Search Results (111)

Rapid building damage assessments are vital for an effective earthquake response. In Japan, traditional Earthquake Damage Certification (EDC) surveys—followed by the issuance of Disaster Victim Certificates (DVCs)—are often inefficient. With advancements in remote sensing technologies and deep learning algorithms, their combined application has been explored for large-scale automated damage assessment. However, the scarcity of remote sensing data on damaged buildings poses significant challenges to this task. In this study, we propose an Uncertainty-Guided Fusion Module (UGFM) integrated into a standard decoder architecture, with a Pyramid Vision Transformer v2 (PVTv2) employed as the encoder. This module leverages uncertainty outputs at each stage to guide the feature fusion process, enhancing the model’s sensitivity to collapsed buildings and increasing its effectiveness under diverse conditions. A training and in-domain testing dataset was constructed using post-earthquake aerial imagery of the severely affected areas in Noto Prefecture. The model approximately achieved a recall of 79% with a precision of 68% for collapsed building extraction on this dataset. We further evaluated the model on an out-of-domain dataset comprising aerial images of Mashiki Town in Kumamoto Prefecture, where it achieved an approximate recall of 66% and a precision of 77%. In a quantitative analysis combining field survey data from Mashiki, the model attained an accuracy exceeding 87% in identifying major damaged buildings, demonstrating that the proposed method offers a reliable solution for initial assessment of major damage and its potential to accelerate DVC issuance in real-world disaster response scenarios. Full article

In the context of sustainable development, improving the durability of engineering materials and the service life of engineering projects is an important path to address engineering sustainability and low-carbon development. This study addresses the durability issues of steel fiber-reinforced cementitious materials (SFRCMs) under the combined action of stray current and chloride ions in metro engineering. Through simulated stray current-accelerated corrosion tests, combined with compressive strength tests and X-ray computed tomography (X-CT) analysis, the effects of steel fiber volume content (0.5%, 1.0%, 1.5%) and electrification duration (0–72 h) on the mechanical properties and corrosion mechanisms were systematically investigated. The results indicate that steel fiber content significantly influences corrosion rate and strength degradation. Specimens with 1.5% fiber content exhibited the highest initial compressive strength (58.43 MPa), but suffered a severe strength loss rate of 37.67% after 72 h of electrification. In contrast, specimens with 1.0% fiber content demonstrated balanced performance, achieving both high initial strength and superior corrosion resistance (19.66% strength loss after 72 h). X-CT analysis revealed that corrosion products initially filled pores during early stages but later induced microcracks in the matrix. Higher fiber content specimens exhibited increased large-pore ratios due to fiber agglomeration, accelerating chloride ion penetration. Furthermore, digital volume correlation (DVC) analysis demonstrated that steel fibers effectively dispersed loads and reduced stress concentration. However, post-corrosion fiber volume loss weakened their crack resistance capacity, highlighting the critical role of fiber integrity in structural durability. Full article

Dense video captioning involves identifying, localizing, and describing multiple events within a video. Capturing temporal and contextual dependencies between events is essential for generating coherent and accurate captions. To effectively capture temporal and contextual dependencies between events, we propose Dense Video Captioning with Dual Contextual, Semantical, and Linguistic Modules (DVC-DCSL), a novel dense video captioning model that integrates contextual, semantic, and linguistic modules. The proposed approach employs two uni-directional LSTMs (forward and backward) to generate distinct captions for each event. A caption selection mechanism then processes these outputs to determine the final caption. In addition, contextual alignment is improved by incorporating visual and textual features from previous video segments into the captioning module, ensuring smoother narrative transitions. Comprehensive experiments conducted using the ActivityNet dataset demonstrate that DVC-DCSL increases the Meteor score from 11.28 to 12.71, representing a 12% improvement over state-of-the-art models in the field of dense video captioning. These results highlight the effectiveness of the proposed approach in improving dense video captioning quality through contextual and linguistic integration. Full article

Accurate measurement uncertainty quantification and its propagation are critical for dimensional compliance in precision manufacturing. This study presents a novel framework that examines the evolution of measurement error along the axial length of CNC-turned components, focusing on spatial and material-specific factors. A systematic experimental comparison was conducted between a manual Digital Vernier Caliper (DVC) and an automated Coordinate Measuring Machine (CMM) using five engineering materials: Aluminum Alloy 6061, Brass C26000, Bronze C51000, Carbon Steel 1020 Annealed, and Stainless Steel 304 Annealed. Dimensional measurements were taken from five consecutive machining zones to capture localized metrological behaviors. The results indicated that the CMM consistently achieved lower expanded uncertainty (as low as 0.00166 mm) and minimal propagated uncertainties (≤0.0038 mm), regardless of material hardness or cutting position. In contrast, the DVC demonstrated significantly higher uncertainty (up to 0.03333 mm) and propagated errors exceeding 0.035 mm, particularly in harder materials and unsupported zones affected by surface degradation and fixture variability. Root-sum-square (RSS) modeling confirmed that manual measurements are more prone to operator-induced error amplification. While the DVC sometimes recorded lower absolute errors, its substantial uncertainty margins hampered measurement reliability. To statistically validate these findings, a two-way ANOVA was performed, confirming that both the measurement system and machining zone significantly impacted uncertainty, as well as their interaction. These results emphasize the importance of material-informed and zone-sensitive metrology, highlighting the advantages of automated systems in sustaining measurement repeatability and dimensional stability in high-precision applications. Full article

Microbially induced concrete deterioration (MID) poses a significant and urgent challenge to urban sewerage systems globally, particularly in tropical coastal regions. Despite the acknowledged importance of biofilms in MICC, limited research on sewer pipe biofilms has hindered a comprehensive understanding of their deterioration mechanisms. To overcome this limitation, our research employed multiple staining techniques and digital volume correlation (DVC) technology, creating a new method to analyze the microstructure of biofilms, precisely identify the components of EPSs, and quantitatively examine MID mechanisms from a microscopic viewpoint. Our results revealed that the biofilm on concrete surfaces regulates the types of amino acids, thereby creating an environment conducive to microbial aggregate survival. Additionally, salinity significantly influences biofilm component distribution, while proteins play a pivotal role in biofilm mechanical stability. Notably, a high salinity fosters microbial migration within the biofilm, exacerbating deterioration. Through this multidimensional inquiry, our study established an advanced echelon of comprehension concerning the intricate mechanisms underpinning MICC. Meanwhile, by peering into the biofilms and elucidating their interplay with concrete, our findings offer profound insights, which can aid in devising strategies to counter urban sewer system deterioration. Full article

Attaining dimensional accuracy in CNC-machined parts is essential for high-precision manufacturing, especially when working with materials that exhibit varying mechanical and thermal characteristics. This research provides a thorough experimental comparison of manual and automated metrological systems, specifically the Digital Vernier Caliper (DVC) and Coordinate Measuring Machine (CMM), as applied to five different engineering alloys through five progressively machined axial zones. The study assesses absolute error, relative error, standard deviation, and measurement repeatability, factoring in material hardness, thermal conductivity, and surface changes due to machining. The results indicate that DVC performance is significantly affected by operator input and surface irregularities, with standard deviations reaching 0.03333 mm for Bronze C51000 and relative errors surpassing 1.02% in the initial zones. Although DVC occasionally showed lower absolute errors (e.g., 0.206 mm for Aluminum 6061), these advantages were countered by greater uncertainty and poor repeatability. In comparison, CMM demonstrated enhanced precision and consistency across all materials, with standard deviations below 0.0035 mm and relative errors being neatly within the 0.005–0.015% range, even with challenging alloys like Stainless Steel 304. Furthermore, a Principal Component Analysis (PCA) was conducted to identify underlying measurement–property relationships. The PCA highlighted clear groupings based on sensitivity to error in manual versus automated methods, facilitating predictive classification of materials according to their metrological reliability. The introduction of multivariate modeling also establishes a new framework for intelligent metrology selection based on material characteristics and machining responses. These results advocate for using CMM in applications requiring precise tolerances in the aerospace, biomedical, and high-end tooling sectors, while suggesting that DVC can serve as an auxiliary tool for less critical evaluations. This study provides practical recommendations for aligning measurement techniques with Industry 4.0’s needs for accuracy, reliability, and data-driven quality assurance. Full article

The global and domestic divisions of labor have had a great influence on the economy and environment in China during the last decade. With the refinement of production processes, national value chains (NVCs) coexist with global value chains (GVCs), enabling regions to participate in dual value chains (DVCs) simultaneously. This study calculates the NVCs and GVCs participation of manufacturing sectors in China’s provinces. On this basis, this research adopts a fixed effects model to analyze the impact of GVCs and NVCs participation and their interaction effect on manufacturing upgrades and green development. The results show, first, that significant regional differences in GVCs participation exist among provinces in China. In comparison, provincial NVCs participation demonstrates fewer regional differences. Second, there are significant sectoral differences of GVCs participation in China’s manufacturing industry—high-tech manufacturing is more embedded than other manufacturing industries. The sectoral differences in NVCs participation are relatively small. Third, GVCs and NVCs participation and their interaction effect have significantly promoted the upgrading and green development of manufacturing sectors in provinces of China, and this impact exhibits significant heterogeneity across regions, industries, and NVCs participation modes. The conclusions of this study provide empirical evidence and policy recommendations for the upgrading and green development of China’s manufacturing industry. Full article

Dense Video Captioning (DVC) represents the cutting edge of advanced multimedia tasks, focusing on generating a series of temporally precise descriptions for events unfolding within a video. In contrast to traditional video captioning, which usually offers a singular summary or caption for an entire video, DVC demands the identification of multiple events within a video, the determination of their exact temporal boundaries, and the production of natural language descriptions for each event. This review paper presents a thorough examination of the latest techniques, datasets, and evaluation protocols in the field of DVC. We categorize and assess existing methodologies, delve into the characteristics, strengths, and limitations of widely utilized datasets, and underscore the challenges and opportunities associated with evaluating DVC models. Furthermore, we pinpoint current research trends, open challenges, and potential avenues for future exploration in this domain. The primary contributions of this review encompass: (1) a comprehensive survey of state-of-the-art DVC techniques, (2) an extensive review of commonly employed datasets, (3) a discussion on evaluation metrics and protocols, and (4) the identification of emerging trends and future directions. Full article

Digital village construction (DVC) is a crucial pathway for increasing farmland productivity, reducing agricultural waste, and ultimately achieving sustainable development goals (SDGs). However, its effects on the sustainable intensification of cultivated land use (SICLU) remain unclear. To bridge this gap, this study investigated the impact effects and mechanisms of DVC on SICLU across 358 counties in China using ordinary least squares and mediating effect models. The results showed the following: (1) DVC and its four sub-indices had significant and positive impacts on SICLU, which were validated through a series of robustness tests. (2) Heterogeneity analysis showed that DVC significantly improved SICLU in the eastern and central regions, as well as in regions with abundant and relatively scarce resource endowments, whereas no such effect was observed in the western region. (3) The relationship between DVC and SICLU was mediated by farmers’ income, technological innovation, and agricultural informatization. These insights highlight the importance of accelerating DVC to enhance SICLU. Full article

Purpose: We assessed the feasibility of qualitative, semiquantitative, and multiparametric quantitative magnetic resonance imaging (MRI) using a three-dimensional (3D) ultrashort echo time (3D-UTE) sequence together with 2D-T2 and 3D-T1 mapping sequences to evaluate normal and pathological discovertebral complexes (DVCs). We assessed the disc (nucleus pulposus [NP] and annulus fibrosus [AF]), vertebral endplate (cartilage endplate [CEP] and growth plate [GP]), and subchondral bone (SB) using a rat model of degenerative disc disease (DDD). We also assessed whether this complete MRI cartography can improve the monitoring of DDD. Methods: DDD was induced by percutaneous disc trituration and collagenase injection of the tail. Then, the animals were imaged at 4.7T. The adjacent disc served as the control. The MRI protocol was performed at baseline and each week (W) postoperatively for 2 weeks. Visual analysis and signal intensity measurements from the 3D-UTE images, as well as T2 and T1 measurements, were carried out in all DVC portions. Histological analysis with hematoxylin–eosin and Masson trichrome staining was performed following euthanization of the rats at 2 weeks and the results were compared to the MRI findings. Results: Complete qualitative identification of the normal zonal anatomy of the DVC, including the AF, CEP, and GP, was achieved using the 3D-UTE sequence. Quantitative measurements of the signal-to-noise ratio in the AF and NP enabled healthy DVCs to be distinguished from surgery-induced DDD, based on an increase in these values post-surgery. The 2D-T2 mapping results showed a significant increase in the T2 values of the AF and a decrease in the values of the NP between the baseline and W1 and W2 postoperatively (p < 0.001). In the 3D-T1 mapping, there was a significant decrease in the T1 values of the AF and NP between baseline and W1 and W2 postoperatively in immature rats (p < 0.01). This variation in T1 and T2 over time was consistent with the results of the 3D-UTE sequence. Conclusions: Use of the 3D-UTE sequence enabled a complete, robust, and reproducible visualization of DVC anatomy in both immature and mature rats under both normal and pathological conditions. The findings were supported quantitatively by the T2 and T1 mapping sequences and histologically. This sequence is therefore of prime interest in spinal imaging and should be regularly be performed. Full article

Statistical characteristics of atmospheric circulation patterns (ACPs) and associated diurnal variation characteristics (DVCs) of precipitation in summer (June–August) from 2015 to 2019 over the complex terrain in northern Xinjiang (NX), northwestern arid region of China, were investigated based on NCEP FNL reanalysis data and Weather Research and Forecasting model simulation data from Nanjing University (WRF-NJU). The results show that six different ACPs (Type 1–6) were identified based on the Simulated ANealing and Diversified RAndomization (SANDRA), exhibiting significant differences in major-influencing synoptic systems and basic meteorological environments. Types 5, 3, and 2 were the most prevalent three patterns, accounting for 21.6%, 19.7%, and 17.7%, respectively. Type 5 mainly occurred in June and July, while Types 3 and 2 mainly occurred in August and July, respectively. From the perspective of DVCs, Type 1 reached its peak at midnight, while Type 5 was most frequent in the afternoon and morning. The overall DVCs of hourly precipitation intensity and frequency demonstrated a unimodal structure, with a peak occurring at around 16 Local Solar Time (LST). Basic meteorological elements in various terrain regions exhibit significant diurnal variation, with marked differences between mountainous and basin areas under different ACPs. In Types 3 and 6, meteorological elements significantly influence precipitation enhancement by promoting the convergence and uplift of low-level wind fields and maintaining high relative humidity (RH). The Altay Mountains region and Western Mountainous regions experience dominant westerly winds under these conditions, while the Junggar Basin and Ili River Valley regions benefit from counterclockwise water vapor transport associated with the Iranian Subtropical High in Type 6, which increases RH. Collectively, these factors facilitate the formation and development of precipitation. Full article

Background: The Beat It program is a clinician-led, community-based group exercise intervention for adults with Type 2 Diabetes Mellitus (T2DM). While previous studies have demonstrated its effectiveness in improving physical and mental health outcomes, this study explores the perspectives of Beat It Trainers to identify key factors contributing to the program’s success and areas for improvement. Methods: Semi-structured interviews were conducted with 11 Accredited Exercise Physiologists who had delivered both in-person and online versions of the program. Interviews were thematically analyzed using inductive approaches. Results: Eight main themes emerged: customization to individual needs, capability building, outcome improvement, affordability, accessibility, sustainability, and a holistic approach delivered in a group setting. Challenges identified included managing group dynamics, maintaining participant commitment in a fully subsidized program, and providing nutrition advice within the trainers’ scope of practice. The program’s adaptability to both in-person and online delivery modes was highlighted as enhancing its accessibility and resilience. Conclusions: This study provides valuable insights into the factors contributing to the success of the Beat It program from the implementers’ perspective. The findings suggest that investing in comprehensive training for facilitators, particularly in group dynamics management, could benefit similar programs. While the program’s fully subsidized structure reduces financial barriers to entry, innovative strategies to enhance participant engagement and perceived value should be explored. The success of the online delivery mode indicates that hybrid models offering both in-person and virtual options could increase accessibility in future supervised, community-based exercise programs for T2DM management. Full article

Grid-forming wind turbines (GFM-WTs) based on virtual synchronous control can support the voltage and frequency of power system by emulating the synchronous generator. The dynamic characteristics of a GFM-WT decided by virtual synchronous control, dq-axis voltage, and current control is significant for small-signal stability analysis. This paper builds a small-signal model of a GFM-WT in active power control (APC) and DC voltage control (DVC) timescale from the perspective of internal voltage. The proposed model describes how the magnitude and phase of the internal voltage are excited by the unbalanced active and reactive power when small disturbances occur. Interactions in different control loops can be identified by the reduced order model. We verify the accuracy of the proposed model in APC and DVC timescales by time domain simulations based on MATLAB/Simulink. Case studies show how the control parameters interact with each other in the two timescales. Full article

Credit risk is a crucial component of daily financial services operations; it measures the likelihood that a borrower will default on a loan, incurring an economic loss. By analysing historical data for assessment of the creditworthiness of a borrower, lenders can reduce credit risk. Data are vital at the core of the credit decision-making processes. Decision-making depends heavily on accurate, complete data, and failure to harness high-quality data would impact credit lenders when assessing the loan applicants’ risk profiles. In this paper, an empirical comparison of the robustness of seven machine learning algorithms to credit risk, namely support vector machines (SVMs), naïve base, decision trees (DT), random forest (RF), gradient boosting (GB), K-nearest neighbour (K-NN), and logistic regression (LR), is carried out using the Lending Club credit data from Kaggle. This task uses seven performance measures, including the F1 Score (recall, accuracy, and precision), ROC-AUC, and HL and MCC metrics. Then, the harnessing of generative adversarial networks (GANs) simulation to enhance the robustness of the single machine learning classifiers for predicting credit risk is proposed. The results show that when GANs imputation is incorporated, the decision tree is the best-performing classifier with an accuracy rate of 93.01%, followed by random forest (92.92%), gradient boosting (92.33%), support vector machine (90.83%), logistic regression (90.76%), and naïve Bayes (89.29%), respectively. The classifier is the worst-performing method with a k-NN (88.68%) accuracy rate. Subsequently, when GANs are optimised, the accuracy rate of the naïve Bayes classifier improves significantly to (90%) accuracy rate. Additionally, the average error rate for these classifiers is over 9%, which implies that the estimates are not far from the actual values. In summary, most individual classifiers are more robust to missing data when GANs are used as an imputation technique. The differences in performance of all seven machine learning algorithms are significant at the 95% level. Full article