Search Results (19,757)

We investigate how symmetries present in datasets affect the structure of the latent space learned by Variational Autoencoders (VAEs). Understanding symmetries in data is essential because symmetries determine the true degrees of freedom, constrain generalization, and provide physically interpretable coordinates. We therefore study whether a standard, non-equivariant VAE can reveal symmetry-induced dimensional reduction directly from data, without imposing the symmetry in the architecture. By training VAEs on data originating from simple mechanical systems and particle collisions, we analyze the organization of the latent space through a relevance measure that identifies the most meaningful latent directions. We show that when symmetries or approximate symmetries are present, the VAE self-organizes its latent space, effectively compressing the data along a reduced number of latent variables. This behavior captures the intrinsic dimensionality determined by the symmetry constraints and reveals hidden relations among the features. Furthermore, we provide a theoretical analysis of a simple toy model, demonstrating how, under idealized conditions, the latent space aligns with the symmetry directions of the data manifold. We illustrate these findings with examples ranging from two-dimensional datasets with $O\left(2\right)$ symmetry to realistic datasets from electron–positron and proton–proton collisions. Our results highlight the potential of unsupervised generative models to expose underlying structures in data and offer a novel approach to symmetry discovery without explicit supervision. Full article

Stance detection has emerged as an essential tool in natural language processing for understanding how individuals express agreement, disagreement, or neutrality toward specific targets in social and online discourse. It plays a crucial role in bilingual and multilingual environments, including English-Arabic social media ecosystems, where differences in language structure, discourse style, and data availability pose significant challenges for reliable stance modelling. Existing approaches often struggle with target awareness, cross-lingual generalization, robustness to noisy user-generated text, and the interpretability of model decisions. This study aims to build a reliable, explainable target-aware bilingual stance-detection framework that generalizes across heterogeneous stance formats and languages without retraining on a dataset specific to the target language. Thus, a unified dual-encoder architecture based on mDeBERTa-v3 is proposed. Cross-language contrastive learning offers an auxiliary training objective to align English and Arabic stance representations in a common semantic space. Robustness-oriented regularization is used to mitigate the effects of informal language, vocabulary variation, and adversarial noise. To promote transparency and trustworthiness, the framework incorporates token-level rationale extraction, enables fine-grained interpretability, and supports analysis of hallucination. The proposed model is tested on a combined bilingual test set and two structurally distinct zero-shot benchmarks: MT-CSD and AraStance. Experimental results show consistent performance, with accuracies of 85.0% and 86.8% and F1-scores of 84.7% and 86.8% on the zero-shot benchmarks, confirming stable performance and realistic generalization. Ultimately, these findings reveal that effective bilingual stance detection can be achieved via explicit target conditioning, cross-lingual alignment, and explainability-driven design. Full article

Urban green spaces (UGSs) are vital ecological infrastructure supporting climate resilience, public health, and environmental equity. Despite UGS’s importance, the distribution of UGS in rapidly growing desert cities is wildly disproportionate, as evidenced by a recent study that links UGS availability with road hierarchy using the Road Buffer Framework. Using Landsat 8-derived UGS (overall accuracy = 0.885; Kappa = 0.853), OpenStreetMap Roads, and WorldPop Population Data, this study found that UGS availability per capita is very low across all road classifications (0.020–0.033 m²/person) and falls significantly short of the World Health Organization’s (WHO) recommendation of 9 m²/person. Primary roads only marginally improved based on distance from roadways (0.026–0.032 m²/person), and secondary roads are experiencing little to no change (0.025–0.026 m²/person). Further, Tertiary roads show the most significant loss, with only 0.022 m²/person available within the 0–300 m buffers containing the most people. In addition, urban green spaces are still significantly inequitable, as demonstrated by Gini coefficient results of >0.80, peaking at 0.895, indicating that UGS availability per capita is substantially below international benchmarks. Therefore, the findings highlight the need of incorporating roadside greening, small park areas, and greenways into our transportation planning efforts to support the UN’s Sustainable Development Goals (SDG) 3, 10, 11, and 13. Full article

Focusing on the construction of a 58-m-diameter double-layer steel space frame dome at the Han Culture Museum Assembly Hall, this study addresses scheme selection and safety control challenges in staggered jacking of large-span spatial structures. A three-dimensional finite element model in MIDAS Gen simulated the three-stage jacking process to compare three temporary support layouts. Numerical evaluation metrics included maximum vertical displacements, peak internal forces, the proportion of members undergoing stress state transitions, and spatio-temporal evolution of stress concentrations. Scheme B demonstrated superior performance, reducing peak vertical displacement by 44% under critical conditions, lowering peak stresses, and enabling more uniform internal force redistribution—effectively mitigating tension–compression cycling and buckling risks. Crucially, only nodal displacements and support elevations were monitored in situ using a 3D system based on magnetic prisms and total stations; no strain or force measurements were conducted due to practical constraints during construction. Monitoring data show good agreement with simulated displacements and support elevations under Scheme B, validating the model’s deformation response. However, localized deviations—including a 29 mm deflection discrepancy and elevation errors up to 28 mm—reveal the influence of uneven boundary conditions, with potential implications for long-term structural behavior. The findings confirm that numerical predictions of deformation are reliable, while internal forces remain unvalidated by field data. The integrated approach of “scheme comparison–construction simulation–full-process displacement monitoring” proves effective for safety control and decision-making in complex jacking operations, offering a transferable framework for similar large-span double-layer space frame projects. Full article

Accurate remaining useful life (RUL) prediction of turbofan engines is critical for aviation safety and maintenance optimization; however, deep learning approaches often lack interpretability and require extensive training data. This study proposes a framework integrating Linear Discriminant Analysis (LDA) with Kalman filtering for turbofan engine prognostics. The methodology projects high-dimensional sensor measurements onto a two-dimensional LDA subspace, where degradation trajectories are tracked using state-space estimation, with RUL predictions derived from distances to learned critical failure boundaries. A health index-based classification scheme partitions engine states into three operational regions: Critical, Warning, and Healthy. Three Kalman filter variants—Linear Kalman Filter (LKF), Extended Kalman Filter (EKF), and Unscented Kalman Filter (UKF)—were compared against an Autoregressive (AR) baseline using the NASA C-MAPSS dataset. Using the Prognostics and Health Management 2008 asymmetric scoring function, UKF achieved the best performance with a Score of 552572, representing a 54.9% improvement over AR (1224299), indicating substantially fewer late predictions. While RMSE values remained comparable across methods (36–37 cycles), the Kalman filter variants demonstrated meaningful improvements in avoiding dangerous late predictions critical for safety-oriented maintenance scheduling. EKF also demonstrated substantial improvement with 36.1% Score reduction. Classification accuracy improved from 70.72% (AR) to 73.27% (UKF). The proposed LDA–Kalman framework provides a computationally efficient and geometrically interpretable alternative to deep learning methods for real-time engine health monitoring. Full article

Aboveground storage tanks are used to store various fluids and chemicals for many industrial purposes. According to API standard 653, the structural integrity of these tanks must be regularly assessed. The U.S. EPA requires each operator to have a Spill Prevention, Control and Countermeasure Plan (SPCC) for aboveground storage containers. The accepted practice for inspection of these tanks, particularly the tank bottoms, requires removing the tank from service, emptying the tank, and interior entry for direct inspection of the structure. The required inspection operations are hazardous due to the chemicals themselves as well as the requirement to operate within confined spaces. An inspection from outside the tank would have significant cost and time benefits and would provide a large reduction in the risks faced by inspection personnel. Guided wave (GW) testing is a promising candidate for screening of storage tank walls and bottoms from the tank exterior due to the ability of GWs to propagate over long distances from a fixed probe location. The lowest-order transverse-motion guided wave modes (e.g., torsional vibrations in pipes) are a good choice for long-range inspection because this mode is not dispersive; therefore, the wave packets do not spread out in time. A common weakness of guided wave inspection is the complexity of report generation in the presence of multiple geometry features in the structure, such as welds, welded plate corners, attachments and so on. In some cases, these features cause generation of non-relevant indications caused by mode conversion. Another significant challenge in applying GW testing is development of probes with high-enough signal amplitudes and relatively small footprints to allow them to be mounted on short tank bottom extensions. In this paper, a new generation of magnetostrictive transducers will be presented. The transducers are based on the reversed Wiedemann effect and can generate shear horizontal mode guided waves over a wide frequency range (20–150 kHz) with SNRs in excess of 50 dB. The recently developed SwRI MST 8 × 8 probe contains an array of eight pairs of individual magnetostrictive transducers (MsTs). The data acquisition hardware allows acquisition using Full Matrix Capture (FMC) and analysis software reporting of anomalies based on Total Focusing Method (TFM) image reconstruction. This novel inspection package allows generation of reports that map out corrosion locations and provide estimates of defect widths. Case studies of this technology on actual storage tank walls and bottoms will be presented together with validation of processing methods on mockups with known anomalies and geometry features. Full article

General-purpose large language models excel at open-domain question answering, but in railway operation and maintenance (O&M) scenarios they still suffer from hallucinated knowledge and poor domain adaptation. In practice, railway O&M knowledge mainly arises from two heterogeneous sources: spatio-temporal data such as train trajectories, which are organized along the spatial layout of railway lines, and domain documents such as operating rules, which exhibit varying degrees of structural regularity. Traditional retrieval-augmented generation (RAG) systems usually flatten these multi-source data into a single unstructured text space and perform global retrieval in one embedding space, which easily introduces noisy context and makes it difficult to precisely target knowledge for specific lines, sections, or equipment states. To overcome these limitations, we propose CARAG, a context-aware RAG framework tailored to railway O&M data. CARAG treats domain documents and spatial data as a unified knowledge substrate and builds a spatial knowledge graph with concept and instance levels. On top of this knowledge graph, a GraphReAct-based multi-turn interaction mechanism guides the LLM to reason and act over the concept knowledge graph, dynamically navigating to spatially and semantically relevant candidate regions, within which vector retrieval and instance-level graph retrieval are performed. Experiments show that CARAG significantly outperforms baseline RAG methods on RAGAS metrics, confirming the effectiveness of structure-guided multi-step reasoning for question answering over multi-source heterogeneous railway O&M data. Full article

With the increase in human space activities, there is a significant amount of space debris as well as defunct satellites that seriously threaten the safety of spacecraft in their orbits. The laser ranging technique is one of the most accurate methods of ground-based space target observation. Therefore, it is very meaningful to study efficient tracking and observation methods for defunct satellites and space debris. In this paper, time bias, which is in advance of the actual observation, was added by analyzing the deviation of the orbit prediction such as the time bias and range bias of the runaway space objects. A new method was used for the determination of the TB and RB in real-time tracking and in data processing. The data produced from the observation of the out-of-control targets, such as the Topex satellite and CZ-2C Long March Launch Vehicle, were presented and analyzed. Taking the laser ranging data of the Topex satellite obtained on 19 November 2019, as an example, the result of the first observation circle provided the initial time bias value for the second observation circle, proving that the laser ranging method for runaway space objects is effective. The results of this paper can effectively improve the acquisition efficiency of the defunct satellites. Full article

This article addresses a key evidence gap in urban safety policy in Colombia: the absence of publicly accessible microdata that jointly measure victimization, perception of security, and probability of sanctions among socioeconomically vulnerable residents. It aims to provide a clean, linkable dataset that enables analysis of variations in these issues across demographic and territorial groups in Cali (recently classified as the 29th most dangerous city worldwide, with 1028 and 1065 homicides in 2024 and 2025, respectively). It reports face-to-face survey data collected from 22 July to 16 August 2024, at Sistema de Identificación de Potenciales Beneficiarios de Programas Sociales (SISBEN) service points. The final dataset includes 2139 adults (aged 18–95 years) and combines (i) primary responses on perceived safety (e.g., public space safety and surveillance cameras), perceived likelihood of sanction, victimization, and self-protection measures with (ii) selected sociodemographic and household characteristics drawn from SISBEN IV records. Individual-level linkage was implemented using respondent identification at interviews, yielding an integrated anonymized file suitable for replication and secondary analysis. The dataset enables distributive analyses of insecurity (e.g., by sex, age, and ethnicity—including Afro-descendant populations) within a policy-relevant target group and supports evaluation and targeting of local interventions by providing individual-level indicators. Full article

The Galileo Search and Rescue (SAR) service is the contribution from the European constellation to the international Cospas–Sarsat system. This system uses a variety of space and ground infrastructure to detect and localize distress signals from beacons on the 406 MHz frequency. Satellites in different orbits detect the signals coming from the Earth and transmit them back to Earth stations that route them to the appropriate government authorities. On top of the standard detection and relay service, the Galileo constellation is the first to offer a Return Link Service (RLS) that acknowledges the processing of the distress signal with a Return Link Message (RLM) back to the originating beacon. This RLM is transmitted in the SAR field of the E1 signal I/NAV message, which allocates 20 bits every 2 s page. Therefore, transmitting a short RLM (80 bits) takes four consecutive pages or eight seconds. Moreover, each RLM is transmitted in parallel from two Galileo satellites. The RLS has been active since 2020, avoiding the spotlight of the GNSS community. This paper presents an analysis of the SAR Return Link Messages extracted from more than 3 months of signal-in-space data to investigate the current bandwidth use, monitor the type of SAR usage, and detect anomalies in the service. To extract and parse the Return Link Messages, we have developed and published an open-source Python library called GalileoSARlib on GitHub, which is also detailed in the paper. Full article

Target decomposition is an essential method for the interpretation of polarimetric Synthetic Aperture Radar (SAR). Most current polarimetric target decomposition methods are designed for quad-pol SAR data, while there is a scarcity of methods tailored for dual-pol SAR data, and these methods often struggle to accurately capture the complete scattering components of targets. Compared to quad-pol SAR, space-borne SAR systems more frequently acquire dual-pol SAR data, which offers a wider observation swath and higher resolution. The fast generalized polarimetric target decomposition (FGPTD) method has exhibited excellent target decomposition performance for quad-pol SAR data by searching for the optimal scattering models through nonlinear optimization. To address the core problem of inaccurate scattering component extraction in dual-pol SAR, deep learning is adopted to simulate the nonlinear optimization process of the FGPTD method. Its powerful nonlinear mapping capability enables the model to learn the intrinsic correlation between dual-pol SAR data and the complete scattering components obtained by FGPTD. Therefore, this paper proposes a model and learning-aided target decomposition method for dual-pol SAR. Firstly, FGPTD is performed on existing quad-pol SAR data. Subsequently, a mapping set between dual-pol SAR data and scattering components is constructed. Then, a neural network that integrates residual connections and dilated convolutional kernels is trained using the constructed mapping set. Finally, the well-trained neural network is tested on dual-pol SAR data from other regions and other sensors. Experimental results demonstrate that the proposed method’s target decomposition results are close to those of quad-pol target decomposition and superior to current state-of-the-art dual-pol target decomposition methods. Full article

Guided bone regeneration (GBR) relies on barrier membrane integrity to prevent soft-tissue ingrowth. Although collagen membranes are widely used, their limited longevity can compromise space maintenance, underscoring the need for strategies that enhance membrane stability without impairing the regenerative potential. We hypothesized that thermal denaturation of platelet-poor plasma (PPP), combined with heat-induced modifications of collagen fibrils, could generate a volume-stable, plasma-rich composite that preserves membrane structure and restricts cellular penetration. To test this proof-of-principle concept, collagen membranes were soaked in PPP and either kept at room temperature or subjected to thermal treatment (75 °C/10 min) prior to implantation in rat calvarial defects. Bone regeneration and membrane behavior were evaluated after three weeks using micro-computed tomography (micro-CT) and histology. Micro-CT suggested only minor numerical differences in mineralized tissue between groups; however, these data should not be overinterpreted because micro-CT cannot differentiate mineralization formed within the collagen membrane from mineralization adjacent to it. Consistent with this limitation, histology demonstrated that mineral deposition and early bone formation extended into the structure of room-temperature PPP membranes, whereas mineralized tissue in the thermally treated group was predominantly located outside the membrane, indicating reduced osteoconductive integration within the membrane. Together, these findings support that thermal denaturation of PPP shifts early composite membrane behavior toward barrier-dominant characteristics at the expense of intramembranous mineralization. Full article

Although intensive outpatient programmes (IOPs) are becoming more prevalent, the evidence base, particularly within the UK, remains limited. Given clinicians’ central role in developing, delivering, and adapting these emerging models of care, their perspectives are essential to understanding how IOPs function in practice. This study therefore aims to address a significant gap in the literature by exploring clinicians’ experiences of working with an IOP and the strengths and opportunities arising from this. Fifteen experienced clinicians participated in individual semi-structured interviews after working with the IOP. Open-ended questions guided the discussions, which were recorded and transcribed verbatim. Data were analysed using the six stages of reflexive thematic analysis. The analysis generated three key themes: (1) Tri-directional Collaboration, (2) Creating Space for Change, and (3) Transitions as Turning Points. Clinicians felt that the IOP provided a structure that strengthened and reinforced the therapeutic alliance between parents and clinicians, helped arrest rapid deterioration, and created space for thoughtful planning. Embedding IOPs within stepped-care frameworks may offer an effective and scalable means of expanding system capacity while delivering enhanced, flexible support during periods of heightened risk. However, longitudinal, mixed-methods evaluations are needed to clarify the sustainability of progress post-IOP and to identify predictors of positive transitions. Full article

Deep learning has profoundly transformed remote sensing, yet prevailing architectures like Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) remain constrained by critical trade-offs: CNNs suffer from limited receptive fields, while ViTs grapple with quadratic computational complexity, hindering their scalability for high-resolution remote sensing data. State Space Models (SSMs), particularly the recently proposed Mamba architecture, have emerged as a paradigm-shifting solution, combining linear computational scaling with global context modeling. This survey presents a comprehensive review of Mamba-based methodologies in remote sensing, systematically analyzing about 120 Mamba-based remote sensing studies to construct a holistic taxonomy of innovations and applications. Our contributions are structured across five dimensions: (i) foundational principles of Vision Mamba architectures, (ii) micro-architectural advancements such as adaptive scan strategies and hybrid SSM formulations, (iii) macro-architectural integrations, including CNN–Transformer–Mamba hybrids and frequency-domain adaptations, (iv) rigorous benchmarking against state-of-the-art methods in multiple application tasks, such as object detection, semantic segmentation, change detection, etc. and (v) critical analysis of unresolved challenges with actionable future directions. By bridging the gap between SSM theory and remote sensing practice, this survey establishes Mamba as a transformative framework for remote sensing analysis. To our knowledge, this paper is the first systematic review of Mamba architectures in remote sensing. Our work provides a structured foundation for advancing research in remote sensing systems through SSM-based methods. We curate an open-source GitHub repository to foster community-driven advancements. Full article

Landscape quality significantly impacts residents’ well-being through visual perception, particularly among the elderly who exhibit heightened sensitivity to environmental stimuli. Therefore, this study investigates how landscape configurations influence emotional and physiological responses in older adults under controlled visual conditions. This study selected representative outdoor activity sites in northern Chinese cities and designed five landscape scenarios by adjusting the green coverage ratio (GCR) and landscape composition. Participants (mean age 64.8) reported feelings of pleasure, relaxation, and fatigue while viewing screen-based landscape images, with simultaneous recording of attention-to-interest area (AOIA), pupil diameter range (PD), and electroencephalogram (EEG) data. Research findings reveal a non-linear relationship between the GCR and emotional and physiological responses among elderly populations: when the GCR increased from 18.4% to 38.1%, participants reported significantly heightened feelings of pleasure and relaxation, alongside marked reductions in fatigue-related physiological indicators. However, when the GCR further rose to 48.5%, both reported subjective measures and physiological indicators deteriorated among elderly participants. Under equivalent green coverage conditions, water features within natural settings enhance visual focus on natural elements more effectively than purely green landscapes. Women demonstrated greater sensitivity to changes in the GCR. Correlation analysis further indicated that visual attention among the elderly positively correlated with positive emotions and negatively correlated with fatigue-related physiological responses. This research provides valuable guidance for green space design. Full article