Search Results (642)

Multi-view data, captured from various perspectives, is crucial for training view-invariant human action recognition models, yet its acquisition is hindered by spatio-temporal constraints and high costs. This study aims to develop the Pose Scene EveryWhere (PSEW) framework, which automatically generates temporally consistent, multi-view 3D human action data from a single monocular video. The proposed framework first predicts 3D human parameters from each video frame using a deep learning-based Human Mesh Recovery (HMR) model. Subsequently, it applies tracking, linear interpolation, and Kalman filtering to refine temporal consistency and produce naturalistic motion. The refined human meshes are then reconstructed into a virtual 3D scene by estimating a stable floor plane for alignment, and finally, novel-view videos are rendered using user-defined virtual cameras. As a result, the framework successfully generated multi-view data with realistic, jitter-free motion from a single video input. To assess fidelity to the original motion, we used Root Mean Square Error (RMSE) and Mean Per Joint Position Error (MPJPE) as metrics, achieving low average errors in both 2D (RMSE: 0.172; MPJPE: 0.202) and 3D (RMSE: 0.145; MPJPE: 0.206) space. PSEW provides an efficient, scalable, and low-cost solution that overcomes the limitations of traditional data collection methods, offering a remedy for the scarcity of training data for action recognition models. Full article

Vehicle kilometers traveled (VKT) is a critical metric in transportation and environmental research. However, conventional VKT estimation approaches frequently fail to capture the complexity of route selection and spatiotemporal dynamics of individual road users. This study presents a framework for accurately estimating the total VKT using high-resolution trajectory data obtained from a commercial navigation system. To address the structural limitations of conventional origin destination matrix-based models, such as the modifiable areal unit problem, representative routes were identified based on cumulative travel distance coverage. A novel metric, coverage of estimated travel (CET), was introduced to quantify the explanatory capacity of these routes in approximating total travel distance. Representative routes were selected to maximize CET, and the resulting VKT estimates were validated against national statistical yearbook data. Robustness was further evaluated using mean absolute percentage error, correlation analysis, paired t-tests, and bootstrap-based confidence intervals. The results indicated that as few as five representative routes accounted for over 80% of the total estimated VKT, exhibiting strong agreement with the national statistics after temporal adjustment. These findings demonstrate that trajectory data can serve as a practical alternative to traditional methods, offering higher spatial resolution and enabling dynamic traffic analyses that support transportation policy and environmental planning. Full article

In recent years, the rapid expansion of urban rail transit has significantly improved travel efficiency, yet it has also exacerbated spatial inequality in service coverage. Accessibility, as a fundamental metric for evaluating the equity of service distribution, remains limited by three major shortcomings in current assessment methods: the neglect of actual road network characteristics, reliance on a single static scale, and the absence of quantitative mechanisms to assess accessibility equity. These deficiencies hinder a comprehensive understanding of how equity evolves with the spatiotemporal dynamics of rail systems. To address the aforementioned issues, this study proposes an innovative spatiotemporally dynamic and multi-scale analytical framework for evaluating urban rail accessibility and its equity implications. Specifically, we develop a network-based buffer decay model to refine service population estimation by incorporating realistic walking paths, capturing both distance decay and road network constraints. The framework integrates multiple spatial analytical techniques, including the Gini coefficient, Lorenz curve, global and local spatial autocorrelation, center-of-gravity shift, and standard deviation ellipse, to quantitatively assess the equity and evolutionary patterns of accessibility across multiple spatial scales. Taking the central urban area of Hangzhou as a case study, this research investigates the spatiotemporal patterns and equity changes in metro station accessibility in 2019 and 2023. The results indicate that the expansion of the metro network has partially improved overall accessibility equity: the Gini coefficient at the TAZ (Traffic Analysis Zone) scale decreased from 0.56 to 0.425. Nevertheless, significant inequality remains at finer spatial resolutions (grid-level Gini coefficient = 0.404). In terms of spatial pattern, the core area (e.g., Wulin Square) forms a ‘high-high’ accessibility agglomeration area, while the urban fringe area (e.g., northern Yuhang) presents a ‘low-low’ agglomeration, and the problem of local ‘accessibility depression’ still exists. Additionally, the accessibility centroid has consistently shifted northwestward, and the long axis of the standard deviation ellipse has rotated from an east–west to a northwest-southeast orientation, indicating a growing spatial polarization between core and peripheral zones. The findings suggest that improving equity in urban rail accessibility cannot rely solely on expanding network size; rather, it requires coordinated strategies involving network structure optimization, branch line development, multimodal integration, and the construction of efficient transfer systems to promote more balanced and equitable spatial distribution of rail transit resources citywide. Full article

Background: A single lumbar-mounted inertial sensor offers a practical alternative to optoelectronic systems for gait analysis, simplifying measurements and improving usability in the clinical field. However, its validity can be influenced by sensor placement and signal choice. This study aimed to develop and validate a novel algorithm for estimating spatiotemporal parameters using anteroposterior linear acceleration and angular velocity around the sagittal axis using a single inertial measurement unit (IMU) placed on the lumbar region. The proposed algorithm was validated comparing the parameters computed by the algorithm with the ones computed using a commercial wearable system based on a two-foot-mounted IMU configuration. Thirty healthy subjects underwent a 2 min walk test, and five spatiotemporal parameters were computed using the two methodologies. Study results showed that cadence and gait cycle time exhibited very high agreement, with only a small, statistically significant bias in cadence negligible for practical purposes. In contrast, swing, stance, and double-support parameters showed disagreement due to the presence of systematic proportional errors. This work introduces a novel algorithm for gait event detection and spatiotemporal parameter estimation, addressing uncertainties related to sensor placement, metric models, processing techniques, and signal selection, while avoiding synchronization issues associated with using multiple sensors. Full article

The Gadouras Reservoir, Rhodes Island’s primary water source, experiences recurrent whiting events—milky turbidity from calcium carbonate precipitation—that challenge treatment operations, with impacts compounded by a major 2023 wildfire in this fire-prone Mediterranean setting. To elucidate these dynamics, a pragmatic, multi-source monitoring framework integrates archived Sentinel-2 and Landsat imagery with treatment-plant records (2017–mid-2025). Unitless spectral indices (e.g., AreaBGR) for whiting detection and chlorophyll-a proxies are combined with laboratory measurements of turbidity, pH, total organic carbon, manganese, and hydrological metrics, analyzed via spatiotemporal Hovmöller diagrams, Pearson correlations, and interrupted time-series models. Two seasonal whiting regimes are identified: a biogenic summer mode (southern origin; elevated chlorophyll-a; water temperature > 15 °C; pH > 8.5) and a non-biogenic winter mode (northern inflows). Following the wildfire, the system exhibits characteristics that could be related to possible hypolimnetic anoxia, prolonged whiting, a ~50% rise in organic carbon, and a manganese excursion to ~0.4 mg L⁻¹ at the deeper intake. Crucially, the post-fire period shows a decoupling of AreaBGR from turbidity (r ≈ 0.233 versus ≈ 0.859 pre-fire)—a key diagnostic finding that confirms a fundamental shift in the composition and optical properties of suspended particulates. The manganese spike is best explained by the confluence of a wildfire-induced biogeochemical predisposition (anoxia and Mn mobilization) and a consequential operational decision (relocation to a deeper, Mn-rich intake). This framework establishes diagnostic baselines and thresholds for managing fire-impacted reservoirs, supports the use of remote sensing in data-scarce systems, and informs adaptive operations under increasing climate pressures. Full article

Modern logistics operations require real-time adaptive solutions for three-dimensional bin packing that maintain spatial symmetry and load balance. This paper introduces a time-series-based online 3D packing problem with dual unknown sequences, where containers and items arrive dynamically. The challenge lies in achieving symmetric distribution for stability and optimal space utilization. We propose the Second-Order Dual Pointer Adversarial Network (So-DPAN), a deep reinforcement learning architecture that leverages symmetry principles to decompose spatiotemporal optimization into sequence matching and spatial arrangement sub-problems. The dual pointer mechanism enables efficient item-container pairing, while the second-order structure captures temporal dependencies by maintaining symmetric packing patterns. Our approach considers geometric symmetry for spatial arrangement and temporal symmetry for sequence matching. The Actor-Critic framework uses symmetry-based reward functions to guide learning toward balanced configurations. Experiments demonstrate that So-DPAN outperforms DQN, DDPG, and traditional heuristics in solution quality and efficiency while maintaining superior symmetry metrics in center-of-gravity positioning and load distribution. The algorithm exploits inherent symmetries in packing structure, advancing theoretical understanding through symmetry-aware optimization while providing a deployable framework for Industry 4.0 smart logistics. Full article

Urban rivers, as crucial components of ecosystems, serve multiple functions, including flood control, drainage, and landscape services. However, with the acceleration of urbanization, factors such as industrial wastewater discharge, domestic sewage leakage, and surface runoff pollution have led to increasingly severe degradation of water quality in urban rivers. Unmanned aerial vehicle (UAV) remote sensing technology, with its sub-meter spatial resolution and operational flexibility, demonstrates significant advantages in the detailed monitoring of complex urban water systems. This study proposes an Ensemble Self-Attention Enhanced Mixture Density Network (ESA-MDN), which integrate an ensemble learning framework with a mixture density network and incorporates a self-attention mechanism for feature enhancement. This approach better captures the nonlinear relationships between water quality parameters and remote sensing features, achieving high-precision modeling of water quality parameter distributions. The resulting spatiotemporal distribution maps provide valuable support for pollution source identification and management decision making. The model successfully retrieved five water quality parameters, Chl-a, TSS, COD, TP, and DO, and validation metrics such as R², RMSE, MAE, MSE, MAPE, bias, and slope were utilized. Key metrics for the ESA-MDN test set were as follows: Chl-a (R² = 0.98, RMSE = 0.31), TSS (R² = 0.93, RMSE = 0.27), COD (R² = 0.93, RMSE = 0.39), TP (R² = 0.99, RMSE = 0.02), and DO (R² = 0.88, RMSE = 0.1). The results indicated that ESA-MDN can effectively extract water quality parameters from multispectral remote sensing data, with the generated spatiotemporal water quality distribution maps providing crucial support for pollution source identification and emergency response decision making. Full article

Fodder maize constitutes a key economic crop in Thailand, particularly in the northeastern region, where it significantly contributes to livestock feed production and local economic development. Nevertheless, the planning and management of cultivation areas remain a major challenge, especially in urban and peri-urban agricultural zones, due to the limited availability of spatial data and suitable analytical frameworks. These difficulties are exacerbated in urban settings, where the complexity of land use patterns and high spectral similarity among land cover types hinder accurate classification. The Google Earth Engine (GEE) platform provides an efficient and scalable solution for geospatial data processing, enabling rapid land use classification and spatiotemporal analysis. This study aims to enhance the classification accuracy of fodder maize cultivation areas in Mueang District, Nakhon Ratchasima Province, Thailand—an area characterized by a heterogeneous mix of urban development and agricultural land use. The research integrates GEE with four machine learning algorithms: Random Forest (RF), Support Vector Machine (SVM), Naïve Bayes (NB), and Classification and Regression Trees (CART). Eleven datasets were developed using Sentinel-2 imagery and a combination of biophysical variables, including elevation, slope, Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI), and Modified Normalized Difference Water Index (MNDWI), to classify land use into six categories: fodder maize cultivation, urban and built-up areas, forest, water bodies, paddy fields, and other field crops. Among the 44 classification scenarios evaluated, the highest performance was achieved using Dataset 11—which integrated all spectral and biophysical variables—with the SVM classifier. This model attained an overall accuracy of 97.41% and a Kappa coefficient of 96.97%. Specifically, fodder maize was classified with 100% accuracy in both Producer’s and User’s metrics, as well as a Conditional Kappa of 100%. These findings demonstrate the effectiveness of integrating GEE with machine learning techniques for precise agricultural land classification. This approach also facilitates timely monitoring of land use changes and supports sustainable land management through informed planning, optimized resource allocation, and mitigation of land degradation in urban and peri-urban agricultural landscapes. Full article

Autonomic dysfunction is a key non-motor feature of Parkinson’s disease (PD) and may influence motor performance, particularly gait. While heart rate variability (HRV) has been associated with freezing of gait, its relationship with broader gait parameters remains unclear. The objective was to investigate correlations between resting-state HRV time-domain measures and spatiotemporal gait parameters during comfortable and fast walking in patients with idiopathic PD. Twenty-eight PD patients (mean age 68 ± 9 years) were evaluated at Campus Bio-Medico University Hospital. HRV was recorded at rest using the e-Sense pule™ portable sensor, including the Baevsky’s Stress Index a measure increasing with sympathetic burden. Gait parameters were assessed via the 10 m Timed Up and Go (TUG) test using the Mon4t™ smartphone app at comfortable and fast pace. Clinical data included UPDRS III, MoCA, and disease characteristics. Gait metrics significantly changed between walking conditions. HRV parameters clustered separately from gait metrics but intersected with significant correlations. Higher Stress Index values, reflecting sympathetic dominance, were associated with poorer gait performance, including prolonged transition times, shorter steps, and increased variability (p < 0.001, r = 0.57–0.61). MoCA scores inversely correlated with the Stress Index (r = −0.52, p = 0.004), linking cognitive and autonomic status. UPDRS III and MoCA were related to TUG metrics but not HRV. Time-domain HRV measures, particularly the Stress Index, are significantly associated with spatiotemporal gait features in PD, independent of gait speed. These findings suggest that impaired autonomic regulation contributes to functional mobility deficits in PD and supports the role of HRV as a biomarker in motor assessment. Full article

Droughts significantly impact agriculture, water resources, and ecosystems. Their timely detection is essential for implementing effective mitigation strategies. This study explores the use of multispectral Sentinel-2 remote sensing indices and machine learning techniques to detect drought conditions in three distinct regions of India, such as Jodhpur, Amravati, and Thanjavur, during the Rabi season (October–April). Twelve remote sensing indices were studied to assess different aspects of vegetation health, soil moisture, and water stress, and their possible joint use and influence as indicators of regional drought events. Reference data used to define drought conditions in each region were primarily sourced from official government drought declarations and regional and national news publications, which provide seasonal maps of drought conditions across the country. Based on this information, a district vs. year (3 × 10) ground truth is created, indicating the presence or absence of drought (Drought/No Drought) for each region across the ten-year period. Using this ground truth table, we extended the remote sensing dataset by adding a binary drought label for each observation: 1 for “Drought” and 0 for “No Drought”. The dataset is organized by year (2016–2025) in a two-dimensional format, with indices as columns and observations as rows. Each observation represents a single measurement of the remote sensing indices. This enriched dataset serves as the foundation for training and evaluating machine learning models aimed at classifying drought conditions based on spectral information. The resultant remote sensing dataset was used to predict drought events through various machine learning models, including Random Forest, XGBoost, Bagging Classifier, and Gradient Boosting. Among the models, XGBoost achieved the highest accuracy (84.80%), followed closely by the Bagging Classifier (83.98%) and Random Forest (82.98%). In terms of precision, Bagging Classifier and Random Forest performed comparably (82.31% and 81.45%, respectively), while XGBoost achieved a precision of 81.28%. We applied a seasonal majority voting strategy, assigning a final drought label for each region and Rabi season based on the majority of predicted monthly labels. Using this method, XGBoost and Bagging Classifier achieved $96.67\%$ accuracy, precision, and recall, while Random Forest and Gradient Boosting reached $90\%$ and $83.33\%$, respectively, across all metrics. Shapley Additive Explanation (SHAP) analysis revealed that Normalized Multi-band Drought Index (NMDI) and Day of Season (DOS) consistently emerged as the most influential features in determining model predictions. This finding is supported by the Borda Count and Weighted Sum analysis, which ranked NMDI, and DOS as the top feature across all models. Additionally, Red-edge Chlorophyll Index (RECI), Normalized Difference Water Index (NDWI), Normalized Difference Moisture Index (NDMI), and Ratio Drought Index (RDI) were identified as important features contributing to model performance. These features help reveal the underlying spatiotemporal dynamics of drought indicators, offering interpretable insights into model decisions. To evaluate the impact of feature selection, we further conducted a feature ablation study. We trained each model using different combinations of top features: Top 1, Top 2, Top 3, Top 4, and Top 5. The performance of each model was assessed based on accuracy, precision, and recall. XGBoost demonstrated the best overall performance, especially when using the Top 5 features. Full article

In light of the rapid expansion of maritime trade, the maritime transportation industry has experienced burgeoning growth and complexity. The deployment of trajectory prediction technology is paramount in safeguarding navigational safety. Due to limitations in design complexity and the high costs of data fusion, current deep learning methods struggle to effectively integrate high-level semantic cues, such as vessel type, geographical identifiers, and navigational states, within predictive frameworks. Yet, these data contain abundant information regarding vessel categories or operational scenarios. Inspired by the robust semantic comprehension exhibited by large language models (LLMs) in natural language processing, this study introduces a trajectory prediction method leveraging LLMs. Initially, Automatic Identification System (AIS) data undergoes processing to eliminate incomplete entries, thereby selecting trajectories of high quality. Distinct from prior research that concentrated solely on vessel position and velocity, this study integrates ship identity, spatiotemporal trajectory, and navigational information through prompt engineering, empowering the LLM to extract multidimensional semantic features of trajectories from comprehensive natural language narratives. Thus, the LLM can amalgamate multi-source semantics with zero marginal cost, significantly enhancing its understanding of complex maritime environments. Subsequently, a supervised fine-tuning approach rooted in Low-Rank Adaptation (LoRA) is applied to train the chosen LLMs. This enables rapid adaptation of the LLM to specific maritime areas or vessel classifications by modifying only a limited subset of parameters, thereby appreciably diminishing both data requirements and computational costs. Finally, representative metrics are utilized to evaluate the efficacy of the model training and to benchmark its performance against prevailing advanced models for ship trajectory prediction. The results indicate that the model demonstrates notable performance in short-term predictions fFor instance, with a prediction step of 1 h, the average distance errors for VTLLM and TrAISformer are 5.26 nmi and 6.12 nmi, respectively, resulting in a performance improvement of approximately 14.05%), having identified certain patterns and features, such as linear movements and turns, from the training data. Full article

Accurate sub-seasonal-to-seasonal (S2S) forecasts are critical for mitigating extreme weather impacts and supporting development in West Africa. This study evaluates real-time ECMWF S2S rainfall forecasts during the 2020–2021 West African monsoon (March–October) and uses corresponding hindcasts for comparison. We verify forecasts at 1–4 dekads lead against two satellite-based rainfall datasets (TAMSAT and GPM-IMERG) to cover observational uncertainty. The analysis focuses on spatio-temporal monsoon patterns over the Gulf of Guinea (GoG) and Sahel (SAH). The results show that ECMWF-S2S captures key monsoon features. The forecast skill is generally higher over the Sahel than the GoG, and peaks during the main monsoon period (July–August). Notably, forecasts achieve approximately 80% synchronization with observed rainfall-anomaly timing, indicating that roughly 4 out of 5 dekads have correctly predicted wet/dry phases. Probabilistic evaluation shows strong reliability. The debiased ranked probability skill score (RPSS) is high across thresholds, whereas the average ROC AUC (~0.68) indicates moderate discrimination. However, forecasts tend to under-predict very low rains in the GoG and very high rains in the Sahel. Using multiple datasets and robust metrics helps mitigate observational uncertainty. These results, for the first real-time S2S pilot over West Africa, demonstrate that ECMWF rainfall forecasts are skillful and actionable (especially up to 2–3 dekads ahead), providing confidence for early-warning and planning systems in the region. Full article

Underwater acoustic communication faces significant challenges including limited bandwidth, high propagation delays, severe multipath fading, and stringent energy constraints. While integrated sensing and communication (ISAC) has shown promise in radio frequency systems, its adaptation to underwater environments remains challenging due to the unique acoustic channel characteristics and the inadequacy of traditional delay-based performance metrics that fail to capture the spatio-temporal value of information in dynamic underwater scenarios. This paper presents a comprehensive underwater ISAC framework centered on a novel Spatio-Temporal Information-Theoretic Freshness metric that fundamentally transforms resource allocation from delay minimization to value maximization. Unlike conventional approaches that treat all data equally, our spatio-temporal framework enables intelligent prioritization by recognizing that obstacle detection data directly ahead of an autonomous underwater vehicle (AUV) require immediate processing. Our framework addresses key underwater ISAC challenges through spatio-temporal-guided power allocation, adaptive beamforming, waveform optimization, and cooperative sensing strategies. Multi-agent reinforcement learning algorithms enable coordinated resource allocation and mission-critical information prioritization across heterogeneous networks comprising surface buoys, AUVs, and static sensors. Extensive simulations in realistic Munk profile acoustic environments demonstrate significant performance improvements. The spatio-temporal framework successfully filters spatially irrelevant data, resulting in substantial energy savings for battery-constrained underwater nodes. Full article

Accurate recognition of power grid operation behaviors is crucial for ensuring both safety and operational efficiency in smart grid systems. However, this task presents significant challenges due to dynamic environmental variations, limited labeled training data availability, and the necessity for continuous model adaptation. To overcome these limitations, we propose an Incremental Spatio-temporal Augmented Sampling (ISAS) method for power grid operation behavior recognition. Specifically, we design a spatio-temporal Feature-Enhancement Fusion Module (FEFM) which employs multi-scale spatio-temporal augmented fusion combined with a cross-scale aggregation mechanism, enabling robust feature learning that is resilient to environmental interference. Furthermore, we introduce a Selective Replay Mechanism (SRM) that implements a dual-criteria sample selection strategy based on error variability and feature-space divergence metrics, ensuring optimal memory bank updates that simultaneously maximize information gain while minimizing feature redundancy. Experimental results on the power grid behavior dataset demonstrate significant advantages of the proposed method in recognition robustness and knowledge retention compared to other methods. For example, it achieves an accuracy of 89.80% on sunny days and maintains exceptional continual learning stability with merely 2.74% forgetting rate on three meteorological scenarios. Full article

Accurate sea clutter modeling is crucial for clutter suppression in edge radar processing. On resource-constrained edge radar platforms, spatiotemporal statistics, together with device-level computation and memory limits, hinder the learning of representative clutter features. This study presents a transformer-based generative adversarial model for sea clutter modeling. The core design of this work uses axial attention to factorize self-attention along pulse and range, preserving long-range dependencies under a reduced attention cost. It also introduces a two-dimensional variable-length spatiotemporal window that retains temporal and spatial coherence across observation lengths. Extensive experiments are conducted to verify the efficacy of the proposed method with quantitative criteria, including a cosine similarity score, spectral-parameter error, and amplitude–distribution distances. Compared with CNN-based GAN, the proposed model achieves a high consistency with real clutter in marginal amplitude distributions, spectral characteristics, and spatiotemporal correlation patterns, while incurring a lower cost than standard multi-head self-attention. The experimental results show that the proposed method achieves improvements of 9.22% and 7.8% over the traditional AR and WaveGAN methods in terms of the similarity metric, respectively. Full article