Search Results (3,733)

In the context of privacy-preserving face recognition systems, entropy plays a crucial role in determining the efficiency and security of computational processes. However, existing schemes often encounter challenges such as inefficiency and high entropy in their computational models. To address these issues, we propose a privacy-preserving face recognition method based on the Face Feature Coding Method (FFCM) and symmetric homomorphic encryption, which reduces computational entropy while enhancing system efficiency and ensuring facial privacy protection. Specifically, to accelerate the matching speed during the authentication phase, we construct an N-ary feature tree using a neural network-based FFCM, significantly improving ciphertext search efficiency. Additionally, during authentication, the server computes the cosine similarity of the matched facial features in ciphertext form using lightweight symmetric homomorphic encryption, minimizing entropy in the computation process and reducing overall system complexity. Security analysis indicates that critical template information remains secure and resilient against both passive and active attacks. Experimental results demonstrate that the facial authentication efficiency with FFCM classification is 4% to 6% higher than recent state-of-the-art solutions. This method provides an efficient, secure, and entropy-aware approach for privacy-preserving face recognition, offering substantial improvements in large-scale applications. Full article

This study investigates the aerodynamic and aeroacoustics behavior of automotive cooling modules in both conventional internal combustion engine (ICE) vehicles and electric vehicles (EVs), with a particular focus on installation effects. Numerical simulations based on the Lattice Boltzmann Method (LBM) are conducted to analyze noise generation mechanisms and flow characteristics across four configurations. The study highlights the challenges of adapting classical cooling module components to EV setups, emphasizing the influence of heat exchanger (HE) placement and duct geometry on noise levels and flow dynamics. The results show that the presence of the HE smooths the upstream flow, improves rotor loading distribution and disrupts long, coherent vortical structures, thereby reducing tonal noise. However, the additional resistance introduced by the HE leads to increased rotor loading and enhanced leakage flow through the shroud-rotor gap. Despite these effects, the overall sound pressure level (OASPL) remains largely unchanged, maintaining a similar magnitude and dipolar directivity pattern as the configuration without the HE. In EV modules, the inclusion of ducts introduces significant flow disturbances and localized pressure fluctuations, leading to regions of high flow rate and rotor loading. These non-uniform flow conditions excite duct modes, resulting in troughs and humps in the acoustic spectrum and potentially causing resonance at the blade-passing frequency, which increases the amplitude in the lower frequency range. Analysis of the loading force components reveals that rotor loading is primarily driven by thrust forces, while duct loading is dominated by lateral forces. Across all configurations, fluctuations at the leading and trailing edges of the rotor are observed, originating from the blade tip and extending to approximately mid-span. These fluctuations are more pronounced in the EV module, identifying it as the dominant source of pressure disturbances. The numerical results are validated against experimental data obtained in the anechoic chamber at the University of Sherbrooke and show good agreement. The relative trends are accurately predicted at lower frequencies, with slight over-prediction, and closely match the experimental data at mid-frequencies. Full article

To address the limitations of traditional non-local mean (NLM) denoising algorithms in terms of neighborhood similarity metrics, weight calculation, and computational efficiency, this paper proposed a structural-tensor-driven and dynamic window-based fast non-local mean denoising algorithm with dual kernel weighting. First, a Gaussian–Tukey dual-kernel weighting function was designed to optimize similarity metrics. Then, spatial neighborhood features were adopted. By measuring both grayscale similarity and spatial correlation, the weight distribution rationality was further enhanced. Second, structural tensor eigenvalues were used to quantify regional structural properties. A dynamic window allocation function was designed to adaptively match search window sizes to different image regions. Finally, an integral image acceleration mechanism was proposed, significantly improving algorithm execution efficiency. Experimental results demonstrated that the proposed algorithm achieved both excellent denoising performance and edge/texture preservation capabilities. In high-noise environments, its Peak Signal-to-Noise Ratio (PSNR) outperformed the Gauss kernel non-local mean algorithm by an average of 1.96 dB, while Structural Similarity (SSIM) improved by an average of 5.7%. Moreover, the algorithm’s execution efficiency increased by approximately 7–11 times, indicating strong potential for real-time application in digital image processing. Full article

The aim of this study was to assess the heart rate variability (HRV) at rest and during active orthostatic challenge in long COVID patients with orthostatic intolerance symptoms (dizziness, pre-syncope, and syncope). We performed a cross-sectional, observational, comparative study of 60 subjects of both sexes, aged 18 to 60 years (31 met the criteria of long COVID, 15 were infected individuals without symptoms, and 14 who had neither infection nor symptoms formed the age-matched control group). HRV was obtained from continuous electrocardiograms in a supine position and active standing with spontaneous breathing. The time from SARS-CoV-2 infection to testing in the COVID-19 group was 573 ± 289 days. The resting (supine position) values of SDNN, RMSSD, SD1, and SD2 were lower in long COVID patients than in control participants, while all other HRV indexes were similar between groups. In response to active standing, both groups had similar changes in all HRV indices. In conclusion, an active orthostatic test was not able to exhibit an autonomic dysregulation in these patients with long COVID, suggesting that cardiac autonomic modulation may have recovered due to the long time that elapsed after SARS-CoV-2 infection. Full article

In this article, we present a tool for matching resumes to job posts and vice versa (job post to resumes). With minor modifications, it may also be adapted to other domains where text matching is necessary. This tool may help organizations save time during the hiring process, as well as assist applicants by allowing them to match their resumes to job posts they have selected. To achieve text matching without any model training (zero-shot matching), we constructed dynamic structured prompts that consisted of unstructured and semi-structured job posts and resumes based on specific criteria, and we utilized the Chain of Thought (CoT) technique on the Mistral model (open-mistral-7b). In response, the model generated structured (segmented) job posts and resumes. Then, the job posts and resumes were cleaned and preprocessed. We utilized state-of-the-art sentence similarity models hosted on Hugging face (nomic-embed-text-v1-5 and google-embedding-gemma-300m) through inference endpoints to create sentence embeddings for each resume and job post segment. We used the cosine similarity metric to determine the optimal matching, and the matching operation was applied to eleven different occupations. The results we achieved reached up to 87% accuracy for some of the occupations and underscore the potential of zero-shot techniques in text matching utilizing LLMs. The dataset we used was from indeed.com, and the Spring AI framework was used for the implementation of the tool. Full article

Social media has become the core channel through which people communicate, and the important role of influencer marketing in creating a fan base for brands is widely recognized. Grounded in Source Credibility, Homophily Theory and Signaling Theory, the purpose of this study is to investigate how influencer efficacy affects the fan effect of green food brands under digital social media. This paper adopts a quantitative research method. A cross-sectional survey was conducted on the Wenjuanxing platform and collected 417 valid responses from consumers who had previously purchased green food based on an influencer’s recommendation. A conceptual model was tested through the structural equation modelling procedure. The results showed that professionalism (β = 0.166, p = 0.011), trustworthiness (β = 0.291, p < 0.001), and similarity (β = 0.267, p < 0.001) had positive effects on perceived quality. Furthermore, perceived quality (β = 0.333, p < 0.001) significantly promoted the formation of the brand fan effect and partially mediated the effects of these characteristics of influencers on the brand fan effect. This study provides new insight into the fan effect of green food brands and also provides a theoretical basis for green food companies to accurately match their brands with suitable influencers, enhance the brand fan effect, and rationally formulate operational strategies. Full article

Mutations in the Piezo1 mechanosensor are associated with blood cell anomalies. The objective of our study was to explore the role of Piezo1 in the development and function of the megakaryocyte (MK) lineage. To this end, PF4-Cre mice, bearing Cre recombinase under the control of the Pf4 gene promoter—which drives expression to hematopoietic progenitors and to the MK/platelet lineage—were crossbred with Piezo1-floxed mice to generate Piezo1 knockout (KO) mice. In our results, the hematopoietic stem cell (HSC) count—including Multipotent Progenitors 2 (MPP2) progenitors that give rise to MKs—tended to be augmented in KO mice, while the level of MPP3 progenitors that give rise to white blood cells (WBCs) tended to be reduced, as compared to matching controls. The level of circulating WBCs was significantly reduced in the KO mice compared to controls. In addition, while platelet count was modestly elevated, platelet activation response was reduced in Piezo1 KO mice compared to controls. MK levels and ploidy were similar in both groups. Baseline serum pro-and anti-inflammatory cytokine profiles were also similar in the two experimental groups. However, upon LPS challenge, there was a significant reduction in IL-6 and INF-γ levels in the sera of Piezo1 KO mice compared to controls. Our findings point to an immunoregulatory and thrombotic potential of Piezo1 in relatively rare bone marrow cells, along with an ability to modulate WBC count. Full article

Most visual localization (VL) methods typically assume that keypoints in the query image are detected with the same algorithm as those stored in the reference map. This poses a serious limitation, as new and better detectors may progressively appear, and we would like to ensure the interoperability and coexistence of cameras with heterogeneous detectors in a single map representation. While rebuilding the map with new detectors might seem a solution, it is often impractical, as original images may be unavailable or restricted due to data privacy constraints. In this paper, we address this challenge with two main contributions. First, we introduce and formalize the problem of cross-detector VL, in which the inherent spatial discrepancies between keypoints from different detectors hinder the process of establishing correct correspondences when relying strictly on the similarity of descriptors for matching. Second, we propose CoplaMatch, the first approach to solve this problem by relaxing strict descriptor similarity and imposing geometric coplanarity constraints. The latter is achieved by leveraging 2D homographies between groups of query and map keypoints. This process involves segmenting planar patches, which is performed offline once for the map, and also in the query image, which adds an extra computational overhead to the VL process, although we demonstrated in our experiments that this does not hinder the online applicability. We extensively validate our proposal through experiments in indoor environments using real-world datasets, demonstrating its effectiveness against two state-of-the-art methods by enabling accurate localization in cross-detector scenarios. Additionally, our work validates the feasibility of cross-detector VL and opens a new direction for the long-term usability of feature-based maps. Full article

Background/Objectives: Lactoferrin, a key bioactive component in human milk, may bridge functional gaps in infant formula; however, its long-term effects on growth and the gut microbiota in term infants remain underexplored, particularly in real-world settings. Methods: This real-world evidence (RWE) study assessed the impact of lactoferrin-fortified formula (LF) on infant growth, the gut microbiota, and feeding tolerance compared with control formula (CF) and exclusive breastfeeding (BF). After propensity score matching (PSM) for maternal education level and infant age, 111 matched Chinese infants (37 per group: LF, CF, and BF; age: 6–12 months) were analyzed. Growth was evaluated using WHO Z-scores (WAZ, LAZ, WLZ, and zBMI). The gut microbiota was profiled via 16S rRNA sequencing (n = 81). Feeding challenges were quantified using the Montreal Children’s Hospital Feeding Scale (MCH-FS). Results: The LF group exhibited significantly higher length-for-age Z-scores (LAZ) compared with both the BF and CF groups (p < 0.001), indicating superior linear growth. LF infants also showed reduced MCH-FS scores (18.0 vs. 36.2 in CF; p < 0.001), signifying fewer feeding difficulties. Gut microbiota analysis revealed enrichment of Bifidobacterium breve and butyrate-producing taxa (e.g., Faecalibacterium and Ruminococcaceae), higher alpha diversity, and metabolic divergence, involving enhanced lysine fermentation to acetate/butyrate in LF infants, suggesting a higher level of short-chain fatty acid (SCFA) production. Beta diversity analysis demonstrated that the LF microbiota clustered close to BF. Conclusions: Lactoferrin-fortified formula was associated with improved linear growth and feeding tolerance while shaping a healthy gut microbiota, showing similarities to breastfed infants’ microbiota. These findings support LF fortification as a strategy to improve functional outcomes in formula-fed infants. Full article

Semantic matching is essential for understanding natural language, but traditional models like BERT face challenges with random masking strategies, limiting their ability to capture key information. Additionally, BERT’s sentence vectors may “collapse,” making it difficult to distinguish between different sentences. This paper introduces a domain-finetuned semantic matching framework that uses dynamic masking and contrastive learning techniques to address these issues. The dynamic masking strategy enhances the model’s ability to retain critical information, while contrastive learning improves sentence vector representations using a small amount of unlabeled text. This approach helps the model better align with the needs of various downstream tasks. Experimental results show that after private domain training, the model improves semantic similarity between entities by 16.9%, outperforming existing models. It also demonstrates an 8.0% average improvement in semantic matching for diverse text. Performance metrics such as A@1, A@3, and A@5 are at least 26.1% higher than those of competing models. For newly added entities, the model achieves a 44.3% average improvement, consistently surpassing other models by at least 30%. These results collectively validate the effectiveness and superiority of the proposed framework in domain-specific semantic matching tasks. Full article

Objective: To assess the oncological outcomes of laparoscopic versus open radical hysterectomy (RH) in patients with early-stage cervical cancer using propensity score–matched analysis. Methods: We conducted a retrospective cohort study of 1244 patients who underwent RH with pelvic lymphadenectomy at Chiang Mai University Hospital between 2003 and 2019. Of these, 82 patients received a laparoscopic approach (LAP) and 1162 underwent open radical hysterectomy. Propensity-score matching was performed in a 1:4 ratio using a caliper of 0.2 standard deviations to achieve balance between groups. Overall survival (OS) and progression-free survival (PFS) were analyzed with Kaplan–Meier curves and the log-rank test. Subgroup analysis was conducted based on tumor size (≤ 2 cm vs. > 2 cm). In addition, multivariable Cox proportional hazards models incorporating all relevant clinical and pathological variables were applied to the overall cohort to assess independent predictors of OS and PFS. Results: After matching, 72 LAP RH cases were compared with 279 open RH cases, showing well-balanced baseline features. At 5 years, OS was nearly the same between the LAP and the open groups (95.8% vs. 95.5%; p = 0.95), and PFS was also similar (92.3% vs. 93.8%; p = 0.85). Subgroup analyses demonstrated that LAP RH did not result in a survival disadvantage for tumors ≤ 2 cm or > 2 cm. In multivariable Cox analysis, surgical approach was not an independent predictor of (HR 0.83, 95% CI 0.40–1.71, p = 0.61) or PFS (HR 1.12, 95% CI 0.44–2.84, p = 0.82). Conclusions: In our single-center cohort analyzed using propensity score matching, LAP RH showed long-term oncological outcomes comparable to those of open RH. These results support LAP RH as a safe surgical option for selected patients with early-stage cervical cancer within our setting, where procedures were performed by experienced surgeons following standardized techniques. Further evaluation in diverse clinical contexts is still needed. Full article

Generative artificial intelligence (GenAI) is rapidly permeating research practices, yet knowledge about its use and topical profile remains fragmented across tools and disciplines. In this study, we present a cross-disciplinary map of GenAI research based on the Web of Science Core Collection (as of 4 November 2025) for the ten tool lines with the largest number of publications. We employed a transparent query protocol in the Title (TI) and Topic (TS) fields, using Boolean and proximity operators together with brand-specific exclusion lists. Thematic similarity was estimated with the Jaccard index for the Top–50, Top–100, and Top–200 sets. In parallel, we computed volume and citation metrics using Python and reconstructed a country-level co-authorship network. The corpus comprises 14,418 deduplicated publications. A strong concentration is evident around ChatGPT, which accounts for approximately 80.6% of the total. The year 2025 shows a marked increase in output across all lines. The Jaccard matrices reveal two stable clusters: general-purpose tools (ChatGPT, Gemini, Claude, Copilot) and open-source/developer-led lines (LLaMA, Mistral, Qwen, DeepSeek). Perplexity serves as a bridge between the clusters, while Grok remains the most distinct. The co-authorship network exhibits a dual-core structure anchored in the United States and China. The study contributes to bibliometric research on GenAI by presenting a perspective that combines publication dynamics, citation structures, thematic profiles, and similarity matrices based on the Jaccard algorithm for different tool lines. In practice, it proposes a comparative framework that can help researchers and institutions match GenAI tools to disciplinary contexts and develop transparent, repeatable assessments of their use in scientific activities. Full article

Leuconostoc mesenteroides is a key microorganism in food biotechnology, valued for its production of flavor-forming metabolites and exopolysaccharides, and its inclusion in starter cultures and biocatalytic systems. However, the application of advanced genetic tools to L. mesenteroides remains hindered by multiple barriers, including inefficient DNA transfer, elevated endogenous nuclease activity, and restriction–modification systems sensitive to plasmid methylation patterns. As a result, even widely accepted electroporation methodologies often yield inconsistent or irreproducible transformation results, limiting the strain’s amenability to metabolic engineering and synthetic biology applications. In this study, a reproducible electroporation protocol for the L. mesenteroides strain H32-02 Ksu is developed and experimentally validated. The protocol concept relies on the sequential optimization of key process steps: targeted weakening of the cell wall followed by osmotic protection, the development of a gentle electrical stimulus that ensures membrane permeability without critical damage, and the creation of recovery conditions that minimize loss of viability and degradation of incoming DNA. Matching plasmid methylation to the recipient’s restriction profile proved critical: choosing a source for plasmid DNA production with a compatible methylation pattern dramatically increased the likelihood of successful transformation. In our case, the selection of an E. coli strain with a more suitable methylation profile increased the yield of transformants by 3.5 times. It was also shown that reducing the pulse voltage increase transformant number by 3 times. The combined optimization resulted in an approximately 40-fold increase in transformation efficiency compared to the baseline level and, for the first time, provided consistently reproducible access to transformants of this strain. The highest transformation efficiency was achieved: 8 × 10² CFU µg⁻¹ DNA. The presented approach highlights the strain-specificity of barriers in Leuconostoc and forms a technological basis for constructing strains with desired properties, expressing heterologous enzymes, and subsequently scaling up bioprocesses in food and related industries. The methodological principles embodied in the protocol are potentially transferable to other lactic acid bacteria with similar limitations. Full article

Evaluating the quality of fine-grained 3D shape editing, such as adjusting a vehicle’s roof length or wheelbase, is essential for assessing generative models but remains challenging. Most existing metrics depend on auxiliary regressors or large-scale human evaluations, which may introduce bias, reduce reproducibility, and increase evaluation cost. To address these issues, a reference-free metric for evaluating fine-grained 3D shape editing is proposed. The method is based on the Rich-Attribute Sufficiency Assumption (RASA), which posits that when a geometric attribute set is sufficiently comprehensive, models with the same attribute vector should exhibit nearly identical shapes. Following this assumption, the dataset itself serves as a validation source: each source model is edited to match a small set of target attribute vectors, and the post-editing similarity to the targets reflects the editor’s accuracy and stability. Reproducible indicators are defined, including mean similarity, variation across targets, and calibration with respect to attribute distance. Empirical validation demonstrates the effectiveness of the proposed metric, showing approximately 9% degradation under semantic perturbations and less than 2% variation across different target-sampling settings, confirming both its discriminative sensitivity and robustness. This framework provides a low-cost, regressor-free benchmark for fine-grained editing and establishes its applicability through an explicit assumption and evaluation protocol. Full article

Background/Objectives: Antibiotic shortages are a growing problem which harms patient safety, increases healthcare costs, and accelerates antibiotic resistance. Based on the example of Sweden, the paper aims to illustrate and discuss how an organized collaboration platform can devise several policy interventions against shortages of antibiotics for human use. Methods: We describe how the multi-sectoral collaboration Platinea (Platform for Innovation of Existing Antibiotics) first identified, structured, and prioritized the causes of antibiotic shortages, and then identified and prioritized a set of policy solutions matching these causes. The specific methods applied include expert elicitation, interactive workshops, focus groups, and multi-criteria decision processes. Results: After an overview of about 60 causes of antibiotic shortages, we relate them to 10 prioritized solutions including, e.g., increased inventories, central coordination, integrated IT systems, increased unit prices, yearly fixed payments, and Nordic collaboration in purchasing and production. Conclusions: We propose a process with six steps to devise policy solutions by involving a multi-sectoral stakeholder group: open brainstorming of the problem’s causes, framing them into a clear structure, selecting and prioritizing key causes, matching causes and solutions, and devising prioritization mechanisms about emerging solutions. This approach can be applied to other national contexts and similar policy issues. Full article