Search Results (3,512)

Existing convolution neural network (CNN)-based video synthetic aperture radar (SAR) moving target shadow detectors are difficult to model long-range dependencies, while transformer-based ones often suffer from greater complexity. To handle these issues, this paper proposes MambaShadowDet, a novel lightweight deep learning (DL) detector based on a state space model (SSM), dedicated to high-speed and high-accuracy moving target shadow detection in video SAR images. By introducing SSM with the linear complexity into YOLOv8, MambaShadowDet effectively captures the global feature dependencies while relieving computational load. Specifically, it designs Mamba-Backbone, combining SSM and CNN to effectively extract both global contextual and local spatial information, as well as a slim path aggregation feature pyramid network (Slim-PAFPN) to enhance multi-level feature extraction and further reduce complexity. Abundant experiments on the Sandia National Laboratories (SNL) video SAR data show that MambaShadowDet achieves superior moving target shadow detection performance with a detection accuracy of 80.32% F1 score and an inference speed of 44.44 frames per second (FPS), outperforming existing models in both accuracy and speed. Full article

Flowers, as integral elements of urban landscapes, are critical not only for aesthetic purposes but also for fostering human–nature interactions in green spaces. However, research on flower color preferences has largely been descriptive, and there is a lack of exploration of potential mechanisms influencing flower color preferences, such as economic and social factors. This study created visual samples through precise color adjustment techniques and introduced the L*, a*, and b* parameters from the CIELAB color system to quantify the flower colors of the survey samples, conducting an online survey with 354 Chinese residents. The complex aesthetic process’s driving factors were unveiled through a comprehensive analysis using a Generalized Additive Model (GAM), a piecewise Structural Equation Model (SEM), and linear regression models. The results show that the public’s flower color preference is primarily related to the a* and b* parameters, which represent color dimensions in the CIELAB color space, and it is not significantly related to L* (lightness). Factors such as age, annual household income level (AI), personal income sources (PI), nature experience, and emotional state (TMD) significantly influence color preferences, with emotional state identified as the most critical factor. Lastly, linear regression models further explain the potential mechanism of the influencing factors. This study proposes a framework to assist urban planners in selecting flower colors that resonate with diverse populations, enhancing both the attractiveness of urban green spaces and their potential to promote pro-environmental behavior. By aligning flower color design with public preferences, this study contributes to sustainable urban planning practices aimed at improving human well-being and fostering deeper connections with nature. Full article

Most deep-learning-based vision tasks rely heavily on crowd-labeled data, and a deep neural network (DNN) is usually impacted by the laborious and time-consuming labeling paradigm. Recently, foundation models (FMs) have been presented to learn richer features from multi-modal data. Moreover, a single foundation model enables zero-shot predictions on various vision tasks. The above advantages make foundation models better suited for remote sensing images, where image annotations are more sparse. However, the inherent differences between natural images and remote sensing images hinder the applications of the foundation model. In this context, this paper provides a comprehensive review of common foundation models and domain-specific foundation models for remote sensing, and it summarizes the latest advances in vision foundation models, textually prompted foundation models, visually prompted foundation models, and heterogeneous foundation models. Despite the great potential of foundation models for vision tasks, open challenges concerning data, model, and task impact the performance of remote sensing images and make foundation models far from practical applications. To address open challenges and reduce the performance gap between natural images and remote sensing images, this paper discusses open challenges and suggests potential directions for future advancements. Full article

This work deals with the key constituent behind the existence of superfluid states in ultracold fermionic gases confined in a harmonic trap in 2D, namely, the formation of Cooper pairs in the presence of a Fermi sea in inhomogeneous confinement. For a set of finite-range models representing particle–particle interaction, we first ascertain the simultaneity of the emergence of bound states and the divergence of the s-wave scattering length in 2D as a function of the interaction potential parameters in free space. Then, through the analysis of two particles interacting in 2D harmonic confinement, we evaluate the energy shift with respect to the discrete harmonic oscillator levels for both repulsive and attractive cases. All of these results are the basis for determining the energy gaps of Cooper pairs arising from two particles interacting in the presence of a Fermi sea consisting of particles immersed in a 2D harmonic trap. Full article

This paper introduces a vision-based target recognition and positioning system for UAV mobile landing scenarios, addressing challenges such as target occlusion due to shadows and the loss of the field of view. A novel image preprocessing technique is proposed, utilizing finite adaptive histogram equalization in the HSV color space, to enhance UAV recognition and the detection of markers under shadow conditions. The system incorporates a Kalman filter-based target motion state estimation method and a binocular vision-based depth camera target height estimation method to achieve precise positioning. To tackle the problem of poor controller performance affecting UAV tracking and landing accuracy, a feedforward model predictive control (MPC) algorithm is integrated into a mobile landing control method. This enables the reliable tracking of both stationary and moving targets via the UAV. Additionally, with a consideration of the complexities of real-world flight environments, a mobile tracking and landing control strategy based on airspace division is proposed, significantly enhancing the success rate and safety of UAV mobile landings. The experimental results demonstrate a 100% target recognition success rate and high positioning accuracy, with x and y-axis errors not exceeding 0.01 m in close range, the x-axis relative error not exceeding 0.05 m, and the y-axis error not exceeding 0.03 m in the medium range. In long-range situations, the relative errors for both axes do not exceed 0.05 m. Regarding tracking accuracy, both KF and EKF exhibit good following performance with small steady-state errors when the target is stationary. Under dynamic conditions, EKF outperforms KF with better estimation results and a faster tracking speed. The landing accuracy is within 0.1 m, and the proposed method successfully accomplishes the mobile energy supply mission for the vehicle-mounted UAV system. Full article

With the impending integration of sensing capabilities into new wireless standards such as 6G and 802.11 bf, there is a growing threat to public privacy. Recent studies have revealed that even small-scale activities, like keyboard typing, can be sensed by attackers using Wi-Fi Channel State Information (CSI) as these devices become more common in commercial spaces. This paper aims to model the minimum CSI data rate required to sense activities in the channel and quantifies the detection accuracy of WiFi-based keystroke recognition in relation to the CSI sensing data rate. Our experimental findings using commercial-off-the-shelf hardware suggest that interference can be used as a defence strategy to degrade the CSI data rate and prevent undesirable Wi-Fi sensing attacks. To achieve a reduced data rate, we propose an extension to Bianchi’s model of CSMA/CA systems and establish a new mathematical relationship between channel contention and the available CSI. This proposed relationship was empirically verified, and our contention-based defence strategy was experimentally validated. Experiments show that our contention-based defence strategy increases the chances of evading undesired WiFi-based keystroke recognition by around 70%. By leveraging prior work that shows a degradation in CSI quality with lower transmission rates, we show that conservative interference injection can sufficiently reduce sensing accuracy whilst maintaining channel bandwidth. Full article

In recent years, significant progress has been made in China in the field of deep-sea large yellow croaker (Larimichthys crocea) farming. Compared with the traditional inshore aquaculture model, deep-sea culture of large yellow croaker enjoys a wider growing space with better water quality, thus enhancing fish quality. However, deep-sea aquaculture also faces challenges such as typhoons and strong currents, which often lead to prolonged starvation in fish. Therefore, in order to further promote the technological advancement of large yellow croaker in the field of deep-sea aquaculture, this experiment aimed to investigate the effects of varying starvation durations on the feeding rhythm and physiological state of large yellow croaker. With an initial body mass of 122.62 ± 11.08 g and a body length of (17.9 ± 1.04) cm as the samples, the experiment was divided into five groups, which were starved for 0 d (S0), 2 d (S2), 4 d (S4), 8 d (S8), and 16 d (S16) before resumption of feeding. The results were as follows: under starvation stress for 8 consecutive days, the total duration of feeding gradually decreased in large yellow croaker, but increased at starvation up to 16 days. Each replicate group had 50 large yellow croakers as test subjects, for a total of 750 large yellow croakers. Analyzing the linear regression equations of S0 with S2, S4, S8, and S16 groups, it was found that the trend of rate of change in feeding duration was consistent with the total duration of feeding, i.e., it decreased during 8 days and increased at 16 days. It indicated that the rate of feeding of large yellow croaker was accelerated within 8 days of starvation, while the rate of feeding was slowed down at 16 days of starvation. Furthermore, the blood glucose concentration of large yellow croaker decreased significantly after 8 days of starvation, while it rebounded significantly in the S16 group. Meanwhile, large areas of fatty degeneration were observed in the liver on the 8th day of starvation, followed by extensive hepatocyte necrosis on the 16th day. After resumption of feeding, there was some recovery within 4 days, but hepatocytes were still extensively edematous in the S8 and S16 groups. Meanwhile, the expression of inflammatory factor genes such as IL-1β, IL-10 and TNF-α in the liver increased with the prolongation of starvation time, in which both S8 and S16 groups in the liver were significantly different from the S0 group, and after resumption of feeding, the IL-1β and TNF-α genes of the S8 and S16 groups were significantly different from those of the normal feeding group (p < 0.05), while there was no differentiation for the IL-10 gene. Therefore, based on the results of this study, it is recommended to limit the duration of starvation in the large yellow croaker to no more than 8 days. Full article

Granular soil is a porous medium composed of particles with different sizes and self-similar structures, exhibiting fractal characteristics. It is well established that variations in these fractal properties, such as particle size distribution (PSD), significantly influence the mechanical behavior of the soil. In this paper, a three-dimensional (3D) Discrete Element Method (DEM) is applied to study the mechanical and critical-state behavior of the idealized granular assemblages, in which various PSD shape parameters are considered, including the coefficient of uniformity (C_u), the coefficient of curvature (C_c), and the coefficient of size span (C_s). In addition, the same PSDs but with different mean particle sizes (D₅₀) are also employed in the numerical simulations to examine the particle size effect on the mechanical behavior of the granular media. Numerical triaxial tests are carried out by imposing axial compression under constant mean effective pressure conditions. A unique critical-state stress ratio in p′-q space is observed, indicating that the critical friction angle is independent of the shape of the PSD. However, in the e-p′ plane, the critical state line (CSL) shifts downward and rotates counterclockwise, as the grading becomes more widely distributed, i.e., the increasing coefficient of span (C_s). Additionally, a decrease in the coefficient of curvature (C_c) would also move the CSL downward but with negligible rotation. However, it is found that the variations in the mean particle size (D₅₀) and coefficient of uniformity (C_u) do not affect the position of the CSL in the e-p′ plane. The numerical findings may shed some light on the development of constitutive models of sand that undergo variations in the grading due to crushing and erosion, and address fractal problems related to micro-mechanics in soils. Full article

Obstructive sleep apnea (OSA) represents a significant health concern. While polysomnography (PSG) remains the gold standard, its resource-intensive nature has encouraged the exploration of further alternative approaches. Most of these were based on the heart rate variability (HRV) analysis, but only a few of them have presented a recurrence-based approach. The present paper addresses this gap by integrating convolutional neural networks (CNNs) with HRV recurrence analysis. Employing three different and publicly available databases from PhysioNet’s official repository (Apnea-ECG, MIT-BIH, and UCD-DB), the presented method was able to expose concealed patterns within the distance matrix of HRV’s phase space, which is discernible at an appropriate level of abstraction through CNNs. Under the challenging context of external validation (MIT-BIH and UCD for training, and Apnea-ECG for testing), the results obtained were comparable to those presented in the state of the art, achieving a peak accuracy of 75%, while maintaining balanced sensitivity and specificity at 74% and 75%, respectively. Moreover, these results obtained by the proposed CNN-based recurrence analysis of HRV also outperformed traditional time–frequency models, which have yielded values of accuracy lower than 65%. Hence, this paper highlights the importance of the proposed method in gaining new insights into OSA’s HRV dynamics, offering a contribution that adds to the existing analytical approaches in the state of the art. Full article

The scarcity of high-quality labeled data poses a challenge to the application of deep learning (DL) in landslide identification from remote sensing (RS) images. Semi-supervised learning (SSL) has emerged as a promising approach to address the issue of low accuracy caused by the limited availability of high-quality labels. Nevertheless, the application of SSL approaches developed for natural images to landslide identification encounters several challenges. This study focuses on two specific challenges: inadequate information extraction from limited unlabeled RS landslide images and the generation of low-quality pseudo-labels. To tackle these challenges, we propose a novel and generic DL framework called hybrid perturbation mean match (HPM-Match). The framework combines dual-branch input perturbation (DIP) and independent triple-stream perturbation (ITP) techniques to enhance model accuracy with limited labels. The DIP generation approach is designed to maximize the utilization of manually pre-defined perturbation spaces while minimizing the introduction of erroneous information during the weak-to-strong consistency learning (WSCL) process. Moreover, the ITP structure unifies input, feature, and model perturbations, thereby broadening the perturbation space and enabling knowledge extraction from unlabeled landslide images across various perspectives. Experimental results demonstrate that HPM-Match has substantial improvements in IoU, with maximum increases of 26.68%, 7.05%, and 12.96% over supervised learning across three datasets with the same label ratio and reduces the number of labels by up to about 70%. Furthermore, HPM-Match strikes a better balance between precision and recall, identifying more landslides than other state-of-the-art (SOTA) SSL approaches. Full article

Many real-world expensive industrial and engineering multi-objective optimization problems (MOPs) are driven by historical, experimental, or simulation data. In such scenarios, due to the expensive cost and time required, we are only left with a small amount of labeled data to perform the optimization. These offline data-driven MOPs are usually solved by multi-objective evolutionary algorithms (MOEAs) with the help of surrogate models constructed from offline historical data. The key challenge in developing these data-driven MOEAs is that they have to replace multiple conflicting fitness functions by approximating these objective functions, which may produce cumulative approximation errors and misguide the search. In order to build a reliable surrogate model from a small amount of multi-output offline data and solve the DDMOPs, we have proposed an adaptive model selection method with a reliable individual-based model management-driven MOEA. The proposed algorithm dynamically selects between DNN and XGBoost by comparing their k-fold cross-validation MAE error, which can capture the true generalization ability of the surrogates on unseen data. Then, the selected surrogate is updated with a reliable individual selection strategy, where the individual who is closest, both in the decision and objective space, to the most preferred solution among labeled offline data is chosen. As a result, these two strategies guide the underlying MOEA to the Pareto optimal solutions. The empirical results of the ZDT and DTLZ benchmark test suite validate the use of the three state-of-the-art offline DDMOEAs, showing that our algorithm is able to achieve highly competitive results in terms of convergence and diversity for 2–3 objectives. Finally, our algorithm is applied to an offline data-driven multi-objective problem—transonic airfoil (RAE 2822) shape optimization—to validate its efficiency on real-world DDMOPs. Full article

The design and industrial innovation of intelligent agricultural machinery and equipment for saline alkali land are important means for comprehensive management and capacity improvement of saline alkali land. The autonomous and unmanned agricultural tractor is the inevitable trend of the development of intelligent machinery and equipment in saline alkali land. As an underactuated system with non-holonomic constraints, the independent trajectory planning and lateral stability control of the tractor-trailer system (TTS) face challenges in saline alkali land. In this study, based on the nonlinear underactuation characteristics of the TTS and the law of passive trailer steering, a dual-trajectory collaborative control model was designed. By solving the TTS kinematic/dynamic state space, a nonlinear leading system that can generate the reference pose of a tractor-trailer was constructed. Based on the intrinsic property of the lateral deviation of the TTS, a collaborative trajectory prediction algorithm that satisfies the time domain and system constraints is proposed. Combining the dual-trajectory independent offset and lateral stability parameter of the TTS, an energy function optimization control parameter was constructed to balance the system trajectory tracking performance and lateral control stability. The experimental results showed good agreement between the predicted trailer trajectory and the collaborative control trajectory, with an average lateral error not exceeding 0.1 m and an average course angle error not exceeding 0.054 rad. This ensures that the dynamic controller designed around the tractor-trailer underactuation system can guarantee the smoothness of the trailer trajectory and the controlling stability of the tractor in saline alkali land. Full article

This paper examines the problem of modeling lithium–sulfur (Li-S) battery discharge dynamics. The importance of this problem stems from the attractive specific energy levels achievable by Li-S batteries, which can be particularly appealing for applications such as aviation electrification. Previous research presents different Li-S battery models, including “zero-dimensional” models that neglect diffusion while using the laws of electrochemistry to represent reduction–oxidation (redox) rates. Zero-dimensional models typically succeed in capturing key features of Li-S battery discharge, including the high plateau, low plateau, and dip point visible in the discharge curves of certain Li-S battery chemistries. However, these models’ use of one state variable to represent the mass of each active species tends to furnish high-order models, with many state variables. This increases the computational complexity of model-based estimation and optimal control. The main contribution of this paper is to develop low-order state-space model of Li-S battery discharge. Specifically, the paper starts with a seventh-order zero-dimensional model of Li-S discharge dynamics, analyzes its discharge behavior, constructs phenomenological second- and third-order models capable of replicating this behavior, and parameterizes these models. The proposed models succeed in capturing battery discharge behavior accurately over a wide range of discharge rates. To the best of our knowledge, these are two of the simplest published models capable of doing so. Full article

Autonomous electric vehicle (AEV) services leverage advanced autonomous driving and electric vehicle technologies to provide innovative, driverless transportation solutions. The biggest challenge faced by AEVs is the limited number of charging stations and long charging times. A critical challenge is maximizing passenger travel satisfaction while reducing the AEV idle time. This involves coordinating passenger transport and charging tasks via leveraging the information from charging stations, passenger transport, and AEV data. There are four important contributions in this paper. Firstly, we introduce an integrated scheduling model that considers both passenger transport and charging tasks. Secondly, we propose a multi-level differentiated charging threshold strategy, which dynamically adjusts the charging threshold based on both AEV battery levels and the availability of charging stations, reducing competition among vehicles and minimizing waiting times. Thirdly, we develop a rapid strategy to optimize the selection of charging stations by combining geographic and deviation distance. Fourthly, we design a new evolutionary algorithm to solve the proposed model, in which a buffer space is introduced to promote diversity within the population. Finally, experimental results show that compared to the existing state-of-the-art scheduling algorithms, the proposed algorithm shortens the running time of scheduling algorithms by 6.72% and reduces the idle driving time of AEVs by 6.53%, which proves the effectiveness and efficiency of the proposed model and algorithm. Full article

Superhydrophobic coatings are beneficial for applications like self-cleaning, anti-corrosion, and drag reduction. In this study, we investigated the impact of surface geometry on the static, dynamic, and sliding contact angles in the Cassie–Baxter state. We used fluoro-silane-treated silicon micro-post patterns fabricated via lithography as model surfaces. By varying the solid fraction (ϕ_s), edge-to-edge spacing (L), and the shape and arrangement of the micro-posts, we examined how these geometric factors influence wetting behavior. Our results show that the solid fraction is the key factor affecting both dynamic and sliding angles, while changes in shape and arrangement had minimal impact. The Cassie–Baxter model accurately predicted receding angles but struggled to predict advancing angles. These insights can guide the development of coatings with enhanced superhydrophobic properties, tailored to achieve higher contact angles and customized for different environmental conditions. Full article