典型文献
Hyperspectral Image Classification Based on A Multi-Scale Weighted Kernel Network
文献摘要：
Recently,many deep learning models have shown excellent performance in hyperspectral image(HSI)classification.Among them,networks with mul-tiple convolution kernels of different sizes have been proved to achieve richer receptive fields and extract more representative features than those with a single convolu-tion kernel.However,in most networks,different-sized convolution kernels are usually used directly on multi-branch structures,and the image features extracted from them are fused directly and simply.In this paper,to fully and adaptively explore the multiscale information in both spectral and spatial domains of HSI,a novel multi-scale weighted kernel network(MSWKNet)based on an adapt-ive receptive field is proposed.First,the original HSI cu-bic patches are transformed to the input features by com-bining the principal component analysis and one-dimen-sional spectral convolution.Then,a three-branch net-work with different convolution kernels is designed to convolve the input features,and adaptively adjust the size of the receptive field through the attention mechanism of each branch.Finally,the features extracted from each branch are fused together for the task of classification.Experiments on three well-known hyperspectral data sets show that MSWKNet outperforms many deep learning networks in HSI classification.
文献关键词：
中图分类号：
作者姓名：
SUN Le;XU Bin;LU Zhenyu
作者机构：
School of Computer Science,Nanjing University of Information Science and Technology,Nanjing 210044,China;Engineering Research Center of Digital Forensics,Ministry of Education,Nanjing University of Information Science and Technology,Nanjing 210044,China;School of Artificial Intelligence,Nanjing University of Information Science and Technology,Nanjing 210044,China
文献出处：
引用格式：
[1]SUN Le;XU Bin;LU Zhenyu-.Hyperspectral Image Classification Based on A Multi-Scale Weighted Kernel Network)[J].电子学报（英文）,2022(05):832-843
A类：
MSWKNet,convolve
B类：
Hyperspectral,Image,Classification,Based,Multi,Scale,Weighted,Kernel,Network,Recently,many,deep,learning,models,have,shown,excellent,performance,hyperspectral,image,HSI,classification,Among,them,networks,tiple,convolution,kernels,different,sizes,been,proved,achieve,richer,receptive,fields,more,representative,features,than,those,single,However,most,sized,are,usually,directly,branch,structures,extracted,from,fused,simply,In,this,paper,fully,adaptively,explore,multiscale,information,both,spatial,domains,novel,weighted,proposed,First,original,cu,bic,patches,transformed,input,by,bining,principal,component,analysis,dimen,sional,Then,three,designed,adjust,through,attention,mechanism,each,Finally,together,task,Experiments,well,known,data,sets,that,outperforms
AB值：
0.522705
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
