Search Results (8,862)

Real-time Unmanned Aerial Vehicle (UAV) detection is a growing research field centered on advanced computer vision and deep learning algorithms. However, the rise of unmanned aerial vehicles (UAVs) has sparked numerous concerns due to their potential for malicious use in illegal activities. To address these concerns, Vision-based object detection approaches for UAVs have recently been developed. Nonetheless, UAV detection in real-world scenarios, such as images with diverse backgrounds and various perspectives, remains underexplored. To fill this gap, we present a new UAV detection dataset called the real-world scenarios dataset (RWSD). This dataset leverages real-world footage and is constructed under challenging conditions, including complex backgrounds, varying UAV sizes, different perspectives, and multiple UAV types. It aims to support the development of robust UAV detection algorithms that can perform well in diverse and realistic conditions. YOLO, a popular one-stage object detection approach, is widely employed for UAV detection across different environments due to its efficiency and simplicity. However, this series of detectors encounters challenges in real-world scenarios, such as excessive computation and suboptimal detection rates. In this study, we propose a lightweight feature enhancement model (LFEM) to address these limitations. Specifically, we base our model on YOLOv5, introducing the Ghost module to improve UAV detection with fewer floating-point operations (FLOPs). Additionally, we incorporate the SIMAM module to enhance feature representation, particularly for real-world scenarios. Extensive experiments on the RWSD, UAVDT, and DOTAv1.0 datasets demonstrate the effectiveness of our approach. Our proposed LFEM achieves an impressive 93.2% mAP₅₀, outperforming baseline models while maintaining a lightweight profile. Comparative and ablation studies further confirm that our algorithm is a promising and efficient solution for practical UAV detection tasks. Full article

Attention-Deficit/Hyperactivity Disorder (ADHD) is a highly prevalent neurodevelopmental condition that is typically identified through behavioral assessments and subjective clinical reports. However, electroencephalography (EEG) offers a cost-effective and non-invasive alternative for capturing neural activity patterns closely associated with this disorder. Despite this potential, EEG-based ADHD classification remains challenged by overfitting, dependence on extensive preprocessing, and limited interpretability. Here, we propose a novel neural architecture that integrates transformer-based temporal attention with Gaussian mixture functional connectivity modeling and a cross-entropy loss regularized through $\alpha$-Rényi mutual information, termed T-GARNet. The multi-scale Gaussian kernel functional connectivity leverages parallel Gaussian kernels to identify complex spatial dependencies, which are further stabilized and regularized by the $\alpha$-Rényi term. This design enables direct modeling of long-range temporal dependencies from raw EEG while enhancing spatial interpretability and reducing feature redundancy. We evaluate T-GARNet on a publicly available ADHD EEG dataset using both leave-one-subject-out (LOSO) and stratified group k-fold cross-validation (SGKF-CV), where groups correspond to control and ADHD, and compare its performance against classical and modern state-of-the-art methods. Results show that T-GARNet achieves competitive or superior performance (82.10% accuracy), particularly under the more challenging SGKF-CV setting, while producing interpretable spatial attention patterns consistent with ADHD-related neurophysiological findings. These results underscore T-GARNet’s potential as a robust and explainable framework for objective EEG-based ADHD detection. Full article

Accurately and efficiently predicting bottomhole pressure (BHP) is of great importance for safe drilling in complex formations. Many researchers have conducted extensive investigations into intelligent BHP prediction techniques. However, the current intelligent models mostly focus on the data-driven relationship between logging parameters and BHP, and less on the influence of the correlation between the logging parameters on the BHP. This paper proposes a real-time prediction framework based on graph neural networks. Our model selects input features based on drilling mechanisms and statistical analyses, and utilizes adaptive learning of the graph based on multivariate time-series parameters to capture the relationship between multivariate logging parameters and BHP. Finally, the model performance is thoroughly analyzed based on field drilling datasets after optimizing model hyperparameters using the Bayesian optimization method. Results indicate that the proposed method performs better in terms of prediction accuracy, captures the inflection points of curve changes better, and is more robust under the new well section. The mean absolute percentage error of the method reaches 1.28% which is reduced by 25% compared with other traditional intelligent models. This study provides a solution for achieving accurate real-time predictions of bottom hole pressure, establishing a solid foundation for the realization of precise pressure control during drilling operations. Full article

Anomaly detection is a critical task in video surveillance, with significant applications in the management and prevention of criminal activities. Traditional convolutional neural networks often struggle with motion modeling and multi-scale feature fusion due to their localized field of view. To address these limitations, this work proposes a Dual-Branch Interactive Feature Fusion Network (DBIFF-Net). DBIFF-Net integrates a CNN branch and a swin transformer branch to extract multi-scale features. To optimize these features for efficient fusion, an interactive fusion module is introduced to efficiently fuse these multi-scale features through skip connections. Then, the temporal shift module is employed to exploit dependencies between video frames, thereby improving the identification of anomalous events. Finally, the channel attention is utilized for decoder to better assist in restoring complex object features in the video. System performance is evaluated on three standard benchmark datasets. DBIFF-Net achieves the area under the receiver operating characteristic (AUC) of 97.7%, 84.5%, and 73.8% on the UCSD ped2, CUHK Avenue, and ShanghaiTech Campus dataset, respectively. Extensive experiments demonstrate that DBIFF-Net outperforms most state-of-the-art methods, validating the effectiveness of our method. Full article

Accurate analysis of trace elements in particulate matter (PM) emitted by brake systems critically depends on the filter selection and handling processes, which can significantly impact analytical results due to contamination and elemental interference from filter elemental composition. This study systematically evaluated two widely used filter types, EMFAB (borosilicate glass microfiber reinforced with PTFE) and Teflon (PTFE), for their suitability in the trace element determination of brake-wear PM₁₀ collected using a tribometer set-up. A total of twenty-three PM₁₀ samples were analyzed, encompassing two different friction materials, to thoroughly assess the performance and analytical implications of each filter type. Filters were tested for their chemical background, handling practicality and potential contamination risk through extensive elemental analysis by inductively coupled plasma–optical emission spectrometry (ICP-OES) and inductively coupled plasma-mass spectrometry (ICP-MS). Additionally, morphological characterization of both filter types was conducted via scanning electron microscopy (SEM) coupled with energy-dispersive X-ray spectroscopy (EDS) to elucidate structural features affecting particle capture and subsequent analytical performance. Significant differences emerged between the two filters regarding elemental interferences: EMFAB filters exhibited substantial background contribution, particularly for alkali and alkaline earth metals (Ca, Na, Mg and K), complicating accurate quantification at trace levels. Conversely, Teflon filters demonstrated considerably lower background but required careful manipulation due to their structural fragility and the necessity to remove supporting rings, potentially introducing analytical variability. Statistical analysis confirmed that the filter material significantly affects elemental quantification, particularly when the collected PM₁₀ mass is limited, highlighting the importance of careful filter selection and handling procedures. Recommendations for optimal analytical practices are provided to minimize contamination risks and enhance reliability in trace element analysis of PM₁₀ emissions. These findings contribute to refining analytical methodologies essential for accurate environmental monitoring and regulatory assessments of vehicular non-exhaust emissions. Full article

Model-Based Systems Engineering (MBSE) is employing systems analysis, design, and development on models of these systems, bringing together different viewpoints, with a step-by-step increase of detail. As such, it replaces traditional document-centric approaches with a methodology that uses structured domain models for information exchange and system representation throughout the engineering lifecycle. MBSE comprises different languages, each with distinct features and approaches. SysML is a widely used language in MBSE, and many tools exist for it. This paper is interested in the complexity of SysML models, as it may yield useful quantitative indicators to assess and predict the complexity of systems modeled in SysML, and, by extension, the complexity of their subsequent development. Two avenues are explored: objective structural metrics applied to the SysML model and assessment of the team experience. The proposed approach is implemented as a Java prototype. Although simpler models are easier to comprehend and modify, they may fail to capture the full scope of system functionality. Conversely, more complex models, though richer in detail, require greater development effort and pose challenges for maintenance and stakeholder communication. Technical and environmental factors are integrated into the complexity assessment to reflect real-world project conditions. A drone-based image acquisition system serves as a case study. Full article

System time delays and nonlinear unmodeled dynamics severely constrain the control performance of the Active Suspension Electro-Hydraulic Servo Actuator (ASEHSA). To tackle these challenges, this paper presents a Dynamic Error Differentiation-based Model-Free Adaptive Control (DE-MFAC) strategy integrated with an Improved Particle Swarm Optimization (IPSO) algorithm. Established under the Model-Free Adaptive Control (MFAC) framework, the DE-MFAC integrates a dynamic error differentiation mechanism and an implicit expression of time delays, thus removing the dependence on a precise system model. The traditional PSO algorithm is improved by incorporating an inertia weight adjustment strategy and a boundary reflection wall strategy, which effectively mitigates the issues of local optima and boundary stagnation. In AMESim 2021, a 1/4 vehicle active suspension electro-hydraulic actuation system model is constructed. To ensure an impartial evaluation of controller performance, the IPSO algorithm is employed to optimize the parameters of the PID, MFAC, and DE-MFAC controllers, respectively. Co-simulations with Simulink 2023b are conducted under two time delay scenarios using a composite square-sine wave signal as the reference. The results indicate that all three IPSO-optimized controllers realize effective position tracking. Among them, the DE-MFAC controller exhibits the optimal performance, demonstrating remarkable advantages in reducing tracking errors and balancing settling time with overshoot. These findings verify the effectiveness of the proposed control strategy, time delay compensation mechanism, and optimization algorithm. Future research will involve validation on a physical ASEHSA platform, further exploration of the method’s applicability and robustness under diverse operating conditions, and extension to other industrial systems with similar nonlinear time delay features. Full article

Infrared and visible image fusion-based object detection is critical for robust environmental perception under adverse conditions, yet existing methods still suffer from insufficient modeling of modality discrepancies and limited adaptivity in their fusion mechanisms. This work proposes an asymmetric spatial–frequency fusion network, AsyFusionNet. The network adopts an asymmetric dual-branch backbone that extends the RGB branch to P5 while truncating the infrared branch at P4, thereby better aligning with the physical characteristics of the two modalities, enhancing feature complementarity, and enabling fine-grained modeling of modality differences. On top of this backbone, a local–global attention fusion (LGAF) module is introduced to model local and global attention in parallel and reorganize them through lightweight convolutions, achieving joint spatial–channel selective enhancement. Modality-specific feature enhancement is further realized via a hierarchical attention module (HAM) in the RGB branch, which employs dynamic kernel selection to emphasize multi-level texture details, and a fourier spatial spectral modulation (FS2M) module in the infrared branch, which more effectively captures global thermal radiation patterns. Extensive experiments on the $M^3$FD and VEDAI datasets demonstrate that AsyFusionNet attains $86.3\%$ and $54.1\%$${\mathrm{mAP}}_{50}$, respectively, surpassing the baseline by $8.8$ and $6.4$ points (approximately $11.4\%$ and $13.4\%$ relative gains) while maintaining real-time inference speed. Full article

The Colombian Orinoquia was shaped within a tectonic and sedimentary framework linked to the uplift of the Andean cordilleras during the Oligocene–Miocene. This orogenic event generated two tectonic fronts and facilitated extensive fluvial sedimentation across a broad alluvial geosyncline. The present geomorphological configuration reflects the cumulative interaction of tectonic and erosional processes with Quaternary climatic dynamics, which together produced complex landscape assemblages characterized by plains with distinctive drainage patterns. To delineate and characterize geomorphological units, we employed multidimensional imagery and Machine Learning techniques within the Google Earth Engine platform. The classification model integrated dual polarizations of synthetic aperture radar (L-band) with key topographic variables including elevation, slope, aspect, convexity, and roughness. The analysis identified three major physiographic units: (i) the Foothills and the Floodplain, both dominated by fluvial environments; (ii) the High plains and Serranía de La Macarena (Macarena Mountain Range), where denudational processes predominate; and (iii) localized aeolian environments embedded within the Floodplain. These contrasting dynamics have generated a broad spectrum of landforms, ranging from terraces and alluvial fans in the Foothills to hills and other erosional features in La Macarena. The Floodplain, developed over a sedimentary depression, illustrates the combined action of fluvial and aeolian processes, whereas the High plains is characterized by rolling plains and peneplains formed through the uplift and erosion of Tertiary sediments. Such geomorphic heterogeneity underscores the interplay between tectonic activity, climatic forcing, and surface processes in shaping the Orinoquia landscape. The geomorphological classification using Random Forest demonstrated high effectiveness in discriminating units at a regional scale, with accuracy levels supported by confusion matrices and associated Kappa indices. Nevertheless, some degree of classificatory overlap was observed in fluvial environments, likely reflecting their transitional nature and complex sedimentary dynamics. Overall, this methodological approach enhances the objectivity of geomorphological analysis and establishes a replicable framework for assessing landform distribution in tropical sedimentary basins. Full article

Vehicle trajectory prediction is a pivotal technology in intelligent transportation systems. Existing methods encounter challenges in effectively modeling lane topology and dynamic interaction relationships in complex traffic scenarios, limiting prediction accuracy and reliability. This paper presents Lane Interaction Transformer (LITransformer), a lane-informed trajectory prediction framework that builds on spatio–temporal graph attention networks and Transformer-based global aggregation. Rather than introducing entirely new network primitives, LITransformer focuses on two design aspects: (i) a lane topology encoder that fuses geometric and semantic lane features via direction-sensitive, multi-scale dilated graph convolutions, converting vectorized lane data into rich topology-aware representations; and (ii) an Interaction-Aware Graph Attention mechanism (IAGAT) that explicitly models four types of interactions between vehicles and lane infrastructure (V2V, V2N, N2V, N2N), with gating-based fusion of structured road constraints and dynamic spatio–temporal features. The overall architecture employs a Transformer module to aggregate global scene context and a multi-modal decoding head to generate diverse trajectory hypotheses with confidence estimation. Extensive experiments on the Argoverse dataset show that LITransformer achieves a minADE of 0.76 and a minFDE of 1.20, and significantly outperforms representative baselines such as LaneGCN and HiVT. These results demonstrate that explicitly incorporating lane topology and interaction-aware spatio-temporal modeling can significantly improve the accuracy and reliability of vehicle trajectory prediction in complex real-world traffic scenarios. Full article

Wind power forecasting is critical to grid stability and renewable energy integration. However, existing deep learning methods struggle to incorporate semantic domain knowledge from textual information, exhibit limited generalization with scarce training data, and require high computational costs for extensive fine-tuning. Large language models (LLMs) offer a promising solution through their semantic representations, few-shot learning capabilities, and multimodal processing abilities. This paper proposes LLM-WPFNet, a dual-modality fusion framework that integrates frozen pre-trained LLMs with time-series analysis for wind power forecasting. The key insight is encoding temporal patterns as structured textual prompts to enable semantic guidance from frozen LLMs without fine-tuning. LLM-WPFNet employs two parallel encoding branches to extract complementary features from time series and textual prompts, unified through asymmetric multi-head attention fusion that enables selective semantic knowledge transfer from frozen LLM embeddings to enhance temporal representations. By maintaining the LLM frozen, our method achieves computational efficiency while leveraging robust semantic representations. Extensive experiments on four wind farm datasets (36–200 MW) across five prediction horizons (1–24 h) demonstrate that LLM-WPFNet consistently outperforms state-of-the-art baselines by 11% in MAE and RMSE. Notably, with only 10% of training data, it achieves a 17.6% improvement over the best baseline, validating its effectiveness in both standard and data-scarce scenarios. These results highlight the effectiveness and robustness of the dual-modality fusion design in predicting wind power under complex real-world conditions. Full article

Existing salient object detection methods for optical remote sensing images still face certain limitations due to complex background variations, significant scale discrepancies among targets, severe background interference, and diverse topological structures. On the one hand, the feature transmission process often neglects the constraints and complementary effects of high-level features on low-level features, leading to insufficient feature interaction and weakened model representation. On the other hand, decoder architectures generally rely on simple cascaded structures, which fail to adequately exploit and utilize contextual information. To address these challenges, this study proposes a Hierarchical Semantic Interaction Module to enhance salient object detection performance in optical remote sensing scenarios. The module introduces foreground content modeling and a hierarchical semantic interaction mechanism within a multi-scale feature space, reinforcing the synergy and complementarity among features at different levels. This effectively highlights multi-scale and multi-type salient regions in complex backgrounds. Extensive experiments on multiple optical remote sensing datasets demonstrate the effectiveness of the proposed method. Specifically, on the EORSSD dataset, our full model integrating both CA and PA modules improves the max F-measure from 0.8826 to 0.9100 (↑2.74%), increases maxE from 0.9603 to 0.9727 (↑1.24%), and enhances the S-measure from 0.9026 to 0.9295 (↑2.69%) compared with the baseline. These results clearly demonstrate the effectiveness of the proposed modules and verify the robustness and strong generalization capability of our method in complex remote sensing scenarios. Full article

An accurate and reliable extraction of building structures from high-resolution (HR) remote sensing images is an important research topic in 3D cartography and smart city construction. However, despite the strong overall performance of recent deep learning models, limitations remain in handling significant variations in building scales and complex architectural forms, which may lead to inaccurate boundaries or difficulties in extracting small or irregular structures. Therefore, the present study proposes MSA-UNet, a reliable semantic segmentation framework that leverages multiscale feature aggregation and attentive skip connections for an accurate extraction of building footprints. This framework is constructed based on the U-Net architecture, incorporating VGG16 as a replacement for the original encoder structure, which enhances its ability to capture low-discriminative features. To further improve the representation of image buildings with different scales and shapes, a serial coarse-to-fine feature aggregation mechanism was used. Additionally, a novel skip connection was built between the encoder and decoder layers to enable adaptive weights. Furthermore, a dual-attention mechanism, implemented through the convolutional block attention module, was integrated to enhance the focus of the network on building regions. Extensive experiments conducted on the WHU and Inria building datasets validated the effectiveness of MSA-UNet. On the WHU dataset, the model demonstrated a state-of-the-art performance with a mean Intersection over Union (mIoU) of 94.26%, accuracy of 98.32%, F1-score of 96.57%, and mean Pixel accuracy (mPA) of 96.85%, corresponding to gains of 1.41% in mIoU over the baseline U-Net. On the more challenging Inria dataset, MSA-UNet achieved an mIoU of 85.92%, indicating a consistent improvement of up to 1.9% over the baseline U-Net. These results confirmed that MSA-UNet markedly improved the accuracy and boundary integrity of building extraction from HR data, outperforming existing classic models in terms of segmentation quality and robustness. Full article

Water and mud inrush hazards pose significant threats to the safety, environmental stability, and resource efficiency of tunnel construction, representing a critical barrier to the development of sustainable transportation infrastructure. Misjudgment—especially missed detections of severe hazards—can lead to extensive geological disturbance, excessive energy consumption, and severe socio-environmental impacts. However, pre-trained large-scale models still face two major challenges when applied to tunnel hazard classification: limited labeled samples and the high cost associated with misclassifying severe hazards. This study proposes a sustainability-oriented intelligent classification framework that integrates a large-scale pre-trained model with multi-strategy data augmentation to accurately identify hazard levels during tunnel excavation. First, a Synthetic Minority Over-Sampling Technique (SMOTE)-based multi-strategy augmentation method is introduced to expand the training set, mitigate class imbalance, and enhance the model’s ability to recognize rare but critical hazard categories. Second, a deep feature extraction architecture built on the robustly optimized BERT pretraining approach (RoBERTa) is designed to strengthen semantic representation under small-sample conditions. Moreover, a hierarchical weighting mechanism is incorporated into the weighted cross-entropy loss to emphasize the identification of severe hazard levels, thereby ensuring zero missed detections. Experimental results demonstrate that the proposed method achieves an accuracy of 99.26%, representing a 27.96% improvement over the traditional SVM baseline. Importantly, the recall for severe hazards (Levels III and IV) reaches 100%, ensuring zero misjudgment of major hazards. By effectively reducing safety risks, minimizing environmental disruptions, and promoting resilient tunnel construction, this method provides strong support for sustainable and low-impact underground engineering practices. Full article

The increasing prevalence of deep learning models in industry has highlighted the critical need to protect the intellectual property (IP) of these models, especially generative adversarial networks (GANs) capable of synthesizing realistic data. Traditional IP protection methods, such as watermarking model parameters (white-box) or verifying outputs (black-box), are insufficient against non-public misappropriation. To address these limitations, we introduce EGAN (Encrypted GANs), which secures GAN models by embedding a novel self-adversarial mechanism. This mechanism is trained to actively maximize the feature divergence between authorized and unauthorized inputs, thereby intentionally corrupting the outputs from non-key inputs and preventing unauthorized operation. Our methodology utilizes key-based transformations applied to GAN inputs and incorporates a generator loss regularization term to enforce model protection without compromising performance. This technique is compatible with existing watermark-based verification methods. Extensive experimental evaluations reveal that EGAN maintains the generative capabilities of original GAN architectures, including DCGAN, SRGAN, and CycleGAN, while exhibiting robust resistance to common attack strategies such as fine-tuning. Compared with prior work, EGAN provides comprehensive IP protection by ensuring unauthorized users cannot achieve desired outcomes, thus safeguarding both the models and their generated data. Full article