Multi-level Network Research Articles

Study on the evolution of multi-level collaborative innovation networks in China’s cloud manufacturing industry

ABSTRACT CMfg is a new service-oriented manufacturing mode integrating new-generation information technologies, and its development promotes the digital transformation of China’s manufacturing industry. Exploring the evolution trend of CMfg's co-innovation networks is essential to the development of manufacturing digitalisation. The research is carried out from three levels: The evolutionary characteristics of the organisational co-innovation network in the CMfg industry are identified through SNA and THA; the evolution of technological co-innovation network is analysed by using FP-Growth and the textual analysis, to recognise the technology convergence trends; the evolution of the factors affecting regional collaborations are discovered by using QAP model to analyse the regional co-innovation network. Following conclusions are achieved: CMfg co-innovation is centred on state-owned enterprises in the power and telecommunication industries, with University–Industry collaboration as the main form; the current hot spots of CMfg technologies are the Resource Perception-and-Access, the Resource Virtualisation-and-Service; the regional collaboration in China's CMfg industry tends to be partnership-led and multi-polar. It suggests that: In the CMfg industry, more enterprises-supporting policy should be made, while the Virtualisation-and-Service technologies need more attention, and the partnerships among regions should be further enhanced.

A Multi-Level Adaptive Lightweight Net for Damaged Road Marking Detection Based on Knowledge Distillation

To tackle the complexity and limited applicability of high-precision segmentation models for damaged road markings, this study proposes a Multi-level Adaptive Lightweight Network (MALNet) based on knowledge distillation. By incorporating multi-scale dilated convolution and adaptive spatial channel attention fusion modules, the MALNet model significantly enhances the precision, integrity, and robustness of its segmentation branch. Furthermore, it employs an intricate knowledge distillation strategy, channeling rich, layered insights from a teacher model to a student model, thus elevating the latter’s segmentation ability. Concurrently, it streamlines the student model by markedly reducing its parameter count and computational demands, culminating in a segmentation network that is both high-performing and pragmatic. Rigorous testing on three distinct data sets for damaged road marking detection—CDM_P (Collective Damaged road Marking—Public), CDM_H (Collective Damaged road Marking—Highways), and CDM_C (Collective Damaged road Marking—Cityroad)—underscores the MALNet model’s superior segmentation abilities across all damage types, outperforming competing models in accuracy and completeness. Notably, the MALNet model excels in parameter efficiency, computational economy, and throughput. After distillation, the student model’s parameters and computational load decrease to only 31.78% and 27.40% of the teacher model’s, respectively, while processing speeds increase to 1.9 times, demonstrating a significant improvement in lightweight design.

Quantitative analysis of grain size effect on tensile mechanical behavior of granite based on multi-level force chain networks

Quantitative analysis of grain size effect on tensile mechanical behavior of granite based on multi-level force chain networks

Multi-Level Attention with 2D Table-Filling for Joint Entity-Relation Extraction

Joint entity-relation extraction is a fundamental task in the construction of large-scale knowledge graphs. This task relies not only on the semantics of the text span but also on its intricate connections, including classification and structural details that most previous models overlook. In this paper, we propose the incorporation of this information into the learning process. Specifically, we design a novel two-dimensional word-pair tagging method to define the task of entity and relation extraction. This allows type markers to focus on text tokens, gathering information for their corresponding spans. Additionally, we introduce a multi-level attention neural network to enhance its capacity to perceive structure-aware features. Our experiments show that our approach can overcome the limitations of earlier tagging methods and yield more accurate results. We evaluate our model using three different datasets: SciERC, ADE, and CoNLL04. Our model demonstrates competitive performance compared to the state-of-the-art, surpassing other approaches across the majority of evaluated metrics.

Metabolic picture of microbial interaction: chemical crosstalk during co-cultivation between three dominant genera of bacteria and fungi in medicinal plants rhizosphere.

Microbial communities affect several aspects of the earth's ecosystem through their metabolic interaction. The dynamics of this interaction emerge from complex multilevel networks of crosstalk. Elucidation of this interaction could help us to maintain the balance for a sustainable future. To investigate the chemical language among highly abundant microbial genera in the rhizospheres of medicinal plants based on the metabolomic analysis at the interaction level. Coculturing experiments involving three microbial species: Aspergillus (A), Trichoderma (T), and Bacillus (B), representing fungi (A, T) and bacteria (B), respectively. These experiments encompassed various interaction levels, including dual cultures (AB, AT, TB) and triple cultures (ATB). Metabolic profiling by LC-QTOFMS revealed the effect of interaction level on the productivity and diversity of microbial specialized metabolites. The ATB interaction had the richest profile, while the bacterial profile in the monoculture condition had the lowest. Two native compounds of the Aspergillus genus, aspergillic acid and the dipeptide asperopiperazine B, exhibited decreased levels in the presence of the AT interaction and were undetectable in the presence of bacteria during the interaction. Trichodermarin N and Trichodermatide D isolated from Trichoderma species exclusively detected during coexistence with bacteria (TB and ATB). These findings indicate that the presence of Bacillus activates cryptic biosynthetic gene clusters in Trichoderma. The antibacterial activity of mixed culture extracts was stronger than that of the monoculture extracts. The TB extract exhibited strong antifungal activity compared to the monoculture extract and other mixed culture treatments. The elucidation of medicinal plant microbiome interaction chemistry and its effect on the environment will also be of great interest in the context of medicinal plant health Additionally, it sheds light on the content of bioactive constituents,and facilitating the discovery of novel antimicrobials.

Tackling Few-Shot Challenges in Automatic Modulation Recognition: A Multi-Level Comparative Relation Network Combining Class Reconstruction Strategy.

Automatic Modulation Recognition (AMR) is a key technology in the field of cognitive communication, playing a core role in many applications, especially in wireless security issues. Currently, deep learning (DL)-based AMR technology has achieved many research results, greatly promoting the development of AMR technology. However, the few-shot dilemma faced by DL-based AMR methods greatly limits their application in practical scenarios. Therefore, this paper endeavored to address the challenge of AMR with limited data and proposed a novel meta-learning method, the Multi-Level Comparison Relation Network with Class Reconstruction (MCRN-CR). Firstly, the method designs a structure of a multi-level comparison relation network, which involves embedding functions to output their feature maps hierarchically, comprehensively calculating the relation scores between query samples and support samples to determine the modulation category. Secondly, the embedding function integrates a reconstruction module, leveraging an autoencoder for support sample reconstruction, wherein the encoder serves dual purposes as the embedding mechanism. The training regimen incorporates a meta-learning paradigm, harmoniously combining classification and reconstruction losses to refine the model's performance. The experimental results on the RadioML2018 dataset show that our designed method can greatly alleviate the small sample problem in AMR and is superior to existing methods.

MST: Topology-Aware Message Aggregation for Exascale Graph Processing of Traversal-Centric Algorithms

This paper presents MST, a communication-efficient message library for fast graph traversal on exascale clusters. The key idea is to follow the multi-level network topology to perform topology-aware message aggregation, where small messages are gathered and scattered at each level of domain. To facilitate message aggregation, we equip MST with flexible buffer management including active buffer switching and dynamic buffer expansion. We implement MST on the newest-generation Tianhe supercomputer and evaluated its performance using various traversal-centric algorithms on both synthetic trillion-scale graphs and real-world big graphs. The results show that MST-based graph traversal is orders of magnitude faster than that based on Active Messages Library (AML). For the Graph500-BFS benchmark, MST-based Tianhe (with 77.2K nodes) outperforms the Fugaku supercomputer (with 148.5K nodes) by 18.53%, while Fugaku is ranked No. 1 in the latest Graph500-BFS ranking (June 2023). MST also greatly improves graph processing performance on other commercial large-scale computing systems at the National Supercomputing Center in Changsha (NSCC) and WuzhenLight.

Power Quality Transient Disturbance Diagnosis Based on Dynamic Large Convolution Kernel and Multi-Level Feature Fusion Network

Power quality is an important metric for the normal operation of a power system, and the accurate identification of transient signals is of great significance for the improvement of power quality. The diverse types of power system transient signals and strong characteristic coupling brings new challenges to the analysis and identification of power system transient signals. In order to enhance the identification accuracy of transient signals, one method of power system transient signal identification is proposed based on a dynamic large convolution kernel and multilevel feature fusion network. First, the more fine-grained and more informative features of the transient signals are extracted by the dynamic large convolution kernel feature extraction module. Then, the multi-scale local features are adaptively fused by the multilevel feature fusion module. Finally, the fused features are reduced in dimension by the fully connected layer in the classification module and fed into the SoftMax layer for transient signal type detection. The proposed method can effectively improve the small receptive field problem of convolutional neural networks and the lack of ability of Transformer network in extracting local context information. Compared with five other power quality transient disturbance identification models, the experimental results show that the proposed method has better diagnostic accuracy and anti-noise capability.

Common functional mechanisms underlying dynamic brain network changes across five general anesthetics: A rat fMRI study.

Reversible loss of consciousness is the primary therapeutic endpoint of general anesthesia; however, the drug-invariant mechanisms underlying anesthetic-induced unconsciousness are still unclear. This study aimed to investigate the static, dynamic, topological and organizational changes in functional brain network induced by five clinically-used general anesthetics in the rat brain. Male Sprague-Dawley rats (n = 57) were randomly allocated to received propofol, isoflurane, ketamine, dexmedetomidine, or combined isoflurane plus dexmedetomidine anesthesia. Resting-state functional magnetic resonance images were acquired under general anesthesia and analyzed for changes in dynamic functional brain networks compared to the awake state. Different general anesthetics induced distinct patterns of functional connectivity inhibition within brain-wide networks, resulting in multi-level network reorganization primarily by impairing the functional connectivity of cortico-subcortical networks as well as by reducing information transmission capacity, intrinsic connectivity, and network architecture stability of subcortical regions. Conversely, functional connectivity and topological properties were preserved within cortico-cortical networks, albeit with fewer dynamic fluctuations under general anesthesia. Our findings highlighted the effects of different general anesthetics on functional brain network reorganization, which might shed light on the drug-invariant mechanism of anesthetic-induced unconsciousness.

Multi-level attention fusion network assisted by relative entropy alignment for multimodal speech emotion recognition

Multi-level attention fusion network assisted by relative entropy alignment for multimodal speech emotion recognition

Few-shot fault diagnosis of switch machine based on data fusion and balanced regularized prototypical network

The turnout switch machine (TSM) is the critical signal equipment of the interlocking system, which directly affects the efficiency and safety of rail transit. However, the incomplete feature information of single-source data and the scarcity of fault data in real scenes make the existing methods unable to obtain satisfactory diagnostic results. Given the above issues, a few-shot fault diagnosis method based on data fusion and balanced regularized prototypical network (BRPN) is proposed. Firstly, a multi-signal adaptive fusion strategy is proposed to adaptively fuse three currents of the TSM to achieve a comprehensive and accurate expression of fault information. Secondly, a multi-level feature fusion network is designed to enhance the feature extraction ability of the fusion current signal. Furthermore, a prototype distribution balanced regularized strategy is proposed to balance the distribution differences between prototype interclass, so as to improve the diagnostic performance of the BRPN. Finally, the proposed BRPN model is verified by data fusion experiments and different small sample fusion datasets. Compared with other methods, the proposed method shows satisfactory diagnostic results in 1-shot and 5-shot experiments with two conversion processes, and the highest accuracy of fault diagnosis reaches 99.60%, which can provide a solution for solving the scarce fault samples of switch machines in a real-world situation.

MFFC-SDN: multi-level feature fusion codec-based ship detection network in SAR images

ABSTRACT In recent years, with the development of deep learning, SAR image-based ship target-detection technology has become a current research hotspot. SAR images are characterized by complex backgrounds, significant speckle noise interference, and poor interpretability. To address this challenge, this study proposes a new Gaussian circle radius determination strategy and a multi-level feature fusion codec-based ship detection network (MFFC-SDN) for SAR images. Differing from CenterNet’s corner-oriented Gaussian circle strategy, the Gaussian circle strategy proposed in this paper takes the centre point coordinates of the real target as the origin to define a circular region. This approach allows for more precise supervision of the training model. Meanwhile, we propose MFFC-SDN to solve the problem of the poor performance of CenterNet in SAR ship detection. MFFC-SDN is an anchor-free, single-stage target-detection method with a low computation cost. In MFFC-SDN, we introduce a Multilevel Feature Recapture Codec Module (MFR-CM) to enhance the feature extraction capability of the network by using the features of the encoding stage twice because of the easy loss of image features in the lengthy codec process of the CenterNet network backbone. A Residual Attention Pyramid Feature Fusion Module (RA-PFFM) is designed to enhance the feature map and obtain multi-scale features. Experimental results on the SSDD and SAR-ship-dataset show that MFFC-SDN significantly enhances SAR image target-detection performance, effectively addressing detection challenges related to large target size variations and complex environmental conditions. Compared with existing algorithms, MFFC-SDN achieves the highest A P 50 .

DMF2Net: Dynamic multi-level feature fusion network for heterogeneous remote sensing image change detection

With the rapid development of remote sensing data fusion technology, heterogeneous remote sensing image (HRSI) change detection (CD) has become a frontier field. The powerful nonlinear information processing capability of deep neural network provides the possibility for image domain conversion of HRSIs. However, the majority of existing methods rely on stacking network layers to extract deep high-level semantic features to accomplish the mutual conversion of image domains. The effective extraction and utilization of shallow features have not been fully considered. At the same time, the dependence between deep and shallow features has not been deeply explored. Therefore, these algorithms severely restrict the performance of CD. A new HRSI-CD network (DMF2Net) based on multi-level feature fusion is proposed in this paper to address these issues. DMF2Net uses central difference convolution to extract fine-grained features from shallow layers of images. It is capable of capturing the intrinsic detail features of image by aggregating intensity and gradient information. The dynamic multi-level feature fusion method is used to learn the fusion weights from the features, which are then used to guide the organic fusion of shallow and deep semantic features. This can preserve more position and detail features of the image, which can prevent the loss of small data during image conversion and enhance the model’s ability to detect subtle changes. On this basis, a new DMF2Net-based method for detecting change in bi-temporal HRSIs is proposed. We conducted extensive experiments on four publicly available HRSI-CD datasets. The experimental results showed that the proposed CD framework achieves significant improvement over the most advanced methods. The project file of the proposed framework will be obtained from https://github.com/cwlnnu/DMF2Net.

Unsupervised domain adaptation multi-level adversarial learning-based crossing-domain retinal vessel segmentation

BackgroundThe retinal vasculature, a crucial component of the human body, mirrors various illnesses such as cardiovascular disease, glaucoma, and retinopathy. Accurate segmentation of retinal vessels in funduscopic images is essential for diagnosing and understanding these conditions. However, existing segmentation models often struggle with images from different sources, making accurate segmentation in crossing-source fundus images challenging. MethodsTo address the crossing-source segmentation issues, this paper proposes a novel Multi-level Adversarial Learning and Pseudo-label Denoising-based Self-training Framework (MLAL&PDSF). Expanding on our previously proposed Multiscale Context Gating with Breakpoint and Spatial Dual Attention Network (MCG&BSA-Net), MLAL&PDSF introduces a multi-level adversarial network that operates at both the feature and image layers to align distributions between the target and source domains. Additionally, it employs a distance comparison technique to refine pseudo-labels generated during the self-training process. By comparing the distance between the pseudo-labels and the network predictions, the framework identifies and corrects inaccuracies, thus enhancing the accuracy of the fine vessel segmentation. ResultsWe have conducted extensive validation and comparative experiments on the CHASEDB1, STARE, and HRF datasets to evaluate the efficacy of the MLAL&PDSF. The evaluation metrics included the area under the operating characteristic curve (AUC), sensitivity (SE), specificity (SP), accuracy (ACC), and balanced F-score (F1). The performance results from unsupervised domain adaptive segmentation are remarkable: for DRIVE to CHASEDB1, results are AUC: 0.9806, SE: 0.7400, SP: 0.9737, ACC: 0.9874, and F1: 0.8851; for DRIVE to STARE, results are AUC: 0.9827, SE: 0.7944, SP: 0.9651, ACC: 0.9826, and F1: 0.8326. ConclusionThese results demonstrate the effectiveness and robustness of MLAL&PDSF in achieving accurate segmentation results from crossing-domain retinal vessel datasets. The framework lays a solid foundation for further advancements in cross-domain segmentation and enhances the diagnosis and understanding of related diseases.

Multi-level Symmetric Semantic Alignment Network for image–text matching

Image–text matching has attracted much attention as one of the visual-linguistic tasks. Most of the existing methods tend to concentrate on single-level semantic similarity by global embeddings or local alignment, which fail to distinguish the ambiguous image–text instances with highly similar contexts due to insufficient interactions of cross-modal features. To mitigate this problem, in this paper, we propose a novel Multi-level Symmetric Semantic Alignment Network (MSSAN), which fully exploits the multi-level semantic features to capture the complicated correlations between images and text. Specifically, global–global-level alignment is first performed based on visual–textual global representations. Then, considering the local semantic consistency, local–local-level alignment is carried out through inter-modal bidirectional cross-attention module to model fine-grained region–word relations. Moreover, a Multi-granularity Feature Fusion Module (MFFM) is constructed to learn intra-modal distribution of significance and incorporate global concepts into local features to obtain more comprehensive semantic representations, thus achieving global–local-level alignment. Finally, in order to achieve a strong separation of semantically similar samples, we develop a novel generic triplet ranking loss with adaptively updatable margins to train our model. Extensive experimental results on Flickr8K, MSCOCO and Flickr30K datasets demonstrate that our proposed MSSAN is superior to other state-of-the-art methods by a considerable margin.

Local Network Interaction as a Mechanism for Wealth Inequality

Given limited institutional resources, low-income populations often rely on social networks to improve their socioeconomic outcomes. However, it remains in question whether small-scale social interactions could affect large-scale economic inequalities in under-resourced contexts. Here, we leverage population-level data from one of the poorest South African settings to construct a large-scale, geographically defined, inter-household social network. Using a multilevel network model, we show that having social ties in close geographic proximity is associated with stable household asset conditions, while geographically distant ties correlate to changes in asset allocation. Notably, we find that localised network interactions are associated with an increase in wealth inequality at the regional level, demonstrating how macro-level inequality may arise from micro-level social processes. Our findings highlight the importance of understanding complex social connections underpinning inter-household resource dynamics, and raise the potential of large-scale social assistance programs to reduce disparities in resource-ownership by accounting for local social constraints.

A novel multi-level reverse logistics network design optimization model for waste batteries considering facility technology types

The increasing prevalence of electric vehicles (EVs) will lead to a continuous rise in the quantity of waste batteries in the future. The topic regarding reverse logistics network design (RLND) for waste batteries has garnered extensive attention from both the academic and societal realms. However, relevant research in this area has yet to delve deeper. This research proposes a mixed-integer linear programming (MILP) model to optimize the multi-level reverse logistics network (RLN) of waste batteries from the perspective of a circular economy. The model aims to minimize costs as an economic objective and to reduce carbon emissions as an environmental objective. It introduces an expected weighting factor to balance these two conflicting goals. Various technological levels are considered for different processing facilities, thereby enhancing the sustainability of the RLN. The model is applied to a numerical case study in the Nanjing Metropolitan Area of China, obtaining optimal network facility location schemes, technological configuration plans, and transportation schemes between facilities. The study emphasizes the significance of selecting different technology types for facilities, as rational process configurations effectively reduce recycling operational costs and decrease network carbon emissions. Sensitivity analysis emphasizes the impact of uncertain factors on RLND, highlighting the imperative need for a forward-thinking approach to ensure long-term sustainable development for enterprises. Managerial insights aim to guide enterprises and government departments, encouraging the adoption of higher technology types to achieve the dual goals of increased revenue and reduced carbon emissions.

Remote Sensing Image Classification Based on Canny Operator Enhanced Edge Features.

Remote sensing image classification plays a crucial role in the field of remote sensing interpretation. With the exponential growth of multi-source remote sensing data, accurately extracting target features and comprehending target attributes from complex images significantly impacts classification accuracy. To address these challenges, we propose a Canny edge-enhanced multi-level attention feature fusion network (CAF) for remote sensing image classification. The original image is specifically inputted into a convolutional network for the extraction of global features, while increasing the depth of the convolutional layer facilitates feature extraction at various levels. Additionally, to emphasize detailed target features, we employ the Canny operator for edge information extraction and utilize a convolution layer to capture deep edge features. Finally, by leveraging the Attentional Feature Fusion (AFF) network, we fuse global and detailed features to obtain more discriminative representations for scene classification tasks. The performance of our proposed method (CAF) is evaluated through experiments conducted across three openly accessible datasets for classifying scenes in remote sensing images: NWPU-RESISC45, UCM, and MSTAR. The experimental findings indicate that our approach based on incorporating edge detail information outperforms methods relying solely on global feature-based classifications.

A Multi-Level Deep Neural Network-Based Tourism Supply Chain Risk Management Study

With the rapid advancement of the tourism, the capital demand of tourism enterprises has gradually risen, but the confusion of market management has increased the difficulty of risk assessment of tourism enterprises by financial institutions, which has led to difficulties in financing tourism enterprises and seriously hindered their development. On this basis, this research first analyses the risk structure of the tourism supply chain (TSC) from the financial aspect and establishes a relevant risk assessment system. After confirming the assessment indexes, the data is dativized and normalized, and finally a multi-level deep neural network (DNN) is used to construct a TSC risk prediction model to calculate the transformed indexes and assess the risk degree of the enterprise according to the results. The experimental results indicate that the model has the best performance when the H-Net hidden layer is 3 layers and the L-Net hidden layer is 4 layers, and its accuracy reaches 93.35%, sensitivity reaches 84.13%, convergence starts at 25 iterations, the final loss value is only 0.8, and the predicted and the real value error is within 2.5%. Therefore, the multi-level DNN model constructed in this experiment has certain application value in TSC risk management.

Spatial-temporal memory enhanced multi-level attention network for origin-destination demand prediction

Origin-destination demand prediction is a critical task in the field of intelligent transportation systems. However, accurately modeling the complex spatial-temporal dependencies presents significant challenges, which arises from various factors, including spatial, temporal, and external influences such as geographical features, weather conditions, and traffic incidents. Moreover, capturing multi-scale dependencies of local and global spatial dependencies, as well as short and long-term temporal dependencies, further complicates the task. To address these challenges, a novel framework called the Spatial-Temporal Memory Enhanced Multi-Level Attention Network (ST-MEN) is proposed. The framework consists of several key components. Firstly, an external attention mechanism is incorporated to efficiently process external factors into the prediction process. Secondly, a dynamic spatial feature extraction module is designed that effectively captures the spatial dependencies among nodes. By incorporating two skip-connections, this module preserves the original node information while aggregating information from other nodes. Finally, a temporal feature extraction module is proposed that captures both continuous and discrete temporal dependencies using a hierarchical memory network. In addition, multi-scale features cascade fusion is incorporated to enhance the performance of the proposed model. To evaluate the effectiveness of the proposed model, extensively experiments are conducted on two real-world datasets. The experimental results demonstrate that the ST-MEN model achieves excellent prediction accuracy, where the maximum improvement can reach to 19.1%.