Electronics

12 pages, 2708 KiB

Open AccessArticle

An Envelope-to-Cycle Difference Compensation Method for eLoran Signals in Seawater Based on a Variable Step Size Least Mean Square Algorithm

by Miao Wu, Liang Liu, Fangneng Li, Bing Zhu, Wenkui Li and Xianzhou Jin

Electronics 2025, 14(3), 597; https://doi.org/10.3390/electronics14030597 (registering DOI) - 3 Feb 2025

Abstract

The dispersion effect of seawater can cause the envelop distortion of underwater eLoran signals, which causes the envelope-to-cycle difference (ECD) to exceed the standard range. Furthermore, it results in incorrect cycle identification and significant positioning errors. However, few studies have focused on the [...] Read more.

The dispersion effect of seawater can cause the envelop distortion of underwater eLoran signals, which causes the envelope-to-cycle difference (ECD) to exceed the standard range. Furthermore, it results in incorrect cycle identification and significant positioning errors. However, few studies have focused on the distortion caused by the dispersion effect. In this study, we propose an accurate underwater eLoran ECD compensation method based on a variable step size least mean square (VSS-LMS) algorithm. First, a systematic modeling approach was employed to investigate the impact of dispersion effects on Loran signals. Second, the VSS-LMS algorithm was introduced to update the filter weight vector in response to discrepancies in the input signal. Finally, the input signal was subjected to an adaptive transversal filtering process, resulting in an output signal that adhered to the specifications of the ECD standard. The efficacy and superiority of the proposed algorithm were demonstrated by experimentation and simulation. When the depth of seawater exceeds 2 m, the ECD value of the original eLoran signal exceeds the standard range, precluding the possibility of cycle identification. However, when the depth of seawater reaches 4 m, the ECD of the signal compensated by the proposed algorithm adaptively compensates for the normal range, thereby enabling the accurate recognition of cycles. Full article

► Show Figures

Figure 1

22 pages, 15628 KiB

Open AccessArticle

A Novel Non-Resonant Full-Bridge Multi-Output Topology for Domestic Induction Heating Applications

by Sezer Aslan, Ulas Oktay and Nihan Altintas

Electronics 2025, 14(3), 596; https://doi.org/10.3390/electronics14030596 (registering DOI) - 3 Feb 2025

Abstract

Induction heating technology plays a significant role in heating applications with its high efficiency, fast response, and precise control ability. Traditional resonant inverter-based systems face problems such as complexity, lack of flexibility, and low efficiency in multi-load situations. To overcome these issues, a [...] Read more.

Induction heating technology plays a significant role in heating applications with its high efficiency, fast response, and precise control ability. Traditional resonant inverter-based systems face problems such as complexity, lack of flexibility, and low efficiency in multi-load situations. To overcome these issues, a new non-resonant full-bridge multiple-output inverter topology using silicon carbide (SiC) semiconductor devices is presented. While the system is simplified by eliminating resonant components, efficiency is increased thanks to SiC devices. In the study, a coil design methodology focusing on coil resistance and inductance is presented to optimize energy transfer and maximize system performance. Load-sensing and advanced frequency-modulation techniques are integrated to provide precise and independent power regulation in multi-loads. Thus, the efficiency of energy distribution and system robustness are increased. The proposed topology offers heating performance that provides homogeneous heat distribution. The developed prototype was proven to operate reliably with high efficiency under different load conditions and was suitably applied for domestic induction heating applications. An efficiency of 96.78% was achieved at a 50 kHz operating frequency and 2000 W power level. Full article

(This article belongs to the Section Power Electronics)

► Show Figures

Figure 1

21 pages, 2637 KiB

Open AccessArticle

Molecular Layer Doping ZnO Films as a Novel Approach to Resistive Oxygen Sensors

by Wojciech Bulowski, Robert P. Socha, Anna Drabczyk, Patryk Kasza, Piotr Panek and Marek Wojnicki

Electronics 2025, 14(3), 595; https://doi.org/10.3390/electronics14030595 (registering DOI) - 2 Feb 2025

Abstract

In the modern world, gas sensors play a crucial role in sectors such as high-tech industries, medicine, and environmental monitoring. Among these fields, oxygen sensors are the most important. There are several types of oxygen sensors, including optical, magnetic, Schottky diode, and resistive [...] Read more.

In the modern world, gas sensors play a crucial role in sectors such as high-tech industries, medicine, and environmental monitoring. Among these fields, oxygen sensors are the most important. There are several types of oxygen sensors, including optical, magnetic, Schottky diode, and resistive (or chemoresistive) ones. Currently, most oxygen-resistive sensors (ORSs) described in the literature are fabricated as thick layers, typically deposited via screen printing, and they operate at high temperatures, often exceeding 700 °C. This work presents a novel approach utilizing atomic layer deposition (ALD) to create very thin layers. Combined with appropriate doping, this method aims to reduce the energy consumption of the sensors by lowering both the mass requiring heating and the operating temperature. The device fabricated using the proposed process demonstrates a response of 88.21 at a relatively low temperature of 450 °C, highlighting its potential in ORS applications based on doped ALD thin films. Full article

(This article belongs to the Special Issue High-Performance Gas Sensors: Materials, Technologies and Applications)

► Show Figures

Figure 1

22 pages, 7052 KiB

Open AccessArticle

DualAD: Exploring Coupled Dual-Branch Networks for Multi-Class Unsupervised Anomaly Detection

by Shiwen He, Yuehan Chen, Liangpeng Wang, Wei Huang, Rong Xu and Yurong Qian

Electronics 2025, 14(3), 594; https://doi.org/10.3390/electronics14030594 (registering DOI) - 2 Feb 2025

Abstract

Anomaly detection (AD) is crucial in various domains such as industrial inspection, medical diagnosis, and video surveillance. Previous advancements in unsupervised AD often necessitated training separate models for different objects, which can be inefficient when dealing with diverse categories in real-world scenarios. This [...] Read more.

Anomaly detection (AD) is crucial in various domains such as industrial inspection, medical diagnosis, and video surveillance. Previous advancements in unsupervised AD often necessitated training separate models for different objects, which can be inefficient when dealing with diverse categories in real-world scenarios. This paper addresses the recently proposed task of multi-class unsupervised anomaly detection (MUAD), which is more practical and challenging. We begin by reviewing the first MUAD framework, UniAD, and analyzing the characteristics of end-to-end feature reconstruction networks that can adapt to various backbone architectures. Building on these insights, we introduce a novel MUAD framework called DualAD. Our approach is based on the innovative design of a Coupled Dual-Branch Network (CDBN), which integrates a Wide–Shallow Network (WSN) with a Narrow–Deep Network (NDN), leveraging the strengths of both to achieve superior performance. We explore a fully transformer-based homogeneous design for the CDBN and introduce a more lightweight heterogeneous CDBN design that integrates a transformer with a Memory-Augmented Multi-Layer Perceptron (MMLP). Experimental results on the MVTec AD and VisA datasets demonstrate that DualAD outperforms the recent state-of-the-art methods and exhibits robust performance across various pre-trained backbone architectures. Full article

(This article belongs to the Section Artificial Intelligence)

16 pages, 467 KiB

Open AccessArticle

A Zero-Shot Framework for Low-Resource Relation Extraction via Distant Supervision and Large Language Models

by Peisheng Han, Geng Liang and Yongfei Wang

Electronics 2025, 14(3), 593; https://doi.org/10.3390/electronics14030593 (registering DOI) - 2 Feb 2025

Abstract

While Large Language Models (LLMs) have significantly advanced various benchmarks in Natural Language Processing (NLP), the challenge of low-resource tasks persists, primarily due to the scarcity of data and difficulties in annotation. This study introduces LoRE, a framework designed for zero-shot relation extraction [...] Read more.

While Large Language Models (LLMs) have significantly advanced various benchmarks in Natural Language Processing (NLP), the challenge of low-resource tasks persists, primarily due to the scarcity of data and difficulties in annotation. This study introduces LoRE, a framework designed for zero-shot relation extraction in low-resource settings, which blends distant supervision with the powerful capabilities of LLMs. LoRE addresses the challenges of data sparsity and noise inherent in traditional distant supervision methods, enabling high-quality relation extraction without requiring extensive labeled data. By leveraging LLMs for zero-shot open information extraction and incorporating heuristic entity and relation alignment with semantic disambiguation, LoRE enhances the accuracy and relevance of the extracted data. Low-resource tasks refer to scenarios where labeled data are extremely limited, making traditional supervised learning approaches impractical. This study aims to develop a robust framework that not only tackles these challenges but also demonstrates the theoretical and practical implications of zero-shot relation extraction. The Chinese Person Relationship Extraction (CPRE) dataset, developed under this framework, demonstrates LoRE’s proficiency in extracting person-related triples. The CPRE dataset consists of 1000 word pairs, capturing diverse semantic relationships. Extensive experiments on the CPRE, IPRE, and DuIE datasets show significant improvements in dataset quality and a reduction in manual annotation efforts. These findings highlight the potential of LoRE to advance both the theoretical understanding and practical applications of relation extraction in low-resource settings. Notably, the performance of LoRE on the manually annotated DuIE dataset attests to the quality of the CPRE dataset, rivaling that of manually curated datasets, and highlights LoRE’s potential for reducing the complexities and costs associated with dataset construction for zero-shot and low-resource tasks. Full article

(This article belongs to the Special Issue Zero-Shot Learning in Natural Language Processing and Its Applications)

27 pages, 18019 KiB

Open AccessArticle

Generalized Multivariate Symplectic Sparsest United Decomposition for Rolling Bearing Fault Diagnosis

by Weikang Sun, Yanfei Liu and Yanfeng Peng

Electronics 2025, 14(3), 592; https://doi.org/10.3390/electronics14030592 (registering DOI) - 2 Feb 2025

Abstract

The non-stationary characteristics of the vibration signals of rolling bearings will be aggravated under variable speed conditions. Meanwhile, multichannel signals can provide a more comprehensive characterization of state information, providing multiple sources of information that facilitate information fusion and enhancement. However, traditional adaptive [...] Read more.

The non-stationary characteristics of the vibration signals of rolling bearings will be aggravated under variable speed conditions. Meanwhile, multichannel signals can provide a more comprehensive characterization of state information, providing multiple sources of information that facilitate information fusion and enhancement. However, traditional adaptive signal decomposition methods generally assume that the frequency information is constant and stationary, and it is difficult to achieve a unified decomposition when dealing with multichannel time-varying signals. Therefore, the intention of this paper is to propose a multichannel signal adaptive decomposition method applicable to variable speed conditions. Specifically, this paper takes advantage of the strong adaptability and robustness of symplectic geometric mode decomposition (SGMD). To improve its applicability to multichannel time-varying signals at variable rotational speeds, a generalized multivariate symplectic sparsest united decomposition (GMSSUD) method is proposed. In GMSSUD, firstly, the completely adaptive projection (CAP) method is employed to achieve a unified representation of the multichannel signals. Then, the generalized demodulation method is introduced to stabilize the signal and subsequently reduce the noise through component screening and reconstruction. Finally, with the new proposed operator as the optimization objective, the constructed sparse filter parameters are optimized to achieve the frequency band segmentation. The analysis results demonstrate that the GMSSUD method possesses higher decomposition precision for multichannel signals with variable speeds and also has a stronger diagnosis ability for variable-speed bearing faults. Full article

► Show Figures

Figure 1

19 pages, 2144 KiB

Open AccessArticle

PDNet by Partial Deep Convolution: A Better Lightweight Detector

by Wei Wang, Yuanze Meng, Han Li, Shun Li, Chenghong Zhang, Guanghui Zhang and Weimin Lei

Electronics 2025, 14(3), 591; https://doi.org/10.3390/electronics14030591 (registering DOI) - 2 Feb 2025

Abstract

Model lightweighting is significant in edge computing and mobile devices. Current studies on fast network design mainly focuses on model computation compression and speedup. Many models aim to compress models by dealing with redundant feature maps. However, most of these methods choose to [...] Read more.

Model lightweighting is significant in edge computing and mobile devices. Current studies on fast network design mainly focuses on model computation compression and speedup. Many models aim to compress models by dealing with redundant feature maps. However, most of these methods choose to preserve the feature maps with simple manipulations and do not effectively reduce redundant feature maps. This paper proposes a new convolution module, PDConv, which compresses redundant feature maps to reduce network complexity and increase network width to maintain accuracy. PDConv (Partial Deep Convolution) outperforms traditional methods in handling redundant feature maps, particularly in deep networks. Its FLOPs are comparable to deep separable convolution but with higher accuracy. This paper proposes PDBottleNeck and PDC2f (Partial Deep CSPDarknet53 to 2-Stage FPN) and build the lightweight network PDNet for experimental validation using the PASCAL VOC dataset. Compared to the popular HorNet, our method achieves an improvement of more than 25% in FLOPs and 1.8% in mAP50:95 accuracy. On the CoCo2017 dataset, our large PDNet achieves a 0.5% improvement in mAP75 and lower FLOPs than the latest RepVit. Full article

(This article belongs to the Section Artificial Intelligence)

► Show Figures

Figure 1

16 pages, 11266 KiB

Open AccessArticle

A Method for Sensing the Internal Current Distribution State of Press-Pack IGBT Devices Based on Hall Sensors

by Xi Yuan, Yirun Ji, Wenqian Yuan, Qing Huai, Zhen Hao and Zhonghao Dongye

Electronics 2025, 14(3), 590; https://doi.org/10.3390/electronics14030590 (registering DOI) - 2 Feb 2025

Abstract

Flexible DC transmission technology is an important support for constructing new power systems. The press-pack IGBT device has the advantages of double-sided heat dissipation and high power density in series and is the core device in the flexible HVDC project. However, the compact [...] Read more.

Flexible DC transmission technology is an important support for constructing new power systems. The press-pack IGBT device has the advantages of double-sided heat dissipation and high power density in series and is the core device in the flexible HVDC project. However, the compact parallel arrangement of multiple chips in the press-pack IGBT (PPI) device easily causes uneven current distribution in the device, which affects its reliability. In this paper, a noninvasive sensing method for the internal current distribution of the PPI device is proposed based on the measurement of the external magnetic field and the numerical analysis of the magnetic field. According to the correlation between the device current distribution and the external magnetic field, a noninvasive current distribution sensing method based on the measurement and numerical analysis of the external magnetic field is proposed, and the placement strategy of the magnetic field measurement points is given. The error of the perception method under ideal conditions is analyzed, the error size and rule are summarized, and the correction methods are given, respectively. Finally, a magnetic field measuring device is designed and verified by experiments. The results can accurately indicate the direction of pressure deviation. Full article

(This article belongs to the Section Industrial Electronics)

► Show Figures

Figure 1

21 pages, 4068 KiB

Open AccessArticle

Three-Dimensional Mesh Character Pose Transfer with Neural Sparse-Softmax Skinning Blending

by Siqi Liu, Mengxiao Yin, Ming Li, Feng Zhan and Bei Hua

Electronics 2025, 14(3), 589; https://doi.org/10.3390/electronics14030589 (registering DOI) - 1 Feb 2025

Abstract

Three-dimensional mesh pose transfer transforms the pose of a source model into the pose of a reference model while preserving the source model’s identity (body detail). It has tremendous potential in computer graphics tasks. Current neural network-based methods primarily focus on extracting pose [...] Read more.

Three-dimensional mesh pose transfer transforms the pose of a source model into the pose of a reference model while preserving the source model’s identity (body detail). It has tremendous potential in computer graphics tasks. Current neural network-based methods primarily focus on extracting pose and body features, not entirely using the articulated body structure of humans and animals. We propose an end-to-end pose transfer network based on skinning deformation to address these issues. This network first extracts skinning weights and model joint features. Then, they are decoded to transfer the source model to a pose similar to the reference model while preserving the features of the source model. During feature extraction, we utilize the features of the k-nearest neighborhoods and one-ring neighborhoods to enable the network to learn the body details of the model better. Additionally, we apply skinning weights and joint features to capture the variation in the source model pose compared to the reference model pose and then use a decoding network to obtain the target model, replacing linear blend skinning. We conducted experiments on datasets such as SMPL, SMAL, FAUST, DYNA, and the MG dataset to provide empirical evidence and demonstrate that our method is the best in quantitative experiments. Our method efficiently transfers poses while better preserving the identity of the source model. Full article

(This article belongs to the Special Issue AI/Machine Learning in Computer Vision/Image Processing and Natural Language Processing)

► Show Figures

Figure 1

11 pages, 4152 KiB

Open AccessArticle

A 5-Transistor CMOS Voltage Reference with Double Supply-Regulation

by Minji Jung, Kyeongmin Min, Hyunwoo Son and Youngwoo Ji

Electronics 2025, 14(3), 588; https://doi.org/10.3390/electronics14030588 (registering DOI) - 1 Feb 2025

Abstract

This paper presents an ultra-low-power CMOS voltage reference designed and verified in an 180 nm standard CMOS technology. To achieve DC and AC supply sensitivity under 0.01%/V and −100 dB, it employs a single transistor and two 2-T cores to improve supply immunity [...] Read more.

This paper presents an ultra-low-power CMOS voltage reference designed and verified in an 180 nm standard CMOS technology. To achieve DC and AC supply sensitivity under 0.01%/V and −100 dB, it employs a single transistor and two 2-T cores to improve supply immunity with minimal overhead, adding only one drain-to-source voltage for the total supply voltage. The proposed design achieves a line sensitivity of 0.0027%/V in a supply voltage range of 0.5 V to 2 V and consumes 630 pW with a supply voltage of 0.5 V. The simulated temperature coefficient is 12 ppm/°C in a temperature range of −40 °C to 150 °C, and the simulated power supply rejection ratio is −100.5 dB at 100 Hz without requiring any output decoupling capacitor. Full article

(This article belongs to the Special Issue New Horizons and Recent Advances of Power Electronics)

► Show Figures

Figure 1

18 pages, 19242 KiB

Open AccessArticle

Feature Fusion Network with Local Information Exchange for Underwater Object Detection

by Xiaopeng Liu, Pengwei Ma and Long Chen

Electronics 2025, 14(3), 587; https://doi.org/10.3390/electronics14030587 (registering DOI) - 1 Feb 2025

Abstract

When using enhanced images for underwater object detection, issues such as detail loss and increased noise often arise, leading to decreased detection efficiency. To address this issue, we propose the Feature Fusion Network with Local Information Exchange (FFNLIE) for underwater object detection. We [...] Read more.

When using enhanced images for underwater object detection, issues such as detail loss and increased noise often arise, leading to decreased detection efficiency. To address this issue, we propose the Feature Fusion Network with Local Information Exchange (FFNLIE) for underwater object detection. We input raw and enhanced images into the Swin Transformer in parallel for feature extraction. Then, we propose a local information exchange module to enhance the feature extraction capability of the Swin Transformer. In order to fully utilize the complementary information of the two images, our feature fusion module consists of two core components: the Discrepancy Information Addition Block (DIAB) and the Common Information Addition Block (CIAB). The DIAB and CIAB are designed by utilizing and modifying cross-attention mechanisms, which can easily extract image discrepancy information and common information. Finally, the fused features are fed into the object detector to perform object detection tasks. The experimental findings demonstrate that the FFNLIE exhibited exceptional performance across four underwater datasets. Full article

17 pages, 3245 KiB

Open AccessArticle

Enhancing Security in Software Design Patterns and Antipatterns: A Framework for LLM-Based Detection

by Roberto Andrade, Jenny Torres and Iván Ortiz-Garcés

Electronics 2025, 14(3), 586; https://doi.org/10.3390/electronics14030586 (registering DOI) - 1 Feb 2025

Abstract

The detection of security vulnerabilities in software design patterns and antipatterns is crucial for maintaining robust and maintainable systems, particularly in dynamic Continuous Integration/Continuous Deployment (CI/CD) environments. Traditional static analysis tools, while effective for identifying isolated issues, often lack contextual awareness, leading to [...] Read more.

The detection of security vulnerabilities in software design patterns and antipatterns is crucial for maintaining robust and maintainable systems, particularly in dynamic Continuous Integration/Continuous Deployment (CI/CD) environments. Traditional static analysis tools, while effective for identifying isolated issues, often lack contextual awareness, leading to missed vulnerabilities and high rates of false positives. This paper introduces a novel framework leveraging Large Language Models (LLMs) to detect and mitigate security risks in design patterns and antipatterns. By analyzing relationships and behavioral dynamics in code, LLMs provide a nuanced, context-aware approach to identifying issues such as unauthorized state changes, insecure communication, and improper data handling. The proposed framework integrates key security heuristics—such as the principles of least privilege and input validation—to enhance LLM performance. An evaluation of the framework demonstrates its potential to outperform traditional tools in terms of accuracy and efficiency, enabling the proactive detection and remediation of vulnerabilities in real time. This study contributes to the field of software engineering by offering an innovative methodology for securing software systems using LLMs, promoting both academic research and practical application in industry settings. Full article

(This article belongs to the Special Issue Recent Advances of Software Engineering)

► Show Figures

Figure 1

17 pages, 1873 KiB

Open AccessArticle

A Blockchain Address Generation Method Based on the Improved SM9 Algorithm

by Ruimin Wang, Haolong Fan and Kezhen Liu

Electronics 2025, 14(3), 585; https://doi.org/10.3390/electronics14030585 (registering DOI) - 1 Feb 2025

Abstract

In recent years, governments have regarded blockchain technology as a key breakthrough in independent innovation of core technologies and have attached great importance to its development. In current blockchain solutions, the elliptic curve digital signature algorithm (ECDSA) is usually combined with the Keccak [...] Read more.

In recent years, governments have regarded blockchain technology as a key breakthrough in independent innovation of core technologies and have attached great importance to its development. In current blockchain solutions, the elliptic curve digital signature algorithm (ECDSA) is usually combined with the Keccak Hash Algorithm 256-bit Variant (keccak256) to generate blockchain addresses. Although the ECDSA algorithm is based on the Public Key Infrastructure (PKI) system, which has prominent issues such as complex deployment processes, high operation and maintenance costs, low efficiency, and difficulties in adapting to a large number of users, it still holds an important position and has a broad application foundation in the field of cryptography. This paper aims to circumvent the existing flaws of the PKI system rather than discarding the strong and well-proven PKI system. Instead, it endeavors to explore the application of the Identity-Based Cryptography (IBC) system in blockchain and reduce dependence on foreign cryptographic algorithms. This study adopts the SM9 algorithm based on the IBC and combines it with the SM3 algorithm to generate blockchain addresses. Considering users’ use, this paper improves the original SM9 algorithm based on the original SM9 algorithm, enabling users to perform functions such as signing and encryption with just a single pair of public and private keys. Experimental results indicate that the time consumed by this proposed scheme in blockchain address generation is 1.29 times that of the existing schemes, and the length of the blockchain addresses generated is the same as that of the existing ones. Full article

► Show Figures

Figure 1

36 pages, 2224 KiB

Open AccessFeature PaperArticle

Multithreaded and GPU-Based Implementations of a Modified Particle Swarm Optimization Algorithm with Application to Solving Large-Scale Systems of Nonlinear Equations

by Bruno Silva, Luiz Guerreiro Lopes and Fábio Mendonça

Electronics 2025, 14(3), 584; https://doi.org/10.3390/electronics14030584 (registering DOI) - 1 Feb 2025

Abstract

This paper presents a novel Graphics Processing Unit (GPU) accelerated implementation of a modified Particle Swarm Optimization (PSO) algorithm specifically designed to solve large-scale Systems of Nonlinear Equations (SNEs). The proposed GPU-based parallel version of the PSO algorithm uses the inherent parallelism of [...] Read more.

This paper presents a novel Graphics Processing Unit (GPU) accelerated implementation of a modified Particle Swarm Optimization (PSO) algorithm specifically designed to solve large-scale Systems of Nonlinear Equations (SNEs). The proposed GPU-based parallel version of the PSO algorithm uses the inherent parallelism of modern hardware architectures. Its performance is compared against both sequential and multithreaded Central Processing Unit (CPU) implementations. The primary objective is to evaluate the efficiency and scalability of PSO across different hardware platforms with a focus on solving large-scale SNEs involving thousands of equations and variables. The GPU-parallelized and multithreaded versions of the algorithm were implemented in the Julia programming language. Performance analyses were conducted on an NVIDIA A100 GPU and an AMD EPYC 7643 CPU. The tests utilized a set of challenging, scalable SNEs with dimensions ranging from 1000 to 5000. Results demonstrate that the GPU accelerated modified PSO substantially outperforms its CPU counterparts, achieving substantial speedups and consistently surpassing the highly optimized multithreaded CPU implementation in terms of computation time and scalability as the problem size increases. Therefore, this work evaluates the trade-offs between different hardware platforms and underscores the potential of GPU-based parallelism for accelerating SNE solvers. Full article

(This article belongs to the Special Issue High-Performance Computing for AI: Architecture, Systems, and Algorithms)

17 pages, 930 KiB

Open AccessArticle

Machine Learning for Predicting Personality and Psychological Symptoms from Behavioral Dynamics

by Min Wang, Ao Xu, Chenxiao Fan and Xiao Sun

Electronics 2025, 14(3), 583; https://doi.org/10.3390/electronics14030583 (registering DOI) - 1 Feb 2025

Abstract

Recent studies have explored correlations between behavioral expressions (facial features, body activity intensities) and personality traits or psychological symptoms, yet systematic research remains limited. Addressing this gap is critical, as understanding these relationships could transform mental health assessment by enabling non-invasive, real-time psychological [...] Read more.

Recent studies have explored correlations between behavioral expressions (facial features, body activity intensities) and personality traits or psychological symptoms, yet systematic research remains limited. Addressing this gap is critical, as understanding these relationships could transform mental health assessment by enabling non-invasive, real-time psychological monitoring. We assessed 167 participants using the SCL-90 and BFI-2 scales. Facial features were extracted via OpenFace 2.0, and body activity intensities via YOLOv7-Pose. Correlation analyses revealed significant links between behavioral expressions and psychological characteristics. Machine learning models (MLP, XGB, GRU) were developed using 56 behavioral features, with performance evaluated by f1-score, precision, recall, and AUC. Key findings showed a negative correlation between phobic anxiety and AU07, AU15, and Gaze-angle-x activity (p < 0.01), and a positive correlation between conscientiousness and AU07, Pose-Rz, and neck activity (p < 0.01). The bimodal GRU model, integrating all features, achieved the highest predictive performance (AUC = 0.736, precision = 0.714, recall = 0.714, f1-score = 0.71). This study highlights the feasibility of using machine learning to predict personality traits and psychological symptoms from facial and body activity data. The results affirm specific associations between behavioral expressions and psychological traits, providing a foundation for precise, personalized mental health assessment tools. This study introduces a novel integration of facial and body data using machine learning, revealing new behavioral–psychological associations and establishing an innovative framework for automated, personalized mental health assessment. Full article

(This article belongs to the Topic Advances in Integrative AI, Machine Learning, and Big Data for Transformative Applications)

13 pages, 523 KiB

Open AccessArticle

A Sliding Window for Data Reuse in Deep Convolution Operations to Reduce Bandwidth Requirements and Resource Utilization

by Yiqi Sun, Yaoyang Ma, Zixuan Chen, Zhiyu Liu, Boxin Chen and Rui Song

Electronics 2025, 14(3), 582; https://doi.org/10.3390/electronics14030582 (registering DOI) - 1 Feb 2025

Abstract

Convolutional Neural Networks (CNNs) have demonstrated high accuracy in applications such as object detection, classification, and image processing. However, convolutional layers account for the majority of computations within CNNs. Typically, these layers are executed on GPUs, resulting in higher-power consumption and hindering lightweight [...] Read more.

Convolutional Neural Networks (CNNs) have demonstrated high accuracy in applications such as object detection, classification, and image processing. However, convolutional layers account for the majority of computations within CNNs. Typically, these layers are executed on GPUs, resulting in higher-power consumption and hindering lightweight deployment. This paper presents a design that deploys convolutional layers on FPGAs with adjustable parameters. In this FPGA deployment, a 4 × 4 3D sliding window is used to traverse the data, reducing bandwidth requirements and facilitating seamless integration with subsequent processing stages. A three-dimensional plane buffer design is proposed, which implements data reuse. Compared to directly inputting the feature map and performing the computation, it reduces the on-chip memory bandwidth requirement by 75%. Additionally, a new addressing strategy is introduced to map 3D feature maps to RAM addresses, eliminating addressing time. Due to the resource-intensive nature of high-level synthesis (HLS) technology, HDL design is used for the convolutional layers. This design achieves an inference speed of 121.36 GOPS at a 16-bit width, providing a 39.10 times increase in performance compared to CPU implementations. Full article

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

► Show Figures

Figure 1

6 pages, 142 KiB

Open AccessEditorial

Security, Privacy, Confidentiality, and Trust in the Blockchain: From Theory to Applications

by Mikolaj Karpinski, Oleksandr Kuznetsov and Roman Oliynykov

Electronics 2025, 14(3), 581; https://doi.org/10.3390/electronics14030581 (registering DOI) - 1 Feb 2025

Abstract

From the financial and medical sectors to various supply chains, most industries have seen a sea change due to the blockchain [...] Full article

(This article belongs to the Special Issue Security, Privacy, Confidentiality and Trust in Blockchain)

20 pages, 596 KiB

Open AccessArticle

Enhancing Hospital Data Security: A Blockchain-Based Protocol for Secure Information Sharing and Recovery

by Jihyeon Ryu and Taeseok Kim

Electronics 2025, 14(3), 580; https://doi.org/10.3390/electronics14030580 (registering DOI) - 1 Feb 2025

Abstract

Hospitals that store sensitive patient medical records have recently faced issues such as the inability to recover medical data and breaches of patient privacy due to hacker attacks. These attacks on medical data often involve ransomware, which obfuscates the entire hospital’s data, making [...] Read more.

Hospitals that store sensitive patient medical records have recently faced issues such as the inability to recover medical data and breaches of patient privacy due to hacker attacks. These attacks on medical data often involve ransomware, which obfuscates the entire hospital’s data, making them inaccessible, and can also occur when hospitals share patient information during transfers of care. In this study, we propose a new authentication protocol to prevent and address such issues within hospital systems. The proposed protocol encrypts medical records on a private blockchain, allowing them to be securely shared among institutions, hospitals, and insurance companies, ensuring data recovery even if a ransomware attack paralyzes the server. Additionally, the protocol facilitates the systematic sharing of patient medical records between hospitals or between hospitals and insurance companies by distributing session keys. In this study, we demonstrate that the proposed protocol provides 11 security properties, including forward and backward secrecy, user untraceability, and resistance to replay attacks. We also evaluate the communication and computational costs, proving that the protocol is feasible for practical use. Full article

(This article belongs to the Special Issue Network Security and Cryptography Applications)

► Show Figures

Figure 1

17 pages, 2627 KiB

Open AccessArticle

A Scalable Sorting Network Based on Hybrid Algorithms for Accelerating Data Sorting

by Xufeng Li, Li Zhou and Yan Zhu

Electronics 2025, 14(3), 579; https://doi.org/10.3390/electronics14030579 (registering DOI) - 1 Feb 2025

Abstract

Sorting in sequential data mining is significantly improved through hardware acceleration, which becomes essential as data volume and complexity increase. This paper presents a scalable hybrid sorting network that maintains or improves performance while reducing computational load and hardware requirements. The network is [...] Read more.

Sorting in sequential data mining is significantly improved through hardware acceleration, which becomes essential as data volume and complexity increase. This paper presents a scalable hybrid sorting network that maintains or improves performance while reducing computational load and hardware requirements. The network is composed of the pre-comparison odd–even sorting network (P-OESN) and the bidirectional insertion sorting network (BISN). A pre-comparison layer is introduced to the original OESN. This layer aims to place larger values in the first half of the input sequence and smaller values in the latter half. The number of iterations is reduced when the P-OESN transitions from fully parallel execution to iterative execution. A novel pipelined BISN architecture is proposed, which leads to enhanced operating frequency and throughput. The experimental results show that the pre-comparison layer reduces the number of iterations by 6% to 50%. Throughput is improved by more than four times, and operating frequency is increased by more than two times due to the pipelined BISN. The proposed hybrid sorting network reduces sorting time or resource usage, while enabling the sorting of large-scale data sets that other methods cannot support. Full article

(This article belongs to the Special Issue Sequence Data Representation Learning and Applications)

Journal Description

Electronics

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI