A Lightweight Convolutional Neural Network with Hierarchical Multi-Scale Feature Fusion for Image Classification

Convolutional neural networks (CNNs) are widely used in image classification tasks, but their increasing model size and computation make them challenging to implement on embedded systems with constrained hardware resources. To address this issue, the MobileNetV1 network was developed, which employs depthwise convolution to reduce network complexity. MobileNetV1 employs a stride of 2 in several convolutional layers to decrease the spatial resolution of feature maps, thereby lowering computational costs. However, this stride setting can lead to a loss of spatial information, particularly affecting the detection and representation of smaller objects or finer details in images. To maintain the trade-off between complexity and model performance, a lightweight convolutional neural network with hierarchical multi-scale feature fusion based on the MobileNetV1 network is proposed. The network consists of two main subnetworks. The first subnetwork uses a depthwise dilated separable convolution (DDSC) layer to learn imaging features with fewer parameters, which results in a lightweight and computationally inexpensive network. Furthermore, depthwise dilated convolution in DDSC layer effectively expands the field of view of filters, allowing them to incorporate a larger context. The second subnetwork is a hierarchical multi-scale feature fusion (HMFF) module that uses parallel multi-resolution branches architecture to process the input feature map in order to extract the multi-scale feature information of the input image. Experimental results on the CIFAR-10, Malaria, and KvasirV1 datasets demonstrate that the proposed method is efficient, reducing the network parameters and computational cost by 65.02% and 39.78%, respectively, while maintaining the network performance compared to the MobileNetV1 baseline.

Optimizing slope safety factor prediction via stacking using sparrow search algorithm for multi-layer machine learning regression models

The safety factor is a crucial quantitative index for evaluating slope stability.However,the traditional calculation methods suffer from unreasonable assumptions,complex soil composition,and inadequate consideration of the influencing factors,leading to large errors in their calculations.Therefore,a stacking ensemble learning model(stacking-SSAOP)based on multi-layer regression algorithm fusion and optimized by the sparrow search algorithm is proposed for predicting the slope safety factor.In this method,the density,cohesion,friction angle,slope angle,slope height,and pore pressure ratio are selected as characteristic parameters from the 210 sets of established slope sample data.Random Forest,Extra Trees,AdaBoost,Bagging,and Support Vector regression are used as the base model(inner loop)to construct the first-level regression algorithm layer,and XGBoost is used as the meta-model(outer loop)to construct the second-level regression algorithm layer and complete the construction of the stacked learning model for improving the model prediction accuracy.The sparrow search algorithm is used to optimize the hyperparameters of the above six regression models and correct the over-and underfitting problems of the single regression model to further improve the prediction accuracy.The mean square error(MSE)of the predicted and true values and the fitting of the data are compared and analyzed.The MSE of the stacking-SSAOP model was found to be smaller than that of the single regression model(MSE=0.03917).Therefore,the former has a higher prediction accuracy and better data fitting.This study innovatively applies the sparrow search algorithm to predict the slope safety factor,showcasing its advantages over traditional methods.Additionally,our proposed stacking-SSAOP model integrates multiple regression algorithms to enhance prediction accuracy.This model not only refines the prediction accuracy of the slope safety factor but also offers a fresh approach to handling the intricate soil composition and other influencing factors,making it a precise and reliable method for slope stability evaluation.This research holds importance for the modernization and digitalization of slope safety assessments.

Risk factors and care of early surgical site infection after primary posterior lumbar interbody fusion

Objectives:To explore the risk factors and nursing measures of early surgical site infection(SSI)after posterior lumbar interbody fusion(PLIF).Methods:A total of 468 patients who received PLIF in our hospital from January 2017 to June 2020 were enrolled into this study.According to the occurrence of early SSI,the patients were divided into two groups,and the general data were analyzed by univariate analysis.Multivariate logistic regression analysis was conducted with the dichotomous variable of whether early SSI occurred and other factors as independent variables to identify the risk factors of early SSI and put forward targeted prevention and nursing measures.Results:Among 468 patients with PLIF,18 patients developed early SSI(3.85%).The proportion of female,age,diabetes mellitus and urinary tract infection(UTI),operation segment,operation time,post-operative drainage volume,and drainage time were significantly higher than those in the uninfected group,with statistical significance(P<0.05),whereas the preoperative albumin and hemoglobin in the infected group were significantly lower than those in the uninfected group,with statistical significance(P<0.05).There was no significant difference between the two groups in the American Society of Anesthesiologists(ASA)grading,body mass index(BMI),complications including cardiovascular and cerebrovascular diseases or hypertension(P>0.05).Logistic regression analysis showed that preoperative diabetes mellitus(OR=2.109,P=0.012)/UTI(OR=1.526,P=0.035),prolonged drainage time(OR=1.639,P=0.029)were risk factors for early SSI.Men(OR=0.736,P=0.027)and albumin level(OR=0.526,P=0.004)were protective factors in reducing early SSI.Conclusions:Women,preoperative diabetes/UTI,hypoproteinemia,and prolonged drainage time are risk factors for early SSI after PLIF.Clinical effective preventive measures should be taken in combination with targeted nursing intervention to reduce the risk of early SSI.

Human Gait Recognition for Biometrics Application Based on Deep Learning Fusion Assisted Framework

The demand for a non-contact biometric approach for candidate identification has grown over the past ten years.Based on the most important biometric application,human gait analysis is a significant research topic in computer vision.Researchers have paid a lot of attention to gait recognition,specifically the identification of people based on their walking patterns,due to its potential to correctly identify people far away.Gait recognition systems have been used in a variety of applications,including security,medical examinations,identity management,and access control.These systems require a complex combination of technical,operational,and definitional considerations.The employment of gait recognition techniques and technologies has produced a number of beneficial and well-liked applications.Thiswork proposes a novel deep learning-based framework for human gait classification in video sequences.This framework’smain challenge is improving the accuracy of accuracy gait classification under varying conditions,such as carrying a bag and changing clothes.The proposed method’s first step is selecting two pre-trained deep learningmodels and training fromscratch using deep transfer learning.Next,deepmodels have been trained using static hyperparameters;however,the learning rate is calculated using the particle swarmoptimization(PSO)algorithm.Then,the best features are selected from both trained models using the Harris Hawks controlled Sine-Cosine optimization algorithm.This algorithm chooses the best features,combined in a novel correlation-based fusion technique.Finally,the fused best features are categorized using medium,bi-layer,and tri-layered neural networks.On the publicly accessible dataset known as the CASIA-B dataset,the experimental process of the suggested technique was carried out,and an improved accuracy of 94.14% was achieved.The achieved accuracy of the proposed method is improved by the recent state-of-the-art techniques that show the significance of this work.

Cellular Automata-based Chloride Ion Diffusion Simulation of Concrete Bridges under Multi-factor Coupling Actions

In order to accurately simulate the diffusion of chloride ion in the existing concrete bridge and acquire the precise chloride ion concentration at given time, a cellular automata （CA）-based model is proposed. The process of chloride ion diffusion is analyzed by the CA-based method and a nonlinear solution of the Fick＇s second law is obtained. Considering the impact of various factors such as stress states, temporal and spatial variability of diffusion parameters and water-cement ratio on the process of chloride ion diffusion, the model of chloride ion diffusion under multi-factor coupling actions is presented. A chloride ion penetrating experiment reported in the literature is used to prove the effectiveness and reasonability of the present method, and a T-type beam is taken as an illustrative example to analyze the process of chloride ion diffusion in practical application. The results indicate that CA-based method can simulate the diffusion of chloride ion in the concrete structures with acceptable precision.

Method of Multi-Mode Sensor Data Fusion with an Adaptive Deep Coupling Convolutional Auto-Encoder

To address the difficulties in fusing multi-mode sensor data for complex industrial machinery, an adaptive deep coupling convolutional auto-encoder (ADCCAE) fusion method was proposed. First, the multi-mode features extracted synchronously by the CCAE were stacked and fed to the multi-channel convolution layers for fusion. Then, the fused data was passed to all connection layers for compression and fed to the Softmax module for classification. Finally, the coupling loss function coefficients and the network parameters were optimized through an adaptive approach using the gray wolf optimization (GWO) algorithm. Experimental comparisons showed that the proposed ADCCAE fusion model was superior to existing models for multi-mode data fusion.

A multi-source data fusion modeling method for debris flow prevention engineering

The Digital Elevation Model(DEM)data of debris flow prevention engineering are the boundary of a debris flow prevention simulation,which provides accurate and reliable DEM data and is a key consideration in debris flow prevention simulations.Thus,this paper proposes a multi-source data fusion method.First,we constructed 3D models of debris flow prevention using virtual reality technology according to the relevant specifications.The 3D spatial data generated by 3D modeling were converted into DEM data for debris flow prevention engineering.Then,the accuracy and applicability of the DEM data were verified by the error analysis testing and fusion testing of the debris flow prevention simulation.Finally,we propose the Levels of Detail algorithm based on the quadtree structure to realize the visualization of a large-scale disaster prevention scene.The test results reveal that the data fusion method controlled the error rate of the DEM data of the debris flow prevention engineering within an allowable range and generated 3D volume data(obj format)to compensate for the deficiency of the DEM data whereby the 3D internal entity space is not expressed.Additionally,the levels of detailed method can dispatch the data of a large-scale debris flow hazard scene in real time to ensure a realistic 3D visualization.In summary,the proposed methods can be applied to the planning of debris flow prevention engineering and to the simulation of the debris flow prevention process.

Fusion of multispectral image and panchromatic image based on NSCT and NMF

A novel fusion method of multispectral image and panchromatic image based on nonsubsampled contourlet transform（NSCT） and non-negative matrix factorization（NMF） is presented,the aim of which is to preserve both spectral and spatial information simultaneously in fused image.NMF is a matrix factorization method,which can extract the local feature by choosing suitable dimension of the feature subspace.Firstly the multispectral image was represented in intensity hue saturation（IHS） system.Then the I component and panchromatic image were decomposed by NSCT.Next we used NMF to learn the feature of both multispectral and panchromatic images＇ low-frequency subbands,and the selection principle of the other coefficients was absolute maximum criterion.Finally the new coefficients were reconstructed to get the fused image.Experiments are carried out and the results are compared with some other methods,which show that the new method performs better in improving the spatial resolution and preserving the feature information than the other existing relative methods.

Data Fusion in Distributed Multi-sensor System

This paper presents a data fusion method in distributed multi-sensor system including GPS and INS sensors’ data processing. First, a residual χ 2 \|test strategy with the corresponding algorithm is designed. Then a coefficient matrices calculation method of the information sharing principle is derived. Finally, the federated Kalman filter is used to combine these independent, parallel, real\|time data. A pseudolite (PL) simulation example is given.

A fusion protein containing murine vascular endothelial growth factor and tissue factor induces thrombogenesis and suppression of tumor growth in a colon carcinoma model

Induction of tumor vasculature occlusion by targeting a thrombogen to newly formed blood vessels in tumor tissues represents an intriguing approach to the eradication of primary solid tumors. In the current study, we construct and express a fusion protein containing vascular endothelial growth factor (VEGF) and tissue factor (TF) to explore whether this fusion protein has the capability of inhibiting tumor growth in a colon carcinoma model. The murine cDNA of VEGF A and TF were amplified by reverse transcriptase polymerase chain reaction (RT-PCR), and then cloned into prokaryotic expression plasmid pQE30 with a linker. The expression product recombinant VEGF-TF (rVEGF-TF) was purified and proved to have comparable enzyme activity to a commercial TF and the capability of specific binding to tumor vessels. Significant decrease of tumor growth was found in the mice administered with rVEGF-TF on Day 6 after initiated rVEGF-TF treatment (P<0.05), and the tumor masses in 2 of 10 mice were almost disappeared on Day 14 after the first treatment. In addition, valid thrombogenesis and tumor necrosis were observed in the tumor tissues injected with rVEGF-TF. Our results demonstrate that occlusion of tumor vasculature with rVEGF-TF is potentially an effective approach for cancer therapy.

STUDY ON THE COAL-ROCK INTERFACE RECOGNITION METHOD BASED ON MULTI-SENSOR DATA FUSION TECHNIQUE

The coal-rock interface recognition method based on multi-sensor data fusiontechnique is put forward because of the localization of single type sensor recognition method. Themeasuring theory based on multi-sensor data fusion technique is analyzed, and hereby the testplatform of recognition system is manufactured. The advantage of data fusion with the fuzzy neuralnetwork (FNN) technique has been probed. The two-level FNN is constructed and data fusion is carriedout. The experiments show that in various conditions the method can always acquire a much higherrecognition rate than normal ones.

Multi-Behavior Fusion Based Potential Field Method for Path Planning of Unmanned Surface Vessel

The problem of the unmanned surface vessel (USV) path planning in static and dynamic obstacle environments is addressed in this paper. Multi-behavior fusion based potential field method is proposed, which contains three behaviors: goal-seeking, boundary-memory following and dynamic-obstacle avoidance. Then, different activation conditions are designed to determine the current behavior. Meanwhile, information on the positions, velocities and the equation of motion for obstacles are detected and calculated by sensor data. Besides, memory information is introduced into the boundary following behavior to enhance cognition capability for the obstacles, and avoid local minima problem caused by the potential field method. Finally, the results of theoretical analysis and simulation show that the collision-free path can be generated for USV within different obstacle environments, and further validated the performance and effectiveness of the presented strategy.

Multi-Focus Image Fusion Based on Wavelet Transformation

In the fusion of image,how to measure the local character and clarity is called activity measurement. According to the problem,the traditional measurement is decided only by the high-frequency detail coefficients, which will make the energy expression insufficient to reflect the local clarity. Therefore,in this paper,a novel construction method for activity measurement is proposed. Firstly,it uses the wavelet decomposition for the fusion resource image, and then utilizes the high and low frequency wavelet coefficients synthetically. Meantime,it takes the normalized variance as the weight of high-frequency energy. Secondly,it calculates the measurement by the weighted energy,which can be used to measure the local character. Finally,the fusion coefficients can be got. In order to illustrate the superiority of this new method,three kinds of assessing indicators are provided. The experiment results show that,comparing with the traditional methods,this new method weakens the fuzzy and promotes the indicator value. Therefore,it has much more advantages for practical application.

MULTI-SOURCE REMOTE SENSING IMAGE FUSION BASED ON SUPPORT VECTOR MACHINE

Remote Sensing image fusion is an effective way to use the large volume of data from multi-source images.This paper introduces a new method of remote sensing image fusion based on support vector machine(SVM),using high spatial resolution data SPIN-2and multi-spectral remote sensing data S POT-4.Firstly,the new method is established by building a model of remote sensing im age fusion based on SVM.Then by using SPIN-2data and SPOT-4data,image classifi-cation fusion is tested.Finally,an evaluation of the fusion result is ma de in two ways.1)From subjectivity assessment,the spatial resolution of the fused i mage is improved compared to the SPOT-4.And it is clearly that the texture of the fused image is distinctive.2)From quantitative analysis,the effect of classification fusion is bett er.As a whole,the re-sult shows that the accuracy of image fusion based on SVMis high and the SVM algorithm can be recommended for app lica-tion in remote sensing image fusion p rocesses.

Performance Validation and Analysis for Multi-Method Fusion Based Image Quality Metrics in A New Image Database

Considering that there is no single full reference image quality assessment method that could give the best performance in all situations, some multi-method fusion metrics were proposed. Machine learning techniques are often involved in such multi-method fusion metrics so that its output would be more consistent with human visual perceptions. On the other hand, the robustness and generalization ability of these multi-method fusion metrics are questioned because of the scarce of images with mean opinion scores. In order to comprehensively validate whether or not the generalization ability of such multi-method fusion IQA metrics are satisfying, we construct a new image database which contains up to 60 reference images. The newly built image database is then used to test the generalization ability of different multi-method fusion IQA metrics. Cross database validation experiment indicates that in our new image database, the performances of all the multi-method fusion IQA metrics have no statistical significant different with some single-method IQA metrics such as FSIM and MAD. In the end, a thorough analysis is given to explain why the performance of multi-method fusion IQA framework drop significantly in cross database validation.

MULTI-SPECTRAL AND HYPERSPECTRAL IMAGE FUSION USING 3-D WAVELET TRANSFORM

Image fusion is performed between one band of multi-spectral image and two bands of hyperspectral image to produce fused image with the same spatial resolution as source multi-spectral image and the same spectral resolution as source hyperspectral image. According to the characteristics and 3-Dimensional (3-D) feature analysis of multi-spectral and hyperspectral image data volume, the new fusion approach using 3-D wavelet based method is proposed. This approach is composed of four major procedures: Spatial and spectral resampling, 3-D wavelet transform, wavelet coefficient inte- gration and 3-D inverse wavelet transform. Especially, a novel method, Ratio Image Based Spectral Resampling (RIBSR) method, is proposed to accomplish data resampling in spectral domain by util- izing the property of ratio image. And a new fusion rule, Average and Substitution (A&S) rule, is employed as the fusion rule to accomplish wavelet coefficient integration. Experimental results illustrate that the fusion approach using 3-D wavelet transform can utilize both spatial and spectral character- istics of source images more adequately and produce fused image with higher quality and fewer artifacts than fusion approach using 2-D wavelet transform. It is also revealed that RIBSR method is capable of interpolating the missing data more effectively and correctly, and A&S rule can integrate coefficients of source images in 3-D wavelet domain to preserve both spatial and spectral features of source images more properly.

Multi-factor analysis of initial poor graft function after orthotopic liver transplantation

BACKGROUND: In the early period of orthotopic liver transplantation (OLT), initial poor graft function (IPGF) is one of the complications which leads to primary graft non-function (PGNF) in serious cases. This study set out to establish the clinical risk factors resulting in IPGF after OLT. METHODS: Eighty cases of OLT were analyzed. The IPGF group consisted of patients with alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST) above 1500 IU/L within 72 hours after OLT, while those in the non-IPGF group had values below 1500 IU/L. Recipient-associated factors before OLT analyzed were age, sex, primary liver disease and Child-Pugh classification; factors analyzed within the peri-operative period were non-heart beating time (NHBT), cold ischemia time (CIT), rewarming ischemic time (RWIT), liver biopsy at the end of cold ischemia; and factors analyzed within 72 hours after OLT were ALT and/or AST values. A logistic regression model was applied to filter the possible factors resulting in IPGF. RESULTS: Donor NHBT, CIT and RWIT were significantly longer in the IPGF group than in the non-IPGF group; in the logistic regression model, NHBT was the risk factor leading to IPGF (P < 0.05), while CIT and RWIT were possible risk factors. In one case in the IPGF group, PGNF appeared with moderate hepatic steatosis. CONCLUSIONS: Longer NHBT is an important risk factor leading to IPGF, while serious steatosis in the donor liver, CIT and RWIT are potential risk factors.

DETECTION METHOD OF SPOT WELDING BASED ON MULTI-INFORMATION FUSION AND FRACTAL

A novel detection method of support vector machine （SVM） based on fractal dimension of signals is presented. And models of SVM are made based on nugget size defects of spot welding. Classification using these trained SVM models is done to signals of spot welding. It is shown from effect of different SVM models that these models with different inputs. In detection of defects, these models with inputs including sound signal have a high percentage of accuracy, the detection accuracy of these models with inputs including voltage signal will reduce. So the SVM models based on fractal dimensions of sound are some optimal nondestructive detection ones. At last a comparison between SVM detection model and ANNS detection model is researched which indicates that SVM is a more effective measure than Artificial neural networks in detection of nugget size defects during spot welding.

The prevalence of azoospermia factor microdeletion on the Y chromosome of Chinese infertile men detected by multi-analyte suspension array technology

Aim： To develop a high-throughput multiplex, fast and simple assay to scan azoospermia factor （AZF） region microdeletions on the Y chromosome and establish the prevalence of Y chromosomal microdeletions in Chinese infertile males with azoospermia or oligozoospermia. Methods： In total, 178 infertile patients with azoospermia （nonobstructed）, 134 infertile patients with oligozoospermia as well as 40 fertile man controls were included in the present study. The samples were screened for AZF microdeletion using optimized multi-analyte suspension array （MASA） technology. Results： Of the 312 patients, 36 （11.5%） were found to have deletions in the AZF region. The rnicrodeletion frequency was 14% （25/178） in the azoospermia group and 8.2% （11/134） in the oligospermia group. Among 36 patients with microdeletions, 19 had deletions in the AZFc region, seven had deletions in AZFa and six had deletions in AZFb. In addition, four patients had both AZFb and AZFc deletions. No deletion in the AZF region was found in the 40 fertile controls. Conclusion： There is a high prevalence of Y chromosomal microdeletions in Chinese infertile males with azoospermia or oligozoospermia. The MASA technology, which has been established in the present study, provides a sensitive and high-throughput method for detecting the deletion of the Y chromosome. And the results suggest that genetic screening should be advised to infertile men before starting assisted reproductive treatments.