Deep Facade Parsing with Occlusions

doi:10.3837/tiis.2022.02.009

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 三维可视计算

	Deep Facade Parsing with Occlusions
	Ma, Wenguang 1; Ma, Wei 1; Xu, Shibiao2
发表期刊	KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS
ISSN	1976-7277
	2022-02-28
卷号	16 期号:2 页码:524-543
通讯作者	Ma, Wei(mawei@bjut.edu.cn)
摘要	Correct facade image parsing is essential to the semantic understanding of outdoor scenes. Unfortunately, there are often various occlusions in front of buildings, which fails many existing methods. In this paper, we propose an end-to-end deep network for facade parsing with occlusions. The network learns to decompose an input image into visible and invisible parts by occlusion reasoning. Then, a context aggregation module is proposed to collect nonlocal cues for semantic segmentation of the visible part. In addition, considering the regularity of man-made buildings, a repetitive pattern completion branch is designed to infer the contents in the invisible regions by referring to the visible part. Finally, the parsing map of the input facade image is generated by fusing the results of the visible and invisible results. Experiments on both synthetic and real datasets demonstrate that the proposed method outperforms state-of-the-art methods in parsing facades with occlusions. Moreover, we applied our method in applications of image inpainting and 3D semantic modeling.
关键词	Facade parsing occlusion repetitive pattern man-made structure
DOI	10.3837/tiis.2022.02.009
关键词[WOS]	WINDOW DETECTION ; SEGMENTATION
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[62176010] ; National Natural Science Foundation of China[61771026] ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR)
项目资助者	National Natural Science Foundation of China ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR)
WOS研究方向	Computer Science ; Telecommunications
WOS类目	Computer Science, Information Systems ; Telecommunications
WOS记录号	WOS:000767305400001
出版者	KSII-KOR SOC INTERNET INFORMATION
引用统计	被引频次：1[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/48076
专题	多模态人工智能系统全国重点实验室_三维可视计算
通讯作者	Ma, Wei
作者单位	1.Beijing Univ Technol, Fac Informat Technol, Beijing 100124, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Ma, Wenguang,Ma, Wei,Xu, Shibiao. Deep Facade Parsing with Occlusions[J]. KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS,2022,16(2):524-543.
APA	Ma, Wenguang,Ma, Wei,&Xu, Shibiao.(2022).Deep Facade Parsing with Occlusions.KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS,16(2),524-543.
MLA	Ma, Wenguang,et al."Deep Facade Parsing with Occlusions".KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS 16.2(2022):524-543.