Low Computational Complexity Research Articles

Exploring systemic effects of the friendship paradox on social network participation

The friendship paradox states that, on average, people have fewer friends than their friends do. This can lead to social perception biases, such as the belief that one’s friends are more socially engaged than oneself. In a recent paper, we investigated the consequences of this type of social comparison for content sharing behavior in online social networks (Medhat and Iyer, The Friendship paradox and social network participation, international workshop on complex networks and their applications, pp 301–315, 2023). We simulated a scenario where people compare the feedback that their own content receives to the feedback that their friends’ content receives and adjust their sharing based on the comparison. If sharing and feedback rates are initially uniform over individuals, then feedback disparities initially depend solely upon the local structural friending paradox. These structurally induced disparities may then induce sharing disparities that further amplify (or potentially, reduce) feedback disparities, triggering cycles of updates of sharing rates and feedback disparities. In our previous work, we observed that monotonic responses to social comparisons, where larger disparities result in greater behavioral changes, led to an asymptotic decline in overall network sharing. In this paper, we extend our earlier work by studying fundamental properties of our friendship-paradox-induced sharing trajectories (FIT) method. We analyze how sharing rates trend when it is dictated by both observed feedback disparities and activity thresholds for collective behavior. We also evaluate the iterative impact of FIT on network nodes, as a measure of node centrality in a network, finding it to have the ability to identify central nodes at a community level and to do so at a lower computational complexity than community detection, a property not matched by commonly used node centrality methods such as betweeness, closeness and degree centrality.

Preload Multi-Objective Optimization Method for Ultrasonic Motors Based on NSGA-II

The preload has a significant impact on the output performance of ultrasonic motors. To achieve optimal output from the ultrasonic motor, this paper proposes a novel multi-objective optimization method for preload, utilizing the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to solve the problem of determining the optimal preload. Firstly, a platform for preload adjustment and testing is established to conduct experimental tests on the motor’s characteristics. Based on the experimental data, the influence of preload on the motor’s output performance is analyzed. Secondly, linear regression models are established for three performance indicators: no-load speed, stall torque, and output efficiency with respect to preload, and the objective functions are fitted accordingly. Finally, the NSGA-II algorithm is used for multi-objective optimization with three objectives to obtain the Pareto front and determine the optimal preload for the ultrasonic motor. Simulation and experimental results show that after preload optimization, the no-load speed increased by 4.93%, stall torque increased by 26.10%, and output efficiency increased by 11.28%. Compared to existing optimization methods, this approach has lower computational complexity and better optimization performance, ensuring higher optimization accuracy and precision. The preload optimization method based on NSGA-II can improve the performance of ultrasonic motors in practical applications.

DGA Domain Detection Based on Transformer and Rapid Selective Kernel Network

Botnets pose a significant challenge in network security by leveraging Domain Generation Algorithms (DGA) to evade traditional security measures. Extracting DGA domain samples is inherently complex, and the current DGA detection models often struggle to capture domain features effectively when facing limited training data. This limitation results in suboptimal detection performance and an imbalance between model accuracy and complexity. To address these challenges, this paper introduces a novel multi-scale feature fusion model that integrates the Transformer architecture with the Rapid Selective Kernel Network (R-SKNet). The proposed model employs the Transformer’s encoder to couple the single-domain character elements with the multiple types of relationships within the global domain block. This paper proposes integrating R-SKNet into DGA detection and developing an efficient channel attention (ECA) module. By enhancing the branch information guidance in the SKNet architecture, the approach achieves adaptive receptive field selection, multi-scale feature capture, and lightweight yet efficient multi-scale convolution. Moreover, the improved Feature Pyramid Network (FPN) architecture, termed EFAM, is utilized to adjust channel weights for outputs at different stages of the backbone network, leading to achieving multi-scale feature fusion. Experimental results demonstrate that, in tasks with limited training samples, the proposed method achieves lower computational complexity and higher detection accuracy compared to mainstream detection models.

MRP-YOLO: An Improved YOLOv8 Algorithm for Steel Surface Defects

The existing detection algorithms are unable to achieve a suitable balance between detection accuracy and inference speed. As the accuracy of the algorithm increases, its complexity also rises, resulting in a decrease in detection speed, which undermines its practicality. This issue is particularly evident in the context of surface defect detection in industrial parts, where low contrast, small target features, difficult feature extraction, and low real-time detection efficiency are prominent challenges. This study proposes a novel method for steel defect detection based on the YOLO v8 algorithm, which improves detection accuracy while maintaining low computational complexity. Firstly, the global background and edge information are adaptively extracted via the MSA-SPPF module in order to obtain a more comprehensive feature representation. Furthermore, the anti-interference ability of the model is enhanced through the deformability of attention and the large convolution kernel characteristics. Secondly, the design of Dynamic Conv and C2f-OREPA enables the model to efficiently reduce the demand for computational resources while maintaining high performance. It is further proposed that the RepHead detection head approximates the multi-branch structure of the original training by a single convolution operation. This approach not only enriches the feature representation but also maintains an efficient inference process. The effectiveness of the improved MRP-YOLO algorithm is verified using the NEU-DET industrial surface defect dataset. The experimental results demonstrate that the mAP of the MRP-YOLO algorithm reaches 75.6%, which is 2.2% higher than that of the YOLOv8n algorithm, while the FLOPs are only 2.3 G higher. It indicates that the detection accuracy is significantly improved with a limited increase in computational complexity.

Quantification of Nearshore Sandbar Seasonal Evolution Based on Drone Pseudo-Bathymetry Time-Lapse Data

Nearshore sandbars are dynamic features that characterize shallow morphobathymetry and vary over a wide range of geometries and temporal lifespans. Nearshore sandbars influence beach geometry by altering the energy of incoming waves; thus, monitoring the evolution of sandbars is a fundamental approach in effective coastal planning. Due to several natural and technical limitations related to shallow seafloor mapping, there is a significant gap in the availability of high-resolution, shallow bathymetric data for monitoring the dynamic behaviour of nearshore sandbars effectively. This study introduces a novel image-processing technique that produces time series of pseudo-bathymetric data by utilizing multi-temporal (monthly) drone imagery, and it provides an assessment of local morphodynamics at a sandy beach in the southeast Mediterranean. The technique is called standardized-ratio bathymetric index (SRBI), and it transforms natural-colour drone imagery to pseudo-bathymetric data by applying an empirical formula used for satellite-derived bathymetry. This technique correlates well with laser altimetry depth measurements; however, it does not require in situ depth data for implementation. The resulting pseudo-bathymetric data allows for extracting cross-shore profiles and delineating the sandbar crest with 4 m horizontal accuracy. Stacking of temporal profiles allowed for the quantification of the sandbar’s crest and trough changes at different alongshore sections. The main findings suggest that the nearshore crescentic sandbar at Episkopi Beach (north Crete) shows strong seasonality regarding net offshore migration that is promoted by enhanced wave action during winter months. In addition, the crescentic sandbar is susceptible to morphology arrestment during prolonged weeks of low wave action. The average migration rate during winter is 10 m.month−1, with some sections exhibiting a maximum of 60 m.month−1. This study (a) offers a novel remote-sensing approach, suitable for nearshore seafloor monitoring with low computational complexity, (b) reveals sandbar geometry and temporal change in superior detail compared to other observational methods, and (c) advances knowledge about nearshore sandbar monitoring in the Mediterranean region.

An improved GM(1,1) model based on weighted MSE and optimal weighted background value and its application

GM(1,1) model is widely used because it does not require a large number of samples and has a low computational complexity and no limitation of statistical assumptions. Common drawback of the GM(1,1) models developed by various techniques for improving the performance is their unsatisfied predicting accuracy although they have satisfied fitting accuracy. The aim of this paper is to develop GM(1,1) model with excellent predicting accuracy rather than fitting one. We proposed an improved GM(1,1) model based on weighted mean squared error (MSE) and optimal weighted background value: OB-WMSE-GM(1,1). To illustrate its effectiveness, it was applied to one simulation example and two application examples. For the exponential function simulation example, the fitting MSE, fitting WMSE and predicting MSE of the proposed GM(1,1) (0.000002, 0.000039 and 0.015485) were much lower than ones of the typical GM(1,1) (0.001492, 0.002606 and 1.144524). For the annual LCD TV output prediction example, the fitting MSE, fitting WMSE and predicting MSE of the proposed GM(1,1) (1.425255, 3.199446 and 132.046775) were much lower than ones of the typical GM (1,1) (48.003290, 111.431942 and 5519.135753). For the crude oil processing volume prediction example, the fitting MAPE, fitting WMRE and predicting MAPE of the proposed GM(1,1) (4.090405, 3.213106 and 3.775669) were lower than ones of the typical GM(1,1) (4.399448, 3.883765 and 5.040509). When the proposed method is properly combined with the other various techniques including metabolic mechanism, residual GM(1,1), original sequence pre-processing, background value reconstruction and initial condition optimization, its performance may be more and more improved.

An attention mechanism-based lightweight UNet for musculoskeletal ultrasound image segmentation.

Accurate musculoseletal ultrasound (MSKUS) image segmentation is crucial for diagnosis and treatment planning. Compared with traditional segmentation methods, deploying deep learning segmentation methods that balance segmentation efficiency, accuracy, and model size on edge devices has greateradvantages. This paper aims to design a MSKUS image segmentation method that has fewer parameters, lower computation complexity and higher segmentationaccuracy. In this study, an attention mechanism-based lightweight UNet (AML-UNet) is designed to segment target muscle regions in MSKUS images. To suppress the transmission of redundant feature, Channel Reconstruction and Spatial Attention Module is designed in the encoding path. In addition, considering the inherent characteristic of MSKUS image, Multiscale Aggregation Module is developed to replace the skip connection architecture of U-Net. Deep supervision is also introduced to the decoding path to refine predicted masks gradually. Our method is evaluated on two MSKUS 2D-image segmentation datasets, including 3917 MSKUS and 1534 images respectively. In the experiments, a five-fold cross-validation method is adopted in ablation experiments and comparison experiments. In addition, Wilcoxon Signed-Rank Test and Bonferroni correction are employed to validate the significance level. 0.01 was used as the statistical significance level in ourpaper. AML-UNet yielded a mIoU of 84.17% and 90.14% on two datasets, representing a 3.38% ( ) and 3.48% ( ) over the Unext model. The number of parameters and FLOPs are only 0.21M and 0.96G, which are 1/34 and 1/29 of those in comparison withUNet. Our proposed model achieved superior results with fewer parameters while maintaining segmentation efficiency and accuracy compared to othermethods.

Semantic segmentation network for mangrove tree species based on UAV remote sensing images

Mangroves are special vegetation that grows in the intertidal zone of the coast and has extremely high ecological and environmental value. Different mangrove species exhibit significant differences in ecological functions and environmental responses, so accurately identifying and distinguishing these species is crucial for ecological protection and monitoring. However, mangrove species recognition faces challenges, such as morphological similarity, environmental complexity, target size variability, and data scarcity. Traditional mangrove monitoring methods mainly rely on expensive and operationally complex multispectral or hyperspectral remote sensing sensors, which have high data processing and storage costs, hindering large-scale application and popularization. Although hyperspectral monitoring is still necessary in certain situations, the low identification accuracy in routine monitoring severely hinders ecological analysis. To address these issues, this paper proposes the UrmsNet segmentation network, aimed at improving identification accuracy in routine monitoring while reducing costs and complexity. It includes an improved lightweight convolution SCConv, an Adaptive Selective Attention Module (ASAM), and a Cross-Layer Feature Fusion Module (CLFFM). ASAM adaptively extracts and fuses features of different mangrove species, enhancing the network’s ability to characterize mangrove species with similar morphology and in complex environments. CLFFM combines shallow details and deep semantic information to ensure accurate segmentation of mangrove boundaries and small targets.Additionally, this paper constructs a high-quality RGB image dataset for mangrove species segmentation to address the data scarcity problem. Compared to traditional methods, our approach is more precise and efficient. While maintaining relatively low parameters and computational complexity (FLOPs), it achieves excellent performance with mIoU and mPA metrics of 92.21% and 95.98%, respectively. This performance is comparable to the latest methods using multispectral or hyperspectral data but significantly reduces cost and complexity. By combining periodic hyperspectral monitoring with UrmsNet-supported routine monitoring, a more comprehensive and efficient mangrove ecological monitoring can be achieved.These research findings provide a new technical approach for large-scale, low-cost monitoring of important ecosystems such as mangroves, with significant theoretical and practical value. Furthermore, UrmsNet also demonstrates excellent performance on LoveDA, Potsdam, and Vaihingen datasets, showing potential for wider application.

Revitalizing Convolutional Network for Image Restoration.

Image restoration aims to reconstruct a high-quality image from its corrupted version, playing essential roles in many scenarios. Recent years have witnessed a paradigm shift in image restoration from convolutional neural networks (CNNs) to Transformer-based models due to their powerful ability to model long-range pixel interactions. In this paper, we explore the potential of CNNs for image restoration and show that the proposed simple convolutional network architecture, termed ConvIR, can perform on par with or better than the Transformer counterparts. By re-examing the characteristics of advanced image restoration algorithms, we discover several key factors leading to the performance improvement of restoration models. This motivates us to develop a novel network for image restoration based on cheap convolution operators. Comprehensive experiments demonstrate that our ConvIR delivers state-of-the-art performance with low computation complexity among 20 benchmark datasets on five representative image restoration tasks, including image dehazing, image motion/defocus deblurring, image deraining, and image desnowing.

Scaled Coherence Factor Based Delay Multiply and Sum Beamforming Algorithm in Microwave Imaging

ABSTRACTMicrowave imaging is a highly promising modality for the early detection of breast cancer. A multitude of image reconstruction algorithms have been devised. Previous methods utilized the coherence factor (CF) to enhance image quality. However, CF‐based methods often oversuppress the desired signals and introduce image artifacts when the signal‐to‐noise ratio is low. This study introduces a scaled coherence factor‐based delay multiply and sum (DMAS) Beamforming technique to achieve good noise suppression while maintaining robustness. The performance of this method was evaluated using datasets derived from an experimental breast phantom. The results show that the proposed method achieves improvements of signal‐to‐clutter ratio by 26%, signal‐to‐mean ratio by 7%, full‐width at half‐maximum by 8%, and localization error by 51% compared to the CF‐DMAS method. The suggested microwave imaging approach for breast cancer detection provides high resolution, contrast, and localization accuracy along with low computational complexity.

HOPE: High-Order Polynomial Expansion of Black-Box Neural Networks.

Despite their remarkable performance, deep neural networks remain mostly "black boxes", suggesting inexplicability and hindering their wide applications in fields requiring making rational decisions. Here we introduce HOPE (High-order Polynomial Expansion), a method for expanding a network into a high-order Taylor polynomial on a reference input. Specifically, we derive the high-order derivative rule for composite functions and extend the rule to neural networks to obtain their high-order derivatives quickly and accurately. From these derivatives, we can then derive the Taylor polynomial of the neural network, which provides an explicit expression of the network's local interpretations. We combine the Taylor polynomials obtained under different reference inputs to obtain the global interpretation of the neural network. Numerical analysis confirms the high accuracy, low computational complexity, and good convergence of the proposed method. Moreover, we demonstrate HOPE's wide applications built on deep learning, including function discovery, fast inference, and feature selection. We compared HOPE with other XAI methods and demonstrated our advantages.

Efficient Risk Assessment of Time-to-Event Targets With Adaptive Information Transfer.

Using informative sources to enhance statistical analysis in target studies has become an increasingly popular research topic. However, cohorts with time-to-event outcomes have not received sufficient attention, and external studies often encounter issues of incomparability due to population heterogeneity and unmeasured risk factors. To improve individualized risk assessments, we propose a novel methodology that adaptively borrows information from multiple incomparable sources. By extracting aggregate statistics through transitional models applied to both the external sources and the target population, we incorporate this information efficiently using the control variate technique. This approach eliminates the need to load individual-level records from sources directly, resulting in low computational complexity and strong privacy protection. Asymptotically, our estimators of both relative and baseline risks are more efficient than traditional results, and the power of covariate effects testing is much enhanced. We demonstrate the practical performance of our method via extensive simulations and a real case study.

Multi-scale perceptual modulation network for low-dose computed tomography denoising.

Low-dose computed tomography (LDCT) reduces radiation exposure, but the introduced noise and artifacts impair its diagnostic accuracy. Convolutional neural networks (CNNs) are widely used for LDCT denoising, but they suffer from a limited receptive field. The use of a larger kernel size can enlarge the receptive field and boost model performance; however, the computational cost of the model greatly increases. We aimed to develop a LDCT denoising CNN with a large receptive field and lower computational complexity. We developed a multi-scale perceptual modulation network (MSPMnet) incorporating a powerful multi-head decomposable convolution (MHDC). To address the high computational complexity of large kernel convolutions, we developed a novel MHDC module that can capture multi-scale features and efficiently expand the receptive field. The MHDC module couples maximum-pooling with three depth-wise convolutions of varying kernel sizes via a channel splitting mechanism, where, unlike conventional CNNs, the two large two-dimensional kernels are each decomposed into a set of cascaded orthogonal one-dimensional kernels to remain lightweight. Further, departing from prior methodologies that apply a uniform kernel size throughout the network, we introduced a receptive field-ramp mechanism that adeptly transitions from local to relatively long-range dependency modeling as the network depth increases, thereby achieving superior performance. The proposed MSPMnet was evaluated on a Mayo Clinic data set with a conventional iterative algorithm, two CNN models, and two Transformer models used for comparison. Compared to the competing baseline methods, the MSPMnet exhibited better performance in both the visual and quantitative assessments. Visually, the MSPMnet preserved the structure, edges, and textures with excellent noise and artifact reduction, generating the denoised images closest to normal-dose computed tomography images. Quantitatively, the MSPMnet had the lowest root mean-square error (RMSE) (8.3094±1.9325) and the highest peak signal-to-noise ratio (PSNR) (33.8525±1.8213 dB), structural similarity index (SSIM) (0.9309±0.0272), and feature similarity index (FSIM) (0.9699±0.0113), demonstrating superior denoising performance. The proposed MSPMnet excelled at LDCT denoising, effectively removing noise and artifacts while preserving edges. Compared to the state-of-the-art CNNs and Transformers, the proposed MSPMnet exhibited superior denoising performance both quantitatively and qualitatively.

Robust and energy-efficient RPL optimization algorithm with scalable deep reinforcement learning for IIoT

The increasing complexity and quantity of the Industrial Internet of Things (IIoT) pose new challenges to the traditional routing protocol for low-power and lossy networks (RPL) in terms of dynamic management, data transmission reliability, and energy efficiency optimization. This paper proposes a scalable deep reinforcement learning (DRL) algorithm with a multi-attention actor double critic model for routing optimization (MADC) to meet the requirements of IIoT for efficient and intelligent routing decisions while improving data transmission reliability and energy efficiency. Specifically, MADC employs the centralized training and decentralized execution (CTDE) learning paradigm to decouple the model’s training and inference tasks, which reduces the difficulty and computational cost of model learning and improves the training efficiency. In addition, a lightweight actor network based on multi-scale convolutional attention mechanism is designed in MADC, which can provide intelligent and real-time decision-making capabilities for resource-constrained nodes with low computational and storage complexities. Moreover, a scalable critic network utilizing multiple attention mechanisms is proposed. It is not only suitable for dynamic and changing network environments but also can more comprehensively and accurately evaluate local observation states, providing more accurate and efficient guidance for model optimization. Furthermore, MADC incorporates a double critic network architecture to mitigate potential overestimation issues during training, thereby ensuring the model’s robustness and reliability. Simulation results demonstrate that MADC outperforms existing RPL optimization algorithms in terms of energy efficiency, data transmission reliability, and adaptability.

Multi-spectral Remote Sensing Image Fusion Method Based on Gradient Moment Matching

Image fusion is a popular research direction in the field of computer vision. Traditional image fusion algorithms can achieve good results in fusing grayscale images, but it is difficult to achieve ideal results in processing multi-spectral images. To address the shortcomings of multi-spectral image fusion, this study proposes a low computational complexity and low latency multi-spectral image fusion model by utilizing a multi-step degree moment matching algorithm and a generative adversarial network for fusion. Through experiments, it was found that the F1 score of the GAN-MMN model on the TinyPerson dataset was 89.79%, with an average recall rate of 89.76%. The GAN-MMN performance was higher than that of the control model. Meanwhile, the GAN-MMN model also exhibited superior performance in high-frequency feature extraction and time delay compared to the control model. According to the experimental results, the proposed multi-spectral remote sensing image fusion model had a high feature extraction effect, and its recall rate and F1 score were better than the control model, so the research model had a certain progressiveness. The proposal of this model gives a new approach for the processing of multi-spectral remote sensing images, effectively promoting the development of the computer vision industry.

Adaptive target localization under uncertainty using Multi-Agent Deep Reinforcement Learning with knowledge transfer

Target localization is a critical task in sensitive applications, where multiple sensing agents communicate and collaborate to identify the target location based on sensor readings. Existing approaches investigated the use of Multi-Agent Deep Reinforcement Learning (MADRL) to tackle target localization. Nevertheless, these methods do not consider practical uncertainties, like false alarms when the target does not exist or when it is unreachable due to environmental complexities. To address these drawbacks, this work proposes a novel MADRL-based method for target localization in uncertain environments. The proposed MADRL method employs Proximal Policy Optimization to optimize the decision-making of sensing agents, which is represented in the form of an actor–critic structure using Convolutional Neural Networks. The observations of the agents are designed in an optimized manner to capture essential information in the environment, and a team-based reward functions is proposed to produce cooperative agents. The MADRL method covers three action dimensionalities that control the agents’ mobility to search the area for the target, detect its existence, and determine its reachability. Using the concept of Transfer Learning, a Deep Learning model builds on the knowledge from the MADRL model to accurately estimating the target location if it is unreachable, resulting in shared representations between the models for faster learning and lower computational complexity. Collectively, the final combined model is capable of searching for the target, determining its existence and reachability, and estimating its location accurately. The proposed method is tested using a radioactive target localization environment and benchmarked against existing methods, showing its efficacy.

Automatic Modulation Recognition Based on Multimodal Information Processing: A New Approach and Application

Automatic modulation recognition (AMR) has wide applications in the fields of wireless communications, radar systems, and intelligent sensor networks. The existing deep learning-based modulation recognition models often focus on temporal features while overlooking the interrelations and spatio-temporal relationships among different types of signals. To overcome these limitations, a hybrid neural network based on a multimodal parallel structure, called the multimodal parallel hybrid neural network (MPHNN), is proposed to improve the recognition accuracy. The algorithm first preprocesses the data by parallelly processing the multimodal forms of the modulated signals before inputting them into the network. Subsequently, by combining Convolutional Neural Networks (CNN) and Bidirectional Gated Recurrent Unit (Bi-GRU) models, the CNN is used to extract spatial features of the received signals, while the Bi-GRU transmits previous state information of the time series to the current state to capture temporal features. Finally, the Convolutional Block Attention Module (CBAM) and Multi-Head Self-Attention (MHSA) are introduced as two attention mechanisms to handle the temporal and spatial correlations of the signals through an attention fusion mechanism, achieving the calibration of the signal feature maps. The effectiveness of this method is validated using various datasets, with the experimental results demonstrating that the proposed approach can fully utilize the information of multimodal signals. The experimental results show that the recognition accuracy of MPHNN on multiple datasets reaches 93.1%, and it has lower computational complexity and fewer parameters than other models.

RCT-YOLOv8: A Tuna Detection Model for Distant-Water Fisheries Based on Improved YOLOv8

With the development of distant-water fisheries, ship fishing and fish catch detection are now vital to modern fishing. Existing manual detection methods are prone to issues such as missed detections and false detections. Deep learning has enabled the deployment of detection models on shipboard devices, offering a new solution. However, many existing models are hindered by large parameters and computational complexity, making them unsuitable for shipboard use due to limited resources and costs onboard ships. To address these challenges, we propose the RCT-YOLOv8 model for tuna catch detection in this paper. Specifically, we adopt YOLOv8 as the base model and replace the network backbone with RepVGG network, which employs re-parameterized convolutions to enhance detection accuracy. Additionally, we incorporate coordinate attention at the end of the backbone to better aggregate channel-wise information. In the neck part, we introduce the contextual transformer (CoT) attention and propose the C2F-CoT model, which combines convolutional neural network with Transformer to capture global features, thereby improving detection accuracy and the effectiveness of feature propagation. We test multiple loss functions and select efficient intersection over union, which is more suitable for our algorithm. Furthermore, to adapt to devices with limited computational resources, we utilize the dependency-graph-based pruning method to compress the network model. Compared to the base network, the pruned model achieves a 9.8% increase in detection accuracy while reducing parameters and computational complexity by 40% and 35.8%, respectively. Compared to various algorithms, the pruned model demonstrates the highest detection accuracy, lowest parameter count, and lowest computational complexity, achieving optimal results at all fronts.

YOLO-SDL: a lightweight wheat grain detection technology based on an improved YOLOv8n model.

Wheat, being a crucial global food crop, holds immense significance for food safety and agricultural economic stability, as the quality and condition of its grains are critical factors. Traditional methods of wheat grain detection are inefficient, and the advancements in deep learning offer a novel solution for fast and accurate grain recognition. This study proposes an improved deep learning model based on YOLOv8n, referred to as YOLO-SDL, aiming to achieve efficient wheat grain detection. A high-quality wheat grain dataset was first constructed, including images of perfect, germinated, diseased, and damaged grains. Multiple data augmentation techniques were employed to enhance the dataset's complexity and diversity. The YOLO-SDL model incorporates the ShuffleNetV2 architecture in its backbone and combines depthwise separable convolutions (DWConv) with the large separable kernel attention (LSKA) mechanism in its neck structure, significantly improving detection speed and accuracy while ensuring the model remains lightweight. The results indicate that YOLO-SDL achieves superior performance in wheat grain detection, balancing lightweight design and performance optimization. The model achieved a P of 0.942, R of 0.903, mAP50 of 0.965, and mAP50-95 of 0.859, with low computational complexity, making it suitable for resource-constrained environments. These findings demonstrate the efficiency of the ShuffleNetV2, DWConv, and LSKA structures. The proposed YOLO-SDL model provides a new technical solution for agricultural automation and serves as a reliable reference for detecting other crops.

Fast DOA Estimation of IGS Noncircular Signals Under Coprime Arrays

Traditional direction-of-arrival (DOA) estimation methods often encounter challenges such as low accuracy and high computational complexity, particularly in complex signal environments with multiple sources. To address these issues, this paper proposes a fast DOA estimation method for improper Gaussian signaling (IGS) non-circular signals, based on a coprime array; this approach involves extending the array model by leveraging the unique properties of non-circular signals. Subsequently, this study introduces the ANCM algorithm, a method designed for fast DOA estimation of IGS non-circular signals within a non-uniform noise background. The proposed ANCM algorithm is then compared to classical algorithms, through simulations in a non-uniform noise environment. Experimental results demonstrate that the ANCM algorithm significantly outperforms traditional methods in such conditions. Therefore, the development of new, efficient DOA estimation methods holds substantial theoretical and practical value.