典型文献
Element-Arrangement Context Network for Facade Parsing
文献摘要：
Facade parsing aims to decompose a building facade image into semantic regions of the facade objects.Consi-dering each architectural element on a facade as a parameterized rectangle,we formulate the facade parsing task as object detection,allowing overlapping and nesting,which will support structural 3D modeling and editing for further applications.In contrast to general object detection,the spatial arrangement regularity and appearance similarity between the facade elements of the same category provide valuable context for accurate element localization.In this paper,we propose to exploit the spatial arrangement regularity and appearance similarity of facade elements in a detection framework.Our element-arrangement context network(EACNet)consists of two unidirectional attention branches,one to capture the column-context and the other to capture row-context to aggregate element-specific features from multiple instances on the facade.We con-duct extensive experiments on four public datasets(ECP,CMP,Graz50,and eTRIMS).The proposed EACNet achieves the highest mIoU(82.1％on ECP,77.35％on Graz50,and 82.3％on eTRIMS)compared with the state-of-the-art methods.Both the quantitative and qualitative evaluation results demonstrate the effectiveness of our dual unidirectional attention branches to parse facade elements.
文献关键词：
中图分类号：
作者姓名：
Yan Tao;Yi-Teng Zhang;Xue-Jin Chen
作者机构：
National Engineering Laboratory for Brain-Inspired Intelligence Technology and Application,University of Science and Technology of China,Hefei 230026,China
文献出处：
引用格式：
[1]Yan Tao;Yi-Teng Zhang;Xue-Jin Chen-.Element-Arrangement Context Network for Facade Parsing)[J].计算机科学技术学报（英文版）,2022(03):652-665
A类：
Facade,facade,Consi,EACNet,Graz50,eTRIMS,parse
B类：
Element,Arrangement,Context,Network,Parsing,parsing,aims,decompose,building,image,into,semantic,regions,objects,dering,each,architectural,parameterized,rectangle,formulate,task,detection,allowing,overlapping,nesting,which,will,support,structural,modeling,editing,further,applications,In,contrast,general,spatial,arrangement,regularity,appearance,similarity,between,elements,same,category,provide,valuable,context,accurate,localization,this,paper,exploit,framework,Our,network,consists,unidirectional,attention,branches,one,capture,column,other,row,aggregate,specific,features,from,multiple,instances,We,duct,extensive,experiments,four,public,datasets,ECP,CMP,proposed,achieves,highest,mIoU,compared,state,art,methods,Both,quantitative,qualitative,evaluation,results,demonstrate,effectiveness,dual
AB值：
0.51901
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。
