Search Results (338)

Symmetry and asymmetry are foundational design principles in artificial intelligence, defining the balance between invariance and adaptability in multimodal learning systems. In audio-visual speaker verification, where speech and lip-motion features are jointly modeled to determine whether two utterances belong to the same individual, these principles govern both fairness and discriminative power. In this work, we analyze how symmetry and asymmetry emerge within a gated-fusion architecture that integrates Time-Delay Neural Networks and Bidirectional Long Short-Term Memory encoders for speech, ResNet-based visual lip encoders, and a shared Conformer-based temporal backbone. Structural symmetry is preserved through weight-sharing across paired utterances and symmetric cosine-based scoring, ensuring verification consistency regardless of input order. In contrast, asymmetry is intentionally introduced through modality-dependent temporal encoding, multi-head attention pooling, and a learnable gating mechanism that dynamically re-weights the contribution of audio and visual streams at each timestep. This controlled asymmetry allows the model to rely on visual cues when speech is noisy, and conversely on speech when lip visibility is degraded, yielding adaptive robustness under cross-modal degradation. Experimental results demonstrate that combining symmetric embedding space design with adaptive asymmetric fusion significantly improves generalization, reducing Equal Error Rate (EER) to 3.419% on VoxCeleb-2 test dataset without sacrificing interpretability. The findings show that symmetry ensures stable and fair decision-making, while learnable asymmetry enables modality awareness together forming a principled foundation for next-generation audio-visual speaker verification systems. Full article

Background/Objectives: This cross-sectional study examined how training age, chronological age, and biological maturity influence motor and perceptual–cognitive performance in youth soccer players, with relevance for health and well-being through sport participation. Methods: Forty-one male athletes (age = 14.86 ± 0.81 years) completed a two-day field-based assessment following a holistic framework integrating motor (sprinting, jumping, and agility) and perceptual–cognitive components (psychomotor speed, visuospatial working memory, and spatial visualization). Biological maturity was estimated using the maturity offset method. Results: Regression analyses showed that biological maturity and training age significantly predicted motor performance, particularly sprinting, jumping, and pre-planned agility, whereas chronological age was not a predictor. In contrast, neither maturity nor training experience influenced perceptual–cognitive skills. Among cognitive measures, only psychomotor speed significantly predicted reactive agility, emphasizing the role of rapid information processing in dynamic, game-specific contexts. Conclusions: Youth soccer training should address both physical and cognitive development through complementary strategies. Physical preparation should be tailored to maturity status to ensure safe and progressive loading, while systematic training of psychomotor speed and decision-making should enhance reactive agility and game intelligence. Integrating maturity and perceptual–cognitive assessments may support individualized development, improved performance, and long-term well-being. Full article

Road anomaly detection is essential for intelligent transportation systems and road maintenance. This work presents a MATLAB-native hybrid Convolutional Neural Network–Bidirectional Long Short-Term Memory (CNN–BiLSTM) framework for context-aware road event detection using multiaxial acceleration and vibration signals. The proposed architecture integrates short-term feature extraction via one-dimensional convolutional layers with bidirectional LSTM-based temporal modeling, enabling simultaneous capture of instantaneous signal morphology and long-range dependencies across driving trajectories. Multiaxial data were acquired at 50 Hz using an AQ-1 On-Board Diagnostics II (OBDII) Data Logger during urban and suburban routes in San Andrés Cholula, Puebla, Mexico. Our hybrid CNN–BiLSTM model achieved a global accuracy of 95.91% and a macro F1-score of 0.959. Per-class F1-scores ranged from 0.932 (none) to 0.981 (pothole), with specificity values above 0.98 for all event categories. Qualitative analysis demonstrates that this architecture outperforms previous CNN-only vibration-based models by approximately 2–3% in macro F1-score while maintaining balanced precision and recall across all event types. Visualization of BiLSTM activations highlights enhanced interpretability and contextual discrimination, particularly for events with similar short-term signatures. Further, the proposed framework’s low computational overhead and compatibility with MATLAB Graphics Processing Unit (GPU) Coder support its feasibility for real-time embedded deployment. These results demonstrate the effectiveness and robustness of our hybrid CNN–BiLSTM approach for road anomaly detection using only acceleration and vibration signals, establishing a validated continuation of previous CNN-based research. Beyond the experimental validation, the proposed framework provides a practical foundation for real-time pavement monitoring systems and can support intelligent transportation applications such as preventive road maintenance, driver assistance, and large-scale deployment on low-power embedded platforms. Full article

Real-time fine-grained human activity recognition (HAR) remains a challenging problem due to rapid spatial–temporal variations, subtle motion differences, and dynamic environmental conditions. Addressing this difficulty, we propose NovAc-DL, a unified deep learning framework designed to accurately classify short human-like actions, specifically, “pour” and “stir” from sequential video data. The framework integrates adaptive time-distributed convolutional encoding with temporal reasoning modules to enable robust recognition under realistic robotic-interaction conditions. A balanced dataset of 2000 videos was curated and processed through a consistent spatiotemporal pipeline. Three architectures, LRCN, CNN-TD, and ConvLSTM, were systematically evaluated. CNN-TD achieved the best performance, reaching 98.68% accuracy with the lowest test loss (0.0236), outperforming the other models in convergence speed, generalization, and computational efficiency. Grad-CAM visualizations further confirm that NovAc-DL reliably attends to motion-salient regions relevant to pouring and stirring gestures. These results establish NovAc-DL as a high-precision real-time-capable solution for deployment in healthcare monitoring, industrial automation, and collaborative robotics. Full article

To address the limitations of sensor deployment and high maintenance costs in smart swine barns, this study proposes a digital twin (DT)-based virtual sensor prediction and visualization method. Spatial constraints and harsh barn environments often cause sensor blackout zones, hindering precise environmental monitoring. To overcome these challenges, a virtual sensor was defined at the central position between Zone 1 and Zone 2, and its data were generated using a hybrid model that combines inverse distance weighting (IDW)-based spatial interpolation with long short-term memory (LSTM)-based time-series prediction. The proposed method was evaluated using 34,992 datasets collected from January to August 2025. Performance analysis demonstrated that the hybrid model achieved high prediction accuracy, particularly for variables with strong spatial heterogeneity, such as carbon dioxide (CO₂) and ammonia (NH₃), with overall coefficients of determination (R²) exceeding 0.95. Furthermore, a Web-based graphics library (WebGL) digital twin visualization environment was developed to intuitively observe spatiotemporal changes in sensor data. The system integrates sensor placement, risk-level assessment, and time-series graphs, thereby supporting users in real-time environmental monitoring and decision-making. This approach improves the precision and reliability of smart barn management and contributes to the stabilization of farm income. Full article

The optimal indoor lighting comfort can enhance physical and mental health and improve work efficiency. The traditional methods for evaluating lighting comfort have problems such as limited data analysis and poor subjectivity. To establish objective criteria, this study proposes a novel method combining deep learning and evoked potentials. This study collected visual evoked potentials across diverse indoor lighting conditions and employed Long Short-Term Memory (LSTM) and Gate Recurrent Unit (GRU) recurrent neural networks to classify temporal evoked electroencephalography data. The experimental results show that both LSTM and GRU achieve higher accuracy than the Feedforward Neural Network. Among them, LSTM performs best, reaching an accuracy of 80.16% while maintaining computational efficiency comparable to GRU. Such effective objective evaluation methods provide a scientific basis for optimizing indoor environments. Full article

Classroom acoustics and noise exposure significantly impact students’ emotional, cognitive, and academic well-being. This study investigates how classroom noise and acoustics affect auditory and cognitive performance among 131 children in three primary schools in northeast Italy. Student performance was assessed using standardised tests evaluating working memory, verbal short and long-term memory, and visuospatial memory. Children were tested under two distinct acoustic conditions: ambient classroom noise and artificially induced noise (comprising a sequence of typical internal and external classroom sounds, intelligible speech, and unintelligible conversations). Prior to testing, hearing threshold was assessed, in order to reveal any existing impairments. Following each experimental session, children rated their perceived effort and fatigue in completing the tests. Acoustic characterisation of empty classrooms was performed using Reverberation Time (T₂₀), Clarity (C₅₀), and Speech Transmission Index (STI), while noise level was measured during all testing phases. Regression analysis was employed to correlate noise levels and reverberation times with class-average performance and perception scores. Results indicate that noise significantly impaired both verbal working memory and visual attention, increasing perceived effort and fatigue. Notably, both ambient and induced noise conditions exhibited comparable adverse effects on attentional and memory task performance. These findings underscore the critical importance of acoustic design in educational environments and provide empirical support for developing classroom acoustic standards. Full article

This study investigates the application of advanced deep learning models to forecast fossil energy prices, a critical factor influencing global economic stability. Unlike previous research, this study conducts a comparative analysis of Gated Recurrent Unit (GRU), Recurrent Neural Network (RNN), Bidirectional Long Short-Term Memory (Bi-LSTM), Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), and Deep Neural Network (DNN) models. The evaluation metrics employed include Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). The results reveal that recurrent architectures, particularly GRU, LSTM, and Bi-LSTM, consistently outperform feedforward and convolutional models, demonstrating superior ability to capture temporal dependencies and nonlinear dynamics in energy markets. In contrast, the RNN and DNN show relatively weaker generalization capabilities. Additionally, visualizations of actual versus predicted prices for each model further emphasize superior forecasting accuracy of recurrent models. The results highlight the potential of deep learning in enhancing investment and policy decisions. Additionally, the results provide significant implications for policymakers and investors by emphasizing the value of accurate energy price forecasting in mitigating market volatility, improving portfolio management, and supporting evidence-based energy policies. Full article

Snoring, a common sleep-disordered breathing phenomenon, impairs sleep quality for both the sufferer and any bed partner. While mild snoring primarily disrupts sleep continuity, severe cases often indicate obstructive sleep apnea (OSA), a disorder affecting 9–17% of the global population, linked to significant comorbidities and socioeconomic burden (see Introduction for supporting data). Here, we propose a low-cost, real-time snoring detection and intervention system that integrates a multiple-spectrum deep learning framework with an Internet of Things (IoT)-enabled smart pillow. The modified Parallel Convolutional Spatiotemporal Network (PCSN) combines three parallel convolutional neural network (CNN) branches processing Constant-Q Transform (CQT), Synchrosqueezing Wavelet Transform (SWT), and Hilbert–Huang Transform (HHT) features with a Long Short-Term Memory (LSTM) network to capture spatial and temporal characteristics of sounds associated with snoring. The smart pillow prototype incorporates two Micro-Electro-Mechanical System (MEMS) microphones, an ESP8266 off-shelf board, a speaker, and two vibration motors for real-time audio acquisition, cloud-based processing via Arduino cloud, and closed-loop haptic/audio feedback that encourages positional changes without fully awakening the snorers. Experiments demonstrated that the modified PCSN model achieves 98.33% accuracy, 99.29% sensitivity, 98.34% specificity, 98.3% recall, and 98.32% F1-score, outperforming existing systems. Hardware costs are under USD 8 and a smartphone app provides authorized users with real-time visualization and secure data access. This solution offers a cost-effective and accurate approach for home-based OSA screening and intervention. Full article

This study introduces a novel mixed reality (MR) TMJ dislocation teaching program developed using HoloLens 2 and collaboration with interdisciplinary teams. The program offers an immersive learning experience, enabling individuals to visualize and interact with detailed 3D temporomandibular joint (TMJ) models and practice different reduction techniques repeatedly. Real-time feedback, combining the visual holographic overlay with mechanical resistance in the physical model, supports the learning process. The 3D-printed skull model provided haptic feedback, strengthened the positive response given by the MR model, and reinforced muscle memory. Despite some challenges related to the learning curve and cost, the program shows promise for practicing uncommon, high-anxiety clinical procedures in medical education. Future research directions include comparisons with traditional teaching methods, evaluating long-term skill retention, and exploring MR applications in other clinical procedures. Overall, this project demonstrates the potential of MR technology to advance medical education and skill acquisition. Full article

Mine water inrush poses a severe threat to coal mine safety, making rapid and accurate identification of water sources essential. Existing methods, including conventional hydrochemical diagrams and machine learning, struggle with high-dimensional, nonlinear hydrogeochemical data characterized by implicit temporal dynamics. This study proposes an intelligent identification model integrating convolutional neural networks (CNNs), long short-term memory (LSTM), and an attention mechanism (CNN–LSTM–Attention). The model employs a CNN to extract local fingerprint features from hydrochemical indicators (K⁺+Na⁺, Ca²⁺, Mg²⁺, Cl⁻, SO₄²⁻, and HCO₃⁻), uses LSTM to model evolutionary patterns, and leverages an attention mechanism to adaptively focus on critical discriminative features. Based on 76 water samples from the Tangjiahui Coal Mine, the model achieved 91% accuracy on the test set, outperforming standalone CNN, LSTM, and CNN–LSTM models. Visualization of attention weights further revealed key diagnostic indicators, enhancing interpretability and bridging data-driven methods with hydrogeochemical mechanisms. This study provides a powerful and interpretable tool for water inrush source identification, supporting the transition toward intelligent and transparent coal mine water hazard prevention. Full article

Objectives: Alexander disease (AxD) is a rare neurodegenerative disorder that represents a group of leukodystrophies with severe disability and premature death, mostly with an infancy/childhood onset. In rare cases of late-onset phenotypes, symptoms are often milder and difficult to diagnose. We present a diagnostic journey of a teenage male patient with a progressive gait disorder starting at the age of 13 years, with a final diagnosis of Alexander disease. Early in the course of the disease, the boy exhibited distinctive cognitive involvement and neuropsychological deterioration characterized by selective impairment of visual and long-term auditory memory, along with a decline in IQ but preserved reasoning abilities. Methods: The patient underwent an extensive neurological diagnostic workup, which included magnetic resonance imaging (MRI) of the brain, spine, and abdomen, as well as electrophysiological, metabolic, and biochemical tests. Numerous specialist consultations were conducted, including genetic, cardiology, ophthalmology, pulmonology, oncohematology, psychological, and speech–language pathology consultations. In addition, a focused literature review was performed using PubMed, Scopus, Web of Science, and Google Scholar with the search terms “Alexander disease,” “GFAP gene,” “late-onset,” “spastic paraplegia” and “GFAP variant p/Gly18Val”. Results: Whole exome sequencing revealed an extremely rare missense GFAP heterozygous variant NM_002055.5: c.54G>T (p/Gly18Val), confirming the diagnosis of AxD. Conclusions: The presented case highlights the importance of whole-exome sequencing in the diagnosis of unexplained otherwise neurological symptoms, such as progressive spastic paraplegia. Full article

This paper presents a federated learning (FL) architecture tailored for anomaly detection in wire arc additive manufacturing (WAAM) that preserves data privacy while enabling secure and distributed model training across heterogeneous process units. WAAM’s inherent process complexity, characterized by high-dimensional and asynchronous sensor streams, including current, voltage, travel speed, and visual bead profiles, necessitates a decentralized learning paradigm capable of handling non-identical client distributions without raw data pooling. To this end, the proposed framework integrates reversible data hiding in the encrypted domain (RDHE) for the secure embedding of sensor-derived features into weld images, enabling confidential parameter transmission and tamper-evident federation. Each client node employs a domain-specific long short-term memory (LSTM)-based classifier trained on locally curated time-series or vision-derived features, with model updates embedded and transmitted securely to a central aggregator. Three FL strategies, FedAvg, FedProx, and FedPer, are systematically evaluated against four robust aggregation techniques, including KRUM, Multi KRUM, and Trimmed Mean, across 100 communication rounds using eight non-independent and identically distributed (non-IID) WAAM clients. Experimental results reveal that FedPer coupled with Trimmed Mean delivers the optimal configuration, achieving maximum F1-score (0.912), area under the curve (AUC) (0.939), and client-wise generalization stability under both geometric and temporal noise. The proposed approach demonstrates near-lossless RDHE encoding (PSNR > 90 dB) and robust convergence across adversarial conditions. By embedding encrypted intelligence within weld imagery and tailoring FL to WAAM-specific signal variability, this study introduces a scalable, secure, and generalizable framework for process monitoring. These findings establish a baseline for federated anomaly detection in metal additive manufacturing, with implications for deploying privacy-preserving intelligence across smart manufacturing (SM) networks. The federated pipeline is backbone-agnostic. We instantiate LSTM clients because the sequences are short (five steps) and edge compute is limited in WAAM. The same pipeline can host Transformer/TCN encoders for longer horizons without changing the FL or security flow. Full article

In the mining industry, operational efficiency, equipment reliability, and mineral quality assessment are paramount for cost-effective and sustainable production. Traditional approaches often address equipment maintenance and quality control as separate challenges, leading to suboptimal operational synergy. This paper proposes a novel artificial intelligence (AI) framework that integrates predictive maintenance with real-time mineral quality assessment through advanced sensor fusion and deep learning. Our model leverages a hybrid architecture, combining Convolutional Neural Networks (CNNs) for analyzing visual and spectral data of iron ore with Long Short-Term Memory (LSTM) networks for processing temporal sensor data (vibration, thermal, acoustic) from critical equipment like crushers and conveyors. A dedicated fusion layer synthesizes these spatial and temporal features to simultaneously predict equipment failure probability and classify mineral quality. Validated on a real-world dataset from active iron ore mines, the system demonstrates a significant 20–30% reduction in projected maintenance downtime and a 15% improvement in mineral classification accuracy compared to baseline models while achieving real-time inference speeds of less than 10 milliseconds. This work underscores the transformative potential of unified AI-driven systems in enhancing the intelligence, resilience, and productivity of modern mining operations. Full article

This study addresses the increasingly complex challenge of multiple-fault diagnosis in modern intelligent automotive electronic systems by proposing an innovative deep learning-based solution. The research integrates Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, and the Transformer architecture to construct a multi-modal fault diagnosis model. By collecting real-world operational data from vehicle electronic systems, including fault samples from key modules such as the Engine Control Unit (ECU), Body Control Module (BCM), and safety systems, a comprehensive dataset comprising 12 major fault types was established. Experimental results demonstrate that the proposed hybrid deep learning model achieves a multiple-fault identification accuracy of 96.8%, representing a 23% performance improvement over traditional diagnostic methods. The integration of Explainable AI (XAI) techniques provides the diagnostic results with visual interpretability, aiding maintenance technicians in understanding the model’s diagnostic logic. The findings of this research can be applied in smart factories, automotive service centers, and on-board diagnostic (OBD) systems, offering significant practical value in enhancing vehicle safety and reducing maintenance costs. Full article