Search Results (491)

This study presents the first systematic investigation into the influence of spectral enhancement techniques on the segmentation accuracy of specific soil and vegetation marks associated with palaeochannels. These marks are often subtle and can be seasonally obscured by vegetation dynamics and soil variability. Spectral enhancement methods, such as spectral indices and statistical aggregations, are routinely applied to improve their visual discriminability and interpretability. Despite recent progress in automated detection workflows, no prior research has rigorously quantified the effects of these enhancement techniques on the performance of deep learning–based segmentation models. This gap at the intersection of remote sensing and AI-driven analysis is critical, as addressing it is essential for improving the accuracy, efficiency, and scalability of subsurface feature detection across large and heterogeneous landscapes. In this study, two state-of-the-art deep learning architectures, U-Net and YOLOv8, were trained and tested to assess the influence of these spectral transformations on model performance, using Sentinel-2 imagery acquired across three seasonal windows. Across all experiments, spectral enhancement techniques led to clear improvements in segmentation accuracy compared with raw multispectral inputs. The multi-temporal Median Visualisation (MV) composite provided the most stable performance overall, achieving mean IoU values of 0.22 ± 0.02 in April, 0.07 ± 0.03 in August, and 0.19 ± 0.03 in November for U-Net, outperforming the full 12-band Sentinel-2 stack, which reached only 0.04, 0.02, and 0.03 in the same periods. FCC and VBB also performed competitively, e.g., FCC reached 0.21 ± 0.02 (April) and VBB 0.18 ± 0.03 (April), showing that compact three-band enhancements consistently exceed the segmentation quality obtained from using all spectral bands. Performance varied with environmental conditions, with April yielding the highest accuracy, while August remained challenging across all methods. These results highlight the importance of seasonally informed spectral preprocessing and establish an empirical benchmark for integrating enhancement techniques into AI-based archaeological and geomorphological prospection workflows. Full article

Background: Type 2 Diabetes Mellitus (T2DM) continues to rise rapidly in Indian communities, affecting millions and posing a major public health challenge. Early identification of risk and timely lifestyle intervention are crucial for prevention. This study aims to develop a lifestyle-driven, fuzzy-enhanced Artificial Neural Network (ANN) model for early T2DM prediction and to design a personalized recommendation framework tailored to the North Indian population. Methods: A comprehensive exploratory data analysis, including statistical significance testing and age-cohort assessment, was conducted to evaluate data quality and identify key lifestyle associations. The ANN model was trained on 1939 lifestyle profiles and classified individuals into four risk categories: low, moderate, high-risk, and diabetic. A monotonic spline-based calibration method was used to refine predicted probabilities. Additionally, a web-based system, the Personalized Care and Intelligence System for Early Diabetes Assessment (PCISEDA), was developed to deliver individualized diet and physical activity recommendations. Cost-effective lifestyle options were curated via a structured web-scraping pipeline. Results: The proposed fuzzy-enhanced ANN model achieved an accuracy of 93.64%, precision of 94.00%, recall of 93.50%, F1-score of 93.50%, and a multiclass ROC–AUC of 94.07%, demonstrating strong discriminative performance. Feature importance analysis revealed age, weight, urination frequency, and thirst as the most influential lifestyle predictors of T2DM risk. The PCISEDA system successfully generated personalized and economically feasible lifestyle recommendations for each risk category. Conclusions: This lifestyle-based AI framework demonstrates substantial potential for early T2DM risk stratification and tailored lifestyle management. The integration of fuzzy calibration and personalized recommendations offers an accurate, scalable, and cost-effective solution that may support diabetes prevention and management in resource-constrained healthcare settings. Full article

The integration of Artificial Intelligence (AI) into higher education offers new opportunities for inclusive and sustainable learning. This study investigates the impact of an AI-enabled microlearning cycle—comprising short instructional videos, formative quizzes, and structured discussions—on student engagement, inclusivity, and academic performance in postgraduate management education. A mixed-methods design was applied across two cohorts (2023, n = 138; 2024, n = 140). Data included: (1) survey responses on engagement, accessibility, and confidence (5-point Likert scale); (2) learning analytics (video views, quiz completion, forum activity); (3) academic results; and (4) qualitative feedback from open-ended questions. Quantitative analyses used Wilcoxon signed-rank tests, regressions, and subgroup comparisons; qualitative data underwent thematic analysis. Findings revealed significant improvements across all dimensions (p < 0.001), with large effect sizes (r = 0.35–0.48). Engagement, accessibility, and confidence increased most, supported by behavioural data showing higher video viewing (+19%), quiz completion (+21%), and forum participation (+65%). Regression analysis indicated that forum contributions (β = 0.39) and video engagement (β = 0.31) were the strongest predictors of grades. Subgroup analysis confirmed equitable outcomes, with non-native English speakers reporting slightly higher accessibility gains. Qualitative themes highlighted interactivity, real-world application, and inclusivity, but also noted quiz-related anxiety and a need for industry tools. The AI-enabled microlearning model enhanced engagement, equity, and academic success, aligning with SDG 4 (Quality Education) and SDG 10 (Reduced Inequalities). By combining Cognitive Load Theory, Kolb’s experiential learning, and Universal Design for Learning, it offers a scalable, pedagogically sustainable framework. Future research should explore emotional impacts, AI co-teaching models, and cross-disciplinary applications. By integrating Kolb’s experiential learning, Universal Design for Learning, and Cognitive Load Theory, this model advances both pedagogical and ecological sustainability. Full article

Co-design in healthcare settings requires teams to utilize each other’s knowledge effectively, but practical guidance and simple methods for observing collaboration are often lacking. We tested whether a lightweight AI assistant that guides the process—and automatically logs who speaks, when, and how work progresses—can make teamwork easier to manage and easier to track. Six four-person teams completed the same five-phase session. The assistant nudged timing, turn-taking, and artifact hand-offs; all interactions were recorded in a shared workspace. We assessed usability and acceptance, expert-rated product quality (technical performance), perceived team performance, and self-rated technical contribution, and we summarized basic log signals of participation and pacing (e.g., turn-taking balance, average turn duration). Analyses were descriptive. All teams finished the protocol with complete logs. Outcomes were favorable (expert ratings averaged 4.18/5; perceived performance 6.14/7; self-rated contribution 4.08/5). Teams with more balanced participation and clearer pacing tended to report better performance, whereas simply having more turns did not. A process-guiding AI assistant can quantify teamwork behaviors as markers of collective intelligence and support reflection in everyday clinical co-design; future work will examine the generalizability of these findings across different sites. Full article

Urban noise is a major environmental concern that affects public health and quality of life, demanding new approaches beyond conventional noise level monitoring. This study investigates the development of an AI-driven Acoustic Event Detection and Classification (AED/C) system designed for urban sound recognition and its integration into smart city application. Using the UrbanSound8K dataset, five acoustic parameters—Mel Frequency Cepstral Coefficients (MFCC), Mel Spectrogram (MS), Spectral Contrast (SC), Tonal Centroid (TC), and Chromagram (Ch)—were mathematically modeled and applied to feature extraction. Their combinations were tested with three classical machine learning algorithms: Support Vector Machines (SVM), Random Forest (RF), Naive Bayes (NB) and a deep learning approach, i.e., Convolutional Neural Networks (CNN). A total of 52 models with the three ML algorithms were analyzed along with 4 models with CNN. The MFCC-based CNN models showed the highest accuracy, achieving up to 92.68% on test data. This achieved accuracy represents approximately +2% improvement compared to prior CNN-based approaches reported in similar studies. Additionally, the number of trained models, 56 in total, exceeds those presented in comparable research, ensuring more robust performance validation and statistical reliability. Real-time validation confirmed the applicability for IoT devices, and a low-cost wireless sensor unit (WSU) was developed with fog and cloud computing for scalable data processing. The constructed WSU demonstrates a cost reduction of at least four times compared to previously developed units, while maintaining good performance, enabling broader deployment potential in smart city applications. The findings demonstrate the potential of AI-based AED/C systems for continuous, source-specific noise classification, supporting sustainable urban planning and improved environmental management in smart cities. Full article

This research presents a solution to the problem of controlling the energy demand and carbon footprint of old buildings, with the focus being on a (heated) building located in Madrid, Spain. A framework that incorporates AI and advanced hybrid ensemble approaches to make very accurate energy consumption predictions was developed and tested using the MATLAB environment. At first, the study evaluated six individual AI models (ANN, RF, XGBoost, RBF, Autoencoder, and Decision Tree) using a dataset of 100 points that were collected from the building’s sensors. Their performance was evaluated with high-quality data, which were ensured to be free of missing values or outliers, and they were prepared using L1/L2 normalization to guarantee optimal model performance. Later, higher accuracy was achieved through combining the models by means of hybrid ensemble techniques (voting, stacking, and blending). The main contribution is the application of a Long Short-Term Memory (LSTM) model for predicting the energy consumption of the building and, very importantly, its carbon footprint over a 30-year period until 2050. Additionally, the proposed methodology provides a structured pathway for existing buildings to progress toward nearly Zero-Energy Building (nZEB) performance by enabling more effective control of their energy demand and operational emissions. The comprehensive assessment of predictive models definitively concludes that the blended ensemble method is the most powerful and accurate forecasting tool, achieving 97% accuracy. A scenario where building heating energy use jumps to 135 by 2050 (a 35% increase above 2020 levels) represents an alarming complete failure to achieve energy efficiency and decarbonization goals, which would fundamentally jeopardize climate targets, energy security, and consumer expenditure. Full article

This study explores the complex psychological mechanisms linking career adaptability to work engagement under AI-driven workplaces. We examine the mediating role of job satisfaction and investigate a key hypothesis: that the adaptive benefits of AI self-efficacy are dampened by the emotional costs associated with AI anxiety. A dual-analytical approach was employed on a sample of 311 young Chinese workers. First, we conducted conditional process analysis using PROCESS Model 11 with 5000 bootstrapped samples to test for conditional indirect effects. Second, we utilized latent variable structural equation modeling for robust validation at the structural level. Analyses were adjusted for demographic and occupational covariates. As a result, the initial PROCESS analysis revealed that the key triple interaction (career adaptability × AI self-efficacy × AI anxiety) was statistically significant in all three test models (e.g., Model 1: b = −0.3509, p = 0.0075). Further analysis showed that the positive moderating effect of AI self-efficacy was contingent on AI anxiety; it was strongest at low AI anxiety and weakest (but still significant) at high AI anxiety. However, the more robust latent variable SEM (CMIN/DF = 1.569, CFI = 0.939, RMSEA = 0.043) revealed a critical separation of effects. The indirect effect operates exclusively through intrinsic job satisfaction, which was significantly predicted by the unified second-order career adaptability factor (b = 1.361, BCa 95% CI [1.023, 1.967]). The path from extrinsic satisfaction to WE was non-significant (b = 0.107, BCa 95% CI [−0.030, 0.250]). Furthermore, the SEM isolated a significant direct positive effect from the unified career adaptability factor to work engagement (b = 0.715, BCa 95% CI [0.385, 1.396]). This study highlights that the adaptability–engagement link operates via two distinct mechanisms: an indirect pathway from a unified career adaptability construct through intrinsic job satisfaction, and a direct pathway from career adaptability to work engagement. While PROCESS analysis suggests that anxiety dampens confidence, our SEM results clarify that this should be interpreted cautiously, as the mediation pathway via extrinsic satisfaction is not robust to measurement error. These findings underscore a multi-faceted mandate for organizations: leaders must not only manage AI anxiety but also foster holistic career adaptability to enhance intrinsic job quality and build direct engagement. Full article

This study presents an optimized Real-Time Detection Transformer (RT-DETRv2) deep learning model for the automated assessment of Tartary buckwheat germination and evaluates the influence of soaking and ultrasonic pretreatments on the germination ratio. Model optimization revealed that image chip size critically affected performance. The 512 × 512-pixel chip size was optimal, providing sufficient image context for detection and achieving a robust F1-score (0.9754 at 24 h, tested with a ResNet-101 backbone). In contrast, smaller chips (e.g., 128 × 128 pixels) caused severe performance degradation (24 h F1 = 0.3626 and 48 h F1 = 0.1211), which occurred because the 128 × 128 chip was too small to capture the entire object, particularly as the elongated and highly variable 48 h sprouts exceeded the chip dimensions. The optimized model, incorporating a ResNet-34 backbone, achieved a peak F1-score of 0.9958 for 24 h germination detection, demonstrating its robustness. The model was applied to assess germination dynamics, indicating that 24 h of treatment with 0.1% CaCl₂ and ultrasound enhanced total polyphenol accumulation (6.42 mg GAE/g). These results demonstrate that RT-DETRv2 enables accurate and efficient automated germination monitoring, providing a promising AI-assisted tool for seed quality evaluation and the optimization of agricultural pretreatments. Full article

The rapid digitalization of higher education has transformed how students learn, collaborate, and engage with emerging technologies such as artificial intelligence (AI). While earlier research emphasized generational or academic-level differences in digital behavior, recent evidence suggests convergence in technology use. This study explores whether undergraduate and postgraduate students at the Faculty of Economics and Business, University of Maribor, display distinct technology engagement patterns across five constructs: excessive technology use, online collaboration, the use of digital learning tools (E-boards), the use of AI in education, and perceived academic success. A survey among 285 students was analyzed using the non-parametric Mann–Whitney U test due to a non-normal data distribution. The findings showed that postgraduate students did not report higher levels of E-Board use, online collaboration, or perceived academic success. Undergraduate students scored higher on one item related to excessive technology use, but not across the full construct. However, significant differences emerged in AI use, where postgraduate students showed greater confidence and willingness to integrate AI tools. The findings suggest that digital competence and the quality of technology integration, rather than study level, shape students’ learning experiences. Higher education institutions should promote balanced and ethical technology use, strengthen AI literacy, and foster self-regulated learning skills. Full article

Drawing upon the Elaboration Likelihood Model (ELM) and Trust Transfer Theory, this study investigates how AI digital avatars influence consumer trust and purchase intention. Using survey data collected from 378 valid respondents, the proposed model was tested with Partial Least Squares Structural Equation Modeling (PLS-SEM). The results reveal that brand awareness and perceived quality significantly affect purchase intention through product trust, while social endorsement cues, anthropomorphism, and interaction quality indirectly influence purchase intention through AI Avatar trust. Furthermore, both AI Avatar trust and product trust have significant direct effects on purchase intention, highlighting the critical role of trust in AI-driven persuasion processes. This study validates the dual-route persuasion mechanism of the ELM in AI marketing contexts and extends the application of trust theory to human–AI interactions, offering valuable insights for future research on AI brand endorsement and consumer psychology. Full article

The general spread of water safety awareness and enforcement often masks the escalating risks of emerging contaminants (ECs) that evade standard detection and monitoring techniques. Traditional monitoring infrastructures depend heavily on localized laboratory-based testing, which is expensive, time-consuming, and reliant on specialized infrastructure and skilled personnel. While specific types of ECs and detection technologies have been examined in numerous studies, a significant gap remains in compiling and commenting on this information in a concise framework that incorporates global impact and monitoring strategies. We aimed to compile and highlight the impact ECs have on global water safety and how advanced sensor technologies, when integrated with digital tools such as the Internet of Things (IoT), artificial intelligence (AI), machine learning (ML), geographic information systems (GIS), and cloud-based analytics, can enhance real-time EC detection and monitoring. Recent case studies were reviewed for the assessment of EC types, global contamination, and current state-of-the-art for EC detection and their limitations. An emphasis has been placed on areas that remain unaddressed in the current literature: a cross-disciplinary integration of integrated sensor platforms, multidisciplinary research collaborations, strategic public–private partnerships, and regulatory bodies engagement will be essential in safeguarding public health, protecting aquatic ecosystems, and ensuring the quality and resilience of our water resources worldwide. Full article

Manufacturing companies are increasingly confronted with critical challenges such as a shortage of skilled labor, rising production costs, and ever-stricter quality requirements. These challenges become particularly acute when defect types exhibit high visual variance, making consistent and accurate inspection difficult. Traditionally, visual inspection of high variance errors is performed manually by human operators—a process that is both costly and prone to errors. Consequently, there is a growing interest in replacing human inspection with AI-based visual quality control systems. However, the adoption of such systems is often hindered by limited access to training data, labor-intensive labeling processes, or the absence of real production data during early development stages. To address these challenges, this paper presents a methodology for training AI models using synthetically generated image data. The synthetic images are created using Physically Based Rendering, which enables precise control over rendering parameters and facilitates automated labeling. This approach allows for a systematic analysis of parameter importance and bypasses the need for large real training datasets. As a case study, the focus is on the inspection of laser welds in battery connectors for fully electric vehicles—a particularly demanding application due to the criticality of each weld. The results demonstrates the effectiveness of synthetic data in training robust AI models, thereby providing a scalable and efficient alternative to traditional data acquisition and labeling methods. The trained binary classifier reaches a precision of $0.94$ with a recall of $0.98$ solely trained on synthetic data and tested on real image data. Full article

Building cracks are among the critical building defects, as they can compromise structural integrity, occupant safety and building sustainability. Traditional laborious building inspection methods are cumbersome and erroneous. Computer vision-based crack detection relies on image recognition and does not analyse the underlying causes or suggest rectification strategies. This study explores the potential and limitations of multimodal AI models, that integrate image and text modalities for building crack analysis. As a proof-of-concept, the vision–language model, CracksGPT was built upon a fine-tuned MiniGPT-v2. It was trained on custom crack images with text descriptions detailing visual features, possible causes, and rectification options. It was tested on crack images from a building site in Sydney. When provided with an image of a wall crack, CracksGPT can classify crack patterns of vertical, horizontal, diagonal, and stair-step and interpret possible underlying causes with potential rectification strategies. The ROUGE metric was used for language generation quality evaluation followed by a performance evaluation by building inspection experts. The model’s performance is sensitive to input image quality and training data limitations, specifically in complex scenarios, reaffirming the value of expert overseeing. The findings highlight the potential and limitations of multimodal AI for integrating vision–language reasoning into building inspections. Full article

The collection and annotation of large-scale image datasets remains a significant challenge in training vision-based AI models, especially in domains such as industrial automation. In industrial settings, this limitation is especially critical for quality inspection tasks within Flexible Manufacturing Systems and Batch-Size-of-One production, where high variability in components restricts the availability of relevant datasets. This study presents a pipeline for generating photorealistic synthetic images to support automated visual inspection. Rendered images derived from geometric models of manufactured parts are enhanced using a Cycle-Consistent Adversarial Network (CycleGAN), which transfers pixel-level features from real camera images. The pipeline is applied in two scenarios: (1) domain transfer between similar objects for data augmentation, and (2) domain transfer between dissimilar objects to synthesize images before physical production. The generated images are evaluated using mean Average Precision (mAP) and the Turing test, respectively. The pipeline is further validated in two industrial setups: object detection for a pick-and-place task using a Niryo robot, and anomaly detection in products manufactured by a FESTO machine. The successful implementation of the pipeline demonstrates its potential to generate effective training data for vision-based AI in industrial applications and highlights the importance of enhancing domain quality in industrial synthetic data workflows. Full article

Generative AI is bringing revolutionary changes to architectural design. From data-driven and sustainable perspectives, this study introduces scientific data analysis methods to explore the specific application scenarios and effectiveness of generative AI in the early, middle, and late stages of architectural project design, while also examining its potential value in the field of sustainability. The research first synthesizes industry viewpoints through online data analysis. Secondly, it selects three typical practical architectural projects of different scales and types in which the author participated in comparative testing, recording the time, operational processes, and outputs required for schemes generated by the “traditional creative workflow” vs. the “AI-assisted workflow” at various stages. A multi-dimensional evaluation is conducted combining subjective questionnaires and objective performance simulation data. This study finds that generative AI can significantly enhance design efficiency and scheme diversity and guide the construction of sustainability dimensions, but challenges exist in quality control and technology integration. This research will provide an empirical framework and data benchmarks for architectural practitioners, clarifying a new design path of “data-driven–human–machine collaboration–sustainable optimization”, which holds significant reference value for promoting the transformation of the construction industry towards high efficiency and low carbon. Full article