Search Results (8,239)

Urban intersection traffic signals play a crucial role in managing traffic flow and ensuring road safety. However, traditional actuated signal controllers make phase-switching decisions based on limited local traffic information, without leveraging network-wide context from navigation services. In this paper, we propose CATS, a Context-Aware Traffic Signal control system that jointly optimizes intersection signal control and road navigation for Connected and Automated Vehicles (CAVs). CATS integrates two key components: a Best-Combination CTR (BC-CTR) scheme and the Self-Adaptive Interactive Navigation Tool (SAINT). BC-CTR enhances the original Cumulative Travel-Time Responsive (CTR) scheme through a two-step selection procedure: it first identifies the phase with the highest cumulative travel time (CTT) and then selects the compatible phase combination with the greatest group CTT, providing an explicit improvement over the single-combination evaluation of the original CTR that allows for a more accurate response to real-time intersection demand. SAINT provides congestion-aware route guidance via a congestion-contribution step function, directing vehicles away from congested segments while signal timings simultaneously adapt to incoming traffic. Under a 100% CAV penetration setting, SUMO-based simulations across moderate-to-heavy traffic conditions (vehicle inter-arrival times of 5 to 9 s) show that CATS reduces the mean end-to-end travel time by up to 23.72% and improves the throughput by up to 93.19% over three baselines (fixed-time navigation with enhanced signal control, congestion-aware navigation with original signal control, and fixed-time navigation with original signal control), confirming that the co-design of navigation and signal control produces complementary benefits. Full article

Inertial measurement units (IMUs) present promising avenues for performance diagnostics in skateboarding, yet systematic validation of their accuracy and applicability remains limited. This study validates the commercially available Spinnax Freak IMU system in the context of skateboarding, with a focus on selected trick detection and classification, distance measurement, maximal horizontal speed, maximal vertical height of the skateboard and airtime during a jump trick. A total of 23 skateboarders (4 females, 19 males; 27.4 ± 10.9 years) participated in this study. Validation methods included comparisons with established reference systems such as laser ranging for maximal horizontal speed (LAVEG), 2D video analysis for maximal vertical height of the skateboard (Kinovea), light barrier measurements for airtime detection (OptoJump Next), and a fixed metric reference (10 m) for rolling distance measurements. The evaluation was supported by statistical analyses including mean absolute error (MAE), root mean-square error (RMSE), mean absolute percentage error (MAPE), t-tests, Bland–Altman plots, linear regression, and ICC(3,1). The Spinnax Freak system demonstrated high validity in detecting trick events and in providing distance measurements that were statistically equivalent to the reference. Trick classification, maximal horizontal speed, maximal vertical height of the skateboard and airtime showed substantial errors, indicating that these outputs are not reliable for biomechanical interpretation at this point. These findings highlight both the potential and the current constraints of single-sensor setups for field-based motion capture in skateboarding. Future developments should prioritize algorithmic refinement, improved temporal resolution, and optimized event classification to enhance measurement accuracy and expand applicability in biomechanical analysis and automated training documentation in skateboarding. Full article

Short-horizon equity forecasting remains challenging because daily prices are noisy, heavy-tailed, and subject to structural breaks and regime shifts. We develop a fully automated, reproducible, and leakage-controlled multi-model pipeline for daily forecasting with horizon-specific configuration selection. The task is formulated as predicting cumulative H-day log-returns from OHLCV-derived information and converting them to implied price forecasts. All model families share a homologated design: causal feature construction, a strictly chronological split with an explicit purging rule to prevent label-window overlap for multi-day targets, training-only robustification (winsorization and adaptive clipping), and a unified metric suite computed consistently in return and price spaces. The framework benchmarks transparent baselines (zero- and mean-return), gradient-boosted trees (XGBoost), and deep temporal models (LSTM and CNN/TCN). Lookback length $L\in \{60,180,500\}$ is selected via an internal walk-forward procedure on the pre-evaluation block, and final performance is reported on an external hold-out segment (last 15% of instances). Experiments on daily data for MT, DELL, and the S&P 500 index (through 3 February 2026) show that all families achieve similarly strong price-level fit at $H=1$, largely driven by persistence in the price process, while separation across families becomes more visible at $H=5$. However, predictive performance in return space remains weak, with $R^2$ close to zero or negative, and Diebold–Mariano tests do not provide consistent evidence of statistical superiority over naive benchmarks. Under an operational rule that minimizes hold-out RMSE on the price scale, selected models are asset- and horizon-dependent, supporting horizon-wise selection rather than a single global architecture. Overall, the primary contribution lies in the proposed leakage-controlled evaluation and benchmarking framework rather than in demonstrating consistent predictive gains in financial time series forecasting. Full article

The automated production of large-scale labeled datasets from concrete X-ray computed tomography (CT) images is a fundamental prerequisite for training and validating deep learning-based segmentation models. However, existing methods either require extensive manual annotation or rely on domain-specific deep learning models that themselves demand labeled data—a circular dependency. This paper presents a parameter-driven three-class segmentation framework that automatically classifies each pixel in a concrete CT slice into one of three material phases: void (air pores and cracks), coarse aggregate, and mortar matrix, generating annotation masks suitable for large-scale dataset production without manual labeling. The proposed method combines: (1) fixed-threshold void detection calibrated to concrete CT grayscale characteristics; (2) adaptive percentile-based initial segmentation responsive to image-specific statistics; (3) multi-criteria connected component scoring based on area, shape descriptors (circularity, solidity, compactness, extent, aspect ratio), intensity distribution, and boundary gradient; (4) material science-informed size constraints aligned with concrete phase volume fractions; and (5) a material continuity enforcement module that applies topological hole-filling and conditional convex-hull consolidation to eliminate internal contamination within accepted aggregate regions, reducing boundary roughness by 7.6% and recovering misclassified boundary pixels. All parameters are centralized in a configuration file, enabling reproducible batch processing of 224 × 224 pixel CT slices at 0.07–1.12 s per image. Evaluated on 1007 224 × 224 concrete CT patches cropped from 200 representative scan frames, the framework produces three-class segmentation masks with physically consistent void fractions (mean 3.2%), aggregate fractions (mean 32.4%), and mortar fractions (mean 64.4%), all within ranges reported in the concrete CT literature (used as a dataset-scale QC screen, not a validation metric). Primary outputs and the archived image–mask pairs for this work are provided as an 8-bit patch archive. For pixel-wise validation, we report IoU, Dice, and pixel accuracy on an independently labeled subset that can be unambiguously paired with the released predictions: averaged over 57 matched patches, mean pixel accuracy is 88.6%, macro-mean IoU is 74.7%, and macro-mean Dice is 84.9%. The framework provides a fully automated annotation pipeline for dataset production, eliminating manual labeling costs for concrete CT image collections. The generated datasets are suitable for training semantic segmentation networks such as U-Net and its variants. Full article

In recent years, deep learning has been applied more and more to medical image analysis. One such application of deep learning is the automated polyp detection in colonoscopy with the target of reducing miss rates. This study presents a comprehensive evaluation of nine state-of-the-art object detection models for colonic polyp detection: YOLOv8, YOLOv9, YOLOv10, YOLO11, YOLO12, YOLO26, RT-DETR, YOLO-World, and YOLOE. The models were evaluated on three publicly available datasets: CVC-ClinicDB, CVC-ColonDB, and ETIS-LaribPolypDB. All models were trained under standardized conditions using identical hyperparameters and data augmentation strategies to guarantee fair comparison. Performance was evaluated using multiple metrics: mAP@50, mAP@50–95, F1 score, precision, recall, inference time, and computational cost. YOLO11 demonstrated the best overall performance, achieving mAP@50 scores of 0.995, 0.944, and 0.978 on the three datasets respectively, while maintaining the fastest inference time of approximately 150 ms per image and the third-lowest computational cost at 21.3 GFLOPs. Cross-dataset generalization experiments revealed a significant loss of performance, with mAP@50 dropping by 20–40% when models were tested on an unseen dataset, highlighting the challenge of true generalization with limited datasets. Statistical analysis by polyp size showed that while all models achieved F1 scores exceeding 0.95 for large polyps, performance decreased to 0.60–0.85 for small polyps, indicating a limitation in detecting small lesions. The analysis of failure modes showed that missed detections, false positives and boundary errors constitute 60–75% of all failures, suggesting that domain adaptation of object detection models may be required. Full article

With the advancement of 3D digital dentistry, accurate 3D tooth segmentation has become increasingly important in orthodontics and computer-aided diagnosis. However, existing supervised approaches heavily rely on exhaustive face-wise annotations and often exhibit limited generalization across complex clinical meshes. Although self-supervised learning offers a promising alternative to alleviate annotation costs, current paradigms remain challenged by sensitivity to data augmentations, suboptimal representation learning in pure masking schemes, and the complex structural characteristics of dental geometry. To address these limitations, we propose Dental-CMAE, a graph-enhanced hierarchical Contrastive masked AutoEncoder framework tailored for 3D tooth segmentation. The framework incorporates a dual-branch masking strategy that leverages graph-based structural priors to generate distinct corrupted views while preserving intrinsic mesh topology, thereby facilitating robust reconstruction. This is integrated with a feature-level contrastive objective designed to enforce semantic consistency between co-masked regions, which enhances representation discriminability without the requirement for negative sample queues. Additionally, the architecture utilizes a hierarchical multi-scale attention mechanism that partitions feature channels into parallel streams, enabling the simultaneous capture of fine-grained morphological variations and the overarching global dental arch context. Extensive experiments demonstrate that our Dental-CMAE consistently outperforms state-of-the-art fully supervised and self-supervised methods across multiple evaluation metrics. Specifically, our framework achieves an Overall Accuracy (OA) of 95.57%, a mean Intersection-over-Union (mIoU) of 88.14%, and a mean Accuracy (mAcc) of 90.85%. Supported by these quantitative findings, our method validates its effectiveness for robust 3D tooth segmentation, highlighting its strong potential to alleviate annotation bottlenecks and improve the reliability of automated 3D digital dental workflows. Full article

Background/Objectives: Echocardiography reports are essential diagnostic tools, but their complexity and specialized English terminology frequently hinder comprehension for non-specialists and patients. This study addresses these accessibility gaps by developing a resource-efficient large language model (LLM) system designed to translate and summarize English echocardiography results into Traditional Chinese. Methods: To overcome significant hardware constraints, we utilized Quantized Low-Rank Adapter (QLoRA) techniques and the Unsloth acceleration framework to fine-tune LLaMA-3.2-1B and LLaMA-3.2-3B-Instruct models on a single mid-tier GPU. The system employs a dual-stage inference architecture: the first stage provides technical medical translation for clinicians, while the second stage generates simplified, patient-centric educational summaries to enhance health literacy. Results: Evaluation across multiple metrics, including BLEU, ROUGE, METEOR, and Perplexity, demonstrated that the LLaMA-3.2-3B-Instruct model with the AdamW 8-bit optimizer achieved the most stable validation performance, excelling in semantic coherence and structural consistency. A preliminary qualitative error analysis conducted in the Discussion section further identified clinical nuances, such as terminology simplification and minor hallucinations, underscoring the critical necessity of a Human-in-the-Loop verification procedure. Conclusions: These findings validate the feasibility of deploying cutting-edge medical AI in resource-limited clinical environments. While the results reflect validation-only performance on a specialized dataset, the platform offers a scalable foundation for enhancing clinical decision support and health literacy through accessible, automated medical text processing. Full article

Additive manufacturing of concrete offers reduced waste, faster construction, and design freedom, yet effective reinforcement integration remains a major challenge due to weak interlayer bonding and anisotropy. Most prior studies focus on vertical reinforcement, short fibers, or metallic systems, achieving modest flexural improvements (15–60%). This study evaluates horizontal continuous reinforcement using glass fiber mesh and two carbon fiber meshes (ARMO-mesh 200/200 and 500/500) integrated during 3D printing. The methods include extrusion-based printing of small (four-layer) and beam-like (eight-layer) specimens, both printed and cast, followed by three-point flexural and compression tests at 7 and 28 days under vertical and horizontal loading. The results show that ARMO-mesh 500/500 significantly enhances flexural strength—up to 100% over unreinforced controls (e.g., 24.4 kNm vs. 12.2 kNm in small specimens at 28 days) and ~60% over ARMO-mesh 200/200, while glass mesh provides only marginal gains (~12%). Carbon meshes also improve post-cracking toughness and apparent interlayer cohesion. A pronounced size effect reduces nominal strength in larger specimens. These findings demonstrate that wide-format porous carbon meshes offer a scalable, corrosion-resistant solution for load-bearing 3D-printed concrete elements, advancing automated digital construction. Full article

The detection technology for metal surface defects plays a crucial role in improving metal product quality and production efficiency in various manufacturing and 3-D printing factories. Metal defect detection faces scale variation and irregular shapes, which limit the adaptability of general object detection models in industrial scenarios. Deep learning-based methods are widely used for metal surface defect detection due to their strong adaptability and high automation. Yet, their existing studies pay limited attention to adaptability, evaluation, and recommendations across different detection methods for metal surface defects. This work mainly discusses YOLO, R-CNN, and transformers, as well as FPN, and analyzes their applications in metal surface defect detection according to their respective characteristics, to provide guidance for future research. YOLO has advantages in real-time industrial online detection, while R-CNN and transformer models show potential advantages in handling complex defect cases. Additionally, this work summarizes commonly used datasets and evaluation metrics for metal surface defect detection and analyzes the benchmark performance of different types of detection methods. It also discusses future research directions, including the current status and improvement paths of different models in terms of accuracy, real-time performance, and adaptability. Future models should focus on balancing accuracy and real-time performance, exploring new hybrid architectures, and improving adaptability to different metal surface defects to support further development in this field. Full article

Dendrobium Sonia orchid cultivation constitutes a vital commercial industry in Thailand; however, production remains persistently threatened by fungal and bacterial diseases. This study proposes a robust automated framework for orchid disease classification under conditions characterized by high visual uncertainty. A comparative analysis was conducted across four Convolutional Neural Network (CNN) architectures: ResNet-50 and three lightweight counterparts—MobileNetV3-Large, EfficientNetV2-B0, and NASNet-Mobile. All models were optimized using transfer learning, Cosine Decay scheduling, and EarlyStopping on a real-world dataset acquired from commercial orchid farms in Thailand. Experimental results indicate that ResNet-50 attained the highest overall performance (Accuracy: 98.96%, Macro F1: 0.9894, AUC-ROC: 0.9996), while EfficientNetV2-B0 achieved comparable results among the lightweight architectures (Accuracy: 98.47%, Macro F1: 0.9846, AUC-ROC: 0.9985). Importantly, statistical evaluation using the Wilcoxon Signed-Rank Test across five independent trials revealed no statistically significant difference between ResNet-50 and all three lightweight models (p > 0.05). This confirms the practical viability of deploying compact architectures on mobile platforms within smart farming systems without sacrificing diagnostic accuracy. Moreover, integrating Grad-CAM++ enhances interpretability by producing visual explanations that align with expert pathological assessments. This transparency effectively mitigates decision-making ambiguity and strengthens farmer confidence in adopting AI-driven precision agriculture. Full article

Accurate brain tumor segmentation from magnetic resonance imaging (MRI) remains challenging due to heterogeneous tumor morphology, intensity variability, and multi-scale structural complexity. This study proposes a DeepLabV3+-based segmentation framework integrating a VGG19-UNet encoder, Atrous Spatial Pyramid Pooling (ASPP), and low-level feature refinement to simultaneously capture hierarchical semantics and boundary-sensitive spatial details. The architecture enhances receptive field coverage without additional downsampling while preserving fine-grained contour information during reconstruction. Extensive evaluation was conducted on the Figshare Brain Tumor Segmentation (FBTS) dataset and the BraTS 2021 and BraTS 2018 benchmarks, focusing on Whole Tumor segmentation across multiple MRI modalities and tumor grades. Under five-fold cross-validation, the proposed model achieved a mean Dice Similarity Coefficient of 0.9717 and Jaccard Index of 0.9456 on FBTS, with stable and competitive performance across FLAIR, T1, T2, and T1CE modalities in both HGG and LGG cases. Boundary-level analysis further confirmed controlled Hausdorff Distance and low Average Symmetric Surface Distance. Statistical validation and ablation analysis demonstrate consistent improvements over baseline U-Net configurations. The proposed framework provides a robust and computationally efficient solution for automated brain tumor segmentation across heterogeneous datasets. Full article

Accurate time synchronization is essential in distributed electrical signal monitoring, where phase coherence and event correlation depend on precise timing agreement between acquisition nodes. Conventional approaches often rely on a single synchronization source, typically internet-based Network Time Protocol (NTP) or GPS-disciplined clocks, which is impractical in isolated, offline, or cost-sensitive scenarios. This paper introduces an autonomous offline synchronization architecture for multi-node monitoring systems built on Raspberry Pi 5 (RPI5) platforms connected to a private Ethernet network. Instead of depending on one timing method, the system integrates several complementary mechanisms: battery-backed RTC persistence via the J5 interface, deterministic orchestration through systemd services, automated boot time recovery, chrony-managed NTP discipline, and Precision Time Protocol (PTP) hardware timestamping using PTP Hardware Clock (PHC). Synchronization performance is validated through continuous multi-day measurements of long-term stability, inter-node phase coherence, and short-term jitter. Controlled power-loss scenarios are also included to verify recovery behavior. The system maintains sub-microsecond alignment between nodes using only commodity hardware and no external time source. To further confirm inter-node timestamp alignment at the signal level, both hardware-based reference signal injection and software-based synchronized signal emulation are employed, providing ground-truth validation alongside scalable and reproducible evaluation. The results show that low-cost embedded hardware can support reliable, long-duration synchronization in fully offline installations. Full article

Breast cancer histopathological classification is critical for diagnosis and treatment planning, yet manual assessment remains time-consuming and subject to inter-observer variability. Although deep learning approaches have advanced automated analysis, image-level data splitting may introduce data leakage, and spatial-domain architectures lack explicit multi-scale frequency modeling. This study proposes MSWA-ResNet, a Multi-Scale Wavelet Attention Residual Network that embeds recursive discrete wavelet decomposition within residual blocks to enable frequency-aware and scale-aware feature learning. The model is evaluated on the BreakHis dataset using a strict patient-level protocol with 70/30 patient-wise splitting, five-fold stratified cross-validation, ensemble prediction, and hierarchical aggregation from patch to patient level. MSWA-ResNet achieves 96% patient-level accuracy at 100×, 200×, and 400× magnifications, and 92% at 40×, with F1-scores of 0.97 and 0.94, respectively. At 200× and 400×, accuracy improves from 0.92 to 0.96 and F1-score from 0.94 to 0.97 over baseline CNNs while maintaining 11.8–12.1 M parameters and 2.5–4.8 ms inference time. Grad-CAM demonstrates improved localization of diagnostically relevant regions, indicating that explicit multi-scale frequency modeling enhances accurate and interpretable patient-level classification. Full article

Kolmogorov–Arnold Networks (KANs) have recently gained increasing attention as an alternative to conventional neural architectures, mainly because they replace fixed activation functions with learnable univariate mappings defined along network edges. This design not only increases modeling flexibility but also makes it easier to interpret how inputs are transformed within the network while maintaining parameter efficiency. KANs are particularly well suited for sensor-driven systems where transparency, robustness, and computational constraints are critical. This study provides a survey of KAN-based approaches for processing sensor data. A literature review conducted from 2024 to 2026 examined the deployment of KAN models in industrial and mechanical sensing, medical and biomedical sensing, and remote sensing and environmental monitoring, utilizing a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-based methodology. We first revisit the theoretical foundations of KANs and their main architectural variants, including spline-based, polynomial-based, monotonic, and hybrid formulations, to structure the discussion. From a practical standpoint, we then examine how KAN modules are integrated into modern deep learning pipelines, such as convolutional, recurrent, transformer-based, graph-based, and physics-informed architectures. KAN-based models demonstrate comparable predictive performance as conventional machine learning models, while having fewer parameters and more interpretable representations. Several limitations persist, including computational overhead, sensitivity to noisy signals, and resource-constrained device deployment challenges. Real-world sensor systems encounter significant challenges in adopting KAN-based models, including scalability in large-scale sensor networks, integration with hardware architectures, automated model development, resilience to out-of-distribution conditions, and the need for standardized evaluation metrics. Collectively, these observations provide a clearer understanding of the current and potential limitations of KAN-based models, offering practical guidance on the development of interpretable and efficient learning systems for future sensor equipment applications. Full article

Recent automated search methods have improved lower bounds for several Ramsey numbers, but the strongest gains often depend on structured seeding and cell-specific heuristic discovery. This leaves open a more fundamental question: Can a useful search structure be transferred across related Ramsey cells rather than rediscovered independently for each target instance? This work proposes a teacher–student framework for transferable structural search in Ramsey graph construction, inspired by the structure-distillation logic of Physics Structure-Informed Neural Networks (Ψ-NNs). The framework builds compressed structural representations from teacher witnesses and search traces, extracts reusable motifs and relations, and reconstructs transfer candidates. These are refined by balanced search and, for weak R(3, s) cells, by exact small-cell supervision. The framework is evaluated as a proof of concept across five Ramsey cells under transfer, matched-compute, search, ablation, and interpretability settings, including a proportional shift-scaling baseline and a greedy triangle-closing baseline that probe the structure-validity frontier from complementary directions. Supplementary experiments cover seed robustness, budget sensitivity, transfer-neighborhood variation, structural-resolution changes, stronger exact supervision, cross-r teacher pooling, single-teacher configurations, and scaling behavior across graph sizes. The results show that the portfolio version of the framework is the strongest balanced transfer method in the current study, while a structure-dominant oracle achieves stronger witness-shape agreement but worse Ramsey-valid construction. These findings reveal a clear structure-validity frontier and suggest that transferable Ramsey search should be evaluated by how well structural priors survive the validity constraints of new cells. Full article