acid battery for long-term storage. This hierarchical design ensured both immediate power stability and extended endurance under cloudy or low-irradiance conditions. The results demonstrate a practical, low-cost, and modular pathway to couple FPV with hybrid storage for coastal energy resilience, improving yield and maintaining safe operation during adverse weather, and enabling scalable deployment across cage-aquaculture facilities. Full article

(This article belongs to the Special Issue Advanced Energy Storage Technologies for Integration in Sustainable Energy Systems)

► Show Figures

Figure 1

27 pages, 1513 KB

Open AccessArticle

Accurate Fault Classification in Wind Turbines Based on Reduced Feature Learning and RVFLN

by Mehmet Yıldırım and Bilal Gümüş

Electronics 2025, 14(19), 3948; https://doi.org/10.3390/electronics14193948 - 7 Oct 2025

Viewed by 445

Abstract

This paper presents a robust and computationally efficient fault classification framework for wind energy conversion systems (WECS), built upon a Robust Random Vector Functional Link Network (Robust-RVFLN) and validated through real-time simulations on a Real-Time Digital Simulator (RTDS). Unlike existing studies that depend [...] Read more.

This paper presents a robust and computationally efficient fault classification framework for wind energy conversion systems (WECS), built upon a Robust Random Vector Functional Link Network (Robust-RVFLN) and validated through real-time simulations on a Real-Time Digital Simulator (RTDS). Unlike existing studies that depend on high-dimensional feature extraction or purely data-driven deep learning models, our approach leverages a compact set of five statistically significant and physically interpretable features derived from rotor torque, phase current, DC-link voltage, and dq-axis current components. This reduced feature set ensures both high discriminative power and low computational overhead, enabling effective deployment in resource-constrained edge devices and large-scale wind farms. A synthesized dataset representing seven representative fault scenarios—including converter, generator, gearbox, and grid faults—was employed to evaluate the model. Comparative analysis shows that the Robust-RVFLN consistently outperforms conventional classifiers (SVM, ELM) and deep models (CNN, LSTM), delivering accuracy rates of up to 99.85% for grid-side line-to-ground faults and 99.81% for generator faults. Beyond accuracy, evaluation metrics such as precision, recall, and F1-score further validate its robustness under transient operating conditions. By uniting interpretability, scalability, and real-time performance, the proposed framework addresses critical challenges in condition monitoring and predictive maintenance, offering a practical and transferable solution for next-generation renewable energy infrastructures. Full article

(This article belongs to the Special Issue Microgrids: Innovations in Fault Detection, Security, and Addressing Challenges)

► Show Figures

Figure 1

29 pages, 632 KB

Open AccessArticle

ML-PSDFA: A Machine Learning Framework for Synthetic Log Pattern Synthesis in Digital Forensics

by Wafa Alorainy

Electronics 2025, 14(19), 3947; https://doi.org/10.3390/electronics14193947 - 6 Oct 2025

Viewed by 532

Abstract

This study introduces the Machine Learning (ML)-Driven Pattern Synthesis for Digital Forensics in Synthetic Log Analysis (ML-PSDFA) framework to address critical gaps in digital forensics, including the reliance on real-world data, limited pattern diversity, and forensic integration challenges. A key innovation is the [...] Read more.

This study introduces the Machine Learning (ML)-Driven Pattern Synthesis for Digital Forensics in Synthetic Log Analysis (ML-PSDFA) framework to address critical gaps in digital forensics, including the reliance on real-world data, limited pattern diversity, and forensic integration challenges. A key innovation is the introduction of a novel temporal forensics loss

L_{TFL}

in the Synthetic Attack Pattern Generator (SAPG), which enhances the preservation of temporal sequences in synthetic logs that are crucial for forensic analysis. The framework employs the SAPG with hybrid seed data (UNSW-NB15 and CICIDS2017) to create 500,000 synthetic log entries using Google Colab, achieving a realism score of 0.96, a temporal consistency score of 0.90, and an entropy of 4.0. The methodology employs a three-layer architecture that integrates data generation, pattern analysis, and forensic training, utilizing TimeGAN, XGBoost classification with hyperparameter tuning via Optuna, and reinforcement learning (RL) to optimize the extraction of evidence. Due to enhanced synthetic data quality and advanced modeling, the results exhibit an average classification precision of 98.5% (best fold 98.7%) 98.5% (best fold 98.7%), outperforming previously reported approaches. Feature importance analysis highlights timestamps (0.40) and event types (0.30), while the RL workflow reduces false positives by 17% over 1000 episodes, aligning with RL benchmarks. The temporal forensics loss improves the realism score from 0.92 to 0.96 and introduces a temporal consistency score of 0.90, demonstrating enhanced forensic relevance. This work presents a scalable and accessible training platform for legally constrained environments, as well as a novel RL-based evidence extraction method. Limitations include a lack of real-system validation and resource constraints. Future work will explore dynamic reward tuning and simulated benchmarks to enhance precision and generalizability. Full article

(This article belongs to the Special Issue AI and Cybersecurity: Emerging Trends and Key Challenges)

► Show Figures

Figure 1

20 pages, 794 KB

Open AccessEditor’s ChoiceArticle

Replay-Based Domain Incremental Learning for Cross-User Gesture Recognition in Robot Task Allocation

by Kanchon Kanti Podder, Pritom Dutta and Jian Zhang

Electronics 2025, 14(19), 3946; https://doi.org/10.3390/electronics14193946 - 6 Oct 2025

Viewed by 414

Abstract

Reliable gesture interfaces are essential for coordinating distributed robot teams in the field. However, models trained in a single domain often perform poorly when confronted with new users, different sensors, or unfamiliar environments. To address this challenge, we propose a memory-efficient replay-based domain [...] Read more.

Reliable gesture interfaces are essential for coordinating distributed robot teams in the field. However, models trained in a single domain often perform poorly when confronted with new users, different sensors, or unfamiliar environments. To address this challenge, we propose a memory-efficient replay-based domain incremental learning (DIL) framework, ReDIaL, that adapts to sequential domain shifts while minimizing catastrophic forgetting. Our approach employs a frozen encoder to create a stable latent space and a clustering-based exemplar replay strategy to retain compact, representative samples from prior domains under strict memory constraints. We evaluate the framework on a multi-domain air-marshalling gesture recognition task, where an in-house dataset serves as the initial training domain and the NATOPS dataset provides 20 cross-user domains for sequential adaptation. During each adaptation step, training data from the current NATOPS subject is interleaved with stored exemplars to retain prior knowledge while accommodating new knowledge variability. Across 21 sequential domains, our approach attains

97.34 %

accuracy on the domain incremental setting, exceeding pooled fine-tuning (

91.87 %

), incremental fine-tuning (

80.92 %

), and Experience Replay (

94.20 %

) by

+ 5.47

,

+ 16.42

, and

+ 3.14

percentage points, respectively. Performance also approaches the joint-training upper bound (

98.18 %

), which represents the ideal case where data from all domains are available simultaneously. These results demonstrate that memory-efficient latent exemplar replay provides both strong adaptation and robust retention, enabling practical and trustworthy gesture-based human–robot interaction in dynamic real-world deployments. Full article

(This article belongs to the Special Issue Coordination and Communication of Multi-Robot Systems)

► Show Figures

Figure 1

18 pages, 2116 KB

Open AccessArticle

A Markov Chain Replacement Strategy for Surrogate Identifiers: Minimizing Re-Identification Risk While Preserving Text Reuse

by John D. Osborne, Andrew Trotter, Tobias O’Leary, Chris Coffee, Micah D. Cochran, Luis Mansilla-Gonzalez, Akhil Nadimpalli, Alex McAnnally, Abdulateef I. Almudaifer, Jeffrey R. Curtis, Salma M. Aly and Richard E. Kennedy

Electronics 2025, 14(19), 3945; https://doi.org/10.3390/electronics14193945 - 6 Oct 2025

Viewed by 860

Abstract

“Hiding in Plain Sight” (HIPS) strategies for Personal Health Information (PHI) replace PHI with surrogate values to hinder re-identification attempts. We evaluate three different HIPS strategies for PHI replacement, a standard Consistent replacement strategy, a Random replacement strategy, and a novel Markov model [...] Read more.

“Hiding in Plain Sight” (HIPS) strategies for Personal Health Information (PHI) replace PHI with surrogate values to hinder re-identification attempts. We evaluate three different HIPS strategies for PHI replacement, a standard Consistent replacement strategy, a Random replacement strategy, and a novel Markov model strategy. We evaluate the privacy-preserving benefits and relative utility for information extraction of these strategies on both a simulated PHI distribution and real clinical corpora from two different institutions using a range of false negative error rates (FNER). The Markov strategy consistently outperformed the Consistent and Random substitution strategies on both real data and in statistical simulations. Using FNER ranging from 0.1% to 5%, PHI leakage at the document level could be reduced from 27.1% to 0.1% and from 94.2% to 57.7% with the Markov strategy versus the standard Consistent substitution strategy, at 0.1% and 0.5% FNER, respectively. Additionally, we assessed the generated corpora containing synthetic PHI for reuse using a variety of information extraction methods. Results indicate that modern deep learning methods have similar performance on all strategies, but older machine learning techniques can suffer from the change in context. Overall, a Markov surrogate generation strategy substantially reduces the chance of inadvertent PHI release. Full article

(This article belongs to the Special Issue Natural Language Processing Based on Neural Networks and Large Language Models)

► Show Figures

Figure 1

16 pages, 379 KB

Open AccessArticle

Prot-GO: A Parallel Transformer Encoder-Based Fusion Model for Accurately Predicting Gene Ontology (GO) Terms from Full-Scale Protein Sequences

by Azwad Tamir and Jiann-Shiun Yuan

Electronics 2025, 14(19), 3944; https://doi.org/10.3390/electronics14193944 - 6 Oct 2025

Viewed by 448

Abstract

Recent developments in next-generation sequencing technology have led to the creation of extensive, open-source protein databases consisting of hundreds of millions of sequences. To render these sequences applicable in biomedical applications, they must be meticulously annotated by wet lab testing or extracting them [...] Read more.

Recent developments in next-generation sequencing technology have led to the creation of extensive, open-source protein databases consisting of hundreds of millions of sequences. To render these sequences applicable in biomedical applications, they must be meticulously annotated by wet lab testing or extracting them from existing literature. Over the last few years, researchers have developed numerous automatic annotation systems, particularly deep learning models based on machine learning and artificial intelligence, to address this issue. In this work, we propose a transformer-based fusion model capable of predicting Gene Ontology (GO) terms from full-scale protein sequences, achieving state-of-the-art accuracy compared to other contemporary machine learning annotation systems. The approach performs particularly well on clustered split datasets, which comprise training and testing samples originating from distinct distributions that are structurally diverse. This demonstrates that the model is able to understand both short and long term dependencies within the protein’s structure and can capture sequence features that are predictive of the various GO terms. Furthermore, the technique is lightweight and less computationally expensive compared to the benchmark methods, while at the same time unaffected by sequence length, rendering it appropriate for diverse applications with varying sequence lengths. Full article

(This article belongs to the Special Issue Artificial Intelligence and Deep Learning Techniques for Healthcare)

► Show Figures

Figure 1

15 pages, 1557 KB

Open AccessArticle

A Dual-Structured Convolutional Neural Network with an Attention Mechanism for Image Classification

by Yongzhuo Liu, Jiangmei Zhang, Haolin Liu and Yangxin Zhang

Electronics 2025, 14(19), 3943; https://doi.org/10.3390/electronics14193943 - 5 Oct 2025

Viewed by 531

Abstract

This paper presents a dual-structured convolutional neural network (CNN) for image classification, which integrates two parallel branches: CNN-A with spatial attention and CNN-B with channel attention. The spatial attention module in CNN-A dynamically emphasizes discriminative regions by aggregating channel-wise information, while the channel [...] Read more.

This paper presents a dual-structured convolutional neural network (CNN) for image classification, which integrates two parallel branches: CNN-A with spatial attention and CNN-B with channel attention. The spatial attention module in CNN-A dynamically emphasizes discriminative regions by aggregating channel-wise information, while the channel attention mechanism in CNN-B adaptively recalibrates feature channel importance. The extracted features from both branches are fused through concatenation, enhancing the model’s representational capacity by capturing complementary spatial and channel-wise dependencies. Extensive experiments on a 12-class image dataset demonstrate the superiority of the proposed model over state-of-the-art methods, achieving 98.06% accuracy, 96.00% precision, and 98.01% F1-score. Despite a marginally longer training time, the model exhibits robust convergence and generalization, as evidenced by stable loss curves and high per-class recognition rates (>90%). The results validate the efficacy of dual attention mechanisms in improving feature discrimination for complex image classification tasks. Full article

(This article belongs to the Special Issue Advances in Object Tracking and Computer Vision)

► Show Figures

Figure 1

18 pages, 552 KB

Open AccessArticle

A Novel Convolutional Vision Transformer Network for Effective Level-of-Detail Awareness in Digital Twins

by Min-Seo Yang, Ji-Wan Kim and Hyun-Suk Lee

Electronics 2025, 14(19), 3942; https://doi.org/10.3390/electronics14193942 - 4 Oct 2025

Viewed by 622

Abstract

In this paper, we propose a novel integrated model architecture, called a level-of-detail (LoD)-aware convolutional vision transformer network (LCvT). It is designed to enhance digital twin (DT) synchronization by effectively integrating LoD awareness in DTs through hierarchical image classification. LCvT employs a vision [...] Read more.

In this paper, we propose a novel integrated model architecture, called a level-of-detail (LoD)-aware convolutional vision transformer network (LCvT). It is designed to enhance digital twin (DT) synchronization by effectively integrating LoD awareness in DTs through hierarchical image classification. LCvT employs a vision transformer (ViT)-based backbone coupled with dedicated branch networks for each LoD. This integration of ViT and branch networks ensures that key features are accurately detected and tailored to the specific objectives of each detail level while also efficiently extracting common features across all levels. Furthermore, LCvT leverages a coarse-to-fine inference strategy and incorporates an early exit mechanism for each LoD, which significantly reduces computational overhead without compromising accuracy. This design enables a single model to dynamically adapt to varying LoD requirements in real-time, offering substantial improvements in inference time and resource efficiency compared to deploying separate models for each level. Through extensive experiments on benchmark datasets, we demonstrate that LCvT outperforms existing methods in accuracy and efficiency across all LoDs, especially in DT synchronization scenarios where LoD requirements fluctuate dynamically. Full article

(This article belongs to the Special Issue Data-Centric Artificial Intelligence: New Methods for Data Processing)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Electronics, Volume 14, Issue 19 (October-1 2025) – 222 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI