Search Results (33,087)

Seconds count differently for people in danger. We present a real-time streaming pipeline for audio-based detection of hazardous life events affecting life and property. The system operates online rather than as a retrospective analysis tool. Its objective is to reduce the latency between the occurrence of a crime, conflict, or accident and the corresponding response by authorities. The key idea is to map reality as perceived by audio into a written story and question the text via a large language model. The method integrates streaming, zero-shot algorithms in an online decoding mode that convert sound into short, interpretable tokens, which are processed by a lightweight language model. CLAP text–audio prompting identifies agitation, panic, and distress cues, combined with conversational dynamics derived from speaker diarization. Lexical information is obtained through streaming automatic speech recognition, while general audio events are detected by a streaming version of Audio Spectrogram Transformer tagger. Prosodic features are incorporated using pitch- and energy-based rules derived from robust F0 tracking and periodicity measures. The system uses a large language model configured for online decoding and outputs binary (YES/NO) life-threatening risk decisions every two seconds, along with a brief justification and a final session-level verdict. The system emphasizes interpretability and accountability. We evaluate it on a subset of the X-Violence dataset, comprising only real-world videos. We release code, prompts, decision policies, evaluation splits, and example logs to enable the community to replicate, critique, and extend our blueprint. Full article

Accurate ultra-short-term wind power forecasting is vital for the secure and economic operation of power systems with high renewable penetration. Conventional models, however, struggle with multi-scale frequency feature extraction, dynamic cross-variable dependencies, and simultaneously capturing local fluctuations and global trends. This study proposes a novel hybrid framework termed VMD–ALIF–GraphBlock–MLLA–Transformer. A dual-decomposition strategy combining variational mode decomposition and adaptive local iterative filtering first extracts dominant periodic components while suppressing high-frequency noise. An adaptive GraphBlock with MixHop convolution then models structured and time-varying inter-variable dependencies. Finally, a multi-scale linear attention-enhanced Mamba-like module and Transformer encoder jointly capture short- and long-range temporal dynamics. Experiments on a real wind farm dataset with 10-min resolution demonstrate substantial superiority over State-of-the-Art baselines across 1-, 4-, and 8-step forecasting horizons. SHAP analysis further confirms excellent consistency with underlying physical mechanisms. The proposed framework provides a robust, accurate, and highly interpretable solution for intelligent wind power forecasting. Full article

Low-altitude Unmanned Aerial Vehicle (UAV) detection using LiDAR range images faces persistent challenges. These include sparse features for long-range targets, large scale variations caused by viewpoint changes, and severe interference from complex backgrounds. To address these issues, we propose an improved detection framework based on YOLOv10. First, we design a Swin-Conv hybrid module that combines sparse attention with deformable convolution. This module enables the network to focus on informative regions and adapt to target geometry. These capabilities jointly strengthen feature extraction for sparse, long-range targets. Second, we introduce Attentional Feature Fusion (AFF) in the neck to replace naïve feature concatenation. AFF employs multi-scale channel attention to softly select and adaptively weight features from different levels, improving robustness to multi-scale targets. In addition, we systematically study how the viewpoint distribution in the training set affects performance. The results show that moderately increasing the proportion of low-elevation-view samples significantly improves detection accuracy. Experiments on a self-built simulated LiDAR range-image dataset demonstrate that our method achieves 88.96% mAP at 54.2 FPS, which is 4.78 percentage points higher than the baseline. Deployment on the Jetson Orin Nano edge device further validates the model’s potential for real-time applications. The proposed method remains robust under noise and complex backgrounds. The proposed approach achieves an effective balance between detection accuracy and computational efficiency, providing a reliable solution for real-time target detection in complex low-altitude environments. Full article

With smart buildings being widely adopted in urban digital transformation, interactive semantic question answering (QA) systems serve as a crucial bridge between user intent and environmental response. However, they still face substantial challenges in semantic understanding and dynamic reasoning. Most existing systems rely on static frameworks built upon Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG), which suffer from rigid prompt design, breakdowns in multi-step reasoning, and inaccurate generation. To tackle these issues, we propose LoopRAG, a multi-agent RAG architecture that incorporates a Plan–Do–Check–Act (PDCA) closed-loop optimization mechanism. The architecture formulates a dynamic QA pipeline across four stages: task parsing, knowledge extraction, quality evaluation, and policy feedback, and further introduces a semantics-driven prompt reconfiguration algorithm and a heterogeneous knowledge fusion module. These components strengthen multi-source information handling and adaptive reasoning. Experiments on HotpotQA, MultiHop-RAG, and an in-house building QA dataset demonstrate that LoopRAG significantly outperforms conventional RAG systems in key metrics, including context recall of 90%, response relevance of 72%, and answer accuracy of 88%. The results indicate strong robustness and cross-task generalization. This work offers both theoretical foundations and an engineering pathway for constructing trustworthy and scalable semantic QA interaction systems in smart building settings. Full article

Cyberbullying has become a pervasive threat on social media, impacting the safety and wellbeing of users worldwide. Most existing studies focus on monolingual content, limiting their applicability to online environments. This study aims to develop an approach that accurately detects abusive content in bilingual settings. Given the large volume of online content in English and Arabic, we propose a bilingual cyberbullying detection approach designed to deliver efficient, scalable, and robust performance. Several datasets were combined, processed, and augmented before proposing a cyberbullying identification approach. The proposed model (FusionBullyNet) is based on fine-tuning of two transformer models (RoBERTa-base + bert-base-arabertv02-twitter), attention-based fusion, gradually unfreezing the layers, and label smoothing to enhance generalization. The test accuracy of 0.86, F1 scores of 0.83 for bullying and 0.88 for no bullying, and an overall ROC-AUC of 0.929 were achieved with the proposed approach. To assess the robustness of the proposed models, several multilingual models, such as XLM-RoBERTa-Base, Microsoft/mdeberta-v3-base, and google-bert/bert-base-multilingual-cased, were also trained in this study, and all achieved a test accuracy of 0.84. Furthermore, several machine learning models were trained in this study, and Logistic Regression, XGBoost Classifier, and Light GBM Classifier achieved the highest accuracy of 0.82. These results demonstrate that the proposed approach provides a reliable, high-performance solution for cyberbullying detection, contributing to safer online communication environments. Full article

Background: Chest X-ray (CXR) imaging is crucial for diagnosing thoracic abnormalities; however, the rising demand burdens radiologists, particularly in resource-limited settings. Method: We used large-scale, diverse public CXR datasets with noisy labels to train general-purpose deep learning models (ResNet, DenseNet, EfficientNet, and DLAD-10) for multi-label classification of thoracic conditions. Uncertainty quantification was incorporated to assess model reliability. Performance was evaluated on both internal and external validation sets, with analyses of data scale, diversity, and fine-tuning effects. Result: EfficientNet achieved the highest overall area under the receiver operating characteristic curve (0.8944) with improved sensitivity and F1-score. Moreover, as training data volume increased—particularly using multi-source datasets—both diagnostic performance and generalizability were enhanced. Although larger datasets reduced predictive uncertainty, conditions such as tuberculosis remained challenging due to limited high-quality samples. Conclusions: General-purpose deep learning models can achieve robust CXR diagnostic performance when trained on large-scale, diverse public datasets despite noisy labels. However, further targeted strategies are needed for underrepresented conditions. Full article

This study investigates the tri-objective optimization of underground metro station structures, considering structural reliability, life-cycle economic cost, and annualized carbon emissions simultaneously. Using a representative metro station in Guangzhou as a case study, a multi-objective optimization framework is developed. The model defines structural failure probability, discounted life-cycle cost, and average annual carbon emissions as the primary objectives, with decision variables including concrete strength, cover thickness, the use of epoxy-coated reinforcement, and various maintenance/repair strategies. Material quantities are calculated through Building Information Modeling (BIM), while cost–carbon relationships are derived from industry price data and carbon emission factors. An improved multi-objective particle swarm optimization algorithm (OMOPSO) is used to derive the Pareto-optimal front. Case study results show that increasing cover thickness significantly improves durability and reduces carbon emissions with only moderate cost increases. In contrast, epoxy-coated reinforcement is excluded from the Pareto set due to its high cost under the given conditions. To facilitate practical decision-making, a weight-based solution selection method is introduced, and sensitivity analyses are performed to assess the model’s robustness. The study concludes by emphasizing the framework’s applicability and limitations: the findings are specific to the case context and require recalibration for use in other sites or construction practices. This research contributes by integrating durability, cost, and carbon considerations into an engineering-level optimization workflow, providing valuable decision support for sustainable metro station design. Full article

Background: Stenotrophomonas maltophilia is an intrinsically multidrug-resistant, biofilm- forming, non-fermenter increasingly implicated in hospital-acquired infections. Evidence from non-cystic fibrosis populations, especially in the Middle East, remains sparse. Methods: We conducted a retrospective observational cohort study at a tertiary academic center (Al-Khobar, Saudi Arabia) spanning 1 May 2001–30 April 2023. Hospitalized adults (≥18 years) with culture-confirmed, clinically diagnosed S. maltophilia infection and ≥72 h of antibiotic therapy were included. The primary outcome was all-cause mortality (14-day, 30-day, 1-year). Secondary outcomes were clinical response, microbiological eradication, and infection recurrence. Predictors of 30-day mortality were assessed using multivariable logistic regression; 14-day mortality was analyzed by Kaplan–Meier/log-rank according to susceptibility-guided versus alternative therapy. Results: Of 539 patients with positive cultures, 436 met the inclusion criteria. Mean age was 60.5 ± 19.3 years; 62.2% were male. Most infections were hospital-acquired (92.9%); pneumonia composed 64.7% and bloodstream infection 15.4%. Polymicrobial growth occurred in 55.5% (predominantly Gram-negative co-isolation). Susceptibility was 95.1% to trimethoprim–sulfamethoxazole, 76.4% to levofloxacin, and 43.6% to ceftazidime. Mortality at 14 days, 30 days, and 1 year was 22.8%, 37.9%, and 57.2%, respectively. On multivariable modelling, intensive care unit (ICU) admission, leukocytosis, neutrophilia, anemia, and thrombocytopenia independently predicted 30-day mortality. Susceptibility-guided therapy was associated with improved 14-day survival (log-rank p = 0.033). Conclusions: In this large, long-running non-cystic fibrosis cohort, host acuity and early alignment of treatment to susceptibility data were dominant drivers of outcome. High polymicrobial burden and limited reliably active agents underscore the need for meticulous stewardship, robust infection prevention, and cautious interpretation of S. maltophilia antimicrobial susceptibility testing. Full article

Fulfilling social responsibilities within the ESG framework has gradually become a core competitive advantage for sustainable corporate development that also serves to enhance future returns. Charitable donations constitute a crucial method through which corporations fulfill social responsibilities and represent a primary indicator in ESG ratings, ratings that in turn have an impact on future stock market returns. This study, based on data from listed companies on the Shanghai and Shenzhen stock exchanges from 2018 to 2022, employed a fixed effects model to analyze the influence of charitable donations on future returns under ESG rating constraints. The research reveals that ESG rating constraints can reduce speculative charitable donations and help to optimize the peak value of a company’s future returns. After a series of robustness tests, including using the one-period lagged explanatory variable, changing the measurement method of the explained variable, replacing the ESG with the assignment method for value determination, and considering the impact of outliers, the conclusion still holds. Heterogeneity analysis indicates that in state-owned enterprises, companies in a recessionary phase, and industries with lower levels of competition, a decelerating effect of ESG ratings on the impact of charitable donations on future returns dominates. Conversely, for mature companies, ESG ratings accelerate the positive effect of charitable donations on future returns. This paper contributes to the ESG economic consequences literature by offering empirical evidence on corporate social responsibility implementation under sustainability strategies. Full article

Automated radiology report generation from chest X-rays presents a critical challenge in medical imaging. Traditional image-captioning models struggle with clinical specificity and rare pathologies. Recently, contrastive vision language learning has emerged as a robust alternative that learns joint visual–textual representations. However, applying contrastive learning (CL) to radiology remains challenging due to severe data scarcity. Prior work has employed input space augmentation, but these approaches incur computational overhead and risk distorting diagnostic features. This work presents CL with feature space interpolation for retrieval (CLFIR), a novel CL framework operating on learned embeddings. The method generates interpolated pairs in the feature embedding space by mixing original and shuffled embeddings in batches using a mixing coefficient $\lambda \sim U(0.85,0.99)$. This approach increases batch diversity via synthetic samples, addressing the limitations of CL on medical data while preserving diagnostic integrity. Extensive experiments demonstrate state-of-the-art performance across critical clinical validation tasks. For report generation, CLFIR achieves BLEU-1/ROUGE/METEOR scores of 0.51/0.40/0.26 (Indiana university [IU] X-ray) and 0.45/0.34/0.22 (MIMIC-CXR). Moreover, CLFIR excels at image-to-text retrieval with R@1 scores of 4.14% (IU X-ray) and 24.3% (MIMIC-CXR) and achieves 0.65 accuracy in zero-shot classification on the CheXpert5×200 dataset, surpassing the established vision-language models. Full article

This study examines the mobilization of social and cultural capital among first and second-generation Afro-Caribbeans in Canada, focusing specifically on Jamaican and Haitian populations. Employing an analytical model grounded in resistance and identity multi-positionality, this research utilizes Yosso’s theory of cultural wealth as a theoretical framework. Qualitative data were collected through focus groups and an intake survey aimed at exploring the dual objectives of defining Blackness and constructing an in-group Black identity alongside the establishment and contestation of social capital within these groups. The findings reveal a dynamic interplay between resistance and identity, highlighting how marginalized groups leverage their resilience to build robust social networks that challenge hegemonic norms. Significant generational differences were identified in experiences of racism, discrimination, and cultural preservation among the participants. This study contributes to the broader discourse on immigrant integration, social cohesion, and the role of cultural capital in mitigating systemic inequalities. The results underscore the necessity for intersectional approaches to comprehend the complexities of identity formation and social integration in multicultural societies. Moreover, the research emphasizes the critical importance of cultural heritage, identity, and community support as sources of strength and resilience for Afro-Caribbean communities in Canada. Full article

High concentrations of tropospheric ozone (${\mathrm{O}}_3$) in urban areas pose a significant risk to human health. This study proposes an evaluation framework based on the XGBoost algorithm to predict ${\mathrm{O}}_3$ concentration, assessing the model’s capacity for seasonal extrapolation and spatial transferability. The experiment uses hourly air pollution data (${\mathrm{O}}_3$, NO, ${\mathrm{NO}}_2$, and NOx) and meteorological factors (temperature, relative humidity, barometric pressure, wind speed, and wind direction) from six monitoring stations in the Monterrey Metropolitan Area, Mexico (from 22 September 2022 to 21 September 2023). In the preprocessing phase, the datasets were extended via feature engineering, including cyclic variables, rolling windows, and lag features, to capture temporal dynamics. The prediction models were optimized using a random search, with time-series cross-validation to prevent data leakage. The models were evaluated across a concentration range of 0.001 to 0.122 ppm, demonstrating high predictive accuracy, with a coefficient of determination (${\mathrm{R}}^2$) of up to 0.96 and a root-mean-square error (RMSE) of 0.0034 ppm when predicting summer (${\mathrm{O}}_3$) concentrations without prior knowledge. Spatial generalization was robust in residential areas (${\mathrm{R}}^2$ > 0.90), but performance decreased in the industrial corridor (AQMS-NL03). We identified that this decrease is related to local complexity through the quantification of domain shift (Kolmogorov–Smirnov test) and Shapley additive explanations (SHAP) diagnostics, since the model effectively learns atmospheric inertia in stable areas but struggles with the stochastic effects of NOx titration driven by industrial emissions. These findings position the proposed approach as a reliable tool for “virtual detection” while highlighting the crucial role of environmental topology in model implementation. Full article

Remaining Useful Life (RUL) prediction has become a crucial component in predictive maintenance and condition-based operation with the rapid advancement of industrial automation and the increasing complexity of mechanical systems. Although existing deep learning models, such as Long Short-Term Memory (LSTM) networks and conventional Convolutional Neural Networks (CNNs), have demonstrated effectiveness in modeling equipment degradation from multivariate sensor data, they still face several limitations. Recurrent architectures often suffer from vanishing gradients and struggle to capture long-term dependencies, while CNN-based methods typically rely on small convolutional kernels and deterministic feature extractors, limiting their ability to model long-range dependencies and stochastic degradation transitions. To address these challenges, this study proposes a novel hybrid deep learning framework that integrates large-kernel convolutional feature extraction with Markov transition modeling for RUL prediction. Specifically, the large-kernel CNN captures both local and global degradation patterns, while the Markov feature module encodes probabilistic state transitions to characterize the stochastic evolution of equipment health. Furthermore, a lightweight channel attention mechanism is incorporated to adaptively emphasize degradation-sensitive sensor information, thereby enhancing feature discriminability. Extensive experiments conducted on the NASA C-MAPSS turbofan engine dataset demonstrate that the proposed model consistently outperforms conventional CNN, LSTM, and hybrid baselines in terms of Root Mean Square Error (RMSE) and the NASA scoring metric. The results verify that combining deep convolutional representations with probabilistic transition information significantly enhances prediction accuracy and robustness in industrial RUL estimation tasks. Full article

To address the stochastic, nonlinear, and strongly coupled characteristics of multivariate long-term structural response in bridge health monitoring, this study proposes the Informer-SEnet prediction model. The model integrates a Squeeze-and-Excitation (SE) channel attention mechanism into the Informer framework, enabling adaptive recalibration of channel importance to suppress redundant information and enhance key structural response features. A sliding-window strategy is used to construct the datasets, and extensive comparative experiments and ablation studies are conducted on one public bridge-monitoring dataset and two long-term monitoring datasets from real bridges. In the best case, the proposed model achieves improvements of up to 54.67% in MAE, 52.39% in RMSE, and 7.73% in R². Ablation analysis confirms that the SE module substantially strengthens channel-wise feature representation, while the sparse attention and distillation mechanisms are essential for capturing long-range dependencies and improving computational efficiency. Their combined effect yields the optimal predictive performance. Five-fold cross-validation further evaluates the model’s generalization capability. The results show that Informer-SEnet exhibits smaller fluctuations across folds compared with baseline models, demonstrating higher stability and robustness and confirming the reliability of the proposed approach. The improvement in prediction accuracy enables more precise characterization of the structural response evolution under environmental and operational loads, thereby providing a more reliable basis for anomaly detection and early damage warning, and reducing the risk of false alarms and missed detections. The findings offer an efficient and robust deep learning solution to support bridge structural safety assessment and intelligent maintenance decision-making. Full article

Artificial intelligence (AI) has played a pivotal role in enhancing the efficiency of financial technology (Fintech), ultimately contributing to the stability of the banking sector. The advancements in Fintech driven by AI tools are significantly improving risk management within the banking industry. This paper investigates the mediating role of AI in the relationship between Fintech and financial stability in the context of Islamic and conventional banks across selected countries in the Organization of Islamic Cooperation (OIC). It employs structural equation modeling (SEM) to explore the causal linkages across time domains. The results of this research identify that AI is a significant mediator, playing a critical role between Fintech and stability. It either mitigates or amplifies risks, depending on the regulatory framework and implementation practices in place. The analysis indicates that AI has a weak mediating effect in the short run, but a strong mediating effect in the long run between Fintech and stability. This research paper emphasizes the importance of developing robust, forward-thinking policies to leverage the benefits of AI. It also addresses the risks to financial stability in both Islamic and conventional banking systems. Full article