To fully extract and mine the multi-scale features of reservoirs and geologic structures in time/depth and space dimensions, a new 3D multi-scale volumetric curvature （MSVC） methodology is presented in this paper. W...To fully extract and mine the multi-scale features of reservoirs and geologic structures in time/depth and space dimensions, a new 3D multi-scale volumetric curvature （MSVC） methodology is presented in this paper. We also propose a fast algorithm for computing 3D volumetric curvature. In comparison to conventional volumetric curvature attributes, its main improvements and key algorithms introduce multi-frequency components expansion in time-frequency domain and the corresponding multi-scale adaptive differential operator in the wavenumber domain, into the volumetric curvature calculation. This methodology can simultaneously depict seismic multi-scale features in both time and space. Additionally, we use data fusion of volumetric curvatures at various scales to take full advantage of the geologic features and anomalies extracted by curvature measurements at different scales. The 3D MSVC can highlight geologic anomalies and reduce noise at the same time. Thus, it improves the interpretation efficiency of curvature attributes analysis. The 3D MSVC is applied to both land and marine 3D seismic data. The results demonstrate that it can indicate the spatial distribution of reservoirs, detect faults and fracture zones, and identify their multi-scale properties.展开更多
Thermal conductivity is one of the most significant criterion of three-dimensional carbon fiber-reinforced SiC matrix composites(3D C/SiC).Represent volume element(RVE)models of microscale,void/matrix and mesoscale pr...Thermal conductivity is one of the most significant criterion of three-dimensional carbon fiber-reinforced SiC matrix composites(3D C/SiC).Represent volume element(RVE)models of microscale,void/matrix and mesoscale proposed in this work are used to simulate the thermal conductivity behaviors of the 3D C/SiC composites.An entirely new process is introduced to weave the preform with three-dimensional orthogonal architecture.The 3D steady-state analysis step is created for assessing the thermal conductivity behaviors of the composites by applying periodic temperature boundary conditions.Three RVE models of cuboid,hexagonal and fiber random distribution are respectively developed to comparatively study the influence of fiber package pattern on the thermal conductivities at the microscale.Besides,the effect of void morphology on the thermal conductivity of the matrix is analyzed by the void/matrix models.The prediction results at the mesoscale correspond closely to the experimental values.The effect of the porosities and fiber volume fractions on the thermal conductivities is also taken into consideration.The multi-scale models mentioned in this paper can be used to predict the thermal conductivity behaviors of other composites with complex structures.展开更多
[Objective]Leaf diseases significantly affect both the yield and quality of tea throughout the year.To address the issue of inadequate segmentation finesse in the current tea spot segmentation models,a novel diagnosis...[Objective]Leaf diseases significantly affect both the yield and quality of tea throughout the year.To address the issue of inadequate segmentation finesse in the current tea spot segmentation models,a novel diagnosis of the severity of tea spots was proposed in this research,designated as MDC-U-Net3+,to enhance segmentation accuracy on the base framework of U-Net3+.[Methods]Multi-scale feature fusion module(MSFFM)was incorporated into the backbone network of U-Net3+to obtain feature information across multiple receptive fields of diseased spots,thereby reducing the loss of features within the encoder.Dual multi-scale attention(DMSA)was incorporated into the skip connection process to mitigate the segmentation boundary ambiguity issue.This integration facilitates the comprehensive fusion of fine-grained and coarse-grained semantic information at full scale.Furthermore,the segmented mask image was subjected to conditional random fields(CRF)to enhance the optimization of the segmentation results[Results and Discussions]The improved model MDC-U-Net3+achieved a mean pixel accuracy(mPA)of 94.92%,accompanied by a mean Intersection over Union(mIoU)ratio of 90.9%.When compared to the mPA and mIoU of U-Net3+,MDC-U-Net3+model showed improvements of 1.85 and 2.12 percentage points,respectively.These results illustrated a more effective segmentation performance than that achieved by other classical semantic segmentation models.[Conclusions]The methodology presented herein could provide data support for automated disease detection and precise medication,consequently reducing the losses associated with tea diseases.展开更多
针对卷积神经网络在高光谱图像特征提取和分类的过程中,存在空谱特征提取不充分以及网络层数太多引起的参数量大、计算复杂的问题,提出快速三维卷积神经网络(3D-CNN)结合深度可分离卷积(DSC)的轻量型卷积模型。该方法首先利用增量主成...针对卷积神经网络在高光谱图像特征提取和分类的过程中,存在空谱特征提取不充分以及网络层数太多引起的参数量大、计算复杂的问题,提出快速三维卷积神经网络(3D-CNN)结合深度可分离卷积(DSC)的轻量型卷积模型。该方法首先利用增量主成分分析(IPCA)对输入的数据进行降维预处理;其次将输入模型的像素分割成小的重叠的三维小卷积块,在分割的小块上基于中心像素形成地面标签,利用三维核函数进行卷积处理,形成连续的三维特征图,保留空谱特征。用3D-CNN同时提取空谱特征,然后在三维卷积中加入深度可分离卷积对空间特征再次提取,丰富空谱特征的同时减少参数量,从而减少计算时间,分类精度也有所提高。所提模型在Indian Pines、Salinas Scene和University of Pavia公开数据集上验证,并且同其他经典的分类方法进行比较。实验结果表明,该方法不仅能大幅度节省可学习的参数,降低模型复杂度,而且表现出较好的分类性能,其中总体精度(OA)、平均分类精度(AA)和Kappa系数均可达99%以上。展开更多
针对高光谱遥感图像分类中空间信息利用不充分、样本标记数量不足的问题,提出一种基于多尺度3D-CNN和卷积块注意力机制的高光谱图像分类方法。采用特征映射方式从不同感受野充分挖掘并融合高光谱图像的空间特征和光谱特征,对融合后的空...针对高光谱遥感图像分类中空间信息利用不充分、样本标记数量不足的问题,提出一种基于多尺度3D-CNN和卷积块注意力机制的高光谱图像分类方法。采用特征映射方式从不同感受野充分挖掘并融合高光谱图像的空间特征和光谱特征,对融合后的空谱特征进行卷积块注意力机制处理;通过残差思想构建深层网络,采用Dropout方法处理过拟合问题,最后通过Softmax分类器进行分类。在Indian Pines、Pavia University和Salinas Valley 3个高光谱数据集上进行大量实验,分类结果表明:所提方法优于其他经典方法。展开更多
A native organ has heterogeneous structures, sirength, and cell components. It is a big challenge to fabricate organ prototypes with controllable shapes, strength, and cells. Herein, a hybrid method is developed to fa...A native organ has heterogeneous structures, sirength, and cell components. It is a big challenge to fabricate organ prototypes with controllable shapes, strength, and cells. Herein, a hybrid method is developed to fabricate organ prototypes with controlled cell deposition by integrating extrusion-based 3D printing, electrospinning, and 3D bioprinting. Multi-scale sheets were first fabricated by 3D printing and electrospinning;then, all the sheets were assembled into organ prototypes by sol-gel react io n duri ng bioprinting. With this method, macroscale structures fabricated by 3D printing ensure the customized structures and provide mechanical support, nanoscale structures fabricated by electrospinning offer a favorable environment for cell growth, and different types of cells with controllable densities are deposited in accurate locations by bioprinting. The results show that L929 mouse fibroblasts encapsulated in the structures exhibited over 90% survival within 10 days and maintai ned a high proliferation rate. Furthermore, the cells grew in spherical shapes first and then migrated to the nano scale fibers showing stretched morphology. Additionally, a branched vascular structure was successfully fabricated using the presented method. Compared with other methods, this strategy offers an easy way to simultancously realize the shape control, nanolibrous structures, and cell accurate deposition, which will have potemidi applications in tissue cngineering.展开更多
Recognition of dynamic hand gestures in real-time is a difficult task because the system can never know when or from where the gesture starts and ends in a video stream.Many researchers have been working on visionbase...Recognition of dynamic hand gestures in real-time is a difficult task because the system can never know when or from where the gesture starts and ends in a video stream.Many researchers have been working on visionbased gesture recognition due to its various applications.This paper proposes a deep learning architecture based on the combination of a 3D Convolutional Neural Network(3D-CNN)and a Long Short-Term Memory(LSTM)network.The proposed architecture extracts spatial-temporal information from video sequences input while avoiding extensive computation.The 3D-CNN is used for the extraction of spectral and spatial features which are then given to the LSTM network through which classification is carried out.The proposed model is a light-weight architecture with only 3.7 million training parameters.The model has been evaluated on 15 classes from the 20BN-jester dataset available publicly.The model was trained on 2000 video-clips per class which were separated into 80%training and 20%validation sets.An accuracy of 99%and 97%was achieved on training and testing data,respectively.We further show that the combination of 3D-CNN with LSTM gives superior results as compared to MobileNetv2+LSTM.展开更多
The Secondary Air System(SAS)plays an important role in the safe operation and performance of aeroengines.The traditional 1D-3D coupling method loses information when used for secondary air systems,which affects the c...The Secondary Air System(SAS)plays an important role in the safe operation and performance of aeroengines.The traditional 1D-3D coupling method loses information when used for secondary air systems,which affects the calculation accuracy.In this paper,a Cross-dimensional Data Transmission method(CDT)from 3D to 1D is proposed by introducing flow field uniformity into the data transmission.First,a uniformity index was established to quantify the flow field parameter distribution characteristics,and a uniformity index prediction model based on the locally weighted regression method(Lowess)was established to quickly obtain the flow field information.Then,an information selection criterion in 3D to 1D data transmission was established based on the Spearman rank correlation coefficient between the uniformity index and the accuracy of coupling calculation,and the calculation method was automatically determined according to the established criterion.Finally,a modified function was obtained by fitting the ratio of the 3D mass-average parameters to the analytical solution,which are then used to modify the selected parameters at the 1D-3D interface.Taking a typical disk cavity air system as an example,the results show that the calculation accuracy of the CDT method is greatly improved by a relative 53.88%compared with the traditional 1D-3D coupling method.Furthermore,the CDT method achieves a speedup of 2 to 3 orders of magnitude compared to the 3D calculation.展开更多
In this study,a novel multi-physics multi-scale model with the dilute multicomponent phase-field method in three-dimensional(3D)space was developed to investigate the complex microstructure evolu-tion in the molten po...In this study,a novel multi-physics multi-scale model with the dilute multicomponent phase-field method in three-dimensional(3D)space was developed to investigate the complex microstructure evolu-tion in the molten pool during laser welding of Al-Li alloy.To accurately compute mass data within both two and three-dimensional computational domains,three efficient computing methods,including central processing unit parallel computing,adaptive mesh refinement,and moving-frame algorithm,were uti-lized.Emphasis was placed on the distinctive equiaxed-to-columnar-to-equiaxed transition phenomenon that occurs during the entire solidification process of Al-Li alloy laser welding.Simulation results indi-cated that the growth distance of columnar grains that epitaxially grew from the base metal(BM)de-creased as the nucleation rate increased.As the nucleation rate increased,the morphology of the newly formed grains near the fusion boundary(FB)changed from columnar to equiaxed,and newly formed equiaxed grains changed from having high-order dendrites to no obvious dendrite structure.When the nucleation rate was sufficiently high,non-dendritic equiaxed grains could directly form near the FB,and there was nearly no epitaxial growth from the BM.Additionally,simulation results illustrated the com-petition among multiple grains with varying orientations that grow in 3D space near the FB.Finally,how equiaxed grain bands develop was elucidated.The equiaxed band not only hindered the growth of early columnar grains but also some of its grains could grow epitaxially to form new columnar grains.These predicted results were in good agreement with experimental measurements and observations.展开更多
肺癌是长期威胁人类健康的恶性疾病之一,针对传统方法在肺癌CT图像分类中的预处理过程复杂、工作量大的问题,本文提出了基于三维卷积神经网络(3D-CNN)模型的肺部CT图像分类方法。该模型以卷积神经网络模型为基础,并在训练的过程中使用...肺癌是长期威胁人类健康的恶性疾病之一,针对传统方法在肺癌CT图像分类中的预处理过程复杂、工作量大的问题,本文提出了基于三维卷积神经网络(3D-CNN)模型的肺部CT图像分类方法。该模型以卷积神经网络模型为基础,并在训练的过程中使用特定顺序输入策略,还在公开的Kaggle Data Science Bowl 2017数据集上进行了实验。实验表明,该方法对图像的分类准确率达到76%,比采用随机顺序的输入策略时有所提升,能够为肺部病理图像的分类研究提供有价值的参考。展开更多
In lung nodules there is a huge variation in structural properties like Shape, Surface Texture. Even the spatial properties vary, where they can be found attached to lung walls, blood vessels in complex non-homogenous...In lung nodules there is a huge variation in structural properties like Shape, Surface Texture. Even the spatial properties vary, where they can be found attached to lung walls, blood vessels in complex non-homogenous lung structures. Moreover, the nodules are of small size at their early stage of development. This poses a serious challenge to develop a Computer aided diagnosis (CAD) system with better false positive reduction. Hence, to reduce the false positives per scan and to deal with the challenges mentioned, this paper proposes a set of three diverse 3D Attention based CNN architectures (3D ACNN) whose predictions on given low dose Volumetric Computed Tomography (CT) scans are fused to achieve more effective and reliable results. Attention mechanism is employed to selectively concentrate/weigh more on nodule specific features and less weight age over other irrelevant features. By using this attention based mechanism in CNN unlike traditional methods there was a significant gain in the classification performance. Contextual dependencies are also taken into account by giving three patches of different sizes surrounding the nodule as input to the ACNN architectures. The system is trained and validated using a publicly available LUNA16 dataset in a 10 fold cross validation approach where a competition performance metric (CPM) score of 0.931 is achieved. The experimental results demonstrate that either a single patch or a single architecture in a one-to-one fashion that is adopted in earlier methods cannot achieve a better performance and signifies the necessity of fusing different multi patched architectures. Though the proposed system is mainly designed for pulmonary nodule detection it can be easily extended to classification tasks of any other 3D medical diagnostic computed tomography images where there is a huge variation and uncertainty in classification.展开更多
The Unmanned Aerial Vehicle(UAV)-assisted sensing-transmission--computing integrated system plays a vital role in emergency rescue scenarios involving damaged infrastructure.To tackle the challenges of data transmissi...The Unmanned Aerial Vehicle(UAV)-assisted sensing-transmission--computing integrated system plays a vital role in emergency rescue scenarios involving damaged infrastructure.To tackle the challenges of data transmission and enable timely rescue decision-making,we propose DWT-3DRec-an efficient wireless transmission model for 3D scene reconstruction.This model leverages MobileNetV2 to extract image and pose features,which are transmitted through a Dual-path Adaptive Noise Modulation network(DANM).Moreover,we introduce the Gumbel Channel Masking Module(GCMM),which enhances feature extraction and improves reconstruction reliability by mitigating the effects of dynamic noise.At the ground receiver,the Multi-scale Deep Source-Channel Coding for 3D Reconstruction(MDS-3DRecon)framework integrates Deep Joint Source-Channel Coding(DeepJSCC)with Cityscale Neural Radiance Fields(CityNeRF).It adopts a progressive close-view training strategy and incorporates an Adaptive Fusion Module(AFM)to achieve high-precision scene reconstruction.Experimental results demonstrate that DWT-3DRec significantly outperforms the Joint Photographic Experts Group(JPEG)standard in transmitting image and pose data,achieving an average loss as low as 0.0323 and exhibiting strong robustness across a Signal-to-Noise Ratio(SNR)range of 5--20 dB.In large-scale 3D scene reconstruction tasks,MDS-3DRecon surpasses Multum in Parvo Neural Radiance Fields(Mip-NeRF)and Bungee Neural Radiance Field(BungeeNeRF),achieving a Peak Signal-to-Noise Ratio(PSNR)of 24.921 dB and a reconstruction loss of 0.188.Ablation studies further confirm the essential roles of GCMM,DANM,and AFM in enabling highfidelity 3D reconstruction.展开更多
基金supported by the National Natural Science Foundation of China (No. 41004054) Research Fund for the Doctoral Program of Higher Education of China (No. 20105122120002)Natural Science Key Project, Sichuan Provincial Department of Education (No. 092A011)
文摘To fully extract and mine the multi-scale features of reservoirs and geologic structures in time/depth and space dimensions, a new 3D multi-scale volumetric curvature （MSVC） methodology is presented in this paper. We also propose a fast algorithm for computing 3D volumetric curvature. In comparison to conventional volumetric curvature attributes, its main improvements and key algorithms introduce multi-frequency components expansion in time-frequency domain and the corresponding multi-scale adaptive differential operator in the wavenumber domain, into the volumetric curvature calculation. This methodology can simultaneously depict seismic multi-scale features in both time and space. Additionally, we use data fusion of volumetric curvatures at various scales to take full advantage of the geologic features and anomalies extracted by curvature measurements at different scales. The 3D MSVC can highlight geologic anomalies and reduce noise at the same time. Thus, it improves the interpretation efficiency of curvature attributes analysis. The 3D MSVC is applied to both land and marine 3D seismic data. The results demonstrate that it can indicate the spatial distribution of reservoirs, detect faults and fracture zones, and identify their multi-scale properties.
基金Supported by Science Center for Gas Turbine Project of China (Grant No.P2022-B-IV-014-001)Frontier Leading Technology Basic Research Special Project of Jiangsu Province of China (Grant No.BK20212007)the BIT Research and Innovation Promoting Project of China (Grant No.2022YCXZ019)。
文摘Thermal conductivity is one of the most significant criterion of three-dimensional carbon fiber-reinforced SiC matrix composites(3D C/SiC).Represent volume element(RVE)models of microscale,void/matrix and mesoscale proposed in this work are used to simulate the thermal conductivity behaviors of the 3D C/SiC composites.An entirely new process is introduced to weave the preform with three-dimensional orthogonal architecture.The 3D steady-state analysis step is created for assessing the thermal conductivity behaviors of the composites by applying periodic temperature boundary conditions.Three RVE models of cuboid,hexagonal and fiber random distribution are respectively developed to comparatively study the influence of fiber package pattern on the thermal conductivities at the microscale.Besides,the effect of void morphology on the thermal conductivity of the matrix is analyzed by the void/matrix models.The prediction results at the mesoscale correspond closely to the experimental values.The effect of the porosities and fiber volume fractions on the thermal conductivities is also taken into consideration.The multi-scale models mentioned in this paper can be used to predict the thermal conductivity behaviors of other composites with complex structures.
文摘[Objective]Leaf diseases significantly affect both the yield and quality of tea throughout the year.To address the issue of inadequate segmentation finesse in the current tea spot segmentation models,a novel diagnosis of the severity of tea spots was proposed in this research,designated as MDC-U-Net3+,to enhance segmentation accuracy on the base framework of U-Net3+.[Methods]Multi-scale feature fusion module(MSFFM)was incorporated into the backbone network of U-Net3+to obtain feature information across multiple receptive fields of diseased spots,thereby reducing the loss of features within the encoder.Dual multi-scale attention(DMSA)was incorporated into the skip connection process to mitigate the segmentation boundary ambiguity issue.This integration facilitates the comprehensive fusion of fine-grained and coarse-grained semantic information at full scale.Furthermore,the segmented mask image was subjected to conditional random fields(CRF)to enhance the optimization of the segmentation results[Results and Discussions]The improved model MDC-U-Net3+achieved a mean pixel accuracy(mPA)of 94.92%,accompanied by a mean Intersection over Union(mIoU)ratio of 90.9%.When compared to the mPA and mIoU of U-Net3+,MDC-U-Net3+model showed improvements of 1.85 and 2.12 percentage points,respectively.These results illustrated a more effective segmentation performance than that achieved by other classical semantic segmentation models.[Conclusions]The methodology presented herein could provide data support for automated disease detection and precise medication,consequently reducing the losses associated with tea diseases.
文摘针对卷积神经网络在高光谱图像特征提取和分类的过程中,存在空谱特征提取不充分以及网络层数太多引起的参数量大、计算复杂的问题,提出快速三维卷积神经网络(3D-CNN)结合深度可分离卷积(DSC)的轻量型卷积模型。该方法首先利用增量主成分分析(IPCA)对输入的数据进行降维预处理;其次将输入模型的像素分割成小的重叠的三维小卷积块,在分割的小块上基于中心像素形成地面标签,利用三维核函数进行卷积处理,形成连续的三维特征图,保留空谱特征。用3D-CNN同时提取空谱特征,然后在三维卷积中加入深度可分离卷积对空间特征再次提取,丰富空谱特征的同时减少参数量,从而减少计算时间,分类精度也有所提高。所提模型在Indian Pines、Salinas Scene和University of Pavia公开数据集上验证,并且同其他经典的分类方法进行比较。实验结果表明,该方法不仅能大幅度节省可学习的参数,降低模型复杂度,而且表现出较好的分类性能,其中总体精度(OA)、平均分类精度(AA)和Kappa系数均可达99%以上。
文摘针对高光谱遥感图像分类中空间信息利用不充分、样本标记数量不足的问题,提出一种基于多尺度3D-CNN和卷积块注意力机制的高光谱图像分类方法。采用特征映射方式从不同感受野充分挖掘并融合高光谱图像的空间特征和光谱特征,对融合后的空谱特征进行卷积块注意力机制处理;通过残差思想构建深层网络,采用Dropout方法处理过拟合问题,最后通过Softmax分类器进行分类。在Indian Pines、Pavia University和Salinas Valley 3个高光谱数据集上进行大量实验,分类结果表明:所提方法优于其他经典方法。
基金the National Nature Science Foundation of China(Nos.51805474,51622510,U1609207)Science Fund for Creative Research Groups of National Natural Science Foundation of China (No.51821093)China Postdoctoral Science Foundation (No.2017M621915).
文摘A native organ has heterogeneous structures, sirength, and cell components. It is a big challenge to fabricate organ prototypes with controllable shapes, strength, and cells. Herein, a hybrid method is developed to fabricate organ prototypes with controlled cell deposition by integrating extrusion-based 3D printing, electrospinning, and 3D bioprinting. Multi-scale sheets were first fabricated by 3D printing and electrospinning;then, all the sheets were assembled into organ prototypes by sol-gel react io n duri ng bioprinting. With this method, macroscale structures fabricated by 3D printing ensure the customized structures and provide mechanical support, nanoscale structures fabricated by electrospinning offer a favorable environment for cell growth, and different types of cells with controllable densities are deposited in accurate locations by bioprinting. The results show that L929 mouse fibroblasts encapsulated in the structures exhibited over 90% survival within 10 days and maintai ned a high proliferation rate. Furthermore, the cells grew in spherical shapes first and then migrated to the nano scale fibers showing stretched morphology. Additionally, a branched vascular structure was successfully fabricated using the presented method. Compared with other methods, this strategy offers an easy way to simultancously realize the shape control, nanolibrous structures, and cell accurate deposition, which will have potemidi applications in tissue cngineering.
文摘Recognition of dynamic hand gestures in real-time is a difficult task because the system can never know when or from where the gesture starts and ends in a video stream.Many researchers have been working on visionbased gesture recognition due to its various applications.This paper proposes a deep learning architecture based on the combination of a 3D Convolutional Neural Network(3D-CNN)and a Long Short-Term Memory(LSTM)network.The proposed architecture extracts spatial-temporal information from video sequences input while avoiding extensive computation.The 3D-CNN is used for the extraction of spectral and spatial features which are then given to the LSTM network through which classification is carried out.The proposed model is a light-weight architecture with only 3.7 million training parameters.The model has been evaluated on 15 classes from the 20BN-jester dataset available publicly.The model was trained on 2000 video-clips per class which were separated into 80%training and 20%validation sets.An accuracy of 99%and 97%was achieved on training and testing data,respectively.We further show that the combination of 3D-CNN with LSTM gives superior results as compared to MobileNetv2+LSTM.
基金supported by the National Science and Technology Major Project,China(No.2017-III-0010-0036).
文摘The Secondary Air System(SAS)plays an important role in the safe operation and performance of aeroengines.The traditional 1D-3D coupling method loses information when used for secondary air systems,which affects the calculation accuracy.In this paper,a Cross-dimensional Data Transmission method(CDT)from 3D to 1D is proposed by introducing flow field uniformity into the data transmission.First,a uniformity index was established to quantify the flow field parameter distribution characteristics,and a uniformity index prediction model based on the locally weighted regression method(Lowess)was established to quickly obtain the flow field information.Then,an information selection criterion in 3D to 1D data transmission was established based on the Spearman rank correlation coefficient between the uniformity index and the accuracy of coupling calculation,and the calculation method was automatically determined according to the established criterion.Finally,a modified function was obtained by fitting the ratio of the 3D mass-average parameters to the analytical solution,which are then used to modify the selected parameters at the 1D-3D interface.Taking a typical disk cavity air system as an example,the results show that the calculation accuracy of the CDT method is greatly improved by a relative 53.88%compared with the traditional 1D-3D coupling method.Furthermore,the CDT method achieves a speedup of 2 to 3 orders of magnitude compared to the 3D calculation.
基金supported by the National Natural Science Foundation of China(Grant Nos.52075201,U22A20196,52188102)GuangDong Basic and Applied Basic Research Foundation(No.2023A1515010081).
文摘In this study,a novel multi-physics multi-scale model with the dilute multicomponent phase-field method in three-dimensional(3D)space was developed to investigate the complex microstructure evolu-tion in the molten pool during laser welding of Al-Li alloy.To accurately compute mass data within both two and three-dimensional computational domains,three efficient computing methods,including central processing unit parallel computing,adaptive mesh refinement,and moving-frame algorithm,were uti-lized.Emphasis was placed on the distinctive equiaxed-to-columnar-to-equiaxed transition phenomenon that occurs during the entire solidification process of Al-Li alloy laser welding.Simulation results indi-cated that the growth distance of columnar grains that epitaxially grew from the base metal(BM)de-creased as the nucleation rate increased.As the nucleation rate increased,the morphology of the newly formed grains near the fusion boundary(FB)changed from columnar to equiaxed,and newly formed equiaxed grains changed from having high-order dendrites to no obvious dendrite structure.When the nucleation rate was sufficiently high,non-dendritic equiaxed grains could directly form near the FB,and there was nearly no epitaxial growth from the BM.Additionally,simulation results illustrated the com-petition among multiple grains with varying orientations that grow in 3D space near the FB.Finally,how equiaxed grain bands develop was elucidated.The equiaxed band not only hindered the growth of early columnar grains but also some of its grains could grow epitaxially to form new columnar grains.These predicted results were in good agreement with experimental measurements and observations.
文摘肺癌是长期威胁人类健康的恶性疾病之一,针对传统方法在肺癌CT图像分类中的预处理过程复杂、工作量大的问题,本文提出了基于三维卷积神经网络(3D-CNN)模型的肺部CT图像分类方法。该模型以卷积神经网络模型为基础,并在训练的过程中使用特定顺序输入策略,还在公开的Kaggle Data Science Bowl 2017数据集上进行了实验。实验表明,该方法对图像的分类准确率达到76%,比采用随机顺序的输入策略时有所提升,能够为肺部病理图像的分类研究提供有价值的参考。
文摘In lung nodules there is a huge variation in structural properties like Shape, Surface Texture. Even the spatial properties vary, where they can be found attached to lung walls, blood vessels in complex non-homogenous lung structures. Moreover, the nodules are of small size at their early stage of development. This poses a serious challenge to develop a Computer aided diagnosis (CAD) system with better false positive reduction. Hence, to reduce the false positives per scan and to deal with the challenges mentioned, this paper proposes a set of three diverse 3D Attention based CNN architectures (3D ACNN) whose predictions on given low dose Volumetric Computed Tomography (CT) scans are fused to achieve more effective and reliable results. Attention mechanism is employed to selectively concentrate/weigh more on nodule specific features and less weight age over other irrelevant features. By using this attention based mechanism in CNN unlike traditional methods there was a significant gain in the classification performance. Contextual dependencies are also taken into account by giving three patches of different sizes surrounding the nodule as input to the ACNN architectures. The system is trained and validated using a publicly available LUNA16 dataset in a 10 fold cross validation approach where a competition performance metric (CPM) score of 0.931 is achieved. The experimental results demonstrate that either a single patch or a single architecture in a one-to-one fashion that is adopted in earlier methods cannot achieve a better performance and signifies the necessity of fusing different multi patched architectures. Though the proposed system is mainly designed for pulmonary nodule detection it can be easily extended to classification tasks of any other 3D medical diagnostic computed tomography images where there is a huge variation and uncertainty in classification.
基金supported by the National Key Research and Development Program of China(2022YFB4500800)the Applied Basic Research Program Project of Liaoning Province(2023JH2/101300192)+2 种基金the National Natural Science Foundation of China(62032013,62072094)the Fundamental Research Funds for the Central Universities(N2416006,N2416016)Shenyang Science and Technology Plan Project(ZX20250050).
文摘The Unmanned Aerial Vehicle(UAV)-assisted sensing-transmission--computing integrated system plays a vital role in emergency rescue scenarios involving damaged infrastructure.To tackle the challenges of data transmission and enable timely rescue decision-making,we propose DWT-3DRec-an efficient wireless transmission model for 3D scene reconstruction.This model leverages MobileNetV2 to extract image and pose features,which are transmitted through a Dual-path Adaptive Noise Modulation network(DANM).Moreover,we introduce the Gumbel Channel Masking Module(GCMM),which enhances feature extraction and improves reconstruction reliability by mitigating the effects of dynamic noise.At the ground receiver,the Multi-scale Deep Source-Channel Coding for 3D Reconstruction(MDS-3DRecon)framework integrates Deep Joint Source-Channel Coding(DeepJSCC)with Cityscale Neural Radiance Fields(CityNeRF).It adopts a progressive close-view training strategy and incorporates an Adaptive Fusion Module(AFM)to achieve high-precision scene reconstruction.Experimental results demonstrate that DWT-3DRec significantly outperforms the Joint Photographic Experts Group(JPEG)standard in transmitting image and pose data,achieving an average loss as low as 0.0323 and exhibiting strong robustness across a Signal-to-Noise Ratio(SNR)range of 5--20 dB.In large-scale 3D scene reconstruction tasks,MDS-3DRecon surpasses Multum in Parvo Neural Radiance Fields(Mip-NeRF)and Bungee Neural Radiance Field(BungeeNeRF),achieving a Peak Signal-to-Noise Ratio(PSNR)of 24.921 dB and a reconstruction loss of 0.188.Ablation studies further confirm the essential roles of GCMM,DANM,and AFM in enabling highfidelity 3D reconstruction.
