Search Results (5,381)

In the field of electronic support and radar warning, it is necessary to effectively detect and recognize the modulation types of non-cooperative radar signals, especially for radars with Low Probability of Intercept (LPI) waveforms. Multiple intelligent detection and recognition algorithms based on the Transformer architecture have been proposed, achieving good performance even under low signal-to-noise ratio (SNR). However, Transformer-based radar intelligent detection and recognition algorithms have a huge number of parameters coupled with complex structures, which will result in significant power consumption and computational latency when deployed on general computing platforms. To address the above issues, this paper proposes a lightweight design for Transformer-based radar signal intelligent detection and recognition networks. A Lightweight Joint Detection and Modulation Recognition Networks (JDMR-LNet) is designed. To enhance the feature extraction ability of lightweight networks, this paper designed a hybrid model distillation method. The experimental results demonstrate that, compared with the directly trained JDMR-LNet, the accuracy of automatic modulation type recognition of the JDMR-LNet after distillation is increased by 2.37% at −12 dB, and the signal detection is increased by 2.07% at −10 dB. The number of parameters of the JDMR-LNet has also decreased significantly. Compared with the original model, the JDMR-LNet is compressed by 11.18 times. Furthermore, this paper completed FPGA deployment of the JDMR-LNet model, with simulation verifying its functional correctness. Full article

Background and Objectives: Sarcopenia is a common yet underrecognized condition in cancer patients and is associated with increased treatment-related toxicity, functional decline, and poor clinical outcomes. Practical, rapid, and bedside-applicable tools are needed to detect sarcopenia early in routine oncology practice. This study aimed to evaluate the diagnostic value of rectus femoris muscle ultrasonography within an integrated clinical assessment combining handgrip strength and bioelectrical impedance analysis for the detection of sarcopenia in cancer patients. Materials and Methods: In this prospective cross-sectional study, 147 adult patients with malignancy were evaluated using a multimodal sarcopenia assessment framework. Muscle strength was assessed by handgrip dynamometry, muscle mass was estimated using bioelectrical impedance analysis (BIA)-derived appendicular skeletal muscle mass index (ASMI), and muscle morphology was evaluated using ultrasonographic measurements of the rectus femoris and biceps brachii muscles. Sarcopenia was defined and classified according to the EWGSOP2 criteria. Associations between clinical variables, BIA parameters, and ultrasonographic measurements were analyzed. Receiver operating characteristic (ROC) curve analyses were performed to assess the diagnostic performance of muscle ultrasonography for sarcopenia detection. Results: The mean age of the study population was 60.2 ± 11.2 years, and 51% of participants were male. Confirmed sarcopenia was identified in 12.2% of patients, while 27.2% were classified as having probable sarcopenia. Sarcopenic patients were significantly older (68.5 ± 7.6 vs. 59.0 ± 11.2 years, p = 0.001) and had lower handgrip strength (15.8 ± 6.0 vs. 24.3 ± 8.4 kg, p < 0.001) and ASMI values (5.96 ± 0.64 vs. 7.23 ± 1.18 kg/m², p < 0.001). Rectus femoris muscle thickness was significantly reduced in patients with sarcopenia (6.40 ± 1.42 vs. 8.19 ± 2.21 mm, p = 0.001). Rectus femoris muscle thickness demonstrated good diagnostic performance for sarcopenia detection (AUC = 0.752; 95% CI: 0.650–0.853; p = 0.001), with an optimal cut-off value of ≤ 7.59 mm (sensitivity 83.3%, specificity 61.2%). Conclusion: Rectus femoris muscle ultrasonography is a practical, rapid bedside assessment for detecting sarcopenia in cancer patients. When integrated with handgrip strength and BIA, this multimodal approach provides a feasible, radiation-free strategy for early sarcopenia screening in routine oncology practice. Full article

Event-based vision sensors (EVSs), with the core advantages of low latency, high dynamic range, and low data volume, have become one of the research hotspots in the field of computer vision. However, the characteristic of detecting changes in light intensity makes EVSs particularly sensitive to noise, so the large number of noise events in the event stream significantly limits the practical application of EVSs. To address this critical issue, considering the types and characteristics of noise, this paper proposes an event stream denoising algorithm based on local spatiotemporal event quantities and implements it in hardware. To comprehensively evaluate the algorithm’s performance, two metrics based on the probability of real events, Noise Event Ratio (NER) and Event Noise Ratio (ENR), are used to quantify the denoising effect, while hardware resource overhead is assessed in terms of event processing latency and memory usage. Experimental results show that the proposed algorithm achieves an NER of 8.37% and an ENR of 25.10%. Compared to existing denoising algorithms, such as the DWF algorithm, the NER and ENR of this algorithm are reduced by 27.72% and 22.89%, respectively, demonstrating superior denoising performance. On the hardware side, the latency for processing a single event is approximately 110 ns, with a total resource usage of N² memory units. Although the hardware consumption is slightly higher, the algorithm exhibits significant advantages in denoising performance, providing effective support for the engineering application of EVSs. Full article

Obstructive sleep apnea (OSA), defined by repetitive airway obstruction during sleep, is significantly underdiagnosed, mainly due to the resource-intensive and time-consuming nature of sleep assessment technologies. Machine learning analysis of the tracheal breathing sounds (TBS) whilst awake offers an alternative approach for OSA quick screening. This study aimed to address the challenge of wakefulness OSA detection using TBS recorded with an inexpensive microphone in a noisy environment. Data of 247 individuals with various degrees of OSA severity were analyzed. Recorded data were segmented into inspiration and expiration phases, followed by acoustic features extraction, feature reduction, and classification. A two-level ensemble architecture was implemented. Nine sub-classifiers were stratified by anthropometric profiles. Each sub-classifier was constructed as an ensemble of bagged decision trees, with a final prediction via probability-based voting. The proposed algorithm achieved an accuracy of 77.1%, sensitivity of 84.3%, and specificity of 59.9%. Although these results have lower performance than those obtained previously using a high-quality microphone in a quiet room, they demonstrate that acoustic OSA detection whilst awake remains feasible, even in very noisy environments. Nevertheless, microphone quality emerged as a key determinant of classification performance. Full article

Early wildfire smoke detection is critical for preventing small ignitions from escalating into large-scale fires, yet early-stage smoke plumes are often faint, low-contrast, and spatially small. When full-resolution frames are resized to satisfy fixed-input detector architectures and enable efficient batched GPU inference, these subtle cues are further diminished, leading to missed detections and unreliable scores near deployment thresholds. Existing remedies such as multi-scale inference, slicing/tiling, or super-resolution could improve sensitivity, but typically incur substantial overhead from multiple forward passes or added network components, limiting real-time use on resource-constrained platforms. To mitigate these challenges, we propose a composite multi-resolution detection framework that improves sensitivity to small smoke regions while maintaining single-pass inference. Motivated by the fact that most operational wildfire monitoring systems rely on Unmanned Aerial Vehicle (UAV) platforms and mountain-top Closed-Circuit Television (CCTV) systems surveillance, their wide-field imagery typically contains a large sky region above the horizon where early smoke is most likely to first become visible. Accordingly, crop placement is guided by a skyline prior that prioritizes this high-probability sky band while retaining the remaining scene for global context. A dynamic compositing stage stacks a global view with a high-resolution, sky-aligned band into a standard square detector input, preserving context with minimal added cost. Detections from the two views are reconciled via coordinate restoration and non-maximum suppression. For deployment, a lightweight second-stage classifier selectively re-evaluates low-confidence detections to stabilize decisions near a fixed operating threshold without retraining the detector. Compared to the baseline detector, our approach improves detection performance on the Early Smoke dataset, achieving gains of +4.6 percentage points in AP @0.5:0.95, +3.4 percentage points in AP @0.5, +2.9 percentage points in precision, +5.3 percentage points in recall, and +4.3 percentage points in F1-score. Full article

Background/Objectives: Cardiac resynchronization therapy (CRT) is a cornerstone treatment for heart failure with reduced ejection fraction (HFrEF), yet many patients remain symptomatic despite long-term electrical optimization. Although sacubitril/valsartan (ARNI) is central to guideline-directed medical therapy (GDMT), data on its late initiation in patients with chronic CRT are scarce. This study evaluated the impact of delayed ARNI initiation on clinical status, functional capacity, and cardiac remodelling in a real-world CRT population. Methods: We performed a single-centre, retrospective observational study including 76 HFrEF patients with chronic CRT who started ARNI between 2022 and late 2024. Patients underwent standardized assessment at baseline (T0) and after 12 ± 3 months (T1), including clinical evaluation, 12-item Kansas City Cardiomyopathy Questionnaire (KCCQ-12), symptom-limited bicycle exercise testing, and comprehensive echocardiography. The primary endpoint was change in quality of life (QoL). Secondary endpoints included exercise capacity, echocardiographic reverse remodelling, NYHA class, loop diuretic dose, and device-detected arrhythmias. Dose–response and multidimensional response patterns were explored. Results: KCCQ-12 increased from 52.96 ± 16.33 to 75.55 ± 18.12 (Δ +22.59 ± 13.22, p < 0.001), with 89.5% achieving a clinically meaningful improvement. Exercise duration and peak workload improved significantly. LVEF increased from 35.08 ± 6.96% to 43.18 ± 8.42% (Δ +8.11%, p < 0.001), with reductions in left ventricular and atrial volumes. Loop diuretic dose decreased (median −10 mg/day furosemide equivalent, p < 0.001), and 26.3% discontinued diuretics. A lower prevalence of device-detected arrhythmias was observed at follow-up, from 34.2% to 6.6% (p < 0.001). Higher ARNI doses were associated with greater likelihood of clinical, functional, and structural response. Longer CRT duration reduced the probability of structural remodelling but not symptomatic or functional benefit. Conclusions: In patients with long-standing CRT, delayed ARNI initiation was associated with improvements in QoL, exercise capacity, cardiac remodelling, congestion status, and electrical stability. These findings suggest that CRT is not a therapeutic ceiling and that late ARNI initiation remains a valuable component of comprehensive GDMT. Full article

Organophosphorus pesticides (OPs) have been widely employed to increase food production and alleviate the increasingly serious food crisis. However, excessive use of these pesticides has seriously affected human health and even caused death due to significant pesticide residues in food. Therefore, enzymatic electrochemical sensors have been developed to monitor OP residues in food. The electrochemical detection performance of these sensors is determined by the physicochemical properties of electrochemical active materials in their active layers. The definition and classification of OPs are first introduced in this review, then the components of enzymatic electrochemical sensors, including electrodes, electrochemical active layer and bioactive enzyme layer, are analyzed in detail. Furthermore, this review emphatically discusses the recent development of enzymatic electrochemical sensors based on various electrochemical active materials: carbon-based, polymer-based, metal-based, metallic compound-based, metal organic framework-based and covalent organic framework-based materials. Finally, probable research directions for developing enzymatic electrochemical sensors with high sensitivity, excellent stability and good reproducibility are outlined to accelerate rapid, effective and low-cost on-site detection OPs in food. This review is expected to provide inspiration for the design and preparation of the high-performance enzymatic electrochemical sensors. Full article

In this work, an experimental procedure that enables testing of canine detection capabilities is described. The developed testbed includes an experimental setup with six exchangeable detection/sniffing boxes for odour masking experiments as well as an air-conditioned (adjustable temperature and humidity) test environment. This design is used to test the masking effects of petroleum and n-decane in high and low concentrations on the detection probability of targets containing technical-grade TNT on explosive detection dogs. The potential influence of the masking agents at different concentrations was investigated with eight canines and, in total, more than 1250 test runs. Within the limits of this investigation, no negative impact of the masking agent on the canine detection capabilities (probability of the successful detection of the target) could be found. Full article

In communication-centric integrated sensing and communication (ISAC) systems, passive radars exploit existing communication signals of opportunity for sensing. To compute delay-Doppler or range–velocity maps (DDMs and RVMs, respectively), modern orthogonal frequency division multiplexing (OFDM)-based sensing systems use the channel frequency response (CFR) originally estimated in communication receivers for equalization. In OFDM-based passive radars utilizing 4G LTE or 5G NR waveforms, CFR estimation typically relies only on reference signals. However, simulation-based studies that assume a priori knowledge of user data symbols indicate potential performance gains when incorporating user data and other downlink channels. In this work, we present an experimental evaluation of an OFDM-based passive radar that jointly utilizes all commonly present components of the 5G NR downlink waveform: synchronization signals (PSS and SSS), broadcast and control channels (PBCHs and PDCCHs, respectively), data channels (PDSCHs), and reference signals (PBCH DM-RSs, PDCCH DM-RSs, PDSCH DM-RSs, and CSI-RSs). Our results show that utilizing user data from fully occupied 5G downlink signals, under the assumption of full knowledge of PDSCH locations, significantly improves both the probability of detection (POD) and the peak height, measured by the peak-to-noise-floor ratio (PNFR), compared with pilot-only sensing. Since perfect knowledge of the user data payload is not assumed, we estimate the transmission bit error rate (BER) and analyze its impact on sensing performance. Finally, we investigate more realistic scenarios in which only a subset of PDSCH resource element locations is known, as in practical 5G deployments, and evaluate how partial data location knowledge affects the POD and PNFR under different BER conditions. Full article

The Laser Interferometer Space Antenna (LISA) is an ESA mission designed to detect gravitational waves from space. To initiate the science phase, six test masses (TMs) are precisely handled and released into near-perfect free fall by dedicated mechanisms known as the Grabbing, Positioning, and Release Mechanisms (GPRMs). The stringent requirements on the noise level affecting the TMs’ release acceleration are extremely ambitious, motivating the need to experimentally verify the feasibility of achieving such performance. To this end, a dedicated precursor mission, LISA Pathfinder (LPF), flew from 2015 to 2017 to test key technologies. However, during the LPF mission, most release tests exhibited anomalous release velocities, often exceeding the requirements. In addition, the TM repositioning tests also revealed a bi-stable behavior in the TM rotations, which depend on the repositioning direction. This effect is produced by an unexpected non-rectilinear motion of the GPRM end effector, characterized by a micrometric side motion at the reversal of its axial motion. The bi-stable behavior also contributes to a TM-GPRM end effector misalignment, producing unwanted contacts and increasing the probability of a non-compliant TM release. Previous analyses identified asymmetric friction forces in the side-guiding system of the GPRM end effector as the primary cause of this behavior. Starting from the LPF flight experience, the GPRM delta development project in view of LISA led to a redesign of the mechanism architecture, supported by numerical analyses and multi-body models. Since the rectilinearity of the end-effector motion has been identified as critical for flight operation, alternative side-guiding concepts are developed, analyzed, and tested experimentally to evaluate their impact on the overall mechanism performance. The correlation of the models with ground and flight experimental data strengthens the understanding of the guiding system behavior, providing pivotal insights for selecting the GPRM design baseline for LISA. Full article

Acrylate-based self-polishing copolymer antifouling paint particles (SPC-APPs) are persistent micropollutants that act as carriers for biocidal heavy metals, posing significant ecological hazards to aquatic ecosystems. Despite their toxicity, the occurrence, characterization, and metal-leaching risks of SPC-APPs in estuarine environments remain largely understudied. This study investigated the contamination characteristics of SPC-APPs in surface sediments from the Yangtze River Estuary, a hotspot of shipping activity. A multi-technique analytical protocol was employed, combining density separation with scanning electron microscopy–energy-dispersive spectroscopy (SEM-EDS), inductively coupled plasma mass spectrometry (ICP-MS), and pyrolysis–gas chromatography/mass spectrometry (Py-GC/MS) to characterize the morphology, quantify particle abundance, and assess the correlation between SPC-APPs and sedimentary heavy metals. SPC-APPs were ubiquitously detected across all sampling sites, with abundances ranging from (0.82 ± 0.15) × 10³ to (3.65 ± 0.42) × 10³ particles g⁻¹ dry sediment. A distinct distribution property (South Branch > North Branch > offshore shoal) was identified, primarily driven by shipping density and hydrodynamic sorting. Morphologically, particles exhibited irregular, abraded surfaces, with EDS confirming Cu (1.76~5.63 wt%) and Zn (0.27~3.65 wt%) as major metallic components. Py-GC/MS analysis identified specific mass fragments (m/z 41, 69, 87) as diagnostic markers. Strong positive correlations were observed between SPC-APP abundance and sediment Cu (r = 0.82, p < 0.01) and Zn (r = 0.76, p < 0.01) concentrations, indicating that these particles are a primary source of metal contamination. Ecological risk assessment based on sediment quality benchmarks showed that Cu in the South Branch reached 82~91% of the probable effect concentration (PEC), highlighting potential risks to benthic organisms. This study provides critical baseline data on the distribution and speciation of SPC-APPs, underscoring their role as vectors for toxic metals and the need for targeted pollution control in high-shipping-intensity estuarine regions. Full article

In this study, we statistically analyze the performance of a threshold-based multiple optical signal selection scheme (TMOS) for wavelength division multiplexing (WDM) and adaptive coded modulation (ACM); this is achieved using free space optical (FSO) communication between mobile platforms in maritime environments with fog and 3D pointing errors. Specifically, we derive a new closed-form expression for a composite probability density function (PDF) that is more appropriate for applying various algorithms to FSO systems under the combined effects of fog and pointing errors. We then analyze the outage probability, average spectral efficiency (ASE), and bit error rate (BER) performance of the conventional detection techniques (i.e., heterodyne and intensity modulation/direct detection). The derived analytical results were cross-verified using Monte Carlo simulations. The results show that we can obtain a higher ASE performance by applying TMOS-based WDM and ACM and that the probability of the beam being detected in the photodetector increased at a low signal-to-noise ratio, contrary to conventional performance. Furthermore, it has been confirmed that applying WDM and ACM is suitable, particularly in maritime environments where channel conditions frequently change. Full article

Machine vision based on artificial intelligence technology is being actively utilized to reduce defect rates in the photovoltaic module production process. This study aims to propose a machine vision approach using EfficientNet-B0 for defective photovoltaic module detection. In particular, the proposed approach is applied to the electroluminescence (EL) operation, which identifies microcracks in PV modules by using polarization current. The proposed approach extracts low-level structures and local brightness variations, such as busbars, fingers, and cell boundaries, from a single convolutional block. Furthermore, the mobile inverted bottleneck convolution (MBConv) block progressively transforms defect patterns—such as microcracks and dark spots—that appear at various shooting angles into high-level feature representations. The converted image is then processed using global average pooling (GAP), Dropout, and a final fully connected layer (Dense) to calculate the probability of a defective module. A sigmoid activation function is then used to determine whether a PV module is defective. Experiments show that the proposed Efficient-B0-based methodology can stably achieve defect detection accuracy comparable to AlexNet and GoogLeNet, despite its relatively small number of parameters and fast processing speed. Therefore, this study will contribute to increasing the efficiency of EL operation in industrial fields and improving the productivity of PV modules. Full article

Accurate characterization of precipitation in complex terrain is essential for hydrological modeling and climate studies. This study uses daily observations from 156 rain gauges in Sichuan Province (2015–2020) to evaluate two high-resolution satellite products (GSMaP-GNRT and IMERG-Early) and to develop a Transformer-based fusion framework at the gauge scale. All three datasets reproduce the regional seasonal cycle with more rainfall in summer and less in winter. At the daily scale, the fused product attains correlation comparable to GSMaP, while GSMaP and the fusion slightly overestimate precipitation (Bias = 6.24% and 5.21%), and IMERG shows stronger underestimation (Bias = −11.46%). At the monthly scale, the fused dataset achieves the best overall performance in terms of correlation, bias and RMSE. Spatially, the fusion reduces bias and RMSE and yields more homogeneous patterns over Sichuan’s complex terrain. Detection metrics indicate that the fused product increases the probability of detection and slightly improves the critical success index, while the false alarm ratio remains relatively high and comparable to the original products. This implies a gain in event sensitivity and spatial consistency rather than substantially reduced false alarms. Overall, the Transformer-based fusion provides a useful compromise between GSMaP and IMERG, adding value particularly for bias reduction, monthly statistics and event detection. The fused dataset offers a promising input for precipitation monitoring, hydrological simulation and disaster-risk analysis in Sichuan and similar mountainous regions. Full article

Persistent fiscal stress in South African municipalities undermines service delivery, yet practical tools for early detection remain limited. This study predicts one-year-ahead municipal financial distress to support risk-based prioritisation. We develop machine learning models using a 2018/19–2022/23 municipality panel, combining 13 financial health indicators from State of Local Government (SoLG) reports with selected socio-economic variables. Penalised logistic regression is benchmarked against random forest and XGBoost under a leakage-aware, time-ordered split into training, validation, and an out-of-time test year; class imbalance is handled through class weighting. Performance is evaluated using PR-AUC, ROC-AUC, calibration, and a capacity-constrained Top-30 rule. All models outperform a naïve last-year baseline on the out-of-time test (PR-AUC 0.934–0.954; ROC-AUC 0.886–0.923), with bootstrap intervals supporting robustness. Random forest performs best overall, while penalised logistic regression remains competitive. Under the Top-30 rule (12.3% workload), precision is high (precision@30 0.967–1.000) while recall is modest (recall@30 0.186–0.192). SHAP values and logistic odds ratios identify liquidity, solvency, cash coverage, and employment deprivation as key drivers. The Top-30 rule corresponds to an annual intensive monitoring portfolio that is reasonable under constrained staffing and budget capacity in national and provincial oversight units, while probability thresholds are reported as conventional benchmarks rather than as policy triggers. Full article