Search Results (135)

Genetic variability in terrestrial mammals is essential for understanding population and evolutionary dynamics, as well as for establishing effective strategies in conservation biology. This comprehensive review aimed to critically analyze invasive and non-invasive techniques used to assess genetic variability in wild terrestrial mammals. Using the PICO (Population, Intervention, Comparison, Outcome) format and following PRISMA guidelines, a comprehensive literature search was conducted in Web of Science, Scopus and Science Direct databases, including articles published in English from January 2015 to April 2025. Thirty-one experimental studies were selected that met specific criteria related to genetic evaluation using invasive (direct blood or tissue collection) and non-invasive (stool, hair and saliva collection) techniques. The results indicate that invasive techniques provide samples of high genetic quality, albeit with important ethical and animal welfare considerations. In contrast, non-invasive techniques offer less disruptive methods, although they present significant challenges in terms of quantity and purity of DNA obtained, potentially affecting the accuracy and confidence of genetic analysis. Detailed analysis of selected studies showed diverse patterns of heterozygosity and inbreeding coefficients between different taxonomic orders (Carnivora, Artiodactyla, Proboscidea, Primates and Rodentia). In addition, the main anthropogenic threats and current conservation strategies implemented in different species were identified. An overall genetic variability ranging from high to moderate was observed, with large species being more vulnerable to genetic reduction due to changes in habitat and human activities. Rather than a static comparison, our synthesis traces a clear methodological arc from small short tandem repeats (STR, or microsatellites) panels towards SNP-based approaches enabled by next-generation sequencing, including reduced representation (ddRAD), amplicon panels (GT-seq), and hybridisation capture tailored to degraded DNA from hair, faeces, and environmental substrates. Over 2015–2025, study designs shifted from presence/absence and coarse diversity estimates to robust inference of relatedness, assignment, effective population size, and gene flow using hundreds–thousands of SNPs and genotype-likelihood frameworks tolerant of allelic dropout and low coverage. Laboratory practice converged on multi-tube replication, synthetic blocking oligos, and capture-based enrichment; bioinformatics adopted probabilistic genotype calling, error-aware filtering, and replication-based consensus. This review provides a solid basis for optimizing genetic sampling methods, allowing for more ethical and efficient studies. Furthermore, it contributes to strengthening conservation strategies by underlining the importance of adapting the sampling method to the biological and ecological particularities of each species studied. Ultimately, these findings can significantly improve genetic conservation decision-making, benefiting the sustainability and resilience of wild land mammal populations. Full article

Background: Multimodal Deep learning has emerged as a crucial method for automated patient-specific quality assurance (PSQA) in radiotherapy research. Integrating image-based dose matrices with tabular plan complexity metrics enables more accurate prediction of quality indicators, including the Gamma Passing Rate (GPR) and dose difference (DD). However, modality imbalance remains a significant challenge, as tabular encoders often dominate training, suppressing image encoders and reducing model robustness. This issue becomes more pronounced under task heterogeneity, with GPR prediction relying more on tabular data, whereas dose difference prediction (DDP) depends heavily on image features. Methods: We propose BMMQA (Balanced Multi-modal Quality Assurance), a novel framework that achieves modality balance by adjusting modality-specific loss factors to control convergence dynamics. The framework introduces four key innovations: (1) task-specific fusion strategies (softmax-weighted attention for GPR regression and spatial cascading for DD prediction); (2) a balancing mechanism supported by Shapley values to quantify modality contributions; (3) a fast network forward mechanism for efficient computation of different modality combinations; and (4) a modality-contribution-based task weighting scheme for multi-task multimodal learning. A large-scale multimodal dataset comprising 1370 IMRT plans was curated in collaboration with Peking Union Medical College Hospital (PUMCH). Results: Experimental results demonstrate that, under the standard 2%/3 mm GPR criterion, BMMQA outperforms existing fusion baselines. Under the stricter 2%/2 mm criterion, it achieves a 15.7% reduction in mean absolute error (MAE). The framework also enhances robustness in critical failure cases (GPR < 90%) and achieves a peak SSIM of 0.964 in dose distribution prediction. Conclusions: Explicit modality balancing improves predictive accuracy and strengthens clinical trustworthiness by mitigating overreliance on a single modality. This work highlights the importance of addressing modality imbalance for building trustworthy and robust AI systems in PSQA and establishes a pioneering framework for multi-task multimodal learning. Full article

In massive multiple-input and multiple-output (mMIMO) systems, superimposed channel state information (CSI) feedback is developed to improve the occupation of uplink bandwidth resources. Nevertheless, the interference from this superimposed mode degrades the recovery performance of both downlink CSI and uplink data sequences. Although machine learning (ML)-based methods effectively mitigate superimposed interference by leveraging the multi-domain features of downlink CSI, the complex interactions among network model parameters cause a significant burden on system resources. To address these issues, inspired by sensing-assisted communication, we propose a novel superimposed CSI feedback method assisted by inactive sensing information that previously existed but was not utilized at the base station (BS). To the best of our knowledge, this is the first time that inactive sensing information is utilized to enhance superimposed CSI feedback. In this method, a new type of modal data, different from communication data, is developed to aid in interference suppression without requiring additional hardware at the BS. Specifically, the proposed method utilizes location, speed, and path information extracted from sensing devices to derive prior information. Then, based on the derived prior information, denoising processing is applied to both the delay and Doppler dimensions of downlink CSI in the delay—Doppler (DD) domain, significantly enhancing the recovery accuracy. Simulation results demonstrate the performance improvement of downlink CSI and uplink data sequences when compared to both classic and novel superimposed CSI feedback methods. Moreover, against parameter variations, simulation results also validate the robustness of the proposed method. Full article

The increasing integration of renewable energy sources (RES) and rising energy demand have created challenges in maintaining stability in interconnected power systems, particularly in terms of frequency, voltage, and tie-line power. While traditional load frequency control (LFC) and automatic voltage regulation (AVR) strategies have been widely studied, they often fail to address the complexities introduced by RES and nonlinear system dynamics such as boiler dynamics, governor deadband, and generation rate constraints. This study introduces the Arithmetic Optimization Algorithm (AOA)-optimized PI(1+DD) controller, chosen for its ability to effectively optimize control parameters in highly nonlinear and dynamic environments. AOA, a novel metaheuristic technique, was selected due to its robustness, efficiency in exploring large search spaces, and ability to converge to optimal solutions even in the presence of complex system dynamics. The proposed controller outperforms classical methods such as PI, PID, I–P, I–PD, and PI–PD in terms of key performance metrics, achieving a settling time of 7.5 s (compared to 10.5 s for PI), overshoot of 2.8% (compared to 5.2% for PI), rise time of 0.7 s (compared to 1.2 s for PI), and steady-state error of 0.05% (compared to 0.3% for PI). Additionally, sensitivity analysis confirms the robustness of the AOA-optimized controller under ±25% variations in turbine and speed control parameters, as well as in the presence of nonlinearities, demonstrating its potential as a reliable solution for improving grid performance in complex, nonlinear multi-area interconnected power systems. Full article

Stomata play a crucial role in plant immune responses, with their morphological characteristics closely linked to disease resistance. Accurate detection and analysis of stomatal phenotypic parameters are essential for soybean disease resistance research and variety breeding. However, traditional stoma detection methods are challenged by complex backgrounds and leaf vein structures in soybean images. To address these issues, we proposed a Soybean Stoma-YOLO (You Only Look Once) model (SS-YOLO) by incorporating large separable kernel attention (LSKA) in the Spatial Pyramid Pooling-Fast (SPPF) module of YOLOv8 and Deformable Large Kernel Attention (DLKA) in the Neck part. These architectural modifications enhanced YOLOV8′s ability to extract multi-scale and irregular stomatal features, thus improving detection accuracy. Experimental results showed that SS-YOLO achieved a detection accuracy of 98.7%. SS-YOLO can effectively extract the stomatal features (e.g., length, width, area, and orientation) and calculate related indices (e.g., density, area ratio, variance, and distribution). Across different soybean rust disease stages, the variety Dandou21 (DD21) exhibited less variation in length, width, area, and orientation compared with Fudou9 (FD9) and Huaixian5 (HX5). Furthermore, DD21 demonstrated greater uniformity in stomatal distribution (SEve: 1.02–1.08) and a stable stomatal area ratio (0.06–0.09). The analysis results indicate that DD21 maintained stable stomatal morphology with rust disease resistance. This study demonstrates that SS-YOLO significantly improved stoma detection and provided valuable insights into the relationship between stomatal characteristics and soybean disease resistance, offering a novel approach for breeding and plant disease resistance research. Full article

Recently, studies on driver inattention state recognition as an advanced mobility application technology are being actively conducted to prevent traffic accidents caused by driver drowsiness and distraction. The driver inattention state recognition system is a technology that recognizes drowsiness and distraction by using driver behavior, biosignals, and vehicle data characteristics. Existing driver drowsiness detection systems are wearable accessories that have partial occlusion of facial features and light scattering due to changes in internal and external lighting, which results in momentary image resolution degradation, making it difficult to recognize the driver’s condition. In this paper, we propose a transformer based spatial temporal fusion network (STFTransNet) that fuses multi-modality information for improved driver inattention state recognition in images where the driver’s face is partially occluded by wearing accessories and the instantaneous resolution is degraded due to light scattering from changes in lighting in a driving environment. The proposed STFTransNet consists of (i) a mediapipe face mesh-based facial landmark extraction process for facial feature extraction, (ii) an RCN-based two-stream cross-attention process for learning spatial features of driver face and body action images, (iii) a TCN-based temporal feature extraction process for learning temporal features of extracted features, and (iv) an ensemble of spatial and temporal features and a classification process to recognize the final driver state. As a result of the experiment, the proposed STFTransNet achieved an accuracy of 4.56% better than the existing VBFLLFA model in the NTHU-DDD public DB, 3.48% better than the existing InceptionV3 + HRNN model in the StateFarm public DB, and 3.78% better than the existing VBFLLFA model in the YawDD public DB. The proposed STFTransNet is designed as a two-stream network that can input the driver’s face and action images and solves the degradation in driver inattention state recognition performance due to partial facial feature occlusion and light blur through spatial feature and temporal feature fusion. Full article

Background: Electrocardiogram (ECG) signals are crucial for cardiovascular diagnosis, but their analysis face challenges from noise contamination, compression difficulties due to their non-stationary nature, and the inherent complexity of its morphological components, particularly for low-amplitude P- and T-waves obscured by noise. Methodology: This study proposes a novel, multi-stage framework for ECG signal denoising, compressing, and component decomposition. The proposed framework leverages the sparsity of ECG signal to denoise and compress these signals using autoencoder-driven dictionary (AE-DD) with matching pursuit. In this work, a data-driven dictionary was developed using a regularized autoencoder. Appropriate trained weights along with matching pursuit were used to compress the denoised ECG segments. This study explored different weight regularization techniques: L1- and L2-regularization. Results: The proposed framework achieves remarkable performance in simultaneous ECG denoising, compression, and morphological decomposition. The L1-DAE model delivers superior noise suppression (SNR improvement up to 18.6 dB at $-3$ dB input SNR) and near-lossless reconstruction ($\mathrm{MSE}<{10}^{-5}$). The L1-AE dictionary enables high-fidelity compression (CR = 28:1 ratio, $\mathrm{MSE}\approx 0.58\times {10}^{-5}$, PRD = 2.1%), outperforming non-regularized models and traditional dictionaries (DCT/wavelets), while its trained weights naturally decompose into interpretable sub-dictionaries for P-wave, QRS complex, and T-wave enabling precise, label-free analysis of ECG components. Moreover, the learned sub-dictionaries naturally decompose into interpretable P-wave, QRS complex, and T-wave components with high accuracy, yielding strong correlation with the original ECG ($r=0.98$, $r=0.99$, and $r=0.95$, respectively) and very low MSE ($1.93\times {10}^{-5}$, $9.26\times {10}^{-4}$, and $3.38\times {10}^{-4}$, respectively). Conclusions: This study introduces a novel autoencoder-driven framework that simultaneously performs ECG denoising, compression, and morphological decomposition. By leveraging L1-regularized autoencoders with matching pursuit, the method effectively enhances signal quality while enabling direct decomposition of ECG signals into clinically relevant components without additional processing. This unified approach offers significant potential for improving automated ECG analysis and facilitating efficient long-term cardiac monitoring. Full article

The one-step method in reactor physics has become one of the important research directions in recent two decades. Based on the open-source OpenMOC code, the following work was carried out. Firstly, the global–local resonance method with multi-group and continuous neutron libraries was researched and established. Next, based on the 2D and 3D MOC solver, the 2D/1D and the MOC/DD transport methods were realized in OpenMOC. Finally, verification of the transport and resonance methods was conducted using the C5G7 macro benchmark and the VERA micro benchmark. The numerical results demonstrated that the average eigenvalue deviation was 44 pcm and average maximum pin power distribution deviation was 0.37% in the VERA-2 benchmark, which showed the good accuracy of the resonance method. As for the transport method, the 3DMOC method exhibited better accuracy in strong anisotropic cases, but the computational time was 38 times that of the 2D/1D method. Full article

Background/Objectives: Pneumonia is a critical lung infection that demands timely and precise diagnosis, particularly during the evaluation of chest X-ray images. Deep learning is widely used for pneumonia detection but faces challenges such as poor denoising, limited feature diversity, low interpretability, and class imbalance issues. This study aims to develop an optimized ResNet-50 based framework for accurate pneumonia detection. Methods: The proposed approach integrates Multiscale Curvelet Filtering with Directional Denoising (MCF-DD) as a preprocessing step to suppress noise while preserving diagnostic details. Multi-feature fusion is performed by combining deep features extracted from ResNet-50 with handcrafted texture descriptors such as Local Binary Patterns (LBPs), leveraging both semantic and structural information. Precision attention mechanisms are incorporated to enhance interpretability by highlighting diagnostically relevant regions. Results: Validation on the Kaggle chest radiograph dataset demonstrates that the proposed model achieves higher accuracy, sensitivity, specificity, and other performance metrics compared to existing methods. The inclusion of MCF-DD preprocessing, multi-feature fusion, and precision attention contributes to improved robustness and diagnostic reliability. Conclusions: The optimized ResNet-50 framework, enhanced by noise suppression, multi-feature fusion, and attention mechanisms, offers a more accurate and interpretable solution for pneumonia detection from chest X-ray images, addressing key challenges in existing deep learning approaches. Full article

This work presents an 80 Gbps photonics-aided millimeter-wave (mm Wave) wireless communication system employing 16-Quadrature Amplitude Modulation (16-QAM) and a 1 × 2 single-input multiple-output (SIMO) architecture with maximum ratio combining (MRC) to achieve robust 87.5 GHz transmission over 4.6 km. By utilizing polarization-diverse optical heterodyne generation and spatial diversity reception, the system enhances spectral efficiency while addressing the low signal-to-noise ratio (SNR) and channel distortions inherent in long-haul links. A blind equalization scheme combining the constant modulus algorithm (CMA) and decision-directed least mean squares (DD-LMS) filtering enables rapid convergence and suppresses residual inter-symbol interference, effectively mitigating polarization drift and phase noise. The experimental results demonstrate an SNR gain of approximately 3 dB and a significant bit error rate (BER) reduction with MRC compared to single-antenna reception, along with improved SNR performance in multi-antenna configurations. The synergy of photonic mm Wave generation, adaptive spatial diversity, and pilot-free digital signal processing (DSP) establishes a robust framework for high-capacity wireless fronthaul, overcoming atmospheric attenuation and dynamic impairments. This approach highlights the viability of 16-QAM in next-generation ultra-high-speed networks (6G/7G), balancing high data rates with resilient performance under channel degradation. Full article

This study assesses how different anode materials influence neutron production rates (NPRs) in multi-state fusion (MSF) reactors, with a particular focus on the effects of deuterium (D) pre-loading on the anode surface. Three types of mesh anodes were assessed: stainless steel (SS), zirconium (Zr), and D pre-loaded zirconium (ZrD). MSF operates using two electrodes to confine ions to various fusion reactions, including D-D and D-T. The reactor features a negatively biased central cathode and a grounded anode within a vacuum vessel. Neutrons and protons are produced through the application of high voltage (tens of kV) and current (tens of mA) on the system to spark the plasma and start the fusion. Assessments at voltages up to 50 kV and currents up to 30 mA showed that Zr mesh anodes produced higher NPRs than SS ones, reaching 1.912 at 30 kV. This increased performance is attributed to surface fusion processes occurring in the anode. These processes were further modified by the deuterium pre-loading in the ZrD anode, as compared to SS and Zr with 1.832 at 30 kV. The findings suggest that material properties and deuterium pre-loading play significant roles in optimizing the efficiency of MSF reactors and the NPR. Future research may explore the long-term stability and durability of these anode materials under continuous operation conditions to fully harness their potential in fusion energy applications. Full article

During unmanned drilling operations, the velocity of drill cuttings serves as an important indicator of drilling conditions, which necessitates real-time and accurate measurements. To address challenges such as the small size of cuttings, weak feature representations, and complex motion trajectories, we propose a novel velocity measurement method integrating small-object detection and tracking. Specifically, we enhance the multi-scale feature fusion capability of the YOLOv11 detection head by incorporating a lightweight feature extraction module, Ghost Conv, and a feature-aligned fusion module, FA-Concat, resulting in an improved model named YOLOv11-Dd (drilling debris). Furthermore, considering the robustness of the ByteTrack algorithm in retaining low-confidence targets and handling occlusions, we integrate ByteTrack into the tracking phase to enhance tracking stability. A velocity estimation module is introduced to achieve high-precision measurement by mapping the pixel displacement of detection box centers across consecutive frames to physical space. To facilitate model training and performance evaluation, we establish a drill-cutting splash simulation dataset comprising 3787 images, covering a diverse range of ejection angles, velocities, and material types. The experimental results show that the YOLOv11-Dd model achieves a 4.65% improvement in mAP@80 over YOLOv11, reaching 76.04%. For mAP@75–95, it improves by 0.79%, reaching 41.73%. The proposed velocity estimation method achieves an average accuracy of 92.12% in speed measurement tasks, representing a 0.42% improvement compared to the original YOLOv11. Full article

In the earthquakes that occurred in the Pazarcık (Mw = 7.7) and Elbistan (Mw = 7.6) districts of Kahramanmaraş Province on 6 February 2023, many buildings collapsed in the Hayrullah neighborhood of the Onikişubat district. In this study, we investigated whether there was a soil amplification effect on the damage occurring in the Hayrullah neighborhood of the Onikişubat district of Kahramanmaraş Province. Firstly, borehole, SPT, MASW (multi-channel surface wave analysis), microtremor, electrical resistivity tomography (ERT), and vertical electrical sounding (VES) tests were carried out in the field to determine the engineering properties and behavior of soil. Laboratory tests were also conducted using samples obtained from bore holes and field tests. Then, an idealized soil profile was created using the laboratory and field test results, and site dynamic soil behavior analyses were performed on the extracted profile. According to The Turkish Building Code (TBC 2018), the earthquake level DD-2 design spectra of the project site were determined and the average design spectrum was created. Considering the seismicity of the project site and TBC (2018) criteria (according to site-specific faulting, distance, and average shear wave velocity), 11 earthquake ground motion sets were selected and harmonized with DD-2 spectra in short, medium, and long periods. Using scaled motions, the soil profile was excited with 22 different earthquake scenarios and the results were obtained for the equivalent and non-linear models. The analysis showed that the soft soil conditions in the area amplified ground shaking by up to 2.8 times, especially for longer periods (1.0–2.5 s). This level of amplification was consistent with the damage observed in mid- to high-rise buildings, highlighting the important role of local site effects in the structural losses seen during the Kahramanmaraş earthquakes. Full article

Remote sensing images are often degraded by atmospheric haze, which not only reduces image quality but also complicates information extraction, particularly in high-level visual analysis tasks such as object detection and scene classification. State-space models (SSMs) have recently emerged as a powerful paradigm for vision tasks, showing great promise due to their computational efficiency and robust capacity to model global dependencies. However, most existing learning-based dehazing methods lack physical interpretability, leading to weak generalization. Furthermore, they typically rely on spatial features while neglecting crucial frequency domain information, resulting in incomplete feature representation. To address these challenges, we propose ScaleViM-PDD, a novel network that enhances an SSM backbone with two key innovations: a Multi-scale EfficientViM with Physical Decoupling (ScaleViM-P) module and a Dual-Domain Fusion (DD Fusion) module. The ScaleViM-P module synergistically integrates a Physical Decoupling block within a Multi-scale EfficientViM architecture. This design enables the network to mitigate haze interference in a physically grounded manner at each representational scale while simultaneously capturing global contextual information to adaptively handle complex haze distributions. To further address detail loss, the DD Fusion module replaces conventional skip connections by incorporating a novel Frequency Domain Module (FDM) alongside channel and position attention. This allows for a more effective fusion of spatial and frequency features, significantly improving the recovery of fine-grained details, including color and texture information. Extensive experiments on nine publicly available remote sensing datasets demonstrate that ScaleViM-PDD consistently surpasses state-of-the-art baselines in both qualitative and quantitative evaluations, highlighting its strong generalization ability. Full article

The large-scale integration of stochastic renewable energy sources necessitates enhanced dynamic balancing capabilities in power systems, positioning hydropower as a critical balancing asset. Conventional excitation systems utilizing hot-standby dual-redundancy configurations remain susceptible to unit shutdown events caused by regulator failures. To mitigate this vulnerability, this study proposes a peer-to-peer distributed excitation architecture integrating asynchronous traffic shaping (ATS) and Data Distribution Service (DDS) technologies. This architecture utilizes control channels of equal priority and achieves high redundancy through cross-communication between discrete acquisition and computation modules. This research advances three key contributions: (1) design of a peer-to-peer distributed architectural framework; (2) development of a real-time data interaction methodology combining ATS and DDS, incorporating cross-layer parameter mapping, multi-priority queue scheduling, and congestion control mechanisms; (3) experimental validation of system reliability and redundancy through dynamic simulation. The results confirm the architecture’s operational efficacy, delivering both theoretical foundations and practical frameworks for highly reliable excitation systems. Full article