Search Results (38,527)

Facial emotion recognition (FER) plays an important role in healthcare, human–computer interaction, and intelligent security systems. However, despite recent advances, many state-of-the-art FER methods depend on computationally intensive CNN or transformer backbones and large-scale annotated datasets while suffering noticeable performance degradation under cross-dataset evaluation because of domain shift. These limitations hinder practical usage in resource-constrained and real-world environments. To address this issue, we propose Cross-Adaptive Gated Fusion Network (CA-GFNet), a lightweight dual-stream FER framework that explicitly combines shallow structural features with deep semantic representations. The proposed architecture integrates domain-robust gradient-based descriptors with compact deep features extracted from a VGG-based backbone. After face detection and normalization, the structural stream captures fine-grained local appearance cues, whereas the semantic stream encodes high-level facial configurations. The two feature streams are projected into a shared latent space and adaptively fused using a gated fusion mechanism that learns sample-specific weights, allowing the model to prioritize the more reliable feature source under dataset shift. Extensive experiments on KDEF along with zero-shot cross-dataset evaluation on CK+ using a strict train-on-KDEF/test-on-CK+ protocol with subject-independent splits demonstrate the effectiveness of the proposed method. CA-GFNet achieves 99.30% accuracy on KDEF and 98.98% on CK+ while requiring significantly fewer parameters than conventional deep FER models. These results confirm that adaptive gated fusion of shallow and deep features can deliver both high recognition accuracy and strong cross-dataset robustness. Full article

Background/Objectives: Long-term weight maintenance remains challenging with conventional dietary strategies due to various barriers. Time-restricted eating (TRE) has recently attracted attention as a potential approach to improve adherence, but evidence on long-term maintenance is limited. We investigated the 6-month follow-up (6FU) of early time-restricted eating with energy restriction (eTRE + ER), late time-restricted eating with energy restriction (lTRE + ER) and energy restriction alone (ER). Methods: This 6FU included 69 of 93 participants from a previously conducted 3-month intervention (3INT). After the intervention, participants returned to free-living conditions without dietary guidance. Outcomes included adherence, perceived barriers, body composition, blood pressure, cardiometabolic risk factors, metabolic hormones, subjective appetite, and dietary intake. Results: Adherence of at least ≥5 days per week was low: 7.7% (eTRE + ER), 18.2% (lTRE + ER), and 9.5% (ER). Reduced adherence during the 6FU was associated with a partial reversal of improvements in body mass, body composition, cardiometabolic risk factors, metabolic hormones, and subjective appetite observed during the 3INT. Analysis of perceived barriers showed that environmental and psychosocial barriers were significant predictors of changes in body mass during the 6FU, while environmental and behavioral barriers were associated with extension of the eating window. These associations were most pronounced in the eTRE + ER group. Conclusions: During the 6FU, differences between dietary strategies gradually diminished, although some remained clinically meaningful. Long-term adherence was low across all three dietary strategies, with psychosocial, environmental, and behavioral barriers particularly evident in the eTRE + ER group. Further research is needed to confirm long-term adherence before TRE + ER interventions can be widely applied in clinical practice. Full article

To address the topological mismatch between signal space and physical space caused by uneven signal feature distribution in indoor non-line-of-sight and complex topological environments, this paper proposes an indoor positioning algorithm based on Energy-domain Fingerprint Manifold Graph Total Variation (EFM-GTV). To mitigate neighborhood distortion caused by uneven high-dimensional signal feature distribution, a UMAP manifold topology graph construction method based on fuzzy simplicial sets is designed to establish a graph basis consistent with physical space topology. To reduce false matching risks in global search, a physical topology pruning strategy combining Jaccard similarity is proposed, effectively eliminating pseudo-connections. Building upon this foundation, we introduced an optimization model based on graph total variation, reformulating the positioning problem as a graph signal recovery task. This approach effectively overcomes signal fluctuation interference in complex topologies like U-shaped corridors, achieving robust position estimation. Experiments demonstrate that this algorithm effectively leverages manifold structure constraints to correct NLOS errors. On real-world field test datasets, compared to traditional weighted algorithms, the average positioning accuracy improves to 1.4267 m, with maximum positioning error reduced by over 50%, achieving high-precision robust positioning. Full article

Fish host complex intestinal bacterial communities that contribute to a wide range of functions, from nutrient assimilation to modulation of the immune system. Understanding how environmental and host-related factors shape the fish gut microbiota is essential for advancing sustainable aquaculture practices. This study compared the intestinal microbiota of gilthead sea bream (Sparus aurata) between wild and aquaculture populations in western Greece using 16S rRNA gene amplicon sequencing targeting the V3–V4 region, combined with culture-based methods. The analysis was based on a 97% similarity threshold and included 141 gastrointestinal samples of fish collected at two aquaculture facilities and two wild fisheries, representing two different growth phases (150 g and 300 g body weight). High-throughput sequencing data revealed a clear separation of gut microbial communities according to origin (wild vs. aquaculture), geographic location, and body growth phase, with most wild fish groups exhibiting higher microbial diversity than their farmed counterparts, except for group MES_150 which showed similar or lower values. The gut microbiota was dominated by Pseudomonadota (53%), Bacillota (29%), Actinomycetota (7%), Deinococcota (5%), and Bacteroidota (4%). A shared core microbiome, comprising Psychrobacter, Staphylococcus, Geobacillus, Aeromonas, Enterobacter, Pantoea, Bacillus, and Acinetobacter, was detected across all populations. Wild fish were enriched in Psychrobacter, Aeromonas, and Photobacterium, while aquaculture fish displayed higher abundances of Vibrio, Allomeiothermus, and Staphylococcus. Network analysis revealed mostly mutually exclusive interactions in both groups but distinct patterns of co-occurrence, driven mainly by Paenibacillus, Enterobacter, and Staphylococcus in wild samples, and by Vibrio, Aeromonas, and Pseudomonas in farmed fish. Culture-based assays demonstrated greater diversity in wild fish, dominated by Pseudomonas, Staphylococcus, and Vibrio strains, in contrast to the frequent occurrence of Staphylococcus and Psychrobacter in aquaculture samples. The findings suggest that aquaculture practices significantly alter gut microbial community structure and reduce diversity, with potential implications for fish health and disease resistance. The identified core and differentially abundant taxa provide candidates for probiotic development to improve aquaculture sustainability. Full article

Background/objectives: In this study, we aimed to investigate acute infarct volume on magnetic resonance diffusion-weighted imaging (MRI DWI), chronic infarct volume on FLAIR (fluid-attenuated inversion recovery), hypoperfused area volume on PWI (perfusion-weighted imaging), stenosis locations and rates on CT (computerized tomography) angiography, CT angiography collateral scoring, and correlation of background data and etiological factors with neurological clinical findings in patients with acute middle cerebral infarction. Methods: A total of 117 patients with MCA (middle cerebral artery) infarction were hospitalized for diagnosis and treatment after undergoing CT angiography within 9 h of symptom onset. Comparative results of Souza’s collateral score system, MRI parameters, and clinical outcomes were determined. Results: According to the Souza CS system, 23 patients were in the malignant profile and 94 in the good profile. There was a statistically significant difference between the malignant and benign profiles in terms of DWI volume, hypoperfused area volume on PWI sequence, white matter assessment using the Fazekas scale, and supratentorial and infratentorial chronic infarct volume on the FLAIR sequence (p < 0.001, p < 0.001, p < 0.001, p < 0.001, p < 0.001). Conclusions: Patients with a malignant profile on CTA may have a larger infarct volume and worse functional outcome. This should be recognized, and these patients should be followed up more carefully and attentively than those with good collateral scores. Full article

The growing urgency of resource-efficient construction in water-stressed and rapidly urbanizing regions necessitates integrated decision support frameworks that move beyond isolated sustainability metrics. This study operationalizes the water-energy nexus within building design evaluation by developing a structured hybrid multi-criteria decision-making (MCDM) framework tailored to the Indian construction context. Unlike conventional sustainability assessments that treat water and energy independently, the proposed approach integrates life cycle-based water consumption, operational and embodied energy demand, environmental impacts, economic feasibility, and project constraints within a unified analytical hierarchy. A Delphi-validated criterion structure comprising five main criteria and twenty sub-criteria is weighted using the Analytic Hierarchy Process (AHP), and ranked using the VIKOR compromise solution method. To strengthen methodological robustness, ranking outcomes are validated across three independent MCDM logics including TOPSIS, PROMETHEE, and COPRAS. The framework evaluates four representative building strategies aligned with Indian regulatory and certification systems (NBC, ECBC, IGBC/GRIHA, and net-zero water-energy design). Using expert-informed weights derived from a Delphi–AHP involving a panel of experienced practitioners, the VIKOR compromise ranking consistently identifies the net-zero alternative as the most favorable option within the evaluated framework. The results are therefore interpreted as an expert-informed assessment demonstrating the applicability of the proposed decision support methodology rather than as statistically generalizable priorities for the entire Indian construction sector. The study contributes by (i) embedding nexus-based resource interdependence into building-level MCDM modeling, (ii) enhancing transparency through explicit benefit-cost classification and decision matrix disclosure, and (iii) demonstrating ranking stability across multiple validation techniques. The proposed framework provides a transferable methodological approach that can be adapted to different regional contexts through locally derived expert inputs. Full article

Given less than five years remaining until the target year for the first phase of China’s dual carbon goals, this paper studies carbon allowance allocation with an in-depth study of Jiangsu Province due to its significant role in driving the Yangtze River Delta’s pioneering achievement of the dual carbon goals. This study considered 2017 (the intermediate target year) as the base year and incorporated socio-economic data such as population, GDP, and the urbanization rate. Then, methods including the entropy weight method, gravity model and social network analysis were applied to classify Jiangsu’s 95 counties. From a regional coordination perspective, carbon governance clusters were constructed with the Shapley value, based on which spatial heterogeneity patterns were analyzed, and a carbon quota allocation was proposed. The findings reveal that: (1) The dominant factors influencing cross-scale carbon reduction capacity at the county level are natural carbon sink capacity (indicator weight: 0.180) and urbanization rate (indicator weight: 0.145). (2) The correlation between carbon reduction factors among different districts and counties exhibits an uneven spatial pattern. And the spatial configuration exhibits a multi-tiered, network-like distribution. (3) Through conducting spatial analysis and spatial grouping, Jiangsu could be divided into 14 county-level carbon governance alliances, with the number of member counties ranging from 4 to 10 within each alliance. (4) The allocation of carbon quotas in Jiangsu exhibits a distinct descending gradient from the southern to the northern regions, which is coupled with the regional economic geography. This is exemplified by the highest quota in Jiangyin (496.46 Mt) in the south and the lowest in Lianyun (34.90 Mt) in the north. It is concluded that two carbon emission reduction pathways should be established as a priority: (a) Tongshan-Gulou (Xuzhou)-Yunlong-Quanshan-Jiawang and (b) Tianning-Jiangyin-Zhangjiagang-Changshu-Taicang-Kunshan. Full article

Background: Acute kidney injury (AKI) is a common and high-risk complication of hospitalization that frequently occurs in patients with chronic cardiometabolic disease. Although heart failure (HF) and diabetes mellitus (DM) are prevalent among hospitalized adults and may differentially modify AKI-associated outcomes, their joint impact on in-hospital risk profiles and cumulative burden remains incompletely characterized. Methods: We conducted a retrospective analysis of adult hospitalizations complicated by AKI using a nationally representative inpatient database. Hospitalizations were classified into four cardiorenal metabolic phenotypes: AKI alone, AKI with HF, AKI with DM, and AKI with both HF and DM. Primary outcomes included in-hospital mortality, dialysis initiation, and mechanical ventilation. Survey-weighted multivariable logistic regression models incorporating HF, DM, and their interaction were used to estimate adjusted associations and model-based predicted probabilities. Adjusted risks were visualized across outcomes, and a composite burden metric was constructed to summarize cumulative in-hospital adverse events. Results: AKI outcomes varied substantially across cardiorenal metabolic phenotypes. HF was consistently associated with higher adjusted mortality and mechanical ventilation risk, whereas DM alone was associated with lower adjusted mortality. A significant interaction between HF and DM was observed regarding dialysis initiation, with a disproportionately higher adjusted risk when both conditions coexisted. Integrated visualization across outcomes demonstrated distinct risk profiles by phenotype, with the combined HF and DM group exhibiting the highest cumulative burden of adverse in-hospital events. Conclusions: Among hospitalizations complicated by AKI, the underlying cardiorenal metabolic status is associated with marked heterogeneity in in-hospital outcomes. HF appears to be a dominant modifier of AKI-associated risk, while DM exerts outcome-specific effects and synergistically increases the risk of dialysis initiation when combined with HF. These findings highlight the importance of incorporating cardiometabolic context into AKI risk stratification approaches and underscore the value of multidimensional in-hospital assessments. Full article

A multi-dimensional quantitative assessment of high-quality development (HQD) in China’s livestock industry and the identification of its main constraints are essential to understanding its current stage and future direction. Guided by global sustainability targets and the United Nations’ Sustainable Development Goals (SDGs), an evaluation system was constructed by this study. This system integrates five key aspects: product safety, output efficiency, resource conservation, environmental friendliness, and regulatory effectiveness. Using provincial panel data from China for 2013–2022, this research applies the entropy-weighted TOPSIS method, kernel density estimation (KDE), and an obstacle degree model for analysis, the goal is to support food security and foster environmentally sustainable growth. The findings indicate the following: (1) Notable inter-provincial disparities exist in the HQD of China’s livestock industry, revealing a spatial pattern of “leading in the east, stable in the center, and lagging in the west.” (2) The nationwide evolution exhibits a “convergence followed by divergence” pattern: from 2013 to 2017, the primary peak of the KDE rose and its width narrowed; from 2018 to 2022, the primary peak declined and its width widened, indicating that inter-provincial disparities first narrowed and then expanded. At the regional level, the development pattern is characterized by eastern polarization, central stability, and western lock-in. (3) Obstacle factor analysis identifies product safety and environmental friendliness as the principal constraints on HQD in the livestock industry. Addressing these bottlenecks is crucial for ensuring the supply of livestock products (SDG 2: Zero Hunger), promoting resource conservation and green production (SDG 12: Responsible Consumption and Production), and alleviating the ecological and environmental pressures of the livestock industry (SDG 15: Protection of Terrestrial Ecosystems). The challenges related to resources, the environment, and quality safety confronting China’s livestock industry are common among developing countries. Consequently, the evaluation framework established in this study can offer methodological references for relevant nations. Full article

Wheel–rail coupled excitation introduces strong low-frequency modulation, random impact interference, and broadband background noise into the vibration system of locomotive gearboxes, causing early weak bearing fault features to become submerged and making traditional deconvolution methods insufficient for effective enhancement. To address this challenge, this study proposes an adaptive parameter optimization method for MCKD based on the weighted envelope spectrum factor (WESF). WESF integrates the Hoyer index, kurtosis, and envelope spectrum energy to jointly characterize sparsity, impulsiveness, and periodicity of signal components. By using WESF as the fitness function, the sparrow search algorithm (SSA) is employed to simultaneously optimize the key MCKD parameters L, T, and M, enabling optimal enhancement of weak periodic impacts. To further mitigate modal aliasing caused by wheel–rail excitation, the original signal is first adaptively decomposed using successive variational mode decomposition (SVMD), and modes with WESF values above the average are selected for signal reconstruction. The reconstructed signal is subsequently enhanced via SSA–MCKD, and fault characteristic frequencies are extracted using envelope spectrum analysis. Experimental validation using gearbox bearing data collected under 40, 50, and 60 Hz operating conditions shows that the proposed method achieves fault feature coefficient (FFC) values of 12.8%, 7.5%, and 7.2%, respectively—representing an average improvement of approximately 156% compared with traditional methods (average FFC of 3.6%). These results demonstrate that the proposed SVMD–WESF–SSA–MCKD approach can significantly enhance weak periodic impact features under strong background noise and wheel–rail excitation, exhibiting strong practical applicability for engineering implementation. Full article

This study presents an optimization tool inspired by bone remodeling principles to address the high computational costs of truss topology optimization. Additionally, a new structural analysis method based on primary forces is proposed to overcome the kinematic stability problem. The strategy developed to obtain the optimal topology optimizes the initial dense ground structure in two stages. In Phase I, unnecessary members in the system are filtered to determine the “primary candidate members”; in Phase II, the final topology is reached through this refined subset. The algorithm performs an effective search in the design space by simulating biological processes that link the rate of mass change in the bone matrix to mechanical stimuli. Numerical results demonstrate high accuracy, as shown by the analytical solution of the 2D Michell truss, with a difference of 1.02%. The results show high consistency with reference studies, providing, in some cases, alternative topologies with the same weight and stiffness as given in the benchmarks. The proposed method achieves significant improvements in computational efficiency, reducing processing times for larger systems by 10 to over 250 times compared to literature benchmarks. Full article

In this paper, a moving horizon estimation (MHE)-based state estimation problem is studied for asynchronous multi-rate nonlinear systems. First, the asynchronous multi-rate system is transformed into a synchronous system at measurement sampling points through pseudo-measurement synchronization modeling. Secondly, a MHE strategy with a time-discounted quadratic objective is proposed. Under the detectability assumption, the exponential stability of the proposed MHE is established via the Lyapunov method, and the corresponding linear matrix inequality (LMI) constraints are derived. Moreover, to address the model mismatch after synchronization, a deep learning-based framework is proposed to approximate and learn the weighting parameters of the MHE. Then, barrier-function regularization is introduced to enforce the aforementioned LMI feasibility conditions, keeping the learned weights within the feasible region throughout training. Finally, the result is illustrated by a target tracking example. Full article

Background: Post-bariatric abdominoplasty is associated with a high risk of bleeding and wound complications due to extensive tissue resection and impaired tissue quality. Tranexamic acid (TXA) reduces perioperative bleeding in multiple surgical disciplines, but evidence in massive-weight-loss abdominoplasty is limited. The aim of our study was to evaluate the association between perioperative TXA use and bleeding-related and surgical outcomes in post-bariatric abdominoplasty. Methods: This retrospective cohort study included 97 patients undergoing post-bariatric abdominoplasty, of whom 49 received perioperative TXA and 48 did not. The primary outcome was a composite of bleeding-related complications within 30 days, including hematoma, clinically relevant bleeding, or reoperation. Secondary outcomes included overall and specific surgical site complications, drain output and duration, length of hospital stay, and perioperative hemoglobin changes. Multivariable regression analyses adjusted for body mass index, abdominoplasty type, and year of surgery. Results: Bleeding-related complications were significantly lower in the TXA group compared with controls (4.1% vs. 33.3%; unadjusted OR 0.09, 95% CI 0.02–0.40; p < 0.001). This association remained significant after adjustment (adjusted OR 0.13, 95% CI 0.03–0.68; p = 0.016). TXA use was associated with lower cumulative drain output (median 200 vs. 382.5 mL; p < 0.001) and shorter drainage duration (median 4 vs. 5 days; p < 0.001). Overall complications were reduced in the TXA group (42.9% vs. 66.7%; p = 0.025), driven by fewer wound healing disturbances. Hemoglobin changes, seroma, and infection rates were similar between groups. Conclusions: Perioperative TXA use in post-bariatric abdominoplasty is associated with significantly fewer bleeding-related and wound complications without increased adverse effects, supporting its use in this high-risk population. Full article

High-resolution remote sensing (RS) images exhibit complex backgrounds, large intra-class variability, and low inter-class differences, posing substantial challenges for semantic segmentation. Although existing class-level contextual modeling methods partially alleviate these issues, they often overlook the importance of accurate and discriminative class representations and fail to effectively handle hard samples during training. To address these limitations, we propose CRECA-Net, a class representation-enhanced class-aware network designed from two complementary perspectives: class prototype refinement and difficulty-aware learning. Specifically, we introduce a class prototype refinement (CPR) module that improves class representations through pixel selection, confidence-aware contribution weighting, and an inter-class prototype separation loss, yielding more reliable and discriminative class centers. In addition, class-level context aggregation (CLCA) modules capture pixel-to-class prototype correlations via cross-attention to inject class-aware semantics into decoder features, thereby reducing interference from cluttered backgrounds and visually similar categories. Furthermore, a difficulty-aware (DA) loss dynamically estimates pixel-wise difficulty and redistributes the loss weights within each image, gradually shifting the learning focus from easy to hard samples while maintaining training stability. Extensive experiments on two benchmark RS segmentation datasets demonstrate that CRECA-Net consistently outperforms state-of-the-art methods across multiple evaluation metrics. Full article

Adverse weather and low illumination remain major challenges for autonomous driving perception, where semantic segmentation must stay reliable despite severe appearance degradation. In unsupervised domain adaptation without target annotations, self-training is widely used, but it is often limited by the inconsistent quality of teacher-generated pseudo labels across samples, regions, and training stages. This paper presents RaDA, a reliability-aware self-training framework that regulates pseudo supervision at three levels. First, a progressive exposure strategy determines which target images are admitted for training. Second, spatial reliability weighting suppresses gradients from degraded regions while retaining informative supervision. Third, adaptive teacher update scheduling stabilizes pseudo label generation over time. Experiments on real-world adverse driving benchmarks show that RaDA improves robustness, training stability, and cross-dataset generalization compared with strong baselines. Compared with the previous state-of-the-art method MIC, RaDA achieves mIoU gains of 10.6 percentage points on Foggy Zurich and 8.8 percentage points on the Foggy Driving benchmark. These results indicate that explicit reliability regulation can strengthen self-training domain adaptation for semantic segmentation in autonomous driving under challenging environmental conditions. Full article