Multistage Networks Research Articles

An Internet of Things Task Scheduling Framework Based on Agile Virtual Network on Demand Service Model

In order to improve the efficiency of cloud computing resource utilization and avoid the problem of computing resource allocation and scheduling lagging behind load changes, the author proposes a cloud computing resource on-demand allocation and elastic scheduling method based on network load prediction. Firstly, the author takes the network load data of Wikimedia as the research object and proposes an adaptive two-stage multi network model load prediction method based on LSTM, i.e. ATSMNN-LSTM load prediction method). This method can classify the network load data into climbing and descending types based on the trend and characteristics of the input network load data, And adaptively schedule the input network load data to the LSTM load prediction model that matches its type for prediction based on the classification results. The author proposes a maximum cloud service revenue computing resource quantity search algorithm based on network load prediction (i.e. MaxCSPRNWP algorithm), which aims to improve cloud service revenue as the optimization objective. Under the premise of ensuring task service quality and system stability, the algorithm allocates cloud computing resources on demand and flexibly schedules them in advance based on the predicted network load results. The experimental results show that the ATSMNNLSTM load prediction method proposed by the author can obtain more accurate network load prediction results compared to other load prediction methods, and the MaxCSPR-NWP algorithm, which is based on network load prediction and is capable of effectively converting the network load prediction results into the required number of cloud servers, is the maximum cloud service revenue computing resource quantity search algorithm proposed by the author, not only does it achieve the early allocation and scheduling of cloud computing resources, thereby avoiding the impact of lagging behind in computing resource allocation and scheduling due to load changes on the quality of cloud computing task services and resource utilization efficiency, at the same time, it has also achieved on-demand allocation and flexible scheduling of cloud computing resources with the goal of improving cloud service revenue.

S2VSNet: Single stage V-shaped network for image deraining & dehazing

Producing high-quality, noise-free images from noisy or hazy inputs relies on essential tasks such as single image deraining and dehazing. In many advanced multi-stage networks, there is often an imbalance in contextual information, leading to increased complexity. To address these challenges, we propose a simplified method inspired by a U-Net structure, resulting in the “Single-Stage V-Shaped Network” (S2VSNet), capable of handling both deraining and dehazing tasks. A key innovation in our approach is the introduction of a Feature Fusion Module (FFM), which facilitates the sharing of information across multiple scales and hierarchical layers within the encoder-decoder structure. As the network progresses towards deeper layers, the FFM gradually integrates insights from higher levels, ensuring that spatial details are preserved while contextual feature maps are balanced. This integration enhances the image processing capability, producing noise-free, high-quality outputs. To maintain efficiency and reduce system complexity, we replaced or removed several non-essential non-linear activation functions, opting instead for simple multiplication operations. Additionally, we introduced a “Multi-Head Attention Integrated Module” (MHAIM) as an intermediary layer between encoder-decoder levels. This module addresses the limited receptive fields of traditional Convolutional Neural Networks (CNNs), allowing for the capture of more comprehensive feature-map information. Our focus on deraining and dehazing led to extensive experiments on a wide range of synthetic and real-world datasets. To further validate the robustness of our network, we implemented S2VSNet on a low-end edge device, achieving deraining in 2.46 seconds.

MSNet: A Multistage Network for Lightweight Image Dehazing with Content-Guided Attention and Adaptive Encoding

MSNet: A Multistage Network for Lightweight Image Dehazing with Content-Guided Attention and Adaptive Encoding

Design of ant colony planning algorithms for multi-stage grids of distribution networks considering network resilience

Network Destruction Resilience is the ability of a system to maintain good operation in the event of an external attack or internal failure. The resilience of distribution network is crucial to guarantee the reliability of power supply. In this paper, we design a planning algorithm considering network destructibility for the multi-stage ant colony planning problem of distribution networks. The method establishes the multi-stage network planning objective function of distribution network from the perspectives of total investment cost and annual operation cost of multi-node network frame, destruction resistance, active power and reactive network loss of distribution network, etc. Then based on the constraints of the model, the improved ant colony algorithm is used to solve the multi-stage network planning objective function of distribution network, and the results of the anti-termite colony planning of multi-stage network frame of distribution network are obtained. In order to verify the effectiveness of the algorithm, simulation experiments are carried out on real distribution network data. The results show that the proposed ant colony planning algorithm for multi-stage grid frames can effectively improve the destruction resistance of distribution networks, and reduce the total investment cost and annual operation cost of multi-stage grid frames, and reduce the network loss rate data of multi-stage grid frames of distribution networks after application. It provides an effective method for planning the distribution network.

Optimal design of a post-demolition autoclaved aerated concrete (AAC) recycling network using a capacitated, multi-period, and multi-stage warehouse location problem

Autoclaved aerated concrete is a popular building material in constructing one- and two-family houses because of its low thermal conductivity and fire resistance. Since autoclaved aerated concrete production rose significantly in the 1960s and 1970s, increasing post-demolition volumes can be expected in the following decades. However, these are currently landfilled as high-quality recycling options are still to be established.This study develops a new capacitated, multi-period, and multi-stage network model for optimising a Germany autoclaved aerated concrete recycling network. The multi-period character of the model enables the precise consideration of increasing post-demolition volumes by constantly allowing the move of recycling plants or opening new ones throughout the planning horizon. Additionally, the multi-stage formulation facilitates incorporating an optional second recycling step, which involves additional effort and higher revenues. The model aims to find a cost-minimised recycling network and identify optimal network transformations until 2050. Results show that recycling is preferred over landfilling. The optimised recycling network uses large recycling plants for economies of scale and opens new plants in the future to handle the expected increase in post-demolition autoclaved aerated concrete. Transport costs account for the largest share of total costs (50%), while fixed costs reach around 40%, and revenues offset approximately 20% of all costs. The total costs of the network reach about 2200 M€ until 2050, which is 4600 M€ (68%) less than without establishing recycling. The results offer new insights into cost-minimal network structures and their future development to encourage decision-makers to promote autoclaved aerated concrete recycling.

Land price dynamics in response to high-speed rail network characteristics: An empirical analysis

High-speed railway (HSR) has become the backbone of China's rapid and comprehensive urban transportation network system. However, the impact of the network characteristics of HSR on urban land prices has not been studied in sufficient detail. To fill this gap, this study focuses on both the Yangtze River Delta urban agglomeration and the Beijing-Tianjin-Hebei urban agglomeration and explores the linkage relationship between the HSR network and urban land prices. For this analysis, social network analysis and fixed effects model are used. The results show that the overall HSR network density of these two urban agglomerations still has room for improvement. Affected by regional proximity and the direction of HSR lines, HSR presents a spatial pattern of a multi-center–multi-tier network. The “siphon effect” is the main effect of the HSR network characteristics on land prices. However, the weakening degree of spatial spillover effect on siphon effect of the HSR network in the Yangtze River Delta urban agglomeration is stronger than that in the Beijing-Tianjin-Hebei urban agglomeration. Local governments should fully utilize the spatial effects of HSR and enhance the level of coordinated development among urban agglomerations.

Unmatched resilience and fatigue resistance in a novel cellulose-derived ionic gel with hierarchical superstructured synergy for advanced human-computer interaction

Unlike traditional hydrogels, solvent-free ionic gels eliminate leakage and solvent volatilization, making them suitable for applications in electronic skin, sensing, and other fields due to their excellent conductivity and flexibility. However, designing solvent-free ionic gels with exceptional fatigue resistance, high ionic conductivity, and toughness remains a significant challenge. We address this challenge by employing small molecule/polymer heating self-assembly to regulate the synergistic effects of various components on the hierarchical superstructure. A multi-stage network was formed through the free radical polymerization of 3-Ethyl-1-vinyl-1H-imidazol-3-ium bromide ([C2VIm]Br), lipoic acid (LA), acrylic acid (AA), and hydroxypropyl cellulose (HPC), stabilized by dense hydrogen bonding and polymer chain entanglement. The resulting HPC/LA/AA/iLs (HLAI) ionic gel exhibits exceptional properties: excellent flexibility (with an elongation at break of 4222 %), toughness (1.03 MJ/m3), high ionic conductivity (3.29 × 10−2 S/m), and outstanding fatigue resistance (with a resilience of 91.9 % after 1000 load-unloading cycles at 50 % strain). Additionally, it shows low sensitivity to crack propagation (50 % notched sample can be stretched to 9 times without breaking) and good environmental stability, demonstrating excellent sensing sensitivity and stability (GF=3.23). The conductive ionic gel features a sophisticated hierarchical superstructure that synergistically enhances its properties at the molecular level. The integration of dynamic disulfide bonds, achieved through the polymerization of Lipoic acid, imparts remarkable self-healing capabilities and superior fatigue resistance. Additionally, the incorporation of a rigid HPC structure with dense cross-linking points bolsters elasticity, fracture resistance, and resilience. The gel’s ionic conductivity and sensing sensitivity are significantly enhanced by the inclusion of ionic liquids with polymerizable double bonds, making it an ideal material for applications requiring both flexibility and conductivity. This work presents a novel strategy for designing high-performance multi-stage network solvent-free ionic gels, opening up exciting possibilities for biomimetic electronic skin applications.

Heterogeneous feature learning network for multimodal remote sensing image collaborative classification

ABSTRACT While multimodal remote sensing data processing has garnered growing interests in the geoscience field, how to effectively extract, represent, and fuse heterogeneous features remains challenging. In this article, a novel heterogeneous feature learning network (HFLN) architecture is investigated for multimodal remote sensing collaborative classification, which takes both advantages of long-range feature modelling capability of transformer as well as local feature extraction of convolutional neural network (CNN). First, the global spectral dependencies are extracted from hyperspectral images using the spectral transformer structure. Then, local spatial invariant characteristics are extracted from multimodal remote sensing images by convolutional operation. Next, heterogeneous spectral and spatial characteristics, having significant structural and semantic differences, are dynamically integrated through the feature coupling module in an interactive manner. Finally, the multi-stage network architecture is utilized for extracting hierarchical characteristics, where the numbers of feature maps in CNN and heads in transformer are gradually increased to represent more complex features. Four benchmark remote sensing datasets (i.e. hyperspectral images and multimodal remote sensing datasets) are utilized for experimental analysis, and extensive experimental results have certified the progressiveness of the investigated approach.

PMANet: Progressive multi-stage attention networks for skin disease classification

Automated skin disease classification is crucial for the timely diagnosis of skin lesions. However, accurate skin disease classification presents a challenge, given the significant intra-class variation and inter-class similarity among different kinds of skin diseases. Previous studies have attempted to address this issue by identifying the most discriminative part of a lesion, but they tend to overlook the interactions between multi-scale features. In this paper, we propose a Progressive Multi-stage Attention Network (PMANet) to enhance the learning of multi-scale discriminative features, so that the model can gradually localize from stable fine-grained to coarse-grained regions in order to improve the accuracy of disease classification. Specifically, we utilize a progressive multi-stage network to supervise feature and classification, thereby fostering multi-scale information and improving the model's ability to learn intra-class consistent information. Additionally, we propose an enhanced region proposal block that highlights key discriminative features and suppresses background noise of lesions, reinforcing the learning of inter-class discriminative features. Furthermore, we propose a multi-branch feature fusion block that effectively fuses multi-scale lesion features from different stages. Comprehensive experiments conducted on two datasets substantiate the effectiveness and superiority of the proposed method in accurately classifying skin disease.

Multi-stage coarse-to-fine progressive enhancement network for single-image HDR reconstruction

Compared with traditional imaging, high dynamic range (HDR) imaging technology can record scene information more accurately, thereby providing users higher quality of visual experience. Inverse tone mapping is a direct and effective way to realize single-image HDR reconstruction, but it usually suffers from some problems such as detail loss, color deviation and artifacts. To solve the problems, this paper proposes a multi-stage coarse-to-fine progressive enhancement network (named MSPENet) for single-image HDR reconstruction. The entire multi-stage network architecture is designed in a progressive manner to obtain higher-quality HDR images from coarse-to-fine, where a mask mechanism is used to eliminate the effects of over-exposure regions. Specifically, in the first two stages, two asymmetric U-Nets are constructed to learn the multi-scale information of input image and perform coarse reconstruction. In the third stage, a residual network with channel attention mechanism is constructed to learn the fusion of progressively transferred multi-level features and perform fine reconstruction. In addition, a multi-stage progressive detail enhancement mechanism is designed, including progressive gated recurrent unit fusion mechanism and multi-stage feature transfer mechanism. The former fuses the progressively transferred features with coarse HDR features to reduce the error stacking effect caused by multi-stage networks. Meanwhile, the latter fuses early features to supplement the lost information during each stage of feature delivery and combines features from different stages. Extensive experimental results show that the proposed method can reconstruct higher quality HDR images and effectively recover texture and color information in over-exposure regions compared to the state-of-the-art methods.

Receive wireless sensor data through IoT gateway using web client based on border gateway protocol

One of the significant topics in the field of the Internet of Things (IoT) pertains to the interaction and information sharing among people. The utilization of the Border Gateway Protocol (BGP) stack enhances the integration of web protocols and sensor networks, leading to greater accessibility. However, the BGP protocol stack introduces substantial overhead to messages transmitted at each layer, resulting in increased data overhead and energy consumption in networks by several orders of magnitude. This paper proposes a method to reduce the overhead on small and medium-sized packets. In multi-temporal networks utilizing BGP, scheduling and aggregating BGP packets at sensor nodes help achieve specific objectives. Various research methodologies and measures are employed to facilitate this, including request classification, BGP response prioritization within the network, determination of maximum acceptable delay, and overall network management. Synchronization and temporal integration of received messages at sensor nodes are performed, considering the maximum allowable delay for each message and the availability of the destination to process the accumulated messages. The evaluation results of the proposed method demonstrate a significant reduction in energy consumption and network traffic, particularly in monitoring applications within multi-stage networks. The protocol stack used is derived from the BGP standard.

Ternary systems engineered conductive hydrogel with extraordinary strength, environmental adaptability and excellent electrochemical performances for flexible power supply devices

The easy failure, poor environmental adaptability and unsatisfactory electrochemical performances of hydrogels hinder their applications as key components of flexible power supply devices (PSDs). Herein, a PAA-based hydrogel with extraordinary strength and environmental adaptability is designed via a ternary system, consisting of the tannin-modified MXene (TA@MXene), ZnCl2-cellulose and malic acid (MA) electrolyte. The TA@MXene and ZnCl2-cellulose promote the crosslinking of hydrogel via forming multi networks, endowing the hydrogel with 1.9 MPa tensile strength and 620 % stretchability. Furthermore, the hydrogel has 38.4 mS·cm−1 conductivity, thanks to the effective ion transfer channels in the hydrogel. The MA electrolyte provides a stable pH environment via forming an acid ionization system; also, MA and the high-concentration ZnCl2 solution enhance their electrochemical performance at extreme environments. Three typical PSDs were assembled using the resultant hydrogel as electrolyte/electrode. The as-prepared supercapacitors display a high specific capacity (173.5 mAh·g−1), a superior energy density (208.2 Wh·kg−1) and outstanding capacity retention (92.1 % after 5000 cycles); flexible batteries efficiently respond to strain signals, with 0.77 V open-circuit voltage (Voc); the as-assembled TENG has a 110 V Voc (100 % stretching deformation). We present a design strategy for the construction of advanced hydrogels based a ternary system that will promote flexible PSDs towards practical use.

Multi-scale terahertz image reconstruction

This paper proposes a novel terahertz (THz) image recovery algorithm and a new THz image dataset is publicly available. Because of transmission noise, artificial errors and problems with diffraction phenomena are among the main problems that cause degradation of THz images. The point spread function (PSF) in real transmission is first obtained using a terahertz imaging system. Experiments can generate degraded images and clean image datasets for supervised training of neural networks via terahertz PSF. A Depthwise Dense Instantiation Normalization Block (DwDIN Block) has been proposed to recover the datasets. We constructed a powerful multi-stage network (DwDINet) at each scale by utilizing the DwDIN module. For input and output, the results of each stage of the network and the original images are taken as input into the next layer in order to enhance the background information. DwDINet achieves state-of-the-art (SOTA) results on the two main THz datasets. To verify the generality of the model, we performed experiments on various image recovery tasks. According to the experimental results, the proposed model is of great potential for the field of low-level image clarification.

Spacecraft Homography Pose Estimation with Single-Stage Deep Convolutional Neural Network.

Spacecraft pose estimation using computer vision has garnered increasing attention in research areas such as automation system theory, control theory, sensors and instruments, robot technology, and automation software. Confronted with the extreme environment of space, existing spacecraft pose estimation methods are predominantly multi-stage networks with complex operations. In this study, we propose an approach for spacecraft homography pose estimation with a single-stage deep convolutional neural network for the first time. We formulated a homomorphic geometric constraint equation for spacecraft with planar features. Additionally, we employed a single-stage 2D keypoint regression network to obtain homography 2D keypoint coordinates for spacecraft. After decomposition to obtain the rough spacecraft pose based on the homography matrix constructed according to the geometric constraint equation, a loss function based on pixel errors was employed to refine the spacecraft pose. We conducted extensive experiments using widely used spacecraft pose estimation datasets and compared our method with state-of-the-art techniques in the field to demonstrate its effectiveness.

Ghost-Unet: multi-stage network for image deblurring via lightweight subnet learning

Ghost-Unet: multi-stage network for image deblurring via lightweight subnet learning

A Deep Convolutional Encoder–Decoder–Restorer Architecture for Image Deblurring

The accuracy of many computer vision tasks is reduced by blurred images, so deblur is important. More details of the image can be captured by a common multi-stage network, but the computational complexity of this method is higher compared with a single-stage network. However, a single-stage network cannot capture multi-scale information well. To tackle the problem, a novel convolutional encoder–decoder–restorer architecture is proposed. In this architecture, a multi-scale input structure is used in the encoder. Improved supervised attention module is inserted into the encoder for enhanced feature acquisition. In decoder, information supplement block is proposed to fuse multi-scale features. Finally, the fused features are used for image recovery in the restorer. In order to optimise the model in multiple domains, the loss function is calculated separately in the spatial and frequency domains. Our method is compared with existing methods on the GOPRO dataset. In addition, to verify the applications of our proposed method, we conduct experiments on the Real image dataset, the VOC2007 dataset and the LFW dataset. Experimental results show that our proposed method outperforms state-of-the-art deblurring methods and improves the accuracy of different vision tasks.

Image Restoration via Frequency Selection.

Image restoration aims to reconstruct the latent sharp image from its corrupted counterpart. Besides dealing with this long-standing task in the spatial domain, a few approaches seek solutions in the frequency domain by considering the large discrepancy between spectra of sharp/degraded image pairs. However, these algorithms commonly utilize transformation tools, e.g., wavelet transform, to split features into several frequency parts, which is not flexible enough to select the most informative frequency component to recover. In this paper, we exploit a multi-branch and content-aware module to decompose features into separate frequency subbands dynamically and locally, and then accentuate the useful ones via channel-wise attention weights. In addition, to handle large-scale degradation blurs, we propose an extremely simple decoupling and modulation module to enlarge the receptive field via global and window-based average pooling. Furthermore, we merge the paradigm of multi-stage networks into a single U-shaped network to pursue multi-scale receptive fields and improve efficiency. Finally, integrating the above designs into a convolutional backbone, the proposed Frequency Selection Network (FSNet) performs favorably against state-of-the-art algorithms on 20 different benchmark datasets for 6 representative image restoration tasks, including single-image defocus deblurring, image dehazing, image motion deblurring, image desnowing, image deraining, and image denoising.

Analytical learning classifier based on predefined evenly-distributed class centroids

For the pattern recognition, most classification models are solved iteratively, except for Linear LDA, KLDA and ELM etc. In this paper, a nonlinear classification network model based on predefined evenly-distributed class centroids (PEDCC) is proposed. Its analytical solution can be obtained and has good interpretability. Using the characteristics of maximizing the inter-class distance of PEDCC and derivative weighted minimum mean square error loss function to minimize the intra-class distance, we can not only realize the effective nonlinearity of the network, but also obtain the analytical solution of the network weight. Then, the sample is classified based on GDA. In order to further improve the performance of classification, PCA is used to reduces the dimensionality of the original sample, meanwhile, the CReLU activation function are adopted to enhances the expression ability of the features. The network transforms the samples into the higher dimensional feature space through the weighted minimum mean square error, so as to find a better separation hyperplane. In experiments, the feasibility of the network structure is verified from pure linear 𝑾, 𝑾+Tanh, and PCA+𝑾+Tanh respectively on many small data sets and large data sets, and compared with SVM and ELM in terms of training speed and recognition rate. The results show that, in general, this model has advantages on small data set both in recognition accuracy and training speed, while it has advantages in training speed on large data sets. Finally, by introducing a multi-stage network structure based on the latent feature norm, the classifier network can further significantly improve the classification performance, the recognition rate of small data sets is effectively improved and much higher than that of existing methods, while the recognition rate of large data sets is similar to that of SVM.

Multi-Stage Network with Geometric Semantic Attention for Two-View Correspondence Learning.

The removal of outliers is crucial for establishing correspondence between two images. However, when the proportion of outliers reaches nearly 90%, the task becomes highly challenging. Existing methods face limitations in effectively utilizing geometric transformation consistency (GTC) information and incorporating geometric semantic neighboring information. To address these challenges, we propose a Multi-Stage Geometric Semantic Attention (MSGSA) network. The MSGSA network consists of three key modules: the multi-branch (MB) module, the GTC module, and the geometric semantic attention (GSA) module. The MB module, structured with a multi-branch design, facilitates diverse and robust spatial transformations. The GTC module captures transformation consistency information from the preceding stage. The GSA module categorizes input based on the prior stage's output, enabling efficient extraction of geometric semantic information through a graph-based representation and inter-category information interaction using Transformer. Extensive experiments on the YFCC100M and SUN3D datasets demonstrate that MSGSA outperforms current state-of-the-art methods in outlier removal and camera pose estimation, particularly in scenarios with a high prevalence of outliers. Source code is available at https://github.com/shuyuanlin.

An Integrated Analysis of Production, Marketing, and Value Addition in Rose and Gerbera Cultivation and Distribution

This comprehensive study offers an integrated investigation of various elements of rose and gerbera farming, marketing, and value-added operations. First, by evaluating the Cobb-Douglas production function’s coefficients and associated statistics, the production aspect is examined. which reveals strong relationships between inputs such as labour, seedlings, fertilisers, and output yield. Second, the marketing channels used in the distribution of rose and gerbera flowers are identified, with eight separate channels ranging from direct farmer-to-consumer transactions to multi-tiered networks comprising local and Dhaka wholesalers. Third, examine average marketing expenditures for performers or agents in each channel, including expenses for baskets, packing, transportation, storage, waste loss, commission, and other costs. Finally, the net value addition for each agent across various channels is derived by subtracting the selling price from the total marketing costs. The findings shed light on the intricacies of rose and gerbera cultivation and distribution, assisting stakeholders in optimising production methods, marketing approaches, and resource allocation to boost floriculture profits and efficiency in the floriculture industry.