A Doubly Graduated Method for Inference in Markov Random Field\ast

doi:10.1137/20M1383574

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 机器人理论与应用

	A Doubly Graduated Method for Inference in Markov Random Field\ast
	Yang, Xu1,2 ; Liu, Zhi-Yong1,2
发表期刊	SIAM JOURNAL ON IMAGING SCIENCES
ISSN	1936-4954
	2021
卷号	14 期号:3 页码:1354-1373
通讯作者	Yang, Xu(xu.yang@ia.ac.cn)
摘要	Maximum a posteriori (MAP) inference in Markov random field (MRF) lays the foundation for many computer vision tasks, which can be formulated by a binary quadratic programming (BQP) problem. Compared with the discrete methods, the continuous relaxation scheme becomes popular due to its generality and efficiency. However, existing continuous relaxation based MAP algorithms are still limited by two problems, i.e., the highly nonconvex original objective function and the gap between the original BQP problem and the relaxed continuous optimization problem. Targeting the two problems, this paper presents a doubly graduated continuous relaxation algorithm for MAP inference in MRF, which are, respectively, the Gaussian smoothing based graduated nonconvexity process and conditional gradient ascent based graduated projection. Experiments on both synthetic data and real-world images illustrate the algorithm's state-of-the-art performance in objective function optimization and typical computer vision tasks.
关键词	maximum a posteriori Markov random field Gaussian smoothing continuous relaxation
DOI	10.1137/20M1383574
关键词[WOS]	ENERGY MINIMIZATION
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation (NSFC) of China[61973301] ; National Natural Science Foundation (NSFC) of China[61972020] ; National Natural Science Foundation (NSFC) of China[61633009] ; National Key R&D Program of China[2020AAA0108902] ; Beijing Science and Technology Plan Project[Z201100008320029]
项目资助者	National Natural Science Foundation (NSFC) of China ; National Key R&D Program of China ; Beijing Science and Technology Plan Project
WOS研究方向	Computer Science ; Mathematics ; Imaging Science & Photographic Technology
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Software Engineering ; Mathematics, Applied ; Imaging Science & Photographic Technology
WOS记录号	WOS:000735767700006
出版者	SIAM PUBLICATIONS
七大方向——子方向分类	模式识别基础
引用统计
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/47122
专题	多模态人工智能系统全国重点实验室_机器人理论与应用
通讯作者	Yang, Xu
作者单位	1.Chinese Acad Sci, State Key Lab Management & Control Complex Syst, Inst Automat, Beijing, Peoples R China 2.Univ Chinese Acad Sci, Beijing, Peoples R China
第一作者单位	中国科学院自动化研究所
通讯作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Yang, Xu,Liu, Zhi-Yong. A Doubly Graduated Method for Inference in Markov Random Field\ast[J]. SIAM JOURNAL ON IMAGING SCIENCES,2021,14(3):1354-1373.
APA	Yang, Xu,&Liu, Zhi-Yong.(2021).A Doubly Graduated Method for Inference in Markov Random Field\ast.SIAM JOURNAL ON IMAGING SCIENCES,14(3),1354-1373.
MLA	Yang, Xu,et al."A Doubly Graduated Method for Inference in Markov Random Field\ast".SIAM JOURNAL ON IMAGING SCIENCES 14.3(2021):1354-1373.