Search Results (19,889)

Postprandial plasma amino acid (AA) kinetics serve as essential indicators of digestive efficiency and systemic metabolic status in pigs. Traditional kinetic analysis relies on Non-Linear Least Squares (NLS) regression using compartmental models, yet these methods typically demand repeated blood sampling and precise initialization to ensure convergence. In this study, we developed a Physics-Informed Neural Network (PINN) framework by integrating mechanistic Ordinary Differential Equations (ODEs) directly into the deep learning loss function. The framework was evaluated using a benchmark dataset. Specifically, we performed a retrospective analysis by downsampling the original high-frequency data to simulate dense and sparse sampling strategies. The results demonstrate that while both models exhibit high fidelity under dense sampling, PINN maintains superior robustness and predictive accuracy under data-constrained conditions. Under the sparse sampling scenario, PINN reduced the Root Mean Square Error (RMSE) compared to NLS in key metabolic profiles, such as Methionine in the FAA group (p < 0.01) and Lysine in the HYD group (p < 0.05). Unlike NLS, which is sensitive to initial guesses, PINN successfully utilized physical laws as a regularization term to robustly solve the inverse problem, demonstrating superior parameter identification stability and predictive consistency under data-constrained conditions compared to NLS. We concluded that the PINN framework provides a reliable and consistent alternative for modeling the AA dynamics. In the future, it may be possible to reconstruct highly accurate physiological trajectories under optimized sparse sampling conditions. Full article

Quantum Key Distribution (QKD) networks guarantee information-theoretical security of exchanged keys, but key rates are still limited. This makes efficient and adaptive routing a critical challenge, especially in meshed topologies without quantum repeaters. Conventional shortest path routing approaches struggle to cope with dynamic key store filling levels and changes in network topologies, which leads to load imbalance and blocked connections. In this work, we propose an adaptive routing framework based on Deep Reinforcement Learning (DRL) for hop-wise end-to-end routing in unknown meshed QKD networks. The agent leverages Graph Attention Networks (GATs) to process the network states of varying topologies, enabling generalization across previously unseen meshed networks without topology-specific retraining. The agent is trained on random graphs with 10 to 20 nodes and learns a routing policy that explicitly balances key consumption across the network by utilizing a reward function that is based on the entropy of key store filling levels. We evaluate the proposed approach on the 14-node NSFNET topology under time-varying traffic demands. Simulation results demonstrate that the DRL-based routing significantly outperforms hop-based and weighted shortest path benchmarks, achieving up to a 18.7% increase in mean key store filling levels while completely avoiding key store depletion. These results highlight the potential of graph-based DRL methods for scalable, adaptive, and resource-efficient routing in future QKD networks. Full article

This study introduces a novel four-parameter lifetime distribution constructed within the Topp–Leone Power Gompertz framework. Owing to its flexible structure, the proposed model accommodates a wide range of density shapes and hazard-rate patterns, including increasing, decreasing, bathtub-shaped, unimodal, and other non-monotone behaviors. Key distributional properties, including moments, entropy-based measures, quantile-based measures, and order statistics, are derived. Parameter inference is conducted using both likelihood-based and Bayesian approaches, and the finite-sample performance of the re-sulting estimators is assessed via Monte Carlo simulations. The practical relevance of the proposed distribution is illustrated using two real datasets and benchmarked against several competing lifetime models, including the Gompertz, Power Gompertz, Weibull, Topp–Leone Gompertz, Marshall–Olkin Gompertz, and Exponentiated Gompertz distri-butions. Overall, the comparative analyses demonstrate the superior fitting performance of the proposed model, highlighting its effectiveness for complex reliability, survival, and pharmacokinetic data. Full article

Background: Navigational bronchoscopy (NB) enables precise sampling of peripheral and central pulmonary nodules using shape-sensing or electromagnetic guidance. A major challenge is anesthesia-induced atelectasis, which alters lung anatomy, reduces registration accuracy, and is known to lower diagnostic accuracy. To counteract this, ventilatory protocols such as the Ventilatory Strategy to Prevent Atelectasis (VESPA) and the Lung Navigation Ventilation Protocol (LNVP) have been recommended. Their adoption and clinical impact, however, remain uncertain. Methods: We conducted a retrospective review of 224 consecutive NB procedures performed under general anesthesia at a single academic medical center (January 2020–August 2024). Demographic, anesthetic, and ventilatory data were extracted from electronic records. Outcomes included navigational success (ability to reach the lesion) and diagnostic accuracy (concordance between bronchoscopic diagnosis and final clinical diagnosis after follow-up). Ventilatory practices were compared with published VESPA and LNVP recommendations. Results: Navigational success, defined as successful advancement of the bronchoscope to the target lesion with tissue acquisition, was achieved in 89.2% of cases. Overall diagnostic accuracy, defined as concordance between bronchoscopic diagnosis and final clinical diagnosis after follow-up, was 81.7%. Ventilatory management consistently diverged from recommended protocols. Most patients were ventilated with FiO₂ > 0.6, PEEP in the range of 7–10 cm H₂O, and tidal volumes of 300–500 mL. The only recommended maneuver systematically applied was recruitment immediately after intubation. Despite widespread deviation from both VESPA and LNVP, diagnostic performance remained favorable relative to published benchmarks. No major anesthesia-related complications occurred. Conclusions: In this retrospective series, navigational success comparable to published studies that adapted strict ventilation protocols was achieved with also comparable diagnostic accuracy without strict adherence to predefined ventilatory strategies. Recruitment maneuvers may represent the most influential component of current protocols, but institutional factors such as procedural expertise and case volume likely contributed to outcomes. Prospective studies are warranted to determine whether standardized ventilatory protocols are necessary for optimizing NB performance. Full article

The planned inclusion of the steel and electrolytic aluminum sectors into China’s Carbon Emission Allowance (CEA) market—initially limited to thermal power since 2021—will expand its coverage to approximately 70% of national carbon emissions, significantly influencing carbon pricing. This study employs a Variational Mode Decomposition–Support Vector Regression (VMD-SVR) model to forecast carbon price fluctuations under three post-expansion scenarios. The results indicate that, in addition to quota allocations, factors such as sectoral emission scales, the CSI 300 Power Index, and the Shanghai Energy Price Index substantially affect price trends. While market expansion induces a short-term price increase, it also stabilizes prices by reducing volatility. Furthermore, different quota allocation methods yield distinct outcomes: equal allocation facilitates a smoother market transition, whereas benchmarking provides stronger incentives for emissions reductions. Full article

Large Language Models (LLMs) have acquired vast amounts of knowledge during pre-training. However, there are a lot of challenges when it is deployed in real-world applications, such as poor interpretability, hallucinations, and the inability to reference private data. To address these issues, Retrieval-Augmented Generation (RAG) has been proposed. Traditional RAG relying on text-based retrievers often converts documents using Optical Character Recognition (OCR) before retrieval. While testing has revealed that it tends to overlook tables and images contained within the documents. RAG, relying on vision-based retrievers, often loses information on text-dense pages. To address these limitations, we propose DRAG: Dual-channel Retrieval-Augmented Generation for Hybrid-Modal Document Understanding, a novel retrieval paradigm. The DRAG method proposed in this paper primarily comprises two core improvements: first, a parallel dual-channel processing architecture is adopted to separately extract and preserve the visual structural information and deep semantic information of documents, thereby effectively enhancing information integrity; second, a novel dynamic weighted fusion mechanism is proposed to integrate the retrieval results from both channels, enabling precise screening of the most relevant information segments. Empirical results demonstrate that our method achieves Competitive performance across multiple general benchmarks. Furthermore, performance on biomedical datasets (e.g., BioM) specifically highlights its potential in specialized, vertical domains such as elderly care and rehabilitation, where documents are characterized by dense hybrid-modal information. Full article

Synthetic aperture radar (SAR) imagery is widely used for target detection in complex backgrounds and adverse weather conditions. However, high-precision detection of rotated small targets remains challenging due to severe speckle noise, significant scale variations, and the need for robust rotation-aware representations. To address these issues, we propose SAR-DRBNet, a high-precision rotated small-target detection framework built upon YOLOv13. First, we introduce a Detail-Enhanced Oriented Bounding Box detection head (DEOBB), which leverages multi-branch enhanced convolutions to strengthen fine-grained feature extraction and improve oriented bounding box regression, thereby enhancing rotation sensitivity and localization accuracy for small targets. Second, we design a Ck-MultiDilated Reparameterization Block (CkDRB) that captures multi-scale contextual cues and suppresses speckle interference via multi-branch dilated convolutions and an efficient reparameterization strategy. Third, we propose a Dynamic Feature Weaving module (DynWeave) that integrates global–local dual attention with dynamic large-kernel convolutions to adaptively fuse features across scales and orientations, improving robustness in cluttered SAR scenes. Extensive experiments on three widely used SAR rotated object detection benchmarks (HRSID, RSDD-SAR, and DSSDD) demonstrate that SAR-DRBNet achieves a strong balance between detection accuracy and computational efficiency compared with state-of-the-art oriented bounding box detectors, while exhibiting superior cross-dataset generalization. These results indicate that SAR-DRBNet provides an effective and reliable solution for rotated small-target detection in SAR imagery. Full article

In the pursuit of an efficient 6G network that achieves an enhanced capacity with minimal power consumption, reconfigurable intelligent surfaces (RISs) and cell-free (CF) massive multiple-input-multiple-output (MIMO) networks emerge as two key technologies. This paper investigates an active-RIS-aided CF massive MIMO system under imperfect channel state information (CSI) and proposes a two-step optimization algorithm to address the max-min achievable rate problem. Given that the original problem is non-convex, we decompose it into two subproblems, which allows us to optimize the AP transmit beamforming and the RIS reflecting precoding, respectively, in an alternating manner. Simulation results demonstrate the superiority of the proposed scheme over existing benchmarks, achieving significant performance gains in active-RIS-aided CF massive MIMO systems. Full article

Sound source localization in dynamic environments with multiple moving speakers presents significant challenges due to reverberation, noise, and unknown source counts. To address these issues, this paper proposes an integrated deep-learning framework combining spatial spectrum estimation with blind source detection. The method employs a causal convolution–recurrent network (SRP-DPCRN) to extract robust spatial features from multichannel audio signals under adverse acoustic conditions. Subsequently, an iterative attention-based detection network (IASDNet) automatically identifies active sources from the estimated spatial spectrum without requiring prior knowledge of source quantity. Evaluated on both simulated datasets and the real-recorded LOCATA benchmark, the proposed system demonstrates superior performance in multi-source tracking scenarios, achieving an average detection accuracy of 96% with mean angular error below 3.5 degrees. The results confirm that joint optimization of feature learning and source counting provides an effective solution for blind localization in practical applications, significantly outperforming conventional and deep-learning baselines. Full article

Truck platooning enabled by V2X and cooperative driving can reduce aerodynamic drag and consequently decrease fuel consumption and CO₂ emissions. Meanwhile, hub-and-spoke courier networks require strategic decisions on hub locations, allocation, and line-haul routing. This paper introduces an integrated Hub Location-Platoon Routing Problem (HLPRP) that jointly optimizes (i) hub selection and single allocation of spokes; (ii) the departure hubs where platoons are formed; (iii) line-haul (inter-hub) service design and route selection; and (iv) demand routing, while internalizing monetized carbon benefits from platooning. A variable neighborhood search-based simulated annealing solution framework is developed to eliminate duplicated hub pair representations induced by network symmetry. Computational experiments on benchmark and large-scale North China instances demonstrate that the proposed approach consistently produces high-quality solutions within practical runtimes. The results indicate that the optimal network structure is primarily driven by transportation cost trade-offs and is further shaped by platoon-enabling investment and the associated carbon benefit, which concentrates on a subset of high-volume inter-hub corridors. Overall, the proposed framework provides a decision support approach for designing low-carbon courier line-haul networks. Full article

Accurate six-degree-of-freedom (6-DoF) visual localization is a fundamental component for modern mapping and navigation. While recent data-centric approaches have leveraged Novel View Synthesis (NVS) to augment training datasets, these methods typically rely on uniform grid-based sampling of virtual cameras. Such naive placement often yields redundant or weakly informative views, failing to effectively bridge the gap between sparse, unordered captures and dense scene geometry. To address these challenges, we present LEGS (Visual Localization Enhanced by 3D Gaussian Splatting), a trajectory-agnostic synthetic-view augmentation framework. LEGS constructs a joint set of 6-DoF camera pose proposals by integrating a coarse 3D lattice with the Structure-from-Motion (SfM) camera graph, followed by a visibility-aware, coverage-driven selection strategy. By utilizing 3D Gaussian Splatting (3DGS), our framework enables high-throughput, scene-specific synthesis within practical computational budgets. Experiments on standard benchmarks and an in-house dataset demonstrate that LEGS consistently improves pose accuracy and robustness, particularly in scenarios characterized by sparse sampling and co-located viewpoints. Full article

Apple leaf disease identification with high precision is one of the main challenges in precision agriculture. The datasets usually have class imbalance problems and environmental changes, which negatively impact deep learning approaches. In this paper, an ablation study is proposed to test three different scenarios: V1, a hybrid balanced dataset consisting of 10,028 images; V2, an imbalanced dataset as a baseline consisting of 14,582 original images; and V3, a 3× physical augmentation approach based on the 14,582 images. The classification performance of YOLOv11x was benchmarked against three state-of-the-art CNN architectures: ResNet-152, DenseNet-201, and EfficientNet-B1. The methodology incorporates controlled downsampling for dominant classes alongside scenario-based augmentation for minority classes, utilizing CLAHE-based texture enhancement, illumination simulation, and sensor noise generation. All the models were trained for up to 100 epochs under identical experimental conditions, with early stopping based on validation performance and an 80/20 train-validation split. The experimental results demonstrate that the impact of balancing strategies is model-dependent and does not universally improve performance. This highlights the importance of aligning data balancing strategies with architectural characteristics rather than applying uniform resampling approaches. YOLOv11x achieved its peak accuracy of 99.18% within the V3 configuration, marking a +0.62% improvement over the V2 baseline (99.01%). In contrast, EfficientNet-B1 reached its optimal performance in the V2 configuration (98.43%) without additional intervention. While all the models exhibited consistently high AUC values (≥99.94%), DenseNet-201 achieved the highest value (99.97%) across both V2 and V3 configurations. In fine-grained discrimination, the superior performance of YOLOv11x on challenging cases is verified, with only one incorrect classification (Rust to Scab), while ResNet-152 and DenseNet-201 incorrectly classified eight and seven samples, respectively. Degradation sensitivity analysis under controlled Gaussian noise and motion blur indicated that CNN baseline models maintained stable performance. High minority-class reliability, including a 96.20% F1-score for Grey Spot and 100% precision for Mosaic, further demonstrates effective fine-grained discrimination. Results indicate that data preservation with physically inspired augmentation (V3) is better than resampling-based balancing (V1), especially in terms of global accuracy and minority-class performance. Full article

Scalable video coding plays an essential role in supporting heterogeneous devices, network conditions, and application requirements in modern video streaming systems. However, most existing scalable coding approaches primarily optimize human perceptual quality and provide limited support for data privacy, as well as for machine analyses and the integration of heterogeneous sensor data. This limitation motivated the development of adaptive scalable video coding frameworks. The proposed approach is designed to serve both human viewers and automated analysis systems while ensuring high security and compression efficiency. The method adaptively encrypts selected layers during transmission to protect sensitive content without degrading decoding or analysis performance. Experimental evaluations on benchmark datasets demonstrate that the proposed framework achieves superior rate distortion efficiency and reconstruction quality, while also improving machine analysis accuracy compared to existing traditional and learning-based codes. In video surveillance scenarios, where the base layer is preserved for analysis, the proposed scalable human machine coding (SHMC) method outperforms scalable extensions of H.265/High Efficiency Video Coding (HEVC), Scalable High Efficiency Video Coding (SHVC), reducing the average bit-per-pixel (bpp) by 26.38%, 30.76%, and 60.29% at equivalent mean Average Precision (mAP), Peak Signal-to-Noise Ratio (PSNR), and Multi-Scale Structural Similarity (MS-SSIM) levels. These results confirm the effectiveness of integrating scalable video coding with intelligent encryption for secure and efficient video transmission. Full article

Text-attributed graphs (TAGs) couple graph topology with node-level text, but real data often contain spurious edges, missing links, and text–structure mismatch that destabilize learning under scarce labels. We propose CoACL (Coupled Augmentation for Contrastive Learning), a framework that uses LLM semantic supervision to denoise structural and textual information and alleviate data sparsity. CoACL first prunes the candidate edge space using structural similarity and then queries an LLM to discard suspicious edges and confirm plausible links, yielding semantically consistent positive and negative pairs. We further introduce keyword-focused text augmentations and learn coupled representations by optimizing a joint text–graph contrastive objective guided by semantics. Experiment results on Cora, PubMed, and the Open Graph Benchmark Arxiv dataset (OGBN-Arxiv) show that CoACL consistently outperforms strong baselines and yields up to 7.1% absolute improvement in node classification accuracy, with the largest gains in low-label regimes. By constraining LLM evaluation to similarity-based candidates, CoACL targets neighborhood-level noise with controlled cost. Full article

In previous studies, fire detection models that only differentiate between fire and smoke are presented. However, a high false detection rate occurs because of confusion between smoke and fog. In this study, a novel method is proposed to effectively distinguish between smoke and fog. A custom dataset is introduced for detecting fire, smoke, and fog, which offers a novel labeling technique to reduce the misclassification of smoke and fog. A new architecture is proposed that is aimed at enhancing the detection performance. The latest You Only Look Once version 11 (YOLOv11) is used to establish a performance baseline. Further research then focuses on improving this benchmark. To ensure accurate texture differentiation, the dataset is designed to exclude overlapping ground-truth boxes, enabling the trained model to determine object boundaries independently, which is a design approach not found in previous studies. The model design aims to maximize performance and cost efficiency. The performance is improved by adding an image pyramid layer to the existing model to improve large-object detection. Cost efficiency is improved by designing a new module, C5Go, to mitigate the additional computational load introduced by the added pyramid level. Comparative experiments on the proposed custom fire dataset demonstrate that the proposed model improves the detection performance while keeping the additional computational overhead modest. The experimental results show that the proposed model achieves a 12.86% improvement in smoke detection performance compared with YOLOv11 and attains an overall mean average precision (mAP)@50 score of 0.906, reflecting superior performance. The contributions of this work are as follows: (i) we design a three-class dataset comprising fire, smoke, and fog so as not to cause false detections; (ii) we propose a new model structure to improve performance; and (iii) we verify that the proposed method indeed improves performance through the experimental results. Full article