Search Results (8,608)

Many place recognition approaches, which identify previously visited places or locations by matching current sensory data, such as 2D RGB images and 3D point clouds, have been proposed to achieve accurate and robust localization and loop closure detection in global positioning system (GPS)-denied environments. Since visual place recognition (VPR) methods that rely on images captured by camera sensors are highly sensitive to variations in appearance, including changes in lighting, surface color, and shadows, they can lead to poor place recognition accuracy. In contrast, light detection and ranging (LiDAR)-based place recognition (LPR) approaches based on 3D point cloud data that captures the shape and geometric structure of the environment are robust to changes in place appearance and can therefore provide more reliable place recognition results than VPR methods. This work presents an indoor LPR method called PointNetVLAD-based indoor pedestrian localization (PIPL). PIPL is a deep network model that uses PointNetVLAD to learn to extract global descriptors from 3D LiDAR point cloud data. PIPL can recognize places previously visited by a pedestrian using point clouds captured by a low-cost LiDAR sensor on a smartphone in small-scale indoor environments, while PointNetVLAD performs place recognition for vehicles using high-cost LiDAR, GPS, and inertial measurement unit (IMU) sensors in large-scale outdoor areas. For place recognition on 3D point cloud reference maps generated from LiDAR scans, PointNetVLAD exploits the universal transverse mercator (UTM) coordinate system based on GPS and IMU measurements, whereas PIPL uses a virtual coordinate system designed in this study due to the unavailability of GPS indoors. In experiments conducted in campus buildings, PIPL shows significant advantages over NetVLAD (known as a convolutional neural network (CNN)-based VPR method). Particularly in indoor environments with repetitive scenes where geometric structures are preserved and image-based appearance features are sparse or unclear, PIPL achieved $39\%$ higher top-1 accuracy and $10\%$ higher top-3 accuracy compared to NetVLAD. Furthermore, PIPL achieved place recognition accuracy comparable to NetVLAD even with a small number of points in a 3D point cloud and outperformed NetVLAD even with a smaller model training dataset. The experimental results also indicate that PIPL requires over $76\%$ less place retrieval time than NetVLAD while maintaining robust place classification performance. Full article

Unlike reactive systems, mechanical ventilation controlled by CO₂ concentration operates at a target efficiency that dynamically increases whenever the target CO₂ level is exceeded. This approach eliminates the typical ‘dead-time’ and prevents air quality degradation by ensuring the system adjusts its performance immediately in response to concentration changes. In this work, the study focuses on the development and evaluation of data-driven predictive models for near-term indoor CO₂ forecasting that can be integrated into pre-occupancy ventilation strategies, rather than designing a complete control scheme. Experimental data were collected over four months in a 48 m² smart laboratory configured as an open-plan office, where a heterogeneous IoT sensing architecture logged synchronized time-series measurements of CO₂ and microclimate variables (temperature, relative humidity, PM_2.5, TVOCs), together with acoustic noise levels and appliance-level energy consumption used as indirect occupancy-related signals. Raw telemetry was transformed into a 22-feature state vector using a structured feature engineering method incorporating z-score standardization, cyclic time encodings, multi-horizon CO₂ lags, rolling statistics, momentum features, and non-linear interactions to represent temporal autocorrelation and daily periodicity. The study benchmarks multiple regression paradigms, including simple baselines and ensemble methods, and found that an automated multi-level stacked ensemble achieved the highest predictive fidelity for short-term forecasting, with an Mean Absolute Error (MAE) of 32.97 ppm across an observed CO₂ range of 403–2305 ppm, representing improvements of approximately 24% and 43% over Linear Regression and K-Nearest Neighbors (KNN), respectively. Temporal diagnostics showed strong phase alignment with observed CO₂ rises during occupancy transitions and statistically reliable prediction intervals. Five-fold walk-forward cross-validation confirmed the temporal stability of these results, with top models achieving consistent R² values of 0.93–0.95 across Folds 2–5. These results demonstrate that, within a single-room university laboratory setting, historical sensor data from low-cost IoT devices can support accurate short-term CO₂ forecasting, providing a predictive layer that could support future proactive ventilation scheduling aimed at reducing CO₂ lag at the start of occupancy while avoiding unnecessary ventilation runtime. Generalization to other building types and occupancy profiles requires further validation. Full article

This paper examines the effect of a bonus cap on the compensation structure of top earners in the Dutch banking sector. Following concerns that performance-based pay may induce excessive risk-taking, regulators introduced caps on variable compensation. This paper analyzes how such regulation affects the composition of pay. The identification strategy exploits a unique institutional setting in which banks with their statutory seat in the Netherlands are subject to a stricter bonus cap than banks headquartered in other EU countries, while operating in the same market. This paper uses administrative microdata and a difference-in-differences approach to compare compensation outcomes across these groups before and after the introduction of the Dutch bonus cap in 2015. Consistent with the predictions of a principal–agent model of incentive contracting, the hourly variable wage decreases by 23 percent, while the hourly fixed wage component increases by 12 percent. The findings indicate that compensation regulation reshapes the composition of pay. Full article

Proportional, fair scheduling in OFDM-based vehicle-to-everything (V2X) uplink causes the resource-block allocation of each vehicle to vary from slot to slot, yet conventional semantic encoders produce a fixed number of output tokens regardless of the instantaneous channel capacity. When the encoder output exceeds the slot budget, transmitted features are truncated and the resulting federated learning gradient is corrupted—a problem that affected 23% of training rounds for non-line-of-sight vehicles in our experiments. The difficulty is worsened by a spatial pattern common in urban deployments: vehicles at congested intersections suffer the poorest propagation conditions while carrying the training data most relevant to safety, and throughput-driven client selection excludes them in favor of vehicles with strong channels but uninformative scenes. We address both issues within a single framework for OFDM-based V2X federated learning. On the transmission side, a Sensing-Guided Adaptive Modulation (SGAM) module derives a per-slot token budget from the current resource-block allocation and selects tokens through differentiable Gumbel-TopK pruning with a hard capacity clip, so the transmitted token count stays within the slot budget. On the scheduling side, a Channel-Decoupled Federated Learning (CDFL) module partitions clients independently by channel quality and data complexity, selects diverse representatives per partition via facility location optimization, and corrects for partition-size imbalance through inverse propensity weighting during model aggregation. Experiments on NuScenes with 20 non-IID vehicular clients under realistic OFDM channel simulation demonstrate a Macro-F1 of 0.710 (+8.7 points over the Oort-adapted baseline), zero budget violations throughout training, and a 75% reduction in training variance; the worst-class F1 more than doubles relative to FedAvg. Full article

Variable cross-section cement–soil pipe piles are an innovative soft ground improvement technology. They are tubular, special-shaped cement–soil mixing piles characterized by a tapered profile along the pile shaft (larger diameter at the top and smaller at the bottom) and an internal soil core. They offer advantages including reduced material consumption, lower engineering cost, and shorter construction duration. However, the systematic theoretical understanding of their bearing performance remains insufficient. In this study, the bearing mechanism and influencing parameters of variable cross-section pipe piles were systematically investigated via full-scale field tests, numerical simulations, and laboratory model tests. An exponential decay constitutive model considering the strain-softening behavior of cement–soil was developed and implemented through secondary development in the ABAQUS platform for parametric analysis. Laboratory model tests were further conducted to advance the understanding of the bearing mechanism of variable cross-section pipe piles. The results show that the ultimate bearing capacity of the proposed variable cross-section cement–soil pipe pile is approximately 189% higher than that of the conventional ones. The expanded outer diameter and expanded height are the dominant factors affecting the bearing capacity, while the inner diameter and pile length have a comparatively minimal influence: increasing the expanded outer diameter from 0.6 m to 1.2 m and the expanded height from 0 m to 5 m increased the ultimate bearing capacity from 445 kN to 868 kN and 936 kN, respectively. The effective pile length is determined to be 6 m, and the recommended minimum wall thickness of the pipe pile is 1/4 of the inner diameter. Laboratory tests further demonstrated an abrupt change in axial force at the variable section. The findings provide reliable theoretical support for the engineering design and field application of cement–soil variable cross-section pipe piles. Full article

Modern neural networks have achieved remarkable success in visual recognition; however, due to their sensitivity to domain shifts, Unsupervised Domain Adaptation (UDA) remains an open research problem. A key reason for this limitation is that source-trained models rely primarily on texture, lacking the explicit geometric information required for object recognition. To overcome this problem, we introduce CoDA, an object-centric learning framework inspired by infant cognitive development, specifically the process of object individuation. By introducing a geometric prior, our approach employs a physically grounded generation pipeline that uses a textureless “Sculpture Mode” and object isolation to complement textural information with 3D geometric features, capturing shape information that is often ignored during training. To enable robust training from scratch, we further integrate two control mechanisms: a Network Stability Scheduler to orchestrate training progression based on convergence stability, and a Dynamic Top-K Pseudo-Labeling strategy that adapts confidence thresholds for each individual class. Extensive evaluations on three real-world target datasets (VegFru, Fruits-262, and Open Images v7) demonstrate that CoDA, trained on a source dataset of just 12,000 synthetic images, achieves comparable results to (and in specific domains surpasses) ImageNet-pretrained models (leveraging 1.2 million images), significantly outperforming state-of-the-art adversarial and semi-supervised domain adaptation methods. Full article

This study addresses the growing need for structured, measurable organizational development (OD) models amid digital transformation, geopolitical uncertainty, and increasing managerial complexity. Contemporary middle- and top-level managers are expected to ensure productivity, strategic clarity, resilience, and data-driven decision-making; however, existing leadership methodologies are often examined separately and lack integrated evaluation frameworks. The research analyses two prominent approaches: the American Action Science methodology and the Scandinavian (particularly Finnish) consensus-based leadership concept. While Action Science emphasizes explicit reasoning, double-loop learning, accountability, and measurable performance outcomes, the Finnish consensus model prioritizes trust, participation, and relational cohesion. The aim of the study is to develop and empirically test the original digital model LeadWinO (LEADership for WINning Organizations) for evaluating the organizational development activities of middle- and top-level managers. The model was empirically tested on managers in Lithuania. The novelty of the research lies in combining management and informatics perspectives by embedding organizational development evaluation into a digital, indicator-based, and potentially predictive framework. The type of study is quantitative research integrating questionnaire analysis in the case of multi-profile sections. Analytical tool used for data simulation is Feedforward Neural Network for constructing sufficient gapless sets of digitalized data. Research results showed that the American Action Science methodology is most effectively used by managers working in very small and small enterprises in the service and maintenance sectors. The findings are expected to contribute to the operationalization of leadership effectiveness under uncertainty and provide organizations with an auditable structure linking managerial behaviour, decision-making processes, and organizational performance outcomes. Full article

Background: Multimodal large language models (MLLMs) show promise in medical imaging, but their performance is highly dependent on prompt engineering. This study systematically evaluates how different prompting strategies affect diagnostic accuracy in clinical laboratory image interpretation. Methods: We evaluated five MLLMs (ChatGPT-4o, Gemini 2.0 Flash, Claude 3.5 Sonnet, Grok-2, and Perplexity Pro (Claude 3.5 Sonnet)) using 177 proficiency testing images across three domains: blood smears (n = 78), urinalysis (n = 50), and parasitology (n = 49). Three prompting approaches were compared: (1) complex multi-choice prompts with 20 diagnostic options, (2) zero-shot open-ended prompts, and (3) two-step descriptive-reasoning prompts. Images were sourced from the Taiwan Society of Laboratory Medicine external quality assurance archives with expert consensus diagnoses. Results: Zero-shot prompting significantly outperformed complex multi-choice prompts across all models and domains (p < 0.001). With zero-shot prompts, Gemini achieved 78.5% overall accuracy (urinalysis: 92.0%; parasitology: 75.5%; blood smears: 64.1%), representing a 17% improvement over complex prompts. Two-step descriptive-reasoning prompts further improved blood smear accuracy by 8–12% for top-performing models, but showed minimal benefit in urinalysis and parasitology. The re-query mechanism (“please reconsider”) improved urinalysis accuracy by 7.6% but had a negligible effect on blood smears and parasitology. Conclusions: Prompting strategy critically determines MLLM diagnostic performance. Zero-shot approaches with minimal constraints consistently outperform complex multi-choice formats. The remarkable performance of general-purpose models in structured domains like urinalysis (>90% accuracy) demonstrates the considerable progress of multimodal AI. However, complex morphological tasks like blood smear interpretation require either specialized prompting techniques or domain-specific fine-tuning. These findings provide evidence-based guidance for optimizing AI integration in clinical laboratories. Full article

The high accuracy achieved by deep learning techniques has made them indispensable in computer vision applications. However, their substantial memory demands and high computational complexity limit their deployment in resource-constrained environments. To address this challenge, this study introduces a Feature Enhancement Module (FEM) as part of a unified hybrid compression framework that combines mixed-precision quantization and structured pruning to improve model efficiency. Experimental results on the Tiny ImageNet dataset using ResNet50 and MobileNetV3 architectures demonstrate the strong adaptability and scalability of the proposed approach. Compared with state-of-the-art compression methods, the proposed FEM-based framework achieves up to 6% improvement in Top-1 accuracy, while reducing memory usage by 32.26% and improving inference speed by 66%. Furthermore, the ablation study demonstrates that incorporating the FEM module leads to up to 24% improvement over the baseline model, highlighting its effectiveness. The results further show that FEM effectively preserves inter-channel feature representation stability even under aggressive compression, making it well suited for real-time processing and practical Artificial Intelligence (AI) applications. By maintaining semantic richness while significantly reducing computational cost, the proposed method bridges the gap between high-performance deep models and lightweight, deployable solutions. Overall, the FEM-based hybrid compression framework establishes a scalable and architecture-independent foundation for sustainable deep learning in resource-limited environments. Full article

This paper introduces a rigorous class of two-dimensional Markov-switching autoregressive moving-average (2D MS-ARMA) models for spatial lattice data exhibiting regime-dependent dynamics. The switching mechanism is governed by a latent causal Markov random field that drives spatial transitions between regime-specific autoregressive and moving-average structures. We provide sufficient conditions for the existence of a strictly stationary solution through the top Lyapunov exponent associated with a sequence of random matrices obtained from a state-space representation constructed along the lexicographic order. For the first-order bidirectional specification, we derive explicit spectral conditions linking stationarity to the regime-dependent spectral radii. Sufficient conditions ensuring the existence of finite second-order moments are also provided. Parameter estimation is carried out using a variational expectation–maximization (VEM) algorithm based on a mean-field approximation of the posterior distribution of the hidden regimes. The E-step yields closed-form coordinate ascent updates, while the M-step relies on gradient-based numerical optimization with derivatives computed via recursive differentiation. Under increasing-domain asymptotics, we discuss the consistency and asymptotic behavior of the variational estimator. The proposed framework fills a methodological gap between classical one-dimensional Markov-switching ARMA models and spatial autoregressive structures by extending regime-switching theory to multi-indexed processes with rigorous probabilistic foundations. It provides a comprehensive basis for statistical inference, model diagnostics, and prediction in spatially heterogeneous environments. Full article

In response to the technical bottleneck of the Venlo greenhouse’s inability to achieve year-round production due to the high temperature and humidity in the summer in South China, this study took an existing Venlo-type greenhouse in Guangzhou as the research object and constructed a three-dimensional computational fluid dynamics (CFD) model of the greenhouse by comprehensively considering key factors such as solar radiation, thermal radiation, and crop canopy resistance. After on-site experiments, it was verified that, except for the top area of the greenhouse, the temperature deviation between the model simulation values and the measured values was less than 2 °C, and the error rate was less than 5%, confirming the model’s accurate representation of the temperature field distribution within the greenhouse. Based on the characteristics of the temperature and humidity fields revealed by the CFD simulation (canopy temperature gradient K = 0.144 °C/m, maximum temperature difference between upper and lower layers 20 °C), an optimized scheme of “wet curtain fan + salt bath dehumidification equipment” for local cooling and dehumidification of the crop canopy was proposed, and a non-uniform air duct layout was designed according to the temperature gradient characteristics. Field experiments showed that after optimization, the daytime temperature of the crop canopy was mostly controlled within 30 °C, the relative humidity was stably maintained below 80%, and the maximum temperature difference along the length of the greenhouse was reduced from 7 °C to 2 °C, effectively solving the problem of poor cooling and dehumidification effects of the traditional system. This scheme enabled the stable operation and year-round production of Venlo-type greenhouses in South China during the summer, providing technical support and engineering reference for greenhouse environmental control in high-humidity areas. Full article

Large-scale chest X-ray (CXR) datasets often rely on report-derived or weak labels, introducing missing and incorrect annotations that can degrade downstream models and limit trust. We study training-free mislabel detection in multi-label CXRs by scoring neighborhood label consistency in a fixed embedding space. Using the NIH Chest X-ray Kaggle sample (5606 CXRs), we extract intermediate CoAtNet features and obtain 64-dimensional embeddings with a frozen CoAtNet backbone and a lightweight refinement head. On top of these embeddings, we compare kNN consistency baselines with distance weighting and label-set similarity against LPV-DW-CS, clustered prototype voting weighted by distance and cluster support. We evaluate three synthetic label-noise regimes with review budgets matched to the corruption rate: random single-label (5% and 20%), boundary-noise (20% corruption within the lowest-density 20% subset), and disjoint-label replacement (20% within that subset). LPV-DW-CS yields the highest downstream macro-AUROC after filtering top-ranked samples (up to 0.8860), while kNN variants achieve higher Recall@budget at the same review rates (up to 99.44%). An image-only expert Likert review of top-ranked real samples finds substantial label-set inconsistencies (54.1% for LPV-DW-CS-280-A; 60.5% for KNN-DW-LSS), supporting neighborhood-consistency ranking as a practical, training-free tool for targeted dataset auditing. Full article

Marine organisms that episodically aggregate near coastal nuclear power plant water intakes pose a substantial risk to cooling water security. Predicting the spatial distribution of such high-risk species remains challenging because their occurrence is shaped not only by environmental conditions but also by complex trophic interactions. In this study, we model the habitat distribution of three high-risk nektonic species, Dotted gizzard shad (Konosirus punctatus), Japanese swimming crab (Charybdis japonica) and squid (Loligo sp.), in the cooling water intake area of a coastal nuclear power plant in eastern Liaodong Bay using generalized linear models (GLMs) and joint species distribution models (JSDMs). Based on summer surveys conducted in 2024–2025, we explicitly incorporated trophic linkages among target species, their prey, and predators within JSDMs. Model performance was evaluated using cross-validation based on AUC, RMSE, and coefficient of determination (R²). Our results indicate that water depth was the dominant environmental driver for all three species, while chlorophyll-a concentration and distance to the intake exerted species-specific effects. By incorporating interspecific trophic associations and environmental responses, JSDMs showed consistently improved predictive performance relative to GLMs, with approximately 1.5-fold higher R² values and 10–30% lower RMSE, while offering enhanced ecological interpretability. The models revealed strong positive associations between target species and both lower-trophic prey and higher-trophic predators, suggesting that top–down and bottom–up processes jointly regulate aggregation dynamics. This study demonstrates that integrating trophic interactions into species distribution modeling substantially improves predictions of high-risk marine species near coastal infrastructure and provides an ecological basis for proactive management of cooling water intake systems. Full article

The aim of this study was to understand the female swimmers’ annual performance progression and stability between 10 and 18 years across swimming distances and techniques. Data from female Portuguese Top-50 rankings in the short-course pool was extracted from an open access database (swimrankings.net). Performances were grouped by distances (50-, 100- and 200 m) and techniques (freestyle, backstroke, breaststroke and butterfly), totalizing 12 events as performance metrics. A total of 343 swimmers and 3087 performances distributed by nine consecutive competitive seasons were retrospectively assessed. The mean and normative stability were computed for tracking performance trends, while reporting the year-to-year percentage improvement. The differences across distances and techniques were tested with a linear mixed-effects model using intraclass correlation coefficient (ICC). The performance progression was characterized by marked improvements during the early ages (up to 13% yearly) and an emerging plateau around the 15–16 years. The stability patterns varied between events, with the backstroke technique (ICC = 0.13) demonstrating greater consistency of individual differences on developmental trajectories, whereas shorter races (i.e., 50 m; ICC = 0.15) tended to be more stable than 100 m or 200 m (ICC = 0.12). It can be concluded that female swimmers’ performance stabilizes at the 15–16 years of age. Despite reduced differences, the backstroke technique and short distances seem to show a slightly more stable trend in progressing from childhood to adulthood. Full article

Fieldwork carries a high risk of irregular, non-compressible traumatic wounds, which often initiate a vicious cycle of “traumatic bleeding-insect bite-secondary infection”. Conventional dressings cannot combine rapid hemostasis with physical protection against venomous insects, creating an urgent demand for multifunctional field trauma dressings. To solve this problem, this study developed a shape-adaptive bilayer hydrogel that concurrently provides rapid hemostasis, promotes wound repair, and acts as a robust physical barrier. The hydrogel adopts a layered design: the bottom layer (PPTY) achieves autogelation within 3 s upon blood contact, while the top armor protective layer (AP) withstands pressures up to 942 kPa. By incorporating chitosan and sodium citrate into the AP precursor solution, the hydrogel achieved in situ formation within 50 s and developed a stable self-renewing armor layer. The tightly bonded bilayer showed complementary functions. In rat models of femoral artery puncture and tail vein bleeding, PPTY-AP hydrogel significantly reduced blood loss and shortened hemostasis time. Moreover, the hydrogel demonstrated excellent tissue adhesion and moisture retention capacity, promoting full-thickness skin wound healing. This multifunctional, rapidly deployable hydrogel presents a promising solution for field trauma management and offers a new design paradigm for advanced wound dressings. Full article