Semi-global shape-aware attention network for image segmentation and retrieval

doi:10.1016/j.neucom.2022.07.069

	Semi-global shape-aware attention network for image segmentation and retrieval
	Zhang, Pengju1,2 ; Zhu, Jiagang 4; Zhang, Chaofan 3; Rong, Zheng1 ; Wu, Yihong1,2
发表期刊	NEUROCOMPUTING
ISSN	0925-2312
	2022-09-28
卷号	506 页码:369-379
通讯作者	Zhang, Chaofan(zcfan@aiofm.ac.cn) ; Rong, Zheng(zheng.rong@nlpr.ia.ac.cn) ; Wu, Yihong(yhwu@nlpr.ia.ac.cn)
摘要	Non-local operations are usually used to capture long-range dependencies via aggregating global context to each position recently. However, most of the methods cannot preserve object shapes since they only focus on feature similarity but ignore proximity between central and other positions for capturing long-range dependencies, while shape-awareness is beneficial to many computer vision tasks. In this paper, we propose a Semi-Global Shape-aware Network (SGSNet) considering both feature similarity and prox-imity for preserving object shapes when modeling long-range dependencies. A hierarchical way is taken to aggregate global context. In the first level, each position in the whole feature map only aggregates con-textual information in vertical and horizontal directions according to both similarity and proximity. And then the result is input into the second level to do the same operations. By this hierarchical way, each central position gains supports from all other positions, and the combination of similarity and proximity makes each position gain supports mostly from the same semantic object. Moreover, we also propose an efficient algorithm for the aggregation of contextual information, where each of rows and columns in the feature map is treated as a binary tree to reduce similarity computation cost. Experiments on semantic segmentation and image retrieval show that adding SGSNet to existing networks gains solid improve-ments on both accuracy and efficiency.(c) 2022 Elsevier B.V. All rights reserved.
关键词	Attention network Shape -awareness Semantic segmentation Image retrieval
DOI	10.1016/j.neucom.2022.07.069
收录类别	SCI
语种	英语
资助项目	National Natural Science Founda- tion of China[61836015] ; National Natural Science Founda- tion of China[62002359] ; National Natural Science Founda- tion of China[62102395]
项目资助者	National Natural Science Founda- tion of China
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:000843462400012
出版者	ELSEVIER
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/173211/49892
专题	多模态人工智能系统全国重点实验室_机器人视觉
通讯作者	Zhang, Chaofan; Rong, Zheng; Wu, Yihong
作者单位	1.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 3.Chinese Acad Sci, Anhui Inst Opt & Fine Mech, Hefei Inst Phys Sci, Anhui 230031, Peoples R China 4.XForwardAI Technol Co Ltd, Beijing, Peoples R China
第一作者单位	模式识别国家重点实验室
通讯作者单位	模式识别国家重点实验室
推荐引用方式 GB/T 7714	Zhang, Pengju,Zhu, Jiagang,Zhang, Chaofan,et al. Semi-global shape-aware attention network for image segmentation and retrieval[J]. NEUROCOMPUTING,2022,506:369-379.
APA	Zhang, Pengju,Zhu, Jiagang,Zhang, Chaofan,Rong, Zheng,&Wu, Yihong.(2022).Semi-global shape-aware attention network for image segmentation and retrieval.NEUROCOMPUTING,506,369-379.
MLA	Zhang, Pengju,et al."Semi-global shape-aware attention network for image segmentation and retrieval".NEUROCOMPUTING 506(2022):369-379.