Search Results (51)

Traditional median filtering with a fixed window easily leads to edge blurring and adaptive median filtering requires manual presetting of the maximum window parameter and has insufficient retention of details when dealing with high-density salt-and-pepper noise. Aiming at these problems, this paper proposes a hybrid decision-making adaptive median filtering algorithm with dual-window detection in collaboration with particle swarm optimization (PSO). The algorithm quickly locates suspected noise points through a 3 × 3 small window and enhances noise identification accuracy by using a PSO dynamically optimized 5–35-pixel large window. Meanwhile, a hybrid decision-making mechanism based on local statistical properties was introduced to dynamically select median filtering, weighted average based on spatial distance, or pixel preservation strategy to balance noise suppression and detail preservation, and the PSO algorithm was used to automatically find the optimal parameters of the large window’s size to avoid the manual parameter-tuning process. Experiments were conducted on standard grayscale and color images and compared with four traditional methods and two more advanced methods. The experiments showed that the algorithm improved the peak signal-to-noise ratio (PSNR) value by 2–4 dB and the structural similarity index measure (SSIM) metric by 0.05–0.2 under high salt-and-pepper noise density compared with the traditional methods, which effectively improved the contradiction between noise suppression and detail retention in traditional filtering algorithms and provided a highly efficient and intelligent solution for image denoising in high-noise scenarios. Full article

The electroencephalogram (EEG), widely used for measuring the brain’s electrophysiological activity, has been extensively applied in the automatic detection of epileptic seizures. However, several challenges remain unaddressed in prior studies on automated seizure detection: (1) Methods based on CNN and LSTM assume that EEG signals follow a Euclidean structure; (2) Algorithms leveraging graph convolutional networks rely on adjacency matrices constructed with fixed edge weights or predefined connection rules. To address these limitations, we propose a novel algorithm: Dynamic Graph Convolutional Network with Dilated Convolution (DGDCN). By leveraging a spatiotemporal attention mechanism, the proposed model dynamically constructs a task-specific adjacency matrix, which guides the graph convolutional network (GCN) in capturing localized spatial and temporal dependencies among adjacent nodes. Furthermore, a dilated convolutional module is incorporated to expand the receptive field, thereby enabling the model to capture long-range temporal dependencies more effectively. The proposed seizure detection system is evaluated on the TUSZ dataset, achieving AUC values of 88.7% and 90.4% on 12-s and 60-s segments, respectively, demonstrating competitive performance compared to current state-of-the-art methods. Full article

The Gaining–Sharing Knowledge-based (GSK) algorithm is a human-inspired metaheuristic that models how people learn and disseminate knowledge across their lifetime. It has shown promising results across a range of engineering optimization problems. However, one of its major limitations lies in the use of fixed parameters to guide the search process, which often causes the algorithm to get stuck in local optima. To address this challenge, we propose an Auto-Tuning Memory-based Adaptive Local Search (ATMALS) empowered GSK, that is, ATMALS-GSK. This enhanced version of GSK introduces two key improvements: adaptive local search and memory-driven automatic tuning of parameters. Rather than relying on fixed values, ATMALS-GSK continuously adjusts its parameters during the optimization process. This is achieved through a Gaussian distribution mechanism that iteratively updates the likelihood of selecting different parameter values based on their historical impact on the fitness function. This selection process is guided by a weighted moving average that tracks each parameter’s contribution to fitness improvement over time. To further reduce the risk of premature convergence, an adaptive local search strategy is embedded, facilitating the algorithm’s escape from local traps and guiding it toward more optimal regions within the search domain. To validate the effectiveness of the ATMALS-GSK algorithm, it is evaluated on the CEC 2011 and CEC 2017 benchmarks. The results indicate that the ATMALS-GSK algorithm outperforms the original GSK, its variants, and other metaheuristics by delivering greater robustness, quicker convergence, and superior solution quality. Full article

Due to the challenges posed by high resolution, substantial background noise, significant object scale variation, and long-tailed data distribution in remote sensing images, traditional techniques often struggle to maintain both high accuracy and low latency. This paper proposes YOLO11-FSDAT, an advanced object detection framework tailored for remote sensing imagery, which integrates not only modular enhancements but also theoretical and architectural innovations to address these limitations. First, we propose the frequency–spatial feature extraction fusion module (Freq-SpaFEFM), which breaks the conventional paradigm of spatial-domain-dominated feature learning by introducing a multi-branch architecture that fuses frequency- and spatial-domain features in parallel. This design provides a new processing paradigm for multi-scale object detection, particularly enhancing the model’s capability in handling dense and small-object scenarios with complex backgrounds. Second, we introduce the deformable attention-based global–local fusion module (DAGLF), which combines fine-grained local features with global context through deformable attention and residual connections. This enables the model to adaptively capture irregularly oriented objects (e.g., tilted aircraft) and effectively mitigates the issue of information dilution in deep networks. Third, we develop the adaptive threshold focal loss (ATFL), which is the first loss function to systematically address the long-tailed distribution in remote sensing datasets by dynamically adjusting focus based on sample difficulty. Unlike traditional focal loss with fixed hyperparameters, ATFL decouples hard and easy samples and automatically adapts to varying class distributions. Experimental results on the public DOTAv1, SIMD, and DIOR datasets demonstrated that YOLO11-FSDAT achieved 75.22%, 82.79%, and 88.01% mAP, respectively, outperforming baseline YOLOv11n by up to 4.11%. These results confirm the effectiveness, robustness, and broader theoretical value of the proposed framework in addressing key challenges in remote sensing object detection. Full article

This paper studies the problem of time-varying formation-tracking control for a class of nonlinear multi-agent systems. A distributed adaptive controller that avoids the global non-zero minimum eigenvalue is designed for heterogeneous systems in which leaders and followers contain different nonlinear terms, and which relies only on the relative errors between adjacent agents. By adopting the Riccati inequality method, the adaptive adjustment factor in the controller is designed to solve the problem of automatically adjusting relative errors based solely on local information. Unlike existing research on time-varying formations with fixed and switching topologies, the method of jointly connected topological graphs is adopted to enable nonlinear followers to track the trajectories of leaders with different nonlinear terms and simultaneously achieve the control objective of the desired time-varying formation. The stability of the system under the jointly connected graph is proved by the Lyapunov stability proof method. Finally, numerical simulation experiments confirm the effectiveness of the proposed control method. Full article

The Global Navigation Satellite System (GNSS) is widely used in various applications of UAVs (unmanned aerial vehicles) that require precise positioning or navigation. However, GNSS signals can be blocked in specific environments and are susceptible to jamming and spoofing, which will degrade the performance of navigation systems. In this study, a deep learning-based navigation system for the automatic landing of fixed-wing UAVs in GNSS-denied environments is proposed to serve as an alternative navigation system. Most visual-based runway landing systems are typically focused on runway detection and localization while neglecting the issue of integrating the localization solution into flight control and guidance laws to become a complete real-time automatic landing system. This study addresses these problems by combining runway detection and localization methods, YOLOv8 and CNN (convolutional neural network) regression, to demonstrate the robustness of deep learning approaches. Moreover, a line detection method is employed to accurately align the UAV with the runway, effectively resolving issues related to runway contours. In the control phase, the guidance law and controller are designed to ensure the stable flight of the UAV. Based on a deep learning model framework, this study conducts experiments within the simulation environment, verifying system stability under various assumed conditions, thereby avoiding the risks associated with real-world testing. The simulation results demonstrate that the UAV can achieve automatic landing on 3-degree and 5-degree glide slopes, whether it is directly aligned with the runway or deviating from it, with trajectory tracking errors within 10 m. Full article

Federated learning has been widely applied in automatic speech recognition. However, variations in speaker behaviors result in a significant data imbalance across client devices. Conventional federated speech recognition algorithms typically use fixed probabilities to select clients for each round in model training, often overlooking the disparities in data volume among clients. In reality, the substantial differences in data quantity can extend the training duration and compromise the stability of the global model. Moreover, models trained through federated learning on global data often fail to achieve optimal performance for individual local clients. While personalized federated learning strategies hold promise for enhancing model performance, the inherent diversity of speech data makes it challenging to apply state-of-the-art personalized methods effectively to speech recognition tasks. In this paper, a dynamic client selection algorithm is proposed to solve the problem of data disparities among different clients. It can be effectively combined with most federated learning algorithms and dynamically adjusts the selection probabilities of clients based on their dataset size during training. Experimental results demonstrate that this algorithm saved training time by 26% compared to traditional methods on public datasets while maintaining the equivalent model performance. To optimize the personalized federated learning, this paper proposes a novel group-balanced personalization strategy that fine-tunes groupings of clients based on their dataset size. The experimental results show that this algorithm brought a relatively 12% reduction in character error rate, while it did not increase computational costs. In particular, the group-balanced personalization effectively improved the model performance for clients with smaller datasets than local fine-tuning. The combination of dynamic client selection and group-balanced personalization significantly enhanced training efficiency and model performance. Full article

Trend forecasting and early anomaly warnings are important for avoiding aircraft engine failures or accidents. This study proposes a trend forecasting method based on enhanced Slice-level Adaptive Normalization (SAN) using a Long Short-Term Memory (LSTM) neural network under multi-operating conditions. Firstly, a condition recognition technology is constructed to automatically identify the operating conditions based on the predetermined judgment conditions, and vibration signal features are adaptively divided into three typical operating conditions, namely, the idling operating condition, the starting operating condition and the utmost operating condition. The features of original signals are extracted to reduce the impacts of signal fluctuations and noise preliminarily. Secondly, enhanced SAN is used to normalize and denormalize the features to alleviate non-stationary factors. To improve prediction accuracy, an L₁ filter is adopted to extract the trend term of the features, which can effectively reduce the overfitting of SAN to local information. Moreover, the slice length is quantitatively estimated by the fixed points in L₁ filtering, and a tail amendment technology is added to expand the applicable range of enhanced SAN. Finally, an LSTM-based forecasting model is constructed to forecast the normalized data from enhanced SAN, serving as input during denormalization. The final results under different operating conditions are the output from denormalization. The validity of the proposed method is verified using the test data of an aircraft engine. The results show that the proposed method can achieve higher forecasting accuracy compared to other methods. Full article

The current growing interest in lighter-than-air platforms (LTA) has been fueled by the significant development of some enabling technologies, in particular electric motors and on-board electronics. The localization of multiple thrust forces in the layout of the airship, as well as the ability to manage them through automatic control, promises to mitigate the controllability issues connatural to this type of flying craft. Employed on unmanned missions and close to the ground, LTA vehicles now appear to be a technically viable alternative to other unmanned aerial vehicles (UAVs) or low-flying manned machines and are similarly capable of effectively achieving the corresponding mission goals. A key step in establishing the credibility of LTA vehicles as industrial solutions for an end user is an assessment of the economic effort required for producing and operating them. This study presents an analytic approach for evaluating these costs, based on the data available at a preliminary design level for an airship. Three missions currently flown by other types of flying machines were considered, and for each mission the sizing and preliminary design of a LTA platform capable of providing the same mission performance was carried out. Correspondingly, a newly introduced method for the estimation of the cost of a LTA platform was applied. Also, an estimation of the costs currently sustained by operators for each mission was obtained from the available data and with the support of relevant companies, who currently do not fly LTA platforms but operate with more standard flying machines (in particular, multicopter or fixed-wing UAVs or manned helicopters). Finally, the costs corresponding to both currently flying non-LTA vehicles and suitably designed LTA solutions were compared, yielding indications of the emerging economic trade-offs. Full article

For reservoir structural models with obvious nonstationary and heterogeneous characteristics, traditional geostatistical simulation methods tend to produce suboptimal results. Additionally, these methods are computationally resource-intensive in consecutive simulation processes. Thanks to the feature extraction capability of deep learning, the generative adversarial network-based method can overcome the limitations of geostatistical simulation and effectively portray the structural attributes of the reservoir models. However, the fixed receptive fields may restrict the extraction of local geospatial multiscale features, while the gradient anomalies and mode collapse during the training process can cause poor reconstruction. Moreover, the sparsely distributed conditioning data lead to possible noise and artifacts in the simulation results due to its weak constraint ability. Therefore, this paper proposes an improved conditioning spectral normalization generation adversarial network framework (CSNGAN-ASPP) to achieve efficient and automatic reconstruction of reservoir geological bodies under sparse hard data constraints. Specifically, CSNGAN-ASPP features an encoder-decoder type generator with an atrous spatial pyramid pooling (ASPP) structure, which effectively identifies and extracts multi-scale geological features. A spectral normalization strategy is integrated into the discriminator to enhance the network stability. Attention mechanisms are incorporated to focus on the critical features. In addition, a joint loss function is defined to optimize the network parameters and thereby ensure the realism and accuracy of the simulation results. Three types of reservoir model were introduced to validate the reconstruction performance of CSNGAN-ASPP. The results show that they not only accurately conform to conditioning data constraints but also closely match the reference model in terms of spatial variance, channel connectivity, and facies attribute distribution. For the trained CSNGAN-ASPP, multiple corresponding simulation results can be obtained quickly through inputting conditioning data, thus achieving efficient and automatic reservoir geological model reconstruction. Full article

Background/Objectives: The accurate quantification of ground-glass opacities (GGOs) and consolidation volumes has prognostic value in COVID-19 patients. Nevertheless, the accurate manual quantification of the corresponding volumes remains a time-consuming task. Deep learning (DL) has demonstrated good performance in the segmentation of normal lung parenchyma and COVID-19 pneumonia. We introduce a Human-in-the-Loop (HITL) strategy for the segmentation of normal lung parenchyma and COVID-19 pneumonia that is both time efficient and quality effective. Furthermore, we propose a Gaussian Mixture Model (GMM) to classify GGO and consolidation based on a probabilistic characterization and case-sensitive thresholds. Methods: A total of 65 Computed Tomography (CT) scans from 64 patients, acquired between March 2020 and June 2021, were randomly selected. We pretrained a 3D-UNet with an international dataset and implemented a HITL strategy to refine the local dataset with delineations by teams of medical interns, radiology residents, and radiologists. Following each HITL cycle, 3D-UNet was re-trained until the Dice Similarity Coefficients (DSCs) reached the quality criteria set by radiologists (DSC = 0.95/0.8 for the normal lung parenchyma/COVID-19 pneumonia). For the probabilistic characterization, a Gaussian Mixture Model (GMM) was fitted to the Hounsfield Units (HUs) of voxels from the CT scans of patients with COVID-19 pneumonia on the assumption that two distinct populations were superimposed: one for GGO and one for consolidation. Results: Manual delineation of the normal lung parenchyma and COVID-19 pneumonia was performed by seven teams on 65 CT scans from 64 patients (56 ± 16 years old (μ ± σ), 46 males, 62 with reported symptoms). Automated lung/COVID-19 pneumonia segmentation with a DSC > 0.96/0.81 was achieved after three HITL cycles. The HITL strategy improved the DSC by 0.2 and 0.5 for the normal lung parenchyma and COVID-19 pneumonia segmentation, respectively. The distribution of the patient-specific thresholds derived from the GMM yielded a mean of −528.4 ± 99.5 HU (μ ± σ), which is below most of the reported fixed HU thresholds. Conclusions: The HITL strategy allowed for fast and effective annotations, thereby enhancing the quality of segmentation for a local CT dataset. Probabilistic characterization of COVID-19 pneumonia by the GMM enabled patient-specific segmentation of GGO and consolidation. The combination of both approaches is essential to gain confidence in DL approaches in our local environment. The patient-specific probabilistic approach, when combined with the automatic quantification of COVID-19 imaging findings, enhances the understanding of GGO and consolidation during the course of the disease, with the potential to improve the accuracy of clinical predictions. Full article

The use of mobile robots in substations improves maintenance efficiency and ensures the personal safety of staff working at substations, which is a trend in the development of technologies. Strong electric and solid magnetic fields around high-voltage equipment in substations may lead to the breakdown and failure of inspection devices. Therefore, safe operation range measurement and coordinated planning are key factors in ensuring the safe operation of substations. This paper first summarizes the current developments that are occurring in the field of fixed and mobile safe operating range sensing methods for substations, such as ultra-wideband technology, the two-way time flight method, and deep learning image processing algorithms. Secondly, this paper introduces path-planning algorithms based on safety range sensing and analyzes the adaptability of global search methods based on a priori information, local planning algorithms, and sensor information in substation scenarios. Finally, in view of the limitations of the existing range awareness and path-planning methods, we investigate the problems that occur in the dynamic changes in equipment safety zones and the frequent switching of operation scenarios in substations. Furthermore, we explore a new type of barrier and its automatic arrangement system to improve the performance of distance control and path planning in substation scenarios. Full article

UAV communications have seen a rapid rise in the last few years. The drone class of UAV has particularly become more widespread around the world, and illicit behavior using drones has become a problem. Therefore, localization, tracking, and even taking control of drones have also gained interest. Knowing the frequency of a target signal, its position can be determined (as the angle of arrival with respect to a fixed receiver point) using radio frequency-based localization techniques. One such technique is represented by the subspace-based algorithms that offer highly accurate results. This paper presents the implementation of the MUSIC algorithm on an SDR-based system using a uniform circular antenna array and its experimental evaluation in relevant outdoor environments for drone localization. The results show the capability of the system to indicate the AoA of the target signal. The results are compared with the actual direction computed from the log files of the drone application and validated with a professional direction-finding solution (i.e., Narda SignalShark equipped with the automatic direction-finding antenna). Full article

Software bugs are a noteworthy concern for developers and maintainers. When a failure is detected late, it costs more to be fixed. To repair the bug that caused the software failure, the location of the bug must first be known. The process of finding the defective source code elements that led to the failure of the software is called bug localization. Effective approaches for automatically locating bugs using bug reports are highly desirable, as they would reduce bug-fixing time, consequently lowering software maintenance costs. With the increasing size and complexity of software projects, manual bug localization methods have become complex, challenging, and time-consuming tasks, which motivates research on automated bug localization techniques. This paper introduces a novel bug localization model, which works on two levels. The first level localizes the buggy classes using an information retrieval approach and it has two additional sub-phases, namely the class-level feature scoring phase and the class-level final score and ranking phase, which ranks the top buggy classes. The second level localizes the buggy methods inside these classes using an information retrieval approach and it has two sub-phases, which are the method-level feature scoring phase and the method-level final score and ranking phase, which ranks the top buggy methods inside the localized classes. A model is evaluated using an AspectJ dataset, and it can correctly localize and rank more than 350 classes and more than 136 methods. The evaluation results show that the proposed model outperforms several state-of-the-art approaches in terms of the mean reciprocal rank (MRR) metrics and the mean average precision (MAP) in class-level bug localization. Full article

Automatic extraction of building contours from high-resolution images is of great significance in the fields of urban planning, demographics, and disaster assessment. Network models based on convolutional neural network (CNN) and transformer technology have been widely used for semantic segmentation of buildings from high resolution remote sensing images (HRSI). However, the fixed geometric structure and the local receptive field of the convolutional kernel are not good at global feature extraction, and the transformer technique with self-attention mechanism introduces computational redundancies and extracts local feature details poorly in the process of modeling the global contextual information. In this paper, a dual-branch fused reconstructive transformer network, DFRTNet, is proposed for efficient and accurate building extraction. In the encoder, the traditional transformer is reconfigured by designing the local and global feature extraction module (LGFE); the branch of global feature extraction (GFE) performs dynamic range attention (DRA) based on the idea of top-k attention for extracting global features; furthermore, the branch of local feature extraction (LFE) is used to obtain fine-grained features. The multilayer perceptron (MLP) is employed to efficiently fuse the local and global features. In the decoder, a simple channel attention module (CAM) is used in the up-sampling part to enhance channel dimension features. Our network achieved the best segmentation accuracy on both the WHU and Massachusetts building datasets when compared to other mainstream and state-of-the-art methods. Full article