Network Framework Research Articles

Leveraging central-surrounding receptive fields for single image dehazing

Image dehazing is a representative low-level vision task that estimates latent haze-free images from hazy images. In recent years, Vision Transformers (ViT) have shown promising dehazing performance, leveraging their capacity for global perception through long-sequence dependencies in cost of high resource consumption. Therefore, our approach seeks to integrate the utilization of global information into the Convolutional Neural Network (CNN) framework in a more resource-efficient manner. In this paper, we introduce the Adaptive Center-Surround Receptive Field (ACSRF) network architecture inspired by the central-peripheral receptive field in biological vision for single-image haze removal. This leads to a unique receptive field mechanism that effectively combines both central and surrounding information. Our ACRSF addresses this by initially compressing global information and then merging it within the CNN, significantly boosting the capability to integrate local and global information, and effectively handling dominant color tones. Experimental results on four publicly available real-world image dehazing datasets show that our ACRSF outperforms current state-of-the-art methods in recovering global information, especially in dominant color tones. Importantly, this technology demonstrates its effectiveness in realistic scenarios, contributing significantly to improving traffic safety in adverse weather conditions. The code is available at https://github.com/JavanTang/ACSRF.

Innovative fault diagnosis for axial piston pumps: A physics-informed neural network framework predicting pump flow ripple

Innovative fault diagnosis for axial piston pumps: A physics-informed neural network framework predicting pump flow ripple

Multilayer network framework and metrics for table tennis analysis: Integrating network science, entropy, and machine learning

Multilayer network framework and metrics for table tennis analysis: Integrating network science, entropy, and machine learning

Self-knowledge Distillation Enhanced Universal Framework for Physics-Informed Neural Networks

Self-knowledge Distillation Enhanced Universal Framework for Physics-Informed Neural Networks

MUNet: a novel framework for accurate brain tumor segmentation combining UNet and mamba networks

Brain tumors are one of the major health threats to humans, and their complex pathological features and anatomical structures make accurate segmentation and detection crucial. However, existing models based on Transformers and Convolutional Neural Networks (CNNs) still have limitations in medical image processing. While Transformers are proficient in capturing global features, they suffer from high computational complexity and require large amounts of data for training. On the other hand, CNNs perform well in extracting local features but have limited performance when handling global information. To address these issues, this paper proposes a novel network framework, MUNet, which combines the advantages of UNet and Mamba, specifically designed for brain tumor segmentation. MUNet introduces the SD-SSM module, which effectively captures both global and local features of the image through selective scanning and state-space modeling, significantly improving segmentation accuracy. Additionally, we design the SD-Conv structure, which reduces feature redundancy without increasing model parameters, further enhancing computational efficiency. Finally, we propose a new loss function that combines mIoU loss, Dice loss, and Boundary loss, which improves segmentation overlap, similarity, and boundary accuracy from multiple perspectives. Experimental results show that, on the BraTS2020 dataset, MUNet achieves DSC values of 0.835, 0.915, and 0.823 for enhancing tumor (ET), whole tumor (WT), and tumor core (TC), respectively, and Hausdorff95 scores of 2.421, 3.755, and 6.437. On the BraTS2018 dataset, MUNet achieves DSC values of 0.815, 0.901, and 0.815, with Hausdorff95 scores of 4.389, 6.243, and 6.152, all outperforming existing methods and achieving significant performance improvements. Furthermore, when validated on the independent LGG dataset, MUNet demonstrated excellent generalization ability, proving its effectiveness in various medical imaging scenarios. The code is available at https://github.com/Dalin1977331/MUNet.

Moulding intergovernmental relations through tiao-kuai governance: dissecting the multi-scalar politics of intercity railway projects in China

ABSTRACT This article explores the dynamics of intergovernmental relations in China, specifically focusing on the tiao-kuai governance system and its role in regional infrastructural development, using intercity railway projects as a case study. The tiao-kuai system, characterised by vertical (tiao) and horizontal (kuai) governance, illustrates the complex interplay between central authority and local autonomy. Drawing on the Territory, Place, Scale and Network (TPSN) framework, this study analyses how multi-scalar politics shape regional competition and governance strategies, particularly in the Yangtze and Pearl River Delta intercity railway projects. Our findings highlight tensions between national priorities and local interests, as well as negotiation processes between various levels of government. This article underscores the importance of both top-down and bottom-up dynamics in regional development, unfolding how state spatial strategies are operationalised in a hybrid model of governance. This study contributes to the understanding of state space and intergovernmental collaboration in China’s large-scale infrastructure projects.

Retraction Note: Internet of things and distributed security framework for 5G networks

Retraction Note: Internet of things and distributed security framework for 5G networks

Coupled CANN-DEM simulation in solid mechanics

Abstract A general, unified neural network approach as replacement for the finite element method without the need for analytic expressions for material laws is suggested. The complete simulation process from the material characterization to simulations on a structural level takes place in the new neural network framework. The drawback of many conventional analytic expressions of material laws to require large numbers of experiments for parametrization is addressed by an integrated inverse approach.
Specifically, an adaptation of the Deep Energy Method (DEM) is combined with a Constitutive Artificial Neural Network (CANN) and trained on measured displacement fields and prescribed boundary conditions in a coupled procedure. Tests on compressible and incompressible Neo-Hookean solids with up to twelve CANN parameters show high accuracy of the approach and very good generalization of CANNs. A small extent of data is required for robust and reliable training.

RMD-Net: A Deep Learning Framework for Automated IHC Scoring of Lung Cancer IL-24

Immunohistochemical (IHC) detection is crucial in diagnosing lung cancer. Interleukin-24 (IL-24) is a valuable marker in IHC analysis, aiding in tumor characterization and prognostication. However, current manual scoring methods are labor-intensive, imprecise, and subjective, leading to inconsistencies among observers. Automated scoring methods also have limitations, such as poor segmentation and lack of interpretability. In this paper, we introduce RMD-Net, a novel scoring network framework specifically designed for IL-24 scoring in lung cancer. The framework incorporates a regional attention mechanism and a multi-channel scoring network. Initially, diagnostic region identification and segmentation are accomplished by integrating the diagnostic regional spatial attention module into the fully convolutional network. Subsequently, we employ the Adaptive Multi-Thresholding algorithm to derive expert, strong feature description maps. Finally, the attention-guided IHC images and expert feature description maps are fed into a multi-channel scoring network. Its backbone includes feature fusion layers and scoring layers to ensure the accuracy and interpretability of the final result. To the best of our knowledge, this is the first system that directly employs lung cancer IL-24 IHC images as input and combines both expert-derived features and deep-learning abstract features to produce clinical scores. Our dataset is sourced from the Institute of Life Sciences and Bioengineering at Beijing Jiaotong University. The experimental results demonstrate that the proposed method achieves an IL-24 score precision of 89.25%, an F1 score of 89.00, and an accuracy of 95.94%, outperforming other state-of-the-art methods. This contribution has the potential to advance clinical diagnosis and treatment strategies for lung cancer.

Transferable machine learning approach for predicting electronic structures of charged defects

The study of electronic properties of charged defects plays a crucial role in advancing our understanding of how defects influence conductivity, magnetism, and optical behavior in various materials. However, despite its significance, research on large-scale defective systems has been hindered by the high computational cost associated with density functional theory (DFT). In this study, we propose HamGNN-Q, an E(3) equivariant graph neural network framework capable of accurately predicting DFT Hamiltonian matrices for diverse point defects with varying charges, utilizing a unified set of network weights. By incorporating background charge features into the element representation, HamGNN-Q facilitates a direct mapping from structural characteristics and background charges to the electronic Hamiltonian matrix of charged defect systems, obviating the need for DFT calculations. We showcase the model's high precision and transferability by evaluating its performance on GaAs systems encompassing diverse charged defect configurations. Furthermore, we predicted the wave function distribution of polarons induced by defects. We analyzed the node features through principal component analysis, providing physical insights for the interpretability of the HamGNN-Q model. Our approach provides a practical solution for accelerating electronic structure calculations of neutral and charged defects and advancing the design of materials with tailored electronic properties.

Applying Deep Learning for Wildfire Identification: Economical and Accessible Solutions Leveraging Small Datasets

Wildfires significantly impact human health, air quality, visibility, weather, and climate change and cause substantial economic losses. While state and county-operated air quality monitors provide critical insights during wildfires, they are not available in all regions. This highlights the need for affordable, accessible tools that allow the general public to assess air quality impacts. In this study, we apply machine learning with deep neural networks to diagnose air quality rapidly from sky images taken at the Pacific Northwest National Laboratory in Richland, WA, USA. Using a convolutional neural network (CNN) framework, we trained a deep learning model to classify air quality indices based on sky images. By leveraging transfer learning, our approach fine-tunes a pre-trained model on a small dataset of sky images, significantly reducing training time while maintaining high accuracy. Our results demonstrate the potential of deep learning to provide rapid air quality diagnostics during wildfire episodes, offering early warnings to the public and enabling timely mitigation strategies, particularly for vulnerable populations. Additionally, we show that lower respiratory infections pose the highest health risk during acute smoke exposures. Reactive oxygen species (ROS) from wildfire particles further exacerbate health risks by triggering inflammation and other adverse effects.

Network Diffusion-Constrained Variational Generative Models for Investigating the Molecular Dynamics of Brain Connectomes Under Neurodegeneration

Alzheimer’s disease (AD) is a complex and progressive neurodegenerative condition with significant societal impact. Understanding the temporal dynamics of its pathology is essential for advancing therapeutic interventions. Empirical and anatomical evidence indicates that network decoupling occurs as a result of gray matter atrophy. However, the scarcity of longitudinal clinical data presents challenges for computer-based simulations. To address this, a first-principles-based, physics-constrained Bayesian framework is proposed to model time-dependent connectome dynamics during neurodegeneration. This temporal diffusion network framework segments pathological progression into discrete time windows and optimizes connectome distributions for biomarker Bayesian regression, conceptualized as a learning problem. The framework employs a variational autoencoder-like architecture with computational enhancements to stabilize and improve training efficiency. Experimental evaluations demonstrate that the proposed temporal meta-models outperform traditional static diffusion models. The models were evaluated using both synthetic and real-world MRI and PET clinical datasets that measure amyloid beta, tau, and glucose metabolism. The framework successfully distinguishes normative aging from AD pathology. Findings provide novel support for the “decoupling” hypothesis and reveal eigenvalue-based evidence of pathological destabilization in AD. Future optimization of the model, integrated with real-world clinical data, is expected to improve applications in personalized medicine for AD and other neurodegenerative diseases.

Multi-block Directed Radiation Routing Algorithm for Optimized Wireless Sensor Network

In recent years, the axial radiation model has emerged as a pivotal framework for Wireless Sensor Networks (WSN), particularly in enhancing intelligent sensing platforms. This study delves into the WSN structured around the axial radiation model, encompassing critical aspects like model reconstruction, node deployment, and routing optimization. Our focus is on evaluating the performance metrics that influence the responsiveness of these intelligent platforms. We introduce the Multi-block Directed Radiation (MBDR) routing algorithm, designed to extend system operational time and boost data transmission efficiency. Comprehensive experimental analyses demonstrate the significant advancements of MBDR in survivability, data transmission rates, and regional balance within central axis radiation model environments.

Retinal OCT image classification based on MGR-GAN.

Accurately classifying optical coherence tomography (OCT) images is essential for diagnosing and treating ophthalmic diseases. This paper introduces a novel generative adversarial network framework called MGR-GAN. The masked image modeling (MIM) method is integrated into the GAN model's generator, enhancing its ability to synthesize more realistic images by reconstructing them based on unmasked patches. A ResNet-structured discriminator is employed to determine whether the image is generated by the generator. Through the unique game process of the generative adversarial network (GAN) model, the discriminator acquires high-level discriminant features, essential for precise OCT classification. Experimental results demonstrate that MGR-GAN achieves a classification accuracy of 98.4% on the original UCSD dataset. As the trained generator can synthesize OCT images with higher precision, and owing to category imbalances in the UCSD dataset, the generated OCT images are leveraged to address this imbalance. After balancing the UCSD dataset, the classification accuracy further improves to 99%.

Enhancing waste classification accuracy with Channel and Spatial Attention-Based Multiblock Convolutional Network.

Municipal waste classification is significant for effective recycling and waste management processes that involve the classification of diverse municipal waste materials such as paper, glass, plastic, and organic matter using diverse techniques. Yet, this municipal waste classification process faces several challenges, such as high computational complexity, more time consumption, and high variability in the appearance of waste caused by variations in color, type, and degradation level, which makes an inaccurate waste classification process. To overcome these challenges, this research proposes a novel Channel and Spatial Attention-Based Multiblock Convolutional Network for accurately classifying municipal waste that utilizes a unique attention mechanism for enhancing feature learning and waste classification accuracy. In this research, the data augmentation technique is utilized to improve the size and diversity of the images, which creates a new municipal waste image from the existing image. After the augmentation process, the data preprocessing is performed through normalization, resizing, and dividing the images into smaller patches. In the feature extraction phase, each image patch's detailed representation is created by integrating and extracting the image's relevant features from the embedding space. Finally, the proposed Channel and Spatial Attention-Based Multiblock Convolutional Network predicts the types of municipal waste that are represented in each image patch by classifying the input images into diverse kinds of waste categories. The experimental validation exposed that the proposed Channel and Spatial Attention-Based Multiblock Convolutional Network effectively classifies the municipal waste images and achieves a higher accuracy of 98.73%, lower mean absolute error of 0.048, and lower root mean square error of 0.087 when compared to existing municipal waste classification strategies. These results prove that the proposed Channel and Spatial Attention-Based Multiblock Convolutional Network framework is more accurate, reliable, and well-suited to real-time municipal waste classification.

Efficient Estimation for Longitudinal Networks via Adaptive Merging

Longitudinal networks consist of sequences of temporal edges among multiple nodes, where the temporal edges are observed in real-time. They have become ubiquitous with the rise of online social platforms and e-commerce, but largely under-investigated in the literature. In this paper, we propose an efficient estimation framework for longitudinal networks, leveraging strengths of adaptive network merging, tensor decomposition, and point processes. It merges neighboring sparse networks so as to enlarge the number of observed edges and reduce estimation variance, whereas the estimation bias introduced by network merging is controlled by exploiting local temporal structures for adaptive network neighborhood. A projected gradient descent algorithm is proposed to facilitate estimation, where the upper bound of the estimation error in each iteration is established. A thorough analysis is conducted to quantify the asymptotic behavior of the proposed method, which shows that it can significantly reduce the estimation error and also provides a guideline for network merging under various scenarios. We further demonstrate the advantage of the proposed method through extensive numerical experiments on synthetic datasets and a militarized interstate dispute dataset.

Feature Enhancement-Based Few-Shot Bearing Surface Defect Image Classification Method

The application of intelligent bearing surface defect classification based on deep neural networks remains challenging in real factories, due to the scarcity of defect samples. Under real-working conditions, with less training samples, the paper proposes a few-shot bearing defect image classification network which can recognize different bearing surface defects image, including notch, reddish rust, scratching, incising, conformity, pitting and mill scale. Based on general metric learning neural network framework, a local feature extraction layer is designed, which calculates the auto-correlation vector of global feature in a sliding region to enhance detail features. Additionally, a similar feature attention module emphasizes the the regions of similarity between the query set and the class prototype center to overcome the influence of background noise on classification. To validate the effectiveness of the proposed network, comparative experiments were conducted using the benchmark dataset miniImageNet, achieving classification accuracies of 59% in the 5-way 1-shot setting and 76% in the 5-way 5-shot setting respectively. Furthermore, to assess its performance in a real-factory condition, a self-made dataset of bearing defects from a factory was employed. The proposed network achieved a remarkable classification accuracy of 88% in the 5-way 5-shot setting. These experimental results confirm the practical application value of our few-shot bearing surface defect image classification network, demonstrating its ability to accurately recognize various bearing defects with limited training samples.

Estimating the contribution of Arctic sea-ice loss to central Asia temperature anomalies: The case of winter 2020/2021

Abstract Arctic sea-ice cover has declined drastically in recent decades, notably in the Barents-Kara Seas. Previous research has linked low autumn Barents-Kara sea-ice cover to subsequent cold Eurasian winters. This lagged relationship was observed in 2020/2021. Using a causal network framework grounded in known physical mechanisms, we assess how strongly one factor directly influences another (i.e., the causal relationship) and apply this analysis to the anomalies from 2020/2021. We show that although year-to-year Barents-Kara sea-ice variations have only a minor impact on Central Asia temperatures, that causal effect becomes important for attribution and prediction when we consider the large long-term trend in Barents-Kara sea-ice cover. In particular, we find that Central Asia’s 2021 negative winter surface air temperature anomaly (relative to 1980-2020) can be fully explained by the Barents-Kara sea-ice anomaly. We further estimate that Barents-Kara sea-ice loss has more than halved the winter warming over Central Asia over the past 40 years. Hence rather than debating a cooling trend in Central Asia during winter, we propose shifting the focus to the influence of Arctic Amplification on anticipated warming trends. This study illustrates the efficacy of causal network analysis, which has implications for seasonal prediction and attribution of midlatitude winter anomalies.

Evolution of sample-based music authorship network

Sample-based music—characterized by the adoption of extant audio fragments (sampling) in its creation process—plays an essential role in contemporary popular music, fostering inter-generational connections between the creators that have resulted in a rich and diverse sonic landscape. The selection, manipulation, and adoption of samples heavily impact the genre, mood, texture, and finally the distinctive identity of a new musical composition. One could call the samples “cultural genes” of sorts, continually incorporated into new music to contribute to its characteristics. One can then ask how this process has taken place in history and shaped the history of contemporary popular music, which we study in this work. We specifically study the evolution of sample-based music between the years 1980 and 2019, taking cue from the citation network analysis of academic literature that intuitively follows a similar dynamic of the flow of ideas and material from the old works into new ones. First, the community structure in the artist–sample network is identified, and its relationship to distinctive musical styles and flavors is verified. A longitudinal analysis of the passing down of musical styles is then performed based on similarity between communities of distinct eras, identifying continuous temporal developments in music as well as instances of the revival of styles dormant across multiple generations, akin to “genetic atavism.” This study demonstrates the complex nature of cultural evolution using a network framework that is also generally applicable to other creative enterprises.

Combined effect of incentives and coupling in multigames in two-layer networks

Abstract The lack of cooperation can easily result in inequality among members of a society, which provides an increasing gap between individual incomes. To tackle this issue, we introduce an incentive mechanism based on individual strategies and incomes, wherein a portion of the income from defectors is allocated to reward low-income cooperators, aiming to enhance cooperation by improving the equitable distribution of wealth across the entire population. Moreover, previous research has typically employed network structures or game mechanisms characterized by homogeneity. In this study, we present a network framework that more accurately reflects real-world conditions, where agents are engaged in multiple games, including prisoner’s dilemma games in the top-layer and public good games in the down-layer networks. Within this framework, we introduce the concept of ‘external coupling’ which connects agents across different networks as acquaintances, thereby facilitating access to shared datasets. Our results indicate that the combined positive effects of external coupling and incentive mechanism lead to optimal cooperation rates and lower Gini coefficients, demonstrating a negative correlation between cooperation and inequality. From a micro-level perspective, this phenomenon primarily arises from the regular network, whereas suboptimal outcomes are observed within the scale-free network. These observations help to give a deeper insight into the interplay between cooperation and wealth disparity in evolutionary games in large populations.