Traditional Face Research Articles

Artificial intelligence in psychiatric education: Enhancing clinical competence through simulation

ABSTRACT The integration of artificial intelligence (AI) in psychiatric education offers transformative potential to enhance clinical competence through realistic simulations. Traditional educational methods face limitations in replicating complex psychiatric cases, and AI-based tools provide a scalable solution. This narrative review examines current evidence on the efficacy of AI-powered simulations, focusing on their role in skill development, diagnostic accuracy, and safe clinical training. Through a comprehensive literature review of studies from 2010 to 2024, key themes such as AI’s ability to standardize patient encounters, provide instant feedback, and improve student confidence are explored. Findings suggest that AI can enhance psychiatric education by offering consistent, adaptable learning experiences that prepare trainees for real-world complexities. However, challenges such as ethical considerations and accessibility disparities must be addressed for AI to be effectively integrated into psychiatric training. This review provides insights into the future of AI in medical education and its potential impact on training the next generation of psychiatrists.

Residual Channel-attention (RCA) network for remote sensing image scene classification

Abstract High-resolution remote sensing (HRRS) image scene classification has gained increasing importance in recent years, with convolutional neural networks (CNNs) showing particular promise due to their proficiency in extracting spatial features. However, traditional CNNs face significant limitations. Specifically, they struggle to capture complex semantic relationships between objects at varying scales, and they lack the ability to effectively capture long-distance dependencies between features. This limitation is especially problematic in HRRS images, where spatial relationships and semantic content are deeply intertwined. Additionally, traditional CNNs are limited in handling substantial intra-class variation and inter-class similarity, which are common in remote sensing images. To overcome these challenges, we introduce a novel Residual Channel-attention (RCA) network for scene classification. The RCA network introduces a lightweight residual structure to better capture multi-scale spatial features and incorporates a channel attention mechanism that selectively emphasizes relevant feature channels while suppressing irrelevant ones. To further refine the focus on critical image features, we integrate a squeeze-and-excitation (SE) mechanism as a self-attention component, which helps the network prioritize the most informative features and ignore background noise. We evaluated the RCA network on three public datasets: RSSCN7, PatternNet, and EuroSAT, achieving classification accuracies of 97%, 99%, and 96%, respectively. The results demonstrate that superior of the RCA network compared to state-of-the-art strategies in remote sensing image classification. Furthermore, visualization using the Grad-CAM++ algorithm highlights the effectiveness of our channel attention mechanism and underscores the RCA network’s robust feature representation capabilities.

A Dual-Channel and Frequency-Aware Approach for Lightweight Video Instance Segmentation

Video instance segmentation, a key technology for intelligent sensing in visual perception, plays a key role in automated surveillance, robotics, and smart cities. These scenarios rely on real-time and efficient target-tracking capabilities for accurate perception and intelligent analysis of dynamic environments. However, traditional video instance segmentation methods face complex models, high computational overheads, and slow segmentation speeds in time-series feature extraction, especially in resource-constrained environments. To address these challenges, a Dual-Channel and Frequency-Aware Approach for Lightweight Video Instance Segmentation (DCFA-LVIS) is proposed in this paper. In feature extraction, a DCEResNet backbone network structure based on a dual-channel feature enhancement mechanism is designed to improve the model’s accuracy by enhancing the feature extraction and representation capabilities. In instance tracking, a dual-frequency perceptual enhancement network structure is constructed, which uses an independent instance query mechanism to capture temporal information and combines with a frequency-aware attention mechanism to capture instance features on different attention layers of high and low frequencies, respectively, to effectively reduce the complexity of the model, decrease the number of parameters, and improve the segmentation efficiency. Experiments show that the model proposed in this paper achieves state-of-the-art segmentation performance with few parameters on the YouTube-VIS dataset, demonstrating its efficiency and practicality. This method significantly enhances the application efficiency and adaptability of visual perception intelligent sensing technology in video data acquisition and processing, providing strong support for its widespread deployment.

LittleFaceNet: A Small-Sized Face Recognition Method Based on RetinaFace and AdaFace

For surveillance video management in university laboratories, issues such as occlusion and low-resolution face capture often arise. Traditional face recognition algorithms are typically static and rely heavily on clear images, resulting in inaccurate recognition for low-resolution, small-sized faces. To address the challenges of occlusion and low-resolution person identification, this paper proposes a new face recognition framework by reconstructing Retinaface-Resnet and combining it with Quality-Adaptive Margin (adaface). Currently, although there are many target detection algorithms, they all require a large amount of data for training. However, datasets for low-resolution face detection are scarce, leading to poor detection performance of the models. This paper aims to solve Retinaface’s weak face recognition capability in low-resolution scenarios and its potential inaccuracies in face bounding box localization when faces are at extreme angles or partially occluded. To this end, Spatial Depth-wise Separable Convolutions are introduced. Retinaface-Resnet is designed for face detection and localization, while adaface is employed to address low-resolution face recognition by using feature norm approximation to estimate image quality and applying an adaptive margin function. Additionally, a multi-object tracking algorithm is used to solve the problem of moving occlusion. Experimental results demonstrate significant improvements, achieving an accuracy of 96.12% on the WiderFace dataset and a recognition accuracy of 84.36% in practical laboratory applications.

Prediction of Thermodynamic Properties of C60-Based Fullerenols Using Machine Learning.

Traditional machine learning methods face significant challenges in predicting the properties of highly symmetric molecules. In this study, we developed a machine learning model based on graph neural networks (GNNs) to accurately and swiftly predict the thermodynamic and photochemical properties of fullerenols, such as C60(OH)n (n = 1 to 30). First, we established a global method for generating fullerenol isomers through isomer fingerprinting, which can generate all possible isomers or produce diverse structural types on demand. Significantly, by incorporating interpretable descriptors such as atomic labels, bond lengths, and bond angles from highly symmetric isomers, our multilayer GNN model achieved over 90% accuracy in predicting the thermodynamic stability of fullerenols. The model also performed excellently in predicting electronic properties, including the highest occupied molecular orbital (HOMO), lowest unoccupied molecular orbital (LUMO), and the energy gap. Overall, this work demonstrates a new strategy using interpretable descriptors for accurately predicting the properties of highly symmetric structures, offering theoretical chemists a valuable tool for studying these materials.

In Silico Method for ssDNA Aptamer Binding with Aurora Kinase A Protein.

While traditional assay methods face challenges in detecting specific proteins, aptamers, known for their high specificity and affinity, are emerging as a valuable biomarker detection tool. Aurora kinase A (AURKA) plays a role in cell division and influences stem cell reprogramming. In this study, an in silico approach method was conducted for a random ssDNA aptamer sequence selection and its binding with AURKA. The aptamer was designed based on AURKA's structure and nucleic acid sequence, obtained from PDB RCSB. Using RNAfold and RNA composer, we predicted the aptamer's secondary and tertiary structures. Protein-aptamer binding was analyzed via HDOCK and HADDOCK, with 2D interactions visualized in LIGPLOT+ v1.4. Autodock 4 and NAMD 2.3 tools were used to conduct docking and MD simulation studies.

Optimization design and application of library face recognition access control system based on improved PCA.

The application of face recognition technology in Library Access Control System (LACS) has an important impact on improving the security and management efficiency of the library. However, the traditional face recognition methods have some limitations in the face of complex environmental conditions such as illumination and posture change. To solve this problem, an improved method combining the Aggregating Spatial Embeddings for Face Recognition (ASEF) algorithm and Principal Component Analysis (PCA) is proposed. The PCA algorithm is optimized by introducing beta prior and full probability Bayesian model. In addition, the research also integrates K-means Clustering Algorithm (KA) to further improve the accuracy and efficiency of face recognition. The experiment showed that the improved PCA method had an average recognition rate of 92.6%, an average recognition speed of 0.40s, and higher accuracy compared to other related methods, reaching 96%. In practical applications, the system quickly and accurately completes the identification of personnel entry and exit, and improves the efficiency and security of library access management.

Controllable Droplet Splitting with Parallel-Plate Optoelectrowetting Chips.

Recent advances in microfluidic technology highlight electrowetting for its programmability and precision. However, traditional electrowetting chips face limitations in scalability due to fixed electrode sizes. Optoelectrowetting (OEW) offers a solution with light-controlled virtual electrodes, but droplet splitting remains challenging. This study introduces a parallel-plate OEW chip, enhancing droplet manipulation and achieving rapid splitting with light-controlled dark stripe electrodes. We proposed the dimensionless parameter η to characterize droplet splitting, and a mathematical model describes key factors influencing controllable splitting, providing valuable design guidelines for OEW chips.

Data-driven parametrization of molecular mechanics force fields for expansive chemical space coverage.

A force field is a critical component in molecular dynamics simulations for computational drug discovery. It must achieve high accuracy within the constraints of molecular mechanics' (MM) limited functional forms, which offers high computational efficiency. With the rapid expansion of synthetically accessible chemical space, traditional look-up table approaches face significant challenges. In this study, we address this issue using a modern data-driven approach, developing ByteFF, an Amber-compatible force field for drug-like molecules. To create ByteFF, we generated an expansive and highly diverse molecular dataset at the B3LYP-D3(BJ)/DZVP level of theory. This dataset includes 2.4 million optimized molecular fragment geometries with analytical Hessian matrices, along with 3.2 million torsion profiles. We then trained an edge-augmented, symmetry-preserving molecular graph neural network (GNN) on this dataset, employing a carefully optimized training strategy. Our model predicts all bonded and non-bonded MM force field parameters for drug-like molecules simultaneously across a broad chemical space. ByteFF demonstrates state-of-the-art performance on various benchmark datasets, excelling in predicting relaxed geometries, torsional energy profiles, and conformational energies and forces. Its exceptional accuracy and expansive chemical space coverage make ByteFF a valuable tool for multiple stages of computational drug discovery.

Preparation of biodegradable, antibacterial core-spun yarns with braided structures using QAS/PLA micro/nanocomposites

Medical antibacterial textiles play a vital role in tackling the issue of bacterial infection. Traditional surgical sutures face significant challenges due to wound infection caused by bacteria and breakage and scars caused by poor suture strength. Therefore, a new antibacterial and high-strength suture preparation strategy with wide clinical applicability was highly desired. In this study, a biodegradable quaternary ammonium salt (QAS)/polylactic acid (PLA) core-spun yarn with excellent antibacterial and mechanical properties was prepared by conjugated electrospinning technology combined with the braiding process. The antibacterial test results revealed the best overall performance of the PLA micro/nanofiber core-spun yarn with 0.3 wt% QAS antibacterial agent. The antibacterial rate against Escherichia coli and Staphylococcus aureus was 94.49% and 94.00%, respectively, which could effectively solve the problem of wound infection caused by bacteria. In addition, we used the diamond-braided structure to address the poor strength and fragility of the traditional suture strength. The braiding angle of 30° and 45° could effectively enhance the mechanical properties of the yarn, and the breaking strength was also in line with the industry standard. The study proposed that the degradable QAS/PLA micro/nanofiber core-spun yarn, due to its excellent antibacterial and mechanical properties, could find application in medical protection. This provided a new avenue for research into new antibacterial surgical sutures.

Transforming learning: Exploring the frontiers of educational innovation in the 21st Century

The objective of this research is to develop a comprehensive methodology to effectively implement Open Innovation for Competitiveness and Development (ICD) in traditional companies. The aim is to design a methodology that considers the particularities and unique challenges that traditional companies face when adopting open innovation practices. The contribution involves the creation of a structured and systematic approach that covers the stages of the ICD adoption process, from the identification of opportunities to the execution and evaluation of open innovation initiatives.

Talbot effect based sensor measuring grating period change in subwavelength range

Talbot length, the distance between two consecutive self-image planes along the propagation axis for a periodic diffraction object (grating) illuminated by a plane wave, depends on the period of the object and the wavelength of illumination. This property makes the Talbot effect a straightforward technique for measuring the period of a periodic object (grating) by accurately determining the Talbot length for a given illumination wavelength. However, since the Talbot length scale is proportional to the square of the grating period, traditional Talbot techniques face challenges when dealing with smaller grating periods and minor changes in the grating period. Recently, we demonstrated a Fourier transform technique-based Talbot imaging method that allows for controlled Talbot lengths of a periodic object with a constant period and illumination wavelength. Using this method, we successfully measured periods as small as a few micrometers and detected sub-micrometer changes in the periodic object. Furthermore, by measuring the Talbot length of gratings with varying periods imaged through the combination of a thick lens of short focal length and a thin lens of long focal length and large aperture, we determined the effective focal length of the thick lens in close agreement with the theoretical effective focal length of a thick lens in the presence of spherical aberration. These findings establish the Talbot effect as an effective and simple technique for various sensing applications in optics and photonics through the measurement of any physical parameter influencing the Talbot length of a periodic object.

Advances and trends in weed management: A comprehensive review

Weed management is a critical aspect of agricultural practices that can significantly impact crop yield and quality. As traditional methods face challenges such as herbicide resistance and environmental concerns, the agricultural sector is witnessing a shift towards innovative strategies for weed control. This review explores the emerging trends in weed management, focusing on sustainable and efficient approaches. Among them, one prominent trend is the adoption of agro ecological weed management strategies, which combine various control methods such as cultivation techniques, mechanical techniques, biological control, and reasonable herbicide use. This approach minimizes reliance on herbicides while maximizing weed suppression and preserving natural ecosystems. Another significant trend is developing and utilizing precision agriculture technologies for targeted weed control techniques such as satellite imaging, unmanned aerial vehicles (UAVs), and sensor-based systems. These innovations enable farmers to accurately identify and manage weeds, reducing herbicide usage and minimizing environmental impact. Furthermore, the exploration of alternative weed control methods, including thermal, electrical, and microwave-based technologies, is gaining momentum. These non-chemical approaches offer potential solutions to herbicide-resistant weeds and contribute to sustainable agricultural practices. Moreover, the integration of advanced breeding techniques and biotechnology for developing herbicide-resistant crops and enhancing allelopathic traits presents promising avenues for long-term weed management. In conclusion, this review highlights emerging technology for dealing with major problems, including increased understanding of weed biology linked with genomics; novel herbicide-resistant crops and redesigned weed-competing crops; multi-target herbicides; and enhanced biocontrol agents. When combined, these strategies could make up the elements of the next integrated packages designed to impede the emergence of new weed issues.

Integrating Artificial Intelligence with 3D Printing Technology in Healthcare: Sustainable Solutions for Clinical Training Optimization

Integrating Artificial Intelligence (AI) with 3D printing technology offers transformative solutions in healthcare, specifically improving clinical training through precise and customizable replicas of human anatomy. Traditional training methods face challenges such as limited access to high quality models, lack of precision, adaptability, and the environmental impact of resource use. This study employed a mixed methods approach, combining quantitative analysis of model accuracy and input from healthcare professionals, to evaluate the effectiveness of AI optimized 3D printing. The focus was on AI enhanced 3D printing models designed for healthcare training. Traditional methods often lack precision, adaptability, and scalability, limiting their effectiveness in dynamic healthcare scenarios. AI based 3D printing reduces material use by 30\% while providing high quality, customized training models, improving accessibility and sustainability. This research highlights the potential of AI and 3D printing integration to drive technological innovation, align healthcare training with Sustainable Development Goals (SDGs), and promote a more sustainable and efficient future in medical education.

A Lightweight Network with Domain Adaptation for Motor Imagery Recognition

Brain–computer interfaces (BCI) are an effective tool for recognizing motor imagery and have been widely applied in the motor control and assistive operation domains. However, traditional intention-recognition methods face several challenges, such as prolonged training times and limited cross-subject adaptability, which restrict their practical application. This paper proposes an innovative method that combines a lightweight convolutional neural network (CNN) with domain adaptation. A lightweight feature extraction module is designed to extract key features from both the source and target domains, effectively reducing the model’s parameters and improving the real-time performance and computational efficiency. To address differences in sample distributions, a domain adaptation strategy is introduced to optimize the feature alignment. Furthermore, domain adversarial training is employed to promote the learning of domain-invariant features, significantly enhancing the model’s cross-subject generalization ability. The proposed method was evaluated on an fNIRS motor imagery dataset, achieving an average accuracy of 87.76% in a three-class classification task. Additionally, lightweight experiments were conducted from two perspectives: model structure optimization and data feature selection. The results demonstrated the potential advantages of this method for practical applications in motor imagery recognition systems.

THE SOCIAL ROLES OF WOMEN AS PORTRAYED IN SIGIRI POETRY

The Sigiri poetry collection is considered to be the origin of Sinhala verse literature in Sri Lanka. These sigiri verses were reportedly composed between the 7th to 14th Century BC. Most of the poems among Sigiri poems are written in verse style called 'Gee'. Thus, those poems are named ‘Sigiri Gee’. These poems are said to have been written on the Sigiri Mirror wall known as the "Katapath Pawura" by various travelers who came to visit the ruined Sigiriya from different eras. Among these poets were kings, priests, soldiers, etc., as well as women in particular. There are also 14 female poets, among the Sigiri writers, introducing themselves by a name. Sigiri rock paintings inspired most poems by all these poets. However, some female poets express their moods and independent ideologies. Those poems have revealed the lives of women of that era as well as their unique perceptions and thought processes. Also, information about women in Sigiri society has been revealed by the opinions of other poets. The study therefore aims to explore how the social roles of women in Sigiri society are portrayed in Sigiri poetry. This analysis is based on the sociological concept of social roles. The Sigiri poetry has highlighted the challenging lives of women in Sigiri society, their social roles and their unwavering independence in the face of tradition.

CS-FL: Cross-Zone Secure Federated Learning with Blockchain and a Credibility Mechanism

Federated learning enables multiple intelligent devices to collaboratively perform machine learning tasks while preserving local data privacy. However, traditional FL architectures face challenges such as centralization and lack of effective defense mechanisms against malicious nodes, particularly in large-scale edge computing scenarios, which can lead to system instability. To address these challenges, this paper proposes a cross-zone secure federated learning method with blockchain and credibility mechanism, named CS-FL. By constructing a dual-layer blockchain network and introducing a blockchain ledger between zone servers, CS-FL establishes a decentralized trust mechanism for index detection and model aggregation. In node selection, CS-FL considers multiple dimensions, including node quality, communication resources, and historical credibility, and employs a three-stage mechanism that introduces lightweight probe tasks to assess node status before formal FL training, ensuring high-quality nodes participate. Additionally, the credibility incentive mechanism penalizes nodes that bypass probe mechanism and engage in malicious behaviors, effectively mitigating the impact of deceptive attacks. Experimental results show that CS-FL significantly improves the defense performance of FL, reducing attack success rates from 75–85% to below 5–20% when facing different types of threats, and effectively maintaining the training accuracy of the FL model. This demonstrates the potential of CS-FL to enhance the security and stability of FL systems in complex edge computing scenarios.

Tea grading, blending, and matching based on computer vision and deep learning.

Accurate tea blending assessment and sample matching are critical in the tea production process. Traditional methods face efficiency and accuracy challenges, which can be addressed by advances in computer vision and deep learning. This study developed an efficient and non-destructive method for fast tea grading classification, blending ratio evaluation, and sample matching. The method trained a Residual Network (ResNet) model on an enhanced dataset of tea images and used Convolutional Block Attention Module (CBAM) to improve the model's feature-extraction ability. The enhanced grade classification model achieved 95.05% accuracy for oolong tea and 99.13% accuracy for black tea, outperforming other deep-learning models such as EfficientNet, MobileNet, and VGG16. For oolong tea blends, the model demonstrated greater efficiency than manual evaluation with an average absolute error of 2.26%. In black tea sample matching, the model achieved an average error of 3.34%. These results highlight the importance of attention mechanisms in improving the analysis of images with intricate textures. The integration of deep learning and attention modules enhanced the accuracy and efficiency of tea quality evaluation processes effectively. This study underscores the transformative potential of intelligent classification and analysis methods in modernizing tea production, ensuring higher standards of consistency and quality. © 2024 Society of Chemical Industry.

Review on Importance of Ectoparasites in Backyard Chicken Production Systems in Rural Ethiopia

Backyard chicken production plays a significant role in rural Ethiopia, contributing to food security and household income. However, indigenous chickens reared under traditional scavenging systems face numerous challenges, with ectoparasite infestations being a major constraint. These ectoparasites, including fleas, lice, ticks, and mites, lead to health issues, reduced productivity, and economic losses. Despite their impact, ectoparasites in Ethiopian backyard poultry systems remain understudied. This review aims to estimate the prevalence of ectoparasites in backyard chicken production and identify the major species affecting poultry in rural Ethiopia. We examine the status of the village chicken production system, explore the challenges faced by poultry farmers, and highlight the role of ectoparasites in reducing productivity. Key ectoparasites such as Echidnophaga gallinacea (flea), Menacanthus stramineus (louse), Argas persicus (tick), and Dermanyssus gallinae (mite) are discussed in terms of their epidemiology, economic impact, and control methods. The findings underscore the need for targeted interventions to control ectoparasite infestations, improve poultry health, and enhance production systems in rural Ethiopia.

Stream ciphers for digital image transactions by learning quantum true random numbers

Abstract The digital economy drives a surge in online digital image transactions, increasing the risk of data breaches due to extensive image file transmission. Stream ciphers, known for their efficiency compared to block ciphers, have emerged as a preferred choice for encrypting images in such transactions to safeguard transmitted data. Nevertheless, traditional stream cipher algorithms face diverse security threats. To address this challenge, efforts have been devoted to generating stream ciphers by generative adversarial networks (GANs) transforming input style into random patterns. Regrettably, these ciphers face issues in key sensitivity, randomness, and style transformation failures. Quantum true random numbers offer a potential solution but are costly to deploy. To handle this dilemma, we design stream ciphers relied on a neural network random number generator (RNG) using quantum true random numbers for training least squares GANs. Specifically, two fully-connected layers are incorporated into the RNG, avoiding the defects of style transformation in existing GANs-based stream ciphers. Besides, a random number calculation formula is employed to ensure that each decimal place output by the generator contributes to the computation of the random numbers. By doing so, the randomness of GANs is enhanced and the deployment of costly quantum devices is avoided. Experiments reveal that the information entropy of our generated images reaches to 7.9991, the adjacent pixel correlation coefficient of the ciphertext attains -0.0015, the Number of Pixel Change Rate and Unified Average Changing Intensity achieve 99.62% and 33.52%, respectively. These results demonstrate that the designed RNG facilitates randomness, whilst having secure properties applied in stream ciphers.