Search Results (3,682)

Post-war reconstruction requires effective decision-support tools capable of integrating technical, economic, and organizational criteria under conditions of high uncertainty. The evaluation and prioritization of damaged buildings for recovery interventions are critical challenges for reconstruction project management. This study proposes a hybrid decision-support framework for assessing the strategic feasibility of building recovery using a novel Strategic Revitalization Index (SRI). The proposed methodology integrates a hierarchical fuzzy inference system, simulation techniques, and an artificial neural network surrogate model. The fuzzy model aggregates four key evaluation dimensions: technical condition of the building, economic feasibility of recovery actions, project implementation capability, and environmental and social impact. To analyze the model’s behavior and generate training data, a synthetic dataset was created using Latin Hypercube Sampling, covering a wide range of possible reconstruction conditions. The generated dataset was subsequently used to train an artificial neural network capable of approximating the nonlinear mapping implemented by the fuzzy decision model. The obtained results demonstrate high predictive performance of the surrogate model, with R² = 0.976, RMSE = 0.0266, MAE = 0.0133, and MAPE = 4.95%. Scenario analysis further illustrates how different recovery strategies influence SRI values and enables comparison of alternative reconstruction approaches. The proposed framework provides a flexible analytical tool for supporting strategic decision-making in post-war reconstruction projects. By combining fuzzy logic, simulation techniques, and machine learning, the model enables systematic prioritization of recovery strategies and may support large-scale reconstruction planning in post-conflict environments. Full article

The global transition toward carbon-neutral energy solutions has established Solid Oxide Fuel Cells (SOFCs) as a key technology for next-generation power generation. This work presents a comprehensive numerical study and multi-factor statistical analysis of an anode-supported SOFC fueled by synthetic diesel. A three-dimensional computational fluid dynamics model, validated against experimental data, was integrated with a Taguchi L₂₇ orthogonal array to systematically evaluate the influence of six key parameters: temperature, fuel mass flow rate, operating pressure, current load, flow channel configuration, and methane molar fraction. Statistical analysis through the signal-to-noise ratio and analysis of variance identified the operating current as the most significant factor affecting cell voltage, followed by the fuel mass flow rate and temperature. The experiments showed that the highest levels of all factors (except for the current, which had the lowest level) maximize electrochemical performance while maintaining a steam-to-carbon ratio (S/C) within a range of 0.83 to 0.92, calculated based on total carbon content, ensuring sufficient humidification for internal reforming across all tested fuel compositions. Furthermore, a multiple linear regression model was developed as a computationally efficient surrogate, demonstrating exceptional predictive accuracy with an R² of 0.9954 and a mean relative error of 1.76% across independent validation cases. These results provide a robust methodology for rapid design and sensitivity analysis of internal-reforming SOFCs, offering a precise tool for optimizing fuel utilization in high-temperature electrochemical systems. Full article

Track quality is key to ensuring the safety and comfort of passengers and freight in railway systems. However, continuous monitoring is rarely implemented due to its high cost and technical complexity. This paper introduces a methodological framework based on machine learning and optimization algorithms for developing onboard track quality monitoring systems using inertial measurements. The workflow addresses crucial, often overlooked aspects such as sensor location, integrating them with downstream processes. The methodology was validated through its application to longitudinal level quality estimation. Synthetic acceleration signals were generated using multibody simulations under parameter configurations defined through a Design of Experiments framework. A multi-objective optimization approach was applied to determine the optimal combination of sensors, balancing estimation accuracy and efficiency. Among the evaluated models, XGBoost achieved a root mean square error of 0.175 mm on the test set, requiring only two acceleration signals and vehicle speed. The use of features derived from wavelet spectra instead of traditional statistical descriptors reduced the estimation error by approximately 20%. These results demonstrate the feasibility of constructing low-cost, data-driven monitoring systems for track quality assessment and highlight the benefits of a structured methodology integrating data generation, sensor analysis, and learning algorithms. Full article

Generative AI systems trained on synthetic data exhibit progressive degradation known as model collapse. This paper provides a theoretical explanation of this phenomenon using Shannon’s Data Processing Inequality (DPI), modeling iterative synthetic-data training as a Markov chain of lossy transformations. We show that mutual information with respect to the original data distribution must decrease monotonically, yielding qualitative predictions for exponential decay tendencies and indicating that information loss arises from general finite-precision and capacity constraints rather than from any specific architectural mechanism. Building on this analysis, we introduce the AI conceptual theorem, a generalized stability limit for computable systems. The theorem states that any purely computational system that generates outputs iteratively under finite precision, bounded capacity, and without external low-entropy input must experience cumulative information degradation after a finite number of steps. DPI-based collapse emerges as a special case of this broader principle. The framework is intended as a conceptual information-theoretic perspective rather than a fully formalized theory, with several assumptions intentionally simplified to highlight the underlying entropic mechanism. The results should therefore be interpreted as principled limits that motivate further empirical and mathematical investigation rather than as definitive closed-form predictions. Together, DPI and the AI Theorem provide a unified information-theoretic framework for understanding degradation in synthetic training, long-horizon inference, and other iterative computational processes. The resulting predictions are quantitatively falsifiable and offer guidance for designing more stable and information-preserving AI systems. Full article

Cross-domain object detection from optical to Synthetic Aperture Radar (SAR) imagery addresses the challenges of SAR data scarcity and high annotation costs, enabling crucial capabilities for persistent maritime surveillance and reconnaissance. However, the substantial modality gap resulting from distinct imaging mechanisms and severe coherent speckle noise significantly hampers knowledge transfer. Existing Unsupervised Domain Adaptation (UDA) methods, which primarily rely on adversarial feature alignment or static pseudo-labeling, struggle to replicate the physical backscattering properties of SAR data and often fall prey to confirmation bias due to intense background clutter. To overcome these limitations, this paper introduces the Diffusion-Enhanced Mutual Consistency (DEMC) framework. DEMC introduces a novel two-stage adaptation paradigm. The first stage, the Diffusion-Based Domain Alignment (DBDA) module, generates a physics-aware intermediate domain. By integrating step-efficient diffusion generation with physical refinement, this module effectively reduces the cross-modal visual discrepancy while preserving the semantic structure of the optical source. In the second stage, this paper tackles the pervasive issue of pseudo-label noise with the Dual-Student Mutual Verification (DSMV) mechanism. Guided by Cross-Agent Spatial Consensus (CASC) and Adaptive Thresholding (AIT), this mechanism dynamically refines pseudo-labels through geometric overlap validation, effectively recovering faint, low-contrast targets that would typically be discarded by standard thresholds. Extensive evaluations across four benchmark tasks (HRSC2016/ShipRSImageNet to SSDD/HRSID) demonstrate that DEMC establishes a new state-of-the-art. Notably, the framework significantly enhances detection recall and reduces omission errors in complex coastal environments, offering a robust solution for zero-tolerance, all-weather surveillance tasks. Full article

Background: Machine learning-based Artificial Intelligence (AI) models have shown significant potential in the biomedical field, offering promising advances in diagnostics, personalized medicine, and patient care. However, to build these models, we have to deal with important challenges, including (1) the scarcity and low quality of available datasets in many important applications and (2) privacy concerns associated with sensitive patient data. Synthetic data (SD) generation has emerged as a promising strategy to address these challenges, yet many existing approaches struggle to simultaneously preserve privacy and accurately model tabular data, the predominant format in healthcare. Methods: We propose Kernel Density Estimation–K-Nearest Neighbors (KDE-KNN), a privacy-aware tabular data generation method, and evaluate its performance against state-of-the-art techniques. Using sepsis detection as a real-world case study, we assess both data utility and privacy protection. Results: Models trained on KDE-KNN-generated SD outperformed those trained on real data across both internal testing and external validation. In particular, a support vector machine achieved superior performance when trained on SD relative to real data. This gain is likely driven by the balanced class distribution of the synthetic dataset, underscoring KDE-KNN’s utility as an effective data balancing strategy. Consistent performance in external validation further supports the robustness and generalizability of the proposed approach. Privacy evaluation indicated a lower re-identification risk, with a mean distance to closest record of 4.971 between synthetic and real samples, compared with 2.715 among real samples. Conclusions: KDE-KNN effectively captures underlying population distributions while generating high-quality SD that preserve statistical fidelity and protect sensitive information. By balancing the trade-off between utility and privacy, the method produces representative datasets without exposing individual records. These findings position KDE-KNN as a valuable tool for data-scarce and privacy-sensitive applications, with broad potential across healthcare and other data-driven domains. Full article

Background: The development of robust artificial intelligence (AI) models for diagnosing Temporomandibular Disorders (TMDs) is severely constrained by data scarcity and patient privacy regulations. Cone-beam computed tomography (CBCT), the gold standard for assessing osseous changes in the temporomandibular joint (TMJ), inherently contains sensitive biometric facial features, making de-identification difficult without losing critical anatomical information. This study aims to develop and evaluate TMJCTGenerator, a specialized latent diffusion model (LDM) framework designed to synthesize high-fidelity, diverse, and anonymous TMJ CBCT images. We hypothesize that this LDM approach can achieve superior anatomical fidelity and diversity compared to traditional generative adversarial network (GAN)- and variational autoencoder (VAE)-based methods, specifically in capturing fine osseous details within sagittal and coronal views of the mandibular condyle. Methods: A training dataset comprising 348 anonymized CBCT volumes was obtained in this retrospective comparative study to extract high-resolution sagittal and coronal regions of interest of the mandibular condyle. An independent test set of 39 anonymized CBCT volumes was further included. We developed a class-conditional LDM that integrates a pre-trained VAE for perceptual compression with a conditional U-Net for iterative denoising in the latent space. Performance was evaluated via qualitative anatomical fidelity assessment, Fréchet Inception Distance (FID), and a blinded Visual Turing test conducted by experienced clinicians to determine the distinguishability of synthetic images from real data. Results: Qualitative analysis revealed that TMJCTGenerator produced images with superior sharpness and anatomical consistency compared to baseline models, successfully reconstructing fine bone structures essential for diagnosing degenerative joint disease. TMJCTGenerator achieved lower FID scores than both VAE and GAN baselines. In the visual Turing test, clinicians were unable to reliably distinguish the generated images from real scans, and non-inferiority analysis confirmed that the synthetic data were statistically non-inferior to real data. Furthermore, TMJCTGenerator demonstrated the capability to generate diverse pathological conditions, ranging from normal anatomy to severe osteoarthritic changes. Conclusions: The proposed LDM framework effectively addresses the data scarcity and privacy bottlenecks in TMJ AI research by generating realistic, fully anonymous medical imaging data. TMJCTGenerator outperforms traditional generative methods in both visual fidelity and diversity, offering a viable solution for training downstream diagnostic algorithms. The source code and pre-trained models of TMJCTGenerator have been made open-source. Full article

Unmanned aerial vehicle (UAV) imaging systems are increasingly deployed in surveillance, infrastructure monitoring, and smart-city applications, where the integrity of captured visual data is critical. Recent advances in generative models enable highly realistic deepfake manipulations that can compromise aerial sensor streams, particularly under real-world degradations such as motion blur, sensor noise, and compression artifacts. This paper introduces DeepLayer-ID, a degradation-aware multi-domain forensic framework specifically designed for UAV sensing environments. The proposed architecture decomposes forensic evidence into complementary spatial, frequency, and residual domains. A discrete wavelet transform module captures sub-band energy inconsistencies, while high-pass residual filtering isolates sensor pattern anomalies. A lightweight transformer-based fusion mechanism adaptively integrates cross-domain representations to enhance robustness under heterogeneous acquisition conditions. To emulate operational UAV pipelines, we construct a balanced dataset of 1096 aerial frames derived from the VisDrone2019-DET validation subset, incorporating synthetic manipulations and physics-consistent degradations. The experimental results show that DeepLayer-ID achieves 97.8% accuracy and 0.991 AUC, outperforming ResNet-50 (90.9%, 0.942 AUC), XceptionNet (92.4%, 0.957 AUC), and Noiseprint CNN (93.1%, 0.964 AUC). Notably, the model maintains real-time feasibility, with only 5.4 M parameters and 9.8 ms inference latency. These findings demonstrate that structured multi-domain signal decomposition combined with attention-guided fusion provides a robust and computationally efficient solution for deepfake detection in degraded UAV sensing systems. Full article

Ground-penetrating radar (GPR) has applications across many domains, including archaeology, mining, and infrastructure inspection. This review is specifically focused on urban subsurface utility mapping, where accurate detection of buried pipelines, cables, and conduits is critical for excavation safety and infrastructure management. Within this scope, two major barriers are identified: event–utility mismatch and the synthetic–field domain gap. Bibliometric analysis shows increasing reliance on deep learning, yet most methods remain limited to event-level hyperbola detection rather than utility-level inference. In real urban environments, radar responses are often affected by orientation-dependent signatures, clutter, overlapping reflections, and non-utility anomalies, making detected events difficult to map directly to physical infrastructure. In parallel, models trained on synthetic data frequently show limited field generalization because simulated radargrams do not fully reproduce soil heterogeneity, acquisition variability, and system artifacts. The review argues that future progress in urban utility mapping requires a shift toward utility-level reasoning supported by multi-sensor fusion, physics-guided learning, hybrid simulation–field datasets, and uncertainty-aware interpretation. Such advances are essential for making GPR outputs more reliable and actionable in urban engineering practice. Full article

Lightweight Unmanned Aerial Vehicles (UAVs) have limited space, low payload capacity, and constrained power supply capabilities. Therefore, their payloads are constrained by size, weight, and power (SWaP). Thus, designing edge-side signal processing architectures for the payloads of UAVs faces severe challenges. Traditional ASIC design based on manual optimization struggles to meet the demands of low latency and low resource occupancy in edge-side applications. To address this challenge, this paper proposes a signal processing hardware accelerator architecture design framework with algorithm-hardware co-design. The framework employs a cross-level dataflow graph representation to formally capture task characteristics. Reconfigurable dataflow templates and reusable operator IP components are systematically constructed based on this representation. Through multi-objective design space exploration, the framework achieves Pareto-optimal mapping from algorithmic specifications to hardware implementations. Finally, automatic generation of top-level hardware descriptions enables rapid FPGA-based prototyping and functional validation. Taking synthetic aperture radar (SAR) imaging as a study example, compared with non-reconfigurable architectures, this scheme reduces the equivalent gate count by 51.4% without increasing processing latency. Compared with a conventional reconfigurable dataflow architecture, the design improves energy efficiency from 12.8 MS/J to 16.0 MS/J, representing a 25.4% enhancement, while also scaling the supported data processing size by a factor of 4×. It provides a high-performance and scalable hardware acceleration solution for lightweight edge-side computing platforms. Full article

Building reliable vehicle detection models for intelligent transportation systems calls for large, well-annotated datasets—yet gathering and labelling real traffic data remains both costly and labour-intensive. This paper introduces Traffic Synth, an automated pipeline that generates synthetic training datasets by altering real traffic camera images rather than constructing entirely artificial scenes. The system begins by detecting vehicles through instance segmentation and removing them from the frame. It then places new vehicles directly into the cleared regions using diffusion-based inpainting, all while retaining the original road layout, lighting, and camera perspective. Doing so preserves the realistic scene context while broadening the visual variety of vehicles in the dataset. To ensure that the resulting traffic looks physically plausible, we incorporate a lane-aware prompting mechanism that matches each vehicle’s orientation to the direction of travel as seen from the camera. The system further draws on a weighted vehicle brand database that mirrors the makes and colours commonly found on European roads to better match actual deployment conditions. Class-specific mask processing—involving anisotropic scaling and relative dilation—rounds out the pipeline by improving generation quality across different vehicle size categories. The final output is a set of images with automatically generated annotations in a standard object detection format. Full article

Dense retrievers rely heavily on high-quality training triplets, yet existing data construction strategies remain inadequate for reasoning-intensive retrieval tasks involving multi-hop reasoning, entity relation tracing, and implicit evidence composition. Positive samples are often based on shallow semantic relevance and fail to capture explicit reasoning chains, while negative samples are typically sampled from lexical overlap or random candidates and therefore provide limited supervision for learning clear decision boundaries. To address these issues, we propose S-Gens, a structure-aware synthetic data generation framework for enhancing reasoning-intensive dense retrieval. S-Gens uses relation paths in an external knowledge graph to synthesize queries and structurally consistent positive samples, and further constructs semantically similar but structurally inconsistent hard negatives. To improve data reliability, we introduce a Siamese graph neural network-based consistency filtering mechanism. Because S-Gens operates entirely during offline supervision construction, it remains model-agnostic, preserves the original inference architecture, and is complementary to graph-guided retrieval or RAG pipelines that inject structure online. Experiments on five benchmark datasets show that S-Gens consistently improves multiple trainable retrievers, with the largest gains on multi-hop reasoning tasks such as WebQSP and HotpotQA. These results indicate that structure-aware synthetic supervision can effectively improve dense retrieval in reasoning-intensive settings. Full article

Accurate absolute radiometric calibration is critical for ensuring the data quality of spaceborne Synthetic Aperture Radar (SAR) systems and supporting quantitative remote sensing applications. Absolute radiometric calibration generally relies on ground reference targets with known radar cross-section (RCS) deployed at dedicated calibration sites. Such ground-based calibration methods are costly and time-consuming, and calibration frequency is constrained by the distribution of calibration sites and the satellite revisit cycles. Additionally, for specialized SAR missions, such as deep space exploration, deploying calibration equipment on the observed extraterrestrial surface is infeasible. This study proposes a space-based absolute calibration concept using a small calibration satellite carrying a well-characterized reference (e.g., a passive reflector or an active transponder) and flying in formation with the SAR satellite. The relative motion ensures a side-looking acquisition geometry, enabling the SAR to image the accompanying target and derive calibration factors. The overall calibration process is divided into two stages: determination of an in-orbit calibration factor using the calibration satellite, followed by its transformation to accommodate ground imaging conditions. This method effectively isolates the radar system gain to characterize the intrinsic hardware response. Furthermore, by operating entirely in space, it avoids atmospheric and ground-clutter distortions, ensuring a fully space-based, end-to-end calibration process dominated primarily by sensor systematic errors. Moreover, it allows for more frequent and flexible calibration, eliminating reliance on ground calibration sites and infrastructure. The feasibility and advantages of the proposed concept are demonstrated through comprehensive simulations, covering orbit analysis, echo simulation, and image processing. Full article

The ORCI project aims to develop an AI-based decision-support tool to assist air traffic controllers in complex TMA operations, taking Barcelona’s transitions as the primary use case. Using historical radar data, the tool has been trained to predict spacing between consecutive arrivals based on real-time vectoring commands. A data-processing pipeline was developed to clean, classify and validate flight trajectories, and synthetic samples were generated to enable a wider variety of situations. Explainable ML models achieved a mean absolute error of around 0.38 NM, demonstrating strong predictive capability. The results show the potential of ORCI to improve sequencing efficiency and runway throughput. Full article

Motor fault diagnosis under small-sample conditions remains challenging because limited labeled data often cause deep models to overfit and generalize poorly. To address this problem, we propose STR-DDPM, a fault data augmentation framework that combines moving-average-based seasonal–trend–residual decomposition with a denoising diffusion probabilistic model. Specifically, multichannel signals are decomposed into trend, seasonal, and residual components, and class-conditional diffusion modeling is performed only in the residual domain. This design emphasizes fault-related stochastic variations while reducing interference from deterministic structures. To improve generation stability, we adopt velocity prediction and develop an enhanced one-dimensional U-Net with multi-scale convolutions, channel attention, self-attention, and feature-wise linear modulation for controllable conditional generation. Experiments on the University of Ottawa and Paderborn motor fault datasets demonstrate that the proposed method generates samples that are highly consistent with real data and improves diagnostic performance under multiple synthetic-data-assisted settings. These results indicate that STR-DDPM provides an effective and practical solution for data augmentation in data-limited motor fault diagnosis. Full article