结合Pixel2style2Pixel的年龄转化方法

年龄转化在刑侦、人脸识别等领域有着重要作用。常见的年龄转化方法需要使用成对带有年龄注释的数据集进行训练,并且存在生成图像质量低、年龄语义信息不够解耦等问题。针对上述问题,在Pix2style2Pix的训练框架中,引入年龄识别、contextual损失函数,对整体损失函数做出符合年龄转化的改进,提取年龄信息并保证图像质量。改进编码网络配合损失函数对潜在空间的图像进行编辑,提出一种基于Pixel2style2Pixel的年龄转化方法。通过FFHQ、CelebA数据集,对所提方法进行验证,实验结果表明,在不采用成对年龄注释的训练集下,改进后的损失函数能生成更符合期望年龄的转化图像,人脸相似度距离为0.346、FID为45.69、SSIM为0.593 6、PSNR为19.64 dB,均优于对比方法,证明所提方法能够生成高质量、年龄语义高度解耦的转化结果。

Role of transforming growth factor-βin peripheral nerve regeneration

Injuries caused by trauma and neurodegenerative diseases can damage the peripheral nervous system and cause functional deficits.Unlike in the central nervous system,damaged axons in peripheral nerves can be induced to regenerate in response to intrinsic cues after reprogramming or in a growth-promoting microenvironment created by Schwann cells.However,axon regeneration and repair do not automatically result in the restoration of function,which is the ultimate therapeutic goal but also a major clinical challenge.Transforming growth factor(TGF)is a multifunctional cytokine that regulates various biological processes including tissue repair,embryo development,and cell growth and differentiation.There is accumulating evidence that TGF-βfamily proteins participate in peripheral nerve repair through various factors and signaling pathways by regulating the growth and transformation of Schwann cells;recruiting specific immune cells;controlling the permeability of the blood-nerve barrier,thereby stimulating axon growth;and inhibiting remyelination of regenerated axons.TGF-βhas been applied to the treatment of peripheral nerve injury in animal models.In this context,we review the functions of TGF-βin peripheral nerve regeneration and potential clinical applications.

Effect of multiple coulomb scattering on the beam tests of silicon pixel detectors

In the research and development of new silicon pixel detectors,a collimated monoenergetic charged-particle test beam equipped with a high-resolution pixel-beam telescope is crucial for prototype verification and performance evaluation.When the beam energy is low,the effect of multiple Coulomb scattering on the measured resolution of the Device Under Test(DUT)must be considered to accurately evaluate the performance of the pixel chips and detectors.This study aimed to investigate the effect of multiple Coulomb scattering on the measured resolution,particularly at low beam energies.Simulations were conducted using Allpix^(2) to study the effects of multiple Coulomb scattering under different beam energies,material budgets,and telescope layouts.The simulations also provided the minimum energy at which the effect of multiple Coulomb scattering could be ignored.Compared with the results of a five-layer detector system tested with an electron beam at DESY,the simulation results were consistent with the beam test results,confirming the reliability of the simulations.

Efficient single-pixel imaging encrypted transmission based on 3D Arnold transformation

Single-pixel imaging(SPI)can transform 2D or 3D image data into 1D light signals,which offers promising prospects for image compression and transmission.However,during data communication these light signals in public channels will easily draw the attention of eavesdroppers.Here,we introduce an efficient encryption method for SPI data transmission that uses the 3D Arnold transformation to directly disrupt 1D single-pixel light signals and utilizes the elliptic curve encryption algorithm for key transmission.This encryption scheme immediately employs Hadamard patterns to illuminate the scene and then utilizes the 3D Arnold transformation to permutate the 1D light signal of single-pixel detection.Then the transformation parameters serve as the secret key,while the security of key exchange is guaranteed by an elliptic curve-based key exchange mechanism.Compared with existing encryption schemes,both computer simulations and optical experiments have been conducted to demonstrate that the proposed technique not only enhances the security of encryption but also eliminates the need for complicated pattern scrambling rules.Additionally,this approach solves the problem of secure key transmission,thus ensuring the security of information and the quality of the decrypted images.

Roles of transforming growth factor-βsignaling in liver disease

In this editorial we expand the discussion on the article by Zhang et al published in the recent issue of the World Journal of Hepatology.We focus on the diagnostic and therapeutic targets identified on the basis of the current understanding of the molecular mechanisms of liver disease.Transforming growth factor-β(TGF-β)belongs to a structurally related cytokine super family.The family members display different time-and tissue-specific expression patterns associated with autoimmunity,inflammation,fibrosis,and tumorigenesis;and,they participate in the pathogenesis of many diseases.TGF-βand its related signaling pathways have been shown to participate in the progression of liver diseases,such as injury,inflammation,fibrosis,cirrhosis,and cancer.The often studied TGF-β/Smad signaling pathway has been shown to promote or inhibit liver fibrosis under different circumstances.Similarly,the early immature TGF-βmolecule functions as a tumor suppressor,inducing apoptosis;but,its interaction with the mitogenic molecule epidermal growth factor alters this effect,activating anti-apoptotic signals that promote liver cancer development.Overall,TGF-βsignaling displays contradictory effects in different liver disease stages.Therefore,the use of TGF-βand related signaling pathway molecules for diagnosis and treatment of liver diseases remains a challenge and needs further study.In this editorial,we aim to review the evidence for the use of TGF-βsignaling pathway molecules as diagnostic or therapeutic targets for different liver disease stages.

Transforming growth factor-β1 and vascular endothelial growth factor levels in senile acute myeloid leukemia and correlation with prognosis

BACKGROUND Acute myeloid leukemia(AML)is a disease in which immature hematopoietic cells accumulate in the bone marrow and continuously expand,inhibiting hematopoiesis.The treatment and prognosis of this disease have always been unsatisfactory.AIM To investigate the correlation between vascular endothelial growth factor(VEGF)and transforming growth factor-β1(TGFβ1)expression and prognosis in older adults with AML.METHODS This study enrolled 80 patients with AML(AML group),including 36 with complete response(AML-CR),23 with partial response(AML-PR),and 21 with no response(AML-NR).The expression levels of VEGF and TGFβ1 were detected by reverse transcription polymerase chain reaction in bone marrow mononuclear cells isolated from 56 healthy controls.Kaplan-Meier analysis was performed to assess overall survival(OS)and progression-or disease-free survival(DFS).Prognostic risk factors were analyzed using a Cox proportional hazards model.RESULTS The AML group showed a VEGF level of 2.68±0.16.VEGF expression was lower in patients with AML-CR than those with AML-PR or AML-NR(P<0.05).TGFβ1 expression in the AML group was 0.33±0.05.Patients with AML-CR showed a higher TGFβ1 expression than those with AML-PR or AML-NR(P<0.05).VEGF and TGFβ1 expression in patients with AML was significantly correlated with the counts of leukocytes,platelets,hemoglobin,and peripheral blood immature cells(P<0.05);Kaplan-Meier survival analysis revealed that patients with high TGFβ1 expression had better OS and DFS than those with low TGFβ1 expression(P<0.05),whereas patients with low VEGF levels showed better OS and DFS than those with high VEGF levels(P<0.05).VEGF,TGFβ1,and platelet count were identified by the Cox proportional hazards model as independent risk factors for OS(P<0.05),while VEGF,TGFβ1,and white blood cell count were independent risk factors for DFS(P<0.05).CONCLUSION Decreased VEGF expression and increased TGFβ1 expression in patients with AML provide valuable references for determining and individualizing clinical treatment strategies.

Transforming growth factor-beta 1 enhances discharge activity of cortical neurons

Transforming growth factor-beta 1(TGF-β1)has been extensively studied for its pleiotropic effects on central nervous system diseases.The neuroprotective or neurotoxic effects of TGF-β1 in specific brain areas may depend on the pathological process and cell types involved.Voltage-gated sodium channels(VGSCs)are essential ion channels for the generation of action potentials in neurons,and are involved in various neuroexcitation-related diseases.However,the effects of TGF-β1 on the functional properties of VGSCs and firing properties in cortical neurons remain unclear.In this study,we investigated the effects of TGF-β1 on VGSC function and firing properties in primary cortical neurons from mice.We found that TGF-β1 increased VGSC current density in a dose-and time-dependent manner,which was attributable to the upregulation of Nav1.3 expression.Increased VGSC current density and Nav1.3 expression were significantly abolished by preincubation with inhibitors of mitogen-activated protein kinase kinase(PD98059),p38 mitogen-activated protein kinase(SB203580),and Jun NH2-terminal kinase 1/2 inhibitor(SP600125).Interestingly,TGF-β1 significantly increased the firing threshold of action potentials but did not change their firing rate in cortical neurons.These findings suggest that TGF-β1 can increase Nav1.3 expression through activation of the ERK1/2-JNK-MAPK pathway,which leads to a decrease in the firing threshold of action potentials in cortical neurons under pathological conditions.Thus,this contributes to the occurrence and progression of neuroexcitatory-related diseases of the central nervous system.

A study on the treatment of chloasma with moxibustion based on the theory of“Yang transforming Qi,and Yin shaping”

Background:“Yang Transforming Qi and Yin shaping”comes from“Plain Question Yin and Yang should be like the theory”.Chinese medicine believes that the imbalance of yin and yang is the cause of human diseases.One of the strengths of TCM in treating diseases is holistic regulation.The theory of“Yang transforming Qi,and Yin shaping”implies a strong holistic view.Chloasma,one of the common skin diseases,seriously affects the physical and mental health of patients due to its unclear etiology,difficult treatment and easy recurrence.The occurrence and development of chloasma are closely related to internal organs.This paper is based on the theory of internal meridian,“Yang transforming Qi,and Yin shaping”which attaches importance to the concept of“Yang Qi”,from a new understanding of the pathogenesis etiology of chloasma.Methods:The author believes that the incidence of the disease is the deficiency of Yang Qi,which can’t promote the movement of Qi and blood to nourish the face,resulting in tangible material condensation,so proposes that this disease needs to be promoted by Yang transforming Qi and warm to dissipate.Moxibustion belongs to the fire,and its warming effect can dissipate the tangible evil.Results:Combined with clinical practice and the theory of internal meridian,the author elaborated that moxibustion is one of the effective methods for treating chloasma,which provides a new way of thinking for the treatment of chloasma.Conclusion:“Yang Transforming Qi and Yin shaping”can be used to guide the treatment of chloasma.

Pixel 2940nm点阵激光联合他克莫司软膏治疗肢端型白癜风的临床观察

目的:观察Pixel 2 940 nm点阵激光联合他克莫司软膏治疗肢端型白癜风的临床效果。方法:选取梅州市第二中医医院2019年5月-2021年5月收治的肢端型白癜风患者62例为研究对象,按照随机数字表法将其分为治疗组和对照组,各31例。对照组予以他克莫司软膏治疗,治疗组予以Pixel 2 940 nm点阵激光联合他克莫司软膏治疗。两组均以12周为1个疗程,共治疗2个疗程。对比两组临床疗效、不良反应发生率和复发率。结果:治疗12、24周后,治疗组总有效率(45.16%、70.97%)均明显高于对照组(19.35%、41.94%)(P<0.05);治疗期间对照组不良反应发生率为12.90%,明显低于治疗组的41.94%(P<0.05);治疗组复发率为4.55%(1/22),低于对照组的61.54%(8/13)(P<0.05)。结论:Pixel 2 940 nm点阵激光联合他克莫司软膏治疗肢端型白癜风虽会增加不良反应,但见效快,临床疗效好,能够显著降低复发率,值得在临床上推广应用。

Hi'CT:a pixel sensor-based device for ion tomography

Carbon ions,commonly referred to as particle therapy,have become increasingly popular in the last decade.Accurately predicting the range of ions in tissues is important for the precise delivery of doses in heavy-ion radiotherapy.Range uncertainty is currently the largest contributor to dose uncertainty in normal tissues,leading to the use of safety margins in treatment planning.One potential method is the direct relative stopping measurement(RSP)with ions.Heavy-ion CT(Hi′CT),a compact segmented full digital tomography detector using monolithic active pixel sensors,was designed and evaluated using a 430 MeV/u high-energy carbon ion pencil beam in Geant4.The precise position of the individual carbon ion track can be recorded and reconstructed using a 30μm×30μm small pixel pitch size.Two types of customized image reconstruction algorithms were developed,and their performances were evaluated using three different modules of CAT-PHAN 600-series phantoms.The RSP measurement accuracy of the tracking algorithm for different types of materials in the CTP404 module was less than 1%.In terms of spatial resolution,the tracking algorithm could achieve a 20%modulation transfer function normalization value of CTP528 imaging results at 5 lp/cm,which is significantly better than that of the fast imaging algorithm(3 lp/cm).The density resolution obtained using the tracking algorithm of the customized CTP515 was approximately 10.5%.In conclusion,a compact digital Hi'CT system was designed,and its nominal performance was evaluated in a simulation.The RSP resolution and image quality provide potential feasibility for scanning most parts of an adult body or pediatric patient,particularly for head and neck tumor treatment.

基于RoBERTa和图增强Transformer的序列推荐方法

自推荐系统出现以来,有限的数据信息就一直制约着推荐算法的进一步发展。为降低数据稀疏性的影响,增强非评分数据的利用率,基于神经网络的文本推荐模型相继被提出,但主流的卷积或循环神经网络在文本语义理解和长距离关系捕捉方面存在明显劣势。为了更好地挖掘用户与商品之间的深层潜在特征,进一步提高推荐质量,提出一种基于Ro BERTa和图增强Transformer的序列推荐(RGT)模型。引入评论文本数据,首先利用预训练的Ro BERTa模型捕获评论文本中的字词语义特征,初步建模用户的个性化兴趣,然后根据用户与商品的历史交互信息,构建具有时序特性的商品关联图注意力机制网络模型,通过图增强Transformer的方法将图模型学习到的各个商品的特征表示以序列的形式输入Transformer编码层,最后将得到的输出向量与之前捕获的语义表征以及计算得到的商品关联图的全图表征输入全连接层,以捕获用户全局的兴趣偏好,实现用户对商品的预测评分。在3组真实亚马逊公开数据集上的实验结果表明,与Deep FM、Conv MF等经典文本推荐模型相比,RGT模型在均方根误差(RMSE)和平均绝对误差(MAE)2种指标上有显著提升,相较于最优对比模型最高分别提升4.7%和5.3%。

基于Depth-wise卷积和视觉Transformer的图像分类模型

图像分类作为一种常见的视觉识别任务,有着广阔的应用场景。在处理图像分类问题时,传统的方法通常使用卷积神经网络,然而,卷积网络的感受野有限,难以建模图像的全局关系表示,导致分类精度低,难以处理复杂多样的图像数据。为了对全局关系进行建模,一些研究者将Transformer应用于图像分类任务,但为了满足Transformer的序列化和并行化要求,需要将图像分割成大小相等、互不重叠的图像块,破坏了相邻图像数据块之间的局部信息。此外,由于Transformer具有较少的先验知识,模型往往需要在大规模数据集上进行预训练,因此计算复杂度较高。为了同时建模图像相邻块之间的局部信息并充分利用图像的全局信息,提出了一种基于Depth-wise卷积的视觉Transformer(Efficient Pyramid Vision Transformer,EPVT)模型。EPVT模型可以实现以较低的计算成本提取相邻图像块之间的局部和全局信息。EPVT模型主要包含3个关键组件:局部感知模块(Local Perceptron Module,LPM)、空间信息融合模块(Spatial Information Fusion,SIF)和“+卷积前馈神经网络(Convolution Feed-forward Network,CFFN)。LPM模块用于捕获图像的局部相关性;SIF模块用于融合相邻图像块之间的局部信息,并利用不同图像块之间的远距离依赖关系,提升模型的特征表达能力,使模型学习到输出特征在不同维度下的语义信息;CFFN模块用于编码位置信息和重塑张量。在图像分类数据集ImageNet-1K上,所提模型优于现有的同等规模的视觉Transformer分类模型,取得了82.6%的分类准确度,证明了该模型在大规模数据集上具有竞争力。

基于Transformer和自适应特征融合的矿井低照度图像亮度提升和细节增强方法

高质量矿井影像为矿山安全生产提供保障,也有利于提高后续图像分析技术的性能。矿井影像受低照度环境的影响,易出现亮度低,照度不均,颜色失真,细节信息丢失严重等问题。针对上述问题,提出一种基于Transformer和自适应特征融合的矿井低照度图像亮度提升和细节增强方法。基于生成对抗思想搭建生成对抗式主体模型框架,使用目标图像域而非单一参考图像驱动判别器监督生成器的训练,实现对低照度图像的充分增强;基于特征表示学习理论搭建特征编码器,将图像解耦为亮度分量和反射分量,避免图像增强过程中亮度与颜色特征相互影响从而导致颜色失真问题;设计CEM-Transformer Encoder通过捕获全局上下文关系和提取局部区域特征,能够充分提升整体图像亮度并消除局部区域照度不均;在反射分量增强过程中,使用结合CEM-Cross-Transformer Encoder的跳跃连接将低级特征与深层网络处特征进行自适应融合,能够有效避免细节特征丢失,并在编码网络中添加ECA-Net,提高浅层网络的特征提取效率。制作矿井低照度图像数据集为矿井低照度图像增强任务提供数据资源。试验显示,在矿井低照度图像数据集和公共数据集中,与5种先进的低照度图像增强算法相比,该算法增强图像的质量指标PSNR、SSIM、VIF平均提高了16.564%,10.998%,16.226%和14.438%,10.888%,14.948%,证明该算法能够有效提升整体图像亮度,消除照度不均,避免颜色失真和细节丢失,实现矿井低照度图像增强。

CNN-Transformer特征融合多目标跟踪算法

在卷积神经网络(CNN)中,卷积运算能高效地提取目标的局部特征,却难以捕获全局表示;而在视觉Transformer中,注意力机制可以捕获长距离的特征依赖,但会忽略局部特征细节。针对以上问题,提出一种基于CNN-Transformer双分支主干网络进行特征提取和融合的多目标跟踪算法CTMOT(CNN-transformer multi-object tracking)。使用基于CNN和Transformer双分支并行的主干网络分别提取图像的局部和全局特征。使用双向桥接模块(two-way braidge module,TBM)对两种特征进行充分融合。将融合后的特征输入两组并行的解码器进行处理。将解码器输出的检测框和跟踪框进行匹配,完成多目标跟踪任务。在多目标跟踪数据集MOT17、MOT20、KITTI以及UADETRAC上进行评估,CTMOT算法的MOTP和IDs指标在四个数据集上均达到了SOTA效果,MOTA指标分别达到了76.4%、66.3%、92.36%和88.57%,在MOT数据集上与SOTA方法效果相当,在KITTI数据集上达到SOTA效果。由于同时完成目标检测和关联,能够端到端进行目标跟踪,跟踪速度可达35 FPS,表明CTMOT算法在跟踪的实时性和准确性上达到了较好的平衡,具有较大潜力。

基于残差U-Net和自注意力Transformer编码器的磁场预测方法

利用有限元方法对几何结构复杂的电机和变压器进行磁场分析,存在仿真时间长且无法复用的问题。因此,该文提出一种基于残差U-Net和自注意力Transformer编码器的磁场预测方法。首先建立永磁同步电机(PMSM)和非晶合金变压器(AMT)有限元模型,得到深度学习训练所需的数据集;然后将Transformer模块与U-Net模型结合,并引入短残差机制建立ResUnet-Transformer模型,通过预测图像的像素实现磁场预测;最后通过Targeted Dropout算法和动态学习率调整策略对模型进行优化,解决拟合问题并提高预测精度。计算实例证明,ResUnet-Transformer模型在PMSM和AMT数据集上测试集的平均绝对百分比误差(MAPE)均小于1%,且仅需500组样本。该文提出的磁场预测方法能减少实际工况和多工况下精细模拟和拓扑优化的时间和资源消耗,亦是虚拟传感器乃至数字孪生的关键实现方法之一。

基于改进Deformable-DETR的水下图像目标检测方法

针对由于水下复杂环境造成的目标检测效果较差、检测精度较低的问题,基于Deformable-DETR算法提出一种改进的水下目标检测算法Deformable-DETR-DA。使用空间注意力模块结合标准Transformer块设计了一个用于增加模型深度的深度特征金字塔(deep feature pyramid networks,DFPN)模块,将其嵌入到模型中提高模型对深层纹理信息的提取能力。使用注意力引导的方式对原模型中编码器部分进行改进,加强了对特征信息的聚合能力,提高了模型在复杂环境下的检测能力。针对URPC数据集,模型各交并比尺度的平均准确度(average precision,AP)为39.5%,相比原模型提升1%,与一些DETR(detection transformer)类的模型相比,不同目标尺度的平均准确度均有1%~4%左右的提高,表明改进的模型能够很好解决复杂环境的水下目标检测的问题。本文提出的模型可作为其他水下目标检测模型设计的参考。

Simulation study of BESⅢwith stitched CMOS pixel detector using AcTs

The reconstruction of the tracks of charged particles with high precision is crucial for HEP experiments to achieve their physics goals.The BESⅢdrift chamber,which is used as the tracking detector of the BESⅢexperiment,has suffered from aging effects resulting in degraded tracking performance after operation for approximately 15 years.To preserve and enhance the tracking performance of BESⅢ,one of the proposals is to add one layer of a thin cylindrical CMOS pixel sensor based on state-of-the-art stitching technology between the beam pipe and the drift chamber.The improvement in the tracking performance of BESⅢwith such an additional pixel detector compared to that with only the existing drift chamber was studied using the modern common tracking software Acts,which provides a set of detector-agnostic and highly performant tracking algorithms that have demonstrated promising performance for a few high-energy physics and nuclear physics experiments.

Development and preliminary results of a large-pixel two-layer LaBr_(3) Compton camera prototype

Lanthanum bromide(LaBr_(3))crystal has a high energy resolution and time resolution and has been used in Compton cameras(CCs)over the past few decades.However,LaBr_(3) crystal arrays are difficult to process because LaBr_(3) is easy to crack and break;thus,few LaBr_(3)-based CC prototypes have been built.In this study,we designed and fabricated a large-pixel LaBr_(3) CC prototype and evaluated its performance with regard to position,energy,and angular resolution.We used two 10×10 LaBr_(3) crystal arrays with a pixel size of 5 mm×5 mm,silicon photomultipliers(SiPMs),and corresponding decoding circuits to construct our prototype.Additionally,a framework based on a Voronoi diagram and a lookup table was developed for list-mode projection data acquisition.Monte Carlo(MC)simulations based on Geant4 and experiments were conducted to evaluate the performance of our CC prototype.The lateral position resolution was 5 mm,and the maximum deviation in the depth direction was 2.5 and 5 mm for the scatterer and absorber,respectively.The corresponding measured energy resolu-tions were 7.65%and 8.44%,respectively,at 511 keV.The experimental results of ^(137)Cs point-like sources were consistent with the MC simulation results with regard to the spatial positions and full widths at half maximum(FWHMs).The angular resolution of the fabricated prototype was approximately 6°when a point-like ^(137)Cs source was centrally placed at a distance of 5 cm from the scatterer.We proposed and investigated a large-pixel LaBr_(3) CC for the first time and verified its feasibility for use in accurate spatial positioning of radiative sources with a high angular resolution.The proposed CC can satisfy the requirements of radiative source imaging and positioning in the nuclear industry and medical applications.

融合卷积注意力和Transformer架构的行人重识别方法

行人重识别技术是智能安防系统中的重要方法之一,为构建一个适用各种复杂场景的行人重识别模型,基于现有的卷积神经网络和Transformer模型,提出一种融合卷积注意力和Transformer(FCAT)架构的方法,以增强Transformer对局部细节信息的关注。所提方法主要将卷积空间注意力和通道注意力嵌入Transformer架构中,分别加强对图像中重要区域的关注和对重要通道特征的关注,以进一步提高Transformer架构对局部细节特征的提取能力。在3个公开行人重识别数据集上的对比消融实验证明,所提方法在非遮挡数据集上取得了与现有方法相当的结果,在遮挡数据集上的性能得到显著提升。所提方法更加轻量化,在不增加额外计算量和模型参数的情况下,推理速度得到了提升。

基于Transformer和动态3D卷积的多源遥感图像分类

多源遥感数据具有互补性和协同性,近年来,基于深度学习的方法已经在多源遥感图像分类中取得了一定进展,但当前方法仍面临关键难题,如多源遥感图像特征表达不一致,融合困难,基于静态推理范式的神经网络缺乏对不同类别地物的适应性。为解决上述问题,提出了基于跨模态Transformer和多尺度动态3D卷积的多源遥感图像分类模型。为提高多源特征表达的一致性,设计了基于Transformer的融合模块,借助其强大的注意力建模能力挖掘高光谱和LiDAR数据特征之间的相互作用;为提高特征提取方法对不同地物类别的适应性,设计了多尺度动态3D卷积模块,将输入特征的多尺度信息融入卷积核的调制,提高卷积操作对不同地物的适应性。采用多源遥感数据集Houston和Trento对所提方法进行验证,实验结果表明:所提方法在Houston和Trento数据集上总体准确率分别达到94.60%和98.21%,相比MGA-MFN等主流方法,总体准确率分别至少提升0.97%和0.25%,验证了所提方法可有效提升多源遥感图像分类的准确率。