Search Results (731)

Object detection in UAV remote sensing imagery is confronted with three primary challenges: severe scale variation, densely clustered small targets, and constrained computational resources. This work introduces a family of lightweight detection models guided by the “Capacity-Aware Configuration Regularity” and incorporates a Feature-Refinement C2f module to enhance representational efficiency. A dynamic coupling mechanism is identified between detection head capacity and the representational quality of Backbone features, which is further validated through systematic ablation studies spanning three parameter magnitudes. Evaluated on the VisDrone2019 benchmark, the proposed model family exhibits a progressive parameter scaling from 1.67 M to 6.15 M. The nano variant achieves 31.7% mAP₅₀ using only 55% of the parameter budget of YOLOv8n, surpassing it by 0.7 percentage points. The small variant, with a parameter budget comparable to YOLOv8n, attains 36.7% mAP₅₀, exceeding it by 5.7 points. The medium variant reaches 43.1% mAP₅₀ with 58% of the parameters of YOLOv8s, outperforming it by 4.1 points. The improvements are pronounced under the stricter mAP_50–95 metric, where the small variant outperforms YOLOv8n by 3.3 points and the medium variant surpasses YOLOv8s by 2.8 points, demonstrating robust localization accuracy across a wide range of IoU thresholds. This consistent superiority in the accuracy–efficiency trade-off extends to the DIOR dataset, confirming the robust generalization of the proposed models across diverse remote sensing scenarios. Moreover, the uncovered capacity-matching regularity offers transferable methodological guidance for designing lightweight detection models tailored to resource-constrained platforms. Full article

Fracture–vuggy carbonate reservoirs contain discrete caves, fractures, conduits, and vugs, which makes recoverable-reserve evaluation strongly dependent on connected volume rather than on total pore volume alone. This study develops a unit-scale dynamic reserve-updating method for the S48 unit, Tahe Oilfield, by coupling a water-body-corrected material-balance equation, a heterogeneity-corrected waterflood characteristic curve, and iterative geological-model calibration. The main methodological contribution is to convert static fracture–vug architecture into dynamically constrained connected subsystems: the parameter Rwo quantifies connected/injected water volume at the fracture–vug unit scale, whereas the coefficient M corrects the apparent slope of waterflood curves for non-uniform sweep and preferential pathways. The revised workflow was calibrated against pressure, production, injection-response, and history-matched simulation data. Sensitivity analysis indicates that the estimated reserve-utilization degree increased from 48.77% +/− 4.8 percentage points during natural depletion to 74.1% +/− 6.7 percentage points after gas injection, reflecting staged reserve mobilization within the tested uncertainty range. The method is intended for field-scale reserve updating in reservoirs with sufficient pressure-production data; its transferability remains limited by static-model quality, channeling intensity, and the single-unit validation scope of this study. Full article

Large-scale classification of architectural styles in Chinese traditional settlements is important for heritage conservation and geospatial documentation, but scalable deployment remains constrained by the high cost of expert annotation because villages are widely distributed, the imagery is captured from heterogeneous viewpoints, and each architectural tradition exhibits substantial intra-class variation. To address this bottleneck, we propose CTSMatch, a label-efficient semi-supervised framework that combines an ImageNet-pretrained EfficientNetV2 backbone with SoftMatch-based adaptive pseudo-label weighting so that ambiguous but informative unlabeled samples can still contribute to training, thereby reducing reliance on costly expert annotation. We also construct SemiCTS, an extension of the original CTS dataset that adds 4360 unlabeled images. Using only 545 labeled samples, CTSMatch achieves 96.93% accuracy on SemiCTS, outperforming the strongest fully supervised baseline (Dense-TL-Aug) by 2.73 percentage points and two standard semi-supervised baselines (FixMatch and FreeMatch) by 3.06 percentage points. Beyond classification, we further analyze the feature space to examine stylistic lineage through intra-style heterogeneity, inter-style transitions, and outlier detection. The results reveal two broad regional groupings, a northern cluster (Jing, Jin, Su) and a southern cluster (Chuan, Min, Wan), connected by gradual transitions rather than rigid boundaries. Approximately 15% of the samples are identified as atypical cases, including 8.7% comprising regional variants and 6.3% comprising hybrid forms. These findings show that CTSMatch provides a practical label-efficient framework for architectural heritage classification while supporting the interpretable analysis of stylistic diversification and convergence in Chinese traditional settlements. Full article

Background and Objectives: Potential anti-neoplastic effects of resveratrol, which has antioxidant features combined with clomipramine, which has antineoplastic features, or with gemcitabine, used as a nucleoside analog widely used in chemotherapy, were evaluated together and individually on the HL-60 leukemia cells in this in vitro screening study. Materials and Methods: HL-60 cells were treated with gemcitabine, clomipramine, resveratrol, or their combinations at concentrations ranging from 1 to 200 µM. Cell viability was assessed at 24, 48, and 72 h using the trypan blue exclusion method, and results are expressed as a percentage of time-matched untreated controls. Cell proliferation was further evaluated by bromodeoxyuridine (BrdU) immunohistochemical labeling. All experiments were performed in triplicate, and statistical analyses were conducted using one-way analysis of variance (ANOVA) with post hoc comparisons. Results: Gemcitabine markedly reduced HL-60 cell viability at all concentrations and time points (p < 0.001), indicating strong time-dependent cytotoxicity, with a significant drop in BrdU proliferation index at 48 h (p < 0.001). Clomipramine exhibited a biphasic response: high concentrations decreased viability (p < 0.05), while low concentrations allowed partial recovery by 72 h. Resveratrol showed concentration-dependent cytotoxicity, with reduced viability at high concentration and near-control levels at low concentration by 72 h; BrdU indices remained significantly lower than control (p < 0.001). Combination treatments with gemcitabine showed no additive cytotoxic or antiproliferative effects (p > 0.05). A transient enhanced effect was observed in the clomipramine + resveratrol group at 24 h (p < 0.01 vs. clomipramine; p < 0.05 vs. gemcitabine). Conclusions: Gemcitabine, clomipramine, and resveratrol all exhibited inhibitory effects on cell proliferation in HL-60 cell cultures. However, the combination treatments did not show additional cytotoxicity or additive effects. These findings suggest that while each of these compounds individually has the potential to inhibit cell growth, their combined application does not enhance the cytotoxic effects beyond those observed with single treatments. These findings highlight the necessity of a rational approach when considering novel drug combinations. Full article

Point cloud scene flow estimation aims to capture the three-dimensional motion of each point in a sequence of point clouds. Although progress has occurred in this field, existing methods often face significant challenges. In particular, two key issues persist: the absence of corresponding local information from the source point cloud to the target, preventing correct feature matching, and the presence of highly similar adjacent structures in target regions, which leads to ambiguous correspondences due to indistinguishable point features. To address these problems, this paper introduces a novel self-supervised method for point cloud scene flow estimation. Theoretically, we establish a new framework that integrates discriminative feature learning with probabilistic flow refinement. A new network architecture, SetConv++, is designed to learn more discriminative point feature representations, enhancing differentiation in similar structures. Additionally, a refinement module uses the random walk algorithm to adjust initial flow estimates. This approach reconstructs low-confidence flows with high-confidence surrounding ones, reducing missing correspondence issues. Crucially, a new flow smoothing loss term ensures local consistency while suppressing error propagation—a fundamental limitation in existing methods. Through comprehensive experiments on the KITTI Scene Flow dataset, our method demonstrates superior performance. It significantly outperforms existing self-supervised approaches across multiple standard evaluation metrics. Specifically, on the KITTI Scene Flow dataset, our method reduces the Endpoint Error (EPE) by 13.6% (from 0.0411 to 0.0355) and improves Accuracy Strict (AS) by 2.43 percentage points (from 92.68% to 95.11%) compared to baseline self-supervised approaches, while also reducing the outlier rate (Out) by 1.5 percentage points. This advancement not only provides a robust theoretical framework for handling ambiguous correspondences but also enables more reliable and efficient downstream applications—such as autonomous driving perception systems requiring real-time motion accuracy in complex scenes. Full article

Background/Objectives: To evaluate the association between fibromyalgia and dementia, with emphasis on temporal directionality and Alzheimer’s disease-related dementia. Methods: We conducted a nationwide, population-based matched case–control study including 9232 patients with fibromyalgia and 46,160 age- and sex-matched controls. Diagnoses in the Leumit Health Services database are recorded using a hybrid coding scheme that combines ICD-9-CM and WHO ICD-10 codes; the specific codes used to ascertain fibromyalgia and each dementia subtype are listed in the Methods. Outcomes were assessed within two predefined windows: up to 20 years before and up to 10 years after fibromyalgia diagnosis. Alzheimer disease-related dementia was defined as the primary outcome. Multivariable logistic regression was used to estimate odds ratios (ORs) with 95% confidence intervals (CIs). Results: During the 20 years preceding fibromyalgia diagnosis, no increased dementia prevalence was observed; Alzheimer disease-related dementia was less frequent among fibromyalgia patients (0.16% vs. 0.31%; absolute difference −0.15 percentage points; OR 0.52, 95% CI 0.31–0.89). In contrast, during the 10 years following diagnosis, fibromyalgia was associated with a higher prevalence of Alzheimer’s disease-related dementia (1.43% vs. 0.99%; absolute difference +0.44 percentage points; OR 1.45, 95% CI 1.18–1.78). No consistent associations were found for other dementia subtypes, which should be interpreted as exploratory given low event counts. Conclusions: Fibromyalgia is associated with a higher prevalence of Alzheimer’s disease-related dementia in the years following diagnosis, with no evidence of pre-diagnostic elevation. Although this temporal pattern argues against reverse causation, the prevalence-based design and residual confounding preclude causal inference. Fibromyalgia should be regarded as a potential risk marker for subsequent Alzheimer-related neurodegeneration rather than a demonstrated causal factor. Full article

Scene graph generation (SGG) provides a structured representation for visual understanding. However, most existing methods are designed to optimize global triplet recall rather than retrieve relation instances specified by a user query. In query-driven visual relation retrieval, two major challenges arise: the target relation must compete with a highly redundant candidate space, and query semantics are not incorporated before relation classification. To address these challenges, we propose a Query-Driven Candidate Relation Screening (QCRS) module, which injects query semantics into the candidate screening process. Specifically, QCRS encodes the query and candidate visual relation features, and then filters query-relevant candidates through relevance scoring. By reducing interference from irrelevant candidates and avoiding redundant computation, QCRS improves the final exact triplet hit performance and enhances the interpretability of query-specific relations, thereby facilitating query-driven visual relation retrieval. Built upon the strong EGTR baseline, QCRS learns query relevance to prioritize relation instances matching the target query, enabling precise triplet retrieval. Extensive ablation studies and analyses on the VG150 benchmark validate the effectiveness of the proposed approach: when integrated with EGTR, QCRS improves PairR@50 from 61.52% to 80.06% and ETR@50 from 30.54% to 47.07%, achieving absolute gains of over 16 percentage points in both correct object pair retention and end-to-end target relation retrieval performance. Full article

Background: Acute fatigue is frequently experienced during handball training and match play and may impair throwing performance; however, findings across studies are inconsistent. This systematic review and meta-analysis examined the acute effects of fatigue on throwing velocity and accuracy in handball. Methods: PubMed, MEDLINE Complete, CINAHL, and Scopus were searched from inception to 24 January 2026, supplemented by Google Scholar and citation tracking. Eligible studies included handball players exposed to an acute fatigue protocol with throwing-performance outcomes. Random-effects meta-analyses were conducted using standardized mean differences (Hedges’ g), oriented so that negative values indicated worse performance under fatigue. Results: Τen studies met the inclusion criteria for qualitative synthesis. For quantitative synthesis, 10 comparisons contributed to the throwing-velocity meta-analysis and 6 comparisons contributed to the throwing-accuracy meta-analysis. Fatigue showed a small-to-moderate tendency to reduce throwing velocity (g = −0.31, 95% CI −0.65 to 0.03; I² = 77.8%). For throwing accuracy, the pooled estimate suggested a possible decline under fatigue (g = −0.82, 95% CI −1.95 to 0.31), but heterogeneity was very high (I² = 95.8%) and findings were sensitive to influential effects. Conclusions: Acute fatigue showed a small-to-moderate tendency to reduce throwing velocity in handball players, with more consistent impairments observed during jump-shot tasks and after localized upper-limb fatigue protocols. In contrast, no robust conclusion can be drawn for throwing accuracy/precision because heterogeneity was extremely high and studies used substantially different outcome definitions, including hit counts, success percentages, points-based scores, and spatial error. Therefore, accuracy findings should be interpreted with considerable caution. Full article

Background/Objectives: This clinical study assessed the validity of articulating paper for detecting occlusal contacts points, including examining the effects of clinical technique, paper thickness, and the arch. Methods: This cross-sectional test–retest study included 32 adults with natural dentitions. Four occlusal registrations were obtained from each participant using articulating paper with a thickness of 100 or 200 μm, applying one of two different clinical techniques (holding in place or pulling with forceps at the intercuspal position), and scanning the occlusal surfaces of their mandibular and maxillary arches. Silicone registrations were obtained and used as the reference standard. Mandibular and maxillary images were scaled and calibrated spatially, and two new images were created based on the sum or the areas of coincidence between the mandibular and maxillary occlusal scheme. Occlusal contact points on the right posterior teeth were analyzed using computer software. Results: The articulating paper produced true-positive contacts in 81% and false-positive contacts in 15%, regardless of the method used. Considering occlusal contact when marks matched on both arches accounted for 2.0% of false-positive contact points. General linear models with repeated measures revealed that the mandibular arch offered a higher true-positive percentage than the maxillary arch, that the 100 µm-thick paper produced higher false-positive contacts (20.6%) than the 200 µm-thick paper (9.4%), and that the pulling technique had no significant effect. Conclusions: Articulating paper offers good validity when detecting occlusal contact points and can be improved by using 200 µm articulating paper and exploring both arches. Full article

To address the challenges of nonlinearity, strong temporal dependence, and accuracy degradation caused by sudden disturbances in power grid customer service telephone traffic forecasting, this paper proposes a novel forecasting method based on an ensemble model pairing Gated Recurrent Unit (GRU) with Correntropy loss (CL) (called EnsCL-GRU). First, to overcome the sensitivity of the traditional Mean Squared Error (MSE) loss to abnormal spikes and its difficulty in capturing the overall trend consistency of the sequence, a CL is introduced as the loss function for the GRU model. This loss function calculates the normalized Correntropy coefficient between the predicted sequence and the true sequence in the time-delay domain, guiding the model to focus on the overall shape matching of the time series data rather than point-wise error fitting. Furthermore, the gated memory mechanism of the GRU can capture long-term dependencies in the time series, while the CL constrains the consistency of the predicted dynamic trends from the sequence level. This preserves the GRU’s temporal modeling capability while enhancing the model’s response accuracy to sudden disturbances and trend changes. Second, to improve the generalization ability of a single GRU model, an ensemble strategy is employed to train multiple CL-enhanced GRU base models serially. By adaptively adjusting sample weights, the fitting capability for difficult samples (such as telephone traffic spikes) is improved, further improving the model’s robustness. Finally, Bayesian optimization is introduced to automatically search for the optimal hyperparameters of the ensemble model, efficiently approximating the global optimal configuration within a limited number of evaluations. Experimental results demonstrate that the proposed method outperforms traditional approaches. Specifically, compared with the standard GRU model, the proposed method reduces MAPE from 29.15% to 22.61%. It also consistently outperforms the ensemble baseline EnsGRU, achieving a MAPE reduction of 4.73 percentage points. The results indicate that the proposed model significantly improves forecasting accuracy and robustness, particularly under scenarios with nonlinear fluctuations and sudden disturbances, providing reliable support for optimal resource allocation in power grid customer service systems. Full article

Distant supervision relation extraction (DS-RE) provides an efficient way to construct large-scale training data by automatically aligning knowledge base relations with unstructured texts. However, this process inevitably introduces erroneous labels because sentences containing the same entity pair do not always express the corresponding knowledge base relations. To address this problem, this paper proposes a noise-weighted unsupervised denoising framework that integrates sentence-level prior confidence estimation, multi-factor representation learning, fine-grained noise detection, and clustering-based label generation. The framework first estimates noise-aware prior weights by matching sentence instances with semantically similar relation triples. It then incorporates lexical, positional, and entity-type factors to enhance sentence representations. For detected noisy instances, an unsupervised clustering-based label generation module is used to regenerate relation labels rather than directly discarding them. Experimental results on the DSRED dataset show that the proposed method achieves 89.7% Precision, 90.6% Recall, 90.1% F1-score, and a PR-AUC of $0.942\pm 0.004$, outperforming the strongest baseline EFEAPN by 1.7 percentage points in F1-score. Statistical analysis further shows that the PR-AUC improvement remains significant after Bonferroni correction ($p_{\mathrm{adj}}=0.0094$). Module-level ablation experiments, sensitivity analysis, and clustering quality evaluation further verify the effectiveness of the noise weighting and clustering-based label generation modules. Supplementary experiments with Transformer-based encoders and cross-dataset evaluation further show that the main performance gain comes from the proposed denoising framework rather than from a specific sentence encoder. Full article

Vision–language models (VLMs) are increasingly deployed in resource-constrained environments, yet efficient fine-tuning remains challenging because post-training quantization often degrades the effectiveness of low-rank adaptation. This paper revisits that mismatch in the context of MobileVLM1.7B and presents QuantFT-VL, a novel initialization strategy following the quantization phase to seamlessly align with the LoRA technique. The key idea is to initialize LoRA using a low-rank approximation of the quantization residual instead of the default zero-initialization used in QLoRA-style pipelines. After quantizing a pretrained weight matrix W into Q, we compute the residual W − Q and use truncated singular value decomposition to initialize the LoRA factors (A and B) so that the starting adapted weight Q + AB^T better matches the full-precision model. This residual-aware initialization reduces the discrepancy introduced by quantization and leads to faster and more stable optimization. Experiments on six standard VLM benchmarks show that QuantFT-VL consistently improves over QLoRA and recovers performance close to or better than full-precision LoRA in the best setting. On two RTX 3090 GPUs, QuantFT-VL improves the average benchmark score by 3.27 percentage points over QLoRA while preserving the memory and speed advantages of quantized fine-tuning. Full article

Background/Objectives: Objective end-product assessment of microsurgical anastomoses is intensive and partly subjective. Vision-capable large language models (LLMs) may enable standardized image-based scoring, but their agreement with expert assessment remains uncertain. Methods: We studied 200 end-to-end femoral artery anastomoses, performed on chicken legs by novice, intermediate, and experienced microsurgeons. Images were scored independently by two blinded expert panels; disagreements were adjudicated by a third senior reviewer to establish expert consensus. Two LLMs, ChatGPT 5.2 Thinking Extended and Gemini 3.1 Pro, were evaluated using the exact same prompt and rubric. Each image was analyzed three times per model. Final scores were aggregated by median for numeric items and majority vote for categorical items. The primary endpoint was exact-match agreement with expert consensus. Agreement within ±1 was also assessed for numeric items. Agreement was measured using simple percentage agreement, Light’s kappa, and Krippendorff’s alpha; Bland–Altman analysis was used for numeric count items. Results: LLM 1 achieved a higher overall exact-match agreement than LLM 2 (0.659 vs. 0.539). Both models performed better on categorical than numeric items (0.713 vs. 0.610 and 0.651 vs. 0.445, respectively). LLM 1 showed the greatest advantages for gaps, knots, oblique stitches, and wide bites. Krippendorff’s alpha was positive for most endpoints with LLM 1, whereas LLM 2 showed negative values throughout. Allowing a ±1 tolerance for numeric items greatly improved agreement, suggesting only minor counting discrepancies, from 0.610 to 0.900 for LLM 1 and from 0.445 to 0.826 for LLM 2. Conclusions: Under a constrained scoring workflow, LLMs partially approximated intraluminal microsurgical end-product scoring. LLM 1 outperformed LLM 2, but agreement remained insufficient to replace the expert assessment entirely. These models can be assistive tools within a human-in-the-loop framework. Full article

Background: Body mass index (BMI) remains the standard tool for obesity screening; however, it does not account for body fat distribution or visceral adiposity, potentially leading to clinically relevant misclassification—particularly in young adults. Evidence on this issue in healthcare professionals is limited. Objective: To evaluate the extent of obesity misclassification when using BMI compared with alternative anthropometric and body composition indices, and to examine sex-specific associations between lifestyle factors and different adiposity phenotypes in young healthcare professionals. Methods: A large cross-sectional study was conducted in 12,874 medical residents, nursing residents, and age-matched controls (22–30 years). Obesity was defined using BMI (≥30 kg/m²), waist-to-height ratio (WtHR ≥ 0.5), Clínica Universidad de Navarra–Body Adiposity Estimator (CUN-BAE), body fat percentage, and bioimpedance-derived visceral fat. Multivariable logistic regression models adjusted for age, sex, professional group, smoking, physical activity, and Mediterranean diet adherence were fitted separately for each adiposity definition. Sex interaction terms were formally tested. Agreement between indices was assessed using Cohen’s kappa. Results: Obesity prevalence varied substantially according to the index applied and was consistently higher when central or visceral adiposity measures were used. Agreement between BMI and alternative indices was only fair to moderate, with the lowest concordance observed for visceral fat (κ = 0.29; 95% CI 0.26–0.32). Male sex was strongly associated with visceral fat-defined obesity (aOR 4.76; 95% CI 3.82–5.92), while effect sizes were attenuated for BMI-defined obesity (aOR 1.41; 95% CI 1.32–1.51). Significant sex interactions were detected for visceral adiposity, particularly for physical activity (p = 0.001) and smoking (p = 0.002), indicating differential lifestyle associations according to fat distribution phenotype. Conclusions: BMI substantially underestimates clinically relevant central and visceral adiposity in young healthcare professionals. Sex-specific differences were observed in the association between lifestyle behaviors and visceral fat. These findings highlight the limitations of relying exclusively on BMI for obesity screening. Incorporating waist-based or body composition-derived measures may improve early risk identification and support targeted preventive strategies. Full article

Objective: To analyze the relationship between weekly accumulated external load and pre-match neuromuscular performance assessed through the countermovement jump (CMJ), in under-21 (U21) football players across 10 competitive microcycles. Methods: Sixteen U21 football players (age: 18.9 ± 0.42 years; height: 180 ± 6.3 cm; body mass: 78.5 ± 8.5 kg) from a Chilean professional club were monitored over 10 consecutive weeks. In each microcycle, the relationship between changes in neuromuscular performance estimated from CMJ-derived variables and two components of external load was analyzed: (1) weekly accumulated external load and (2) the acute–chronic workload ratio (ACWR). External load variables included total distance (TD), high-speed running distance (HSR), accelerations (ACC), decelerations (DC), and PlayerLoad (PL). CMJ variables included jump height (JH), modified reactive strength index (RSI-mod), and peak eccentric velocity (PEV). Performance changes were calculated as the percentage change (Δ%) between MD + 2 (start of the microcycle) and MD − 1 (pre-match). Pearson or Spearman correlation coefficients were applied depending on data distribution. Results: Significant negative associations were observed between weekly accumulated external load and changes in CMJ performance. Reductions in JH were associated with TD, HSR, ACC, and PL. Similar patterns were found for RSI-mod, while PEV showed a particularly strong association with ACC. Additionally, ACWR demonstrated significant negative relationships with CMJ changes, especially for HSR, ACC, and PL. Conclusions: Higher weekly accumulated external loads and elevated ACWR, particularly in high-intensity metrics such as high-speed running and accelerations, are associated with impaired pre-match neuromuscular performance. Consequently, monitoring CMJ-derived variables alongside external load data is recommended to manage fatigue and optimize match readiness in young football players. Full article