Branching Path Following for Graph Matching

被引:7
作者
Wang, Tao [1 ,2 ]
Ling, Haibin [1 ,3 ]
Lang, Congyan [2 ]
Wu, Jun [2 ]
机构
[1] HiScene Informat Technol, Meitu HiScene Lab, Shanghai, Peoples R China
[2] Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing, Peoples R China
[3] Temple Univ, Dept Comp & Informat Sci, Philadelphia, PA 19122 USA
来源
COMPUTER VISION - ECCV 2016, PT II | 2016年 / 9906卷
基金
美国国家科学基金会;
关键词
Graph matching; Path following; Numerical continuation; Singular point; Branch switching; MODELS;
D O I
10.1007/978-3-319-46475-6_32
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recently, graph matching algorithms utilizing the path following strategy have exhibited state-of-the-art performances. However, the paths computed in these algorithms often contain singular points, which usually hurt the matching performance. To deal with this issue, in this paper we propose a novel path following strategy, named branching path following (BPF), which consequently improves graph matching performance. In particular, we first propose a singular point detector by solving an KKT system, and then design a branch switching method to seek for better paths at singular points. Using BPF, a new graph matching algorithm named BPF-G is developed by applying BPF to a recently proposed path following algorithm named GNCCP (Liu &Qiao 2014). For evaluation, we compare BPF-G with several recently proposed graph matching algorithms on a synthetic dataset and four public bench-mark datasets. Experimental results show that our approach achieves remarkable improvement in matching accuracy and outperforms other algorithms.
引用
收藏
页码:508 / 523
页数:16
相关论文
共 45 条
[1]   Discrete Tabu Search for Graph Matching [J].
Adamczewski, Kamil ;
Suh, Yumin ;
Lee, Kyoung Mu .
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2015, :109-117
[2]  
Allgower E. L., 1990, Numerical Continuation Methods: an Introduction
[3]  
[Anonymous], 2010, ECCV
[4]   Learning Context-Sensitive Shape Similarity by Graph Transduction [J].
Bai, Xiang ;
Yang, Xingwei ;
Latecki, Longin Jan ;
Liu, Wenyu ;
Tu, Zhuowen .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2010, 32 (05) :861-874
[5]   A subspace, interior, and conjugate gradient method for large-scale bound-constrained minimization problems [J].
Branch, MA ;
Coleman, TF ;
Li, YY .
SIAM JOURNAL ON SCIENTIFIC COMPUTING, 1999, 21 (01) :1-23
[6]   Graphical models and point pattern matching [J].
Caetano, Tiberio S. ;
Caelli, Terry ;
Schuurmans, Dale ;
Barone, Dante A. C. .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2006, 28 (10) :1646-1663
[7]   Robust Deformable and Occluded Object Tracking With Dynamic Graph [J].
Cai, Zhaowei ;
Wen, Longyin ;
Lei, Zhen ;
Vasconcelos, Nuno ;
Li, Stan Z. .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2014, 23 (12) :5497-5509
[8]  
Chen C, 2012, IEEE POSITION LOCAT, P1274, DOI 10.1109/PLANS.2012.6236984
[9]   Learning Graphs to Match [J].
Cho, Minsu ;
Alahari, Karteek ;
Ponce, Jean .
2013 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2013, :25-32
[10]   Feature Correspondence and Deformable Object Matching via Agglomerative Correspondence Clustering [J].
Cho, Minsu ;
Lee, Jungmin ;
Lee, Kyoung Mu .
2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2009, :1280-1287