Abstract

Multi-graph matching refers to finding correspondences across graphs, which are traditionally solved by matching all the graphs in a single batch. However in real-world applications, graphs are often collected incrementally, rather than once for all. In this paper, we present an incremental multi-graph matching approach, which deals with the arriving graph utilizing the previous matching results under the global consistency constraint. When a new graph arrives, rather than re-optimizing over all graphs, we propose to partition graphs into subsets with certain topological structure and conduct optimization within each subset. The partitioning procedure is guided by the diversity within partitions and randomness over iterations, and we present an interpretation showing why these two factors are essential. The final matching results are calculated over all subsets via an intersection graph. Extensive experimental results on synthetic and real image datasets show that our algorithm notably improves the efficiency without sacrificing the accuracy.

Original languageEnglish (US)
Title of host publicationComputer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
EditorsVittorio Ferrari, Cristian Sminchisescu, Yair Weiss, Martial Hebert
PublisherSpringer Verlag
Pages142-158
Number of pages17
ISBN (Print)9783030012601
DOIs
StatePublished - Jan 1 2018
Event15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany
Duration: Sep 8 2018Sep 14 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11217 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other15th European Conference on Computer Vision, ECCV 2018
CountryGermany
CityMunich
Period9/8/189/14/18

Fingerprint

Graph Clustering
Graph Matching
Multigraph
Set theory
Randomness
Graph in graph theory
Subset
Graph Partition
Intersection Graphs
Topological Structure
Real-world Applications
Batch
Partitioning
Correspondence
Partition
Iteration
Optimization
Experimental Results

Keywords

  • Determinantal point process
  • Graph clustering
  • Incremental graph matching
  • Multi-graph matching

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Yu, T., Yan, J., Liu, W., & Li, B. (2018). Incremental multi-graph matching via diversity and randomness based graph clustering. In V. Ferrari, C. Sminchisescu, Y. Weiss, & M. Hebert (Eds.), Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings (pp. 142-158). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11217 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-01261-8_9

Incremental multi-graph matching via diversity and randomness based graph clustering. / Yu, Tianshu; Yan, Junchi; Liu, Wei; Li, Baoxin.

Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. ed. / Vittorio Ferrari; Cristian Sminchisescu; Yair Weiss; Martial Hebert. Springer Verlag, 2018. p. 142-158 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11217 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Yu, T, Yan, J, Liu, W & Li, B 2018, Incremental multi-graph matching via diversity and randomness based graph clustering. in V Ferrari, C Sminchisescu, Y Weiss & M Hebert (eds), Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11217 LNCS, Springer Verlag, pp. 142-158, 15th European Conference on Computer Vision, ECCV 2018, Munich, Germany, 9/8/18. https://doi.org/10.1007/978-3-030-01261-8_9
Yu T, Yan J, Liu W, Li B. Incremental multi-graph matching via diversity and randomness based graph clustering. In Ferrari V, Sminchisescu C, Weiss Y, Hebert M, editors, Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. Springer Verlag. 2018. p. 142-158. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-01261-8_9
Yu, Tianshu ; Yan, Junchi ; Liu, Wei ; Li, Baoxin. / Incremental multi-graph matching via diversity and randomness based graph clustering. Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. editor / Vittorio Ferrari ; Cristian Sminchisescu ; Yair Weiss ; Martial Hebert. Springer Verlag, 2018. pp. 142-158 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{de5f6c3464554b0a82d10ec09b89ab6a,
title = "Incremental multi-graph matching via diversity and randomness based graph clustering",
abstract = "Multi-graph matching refers to finding correspondences across graphs, which are traditionally solved by matching all the graphs in a single batch. However in real-world applications, graphs are often collected incrementally, rather than once for all. In this paper, we present an incremental multi-graph matching approach, which deals with the arriving graph utilizing the previous matching results under the global consistency constraint. When a new graph arrives, rather than re-optimizing over all graphs, we propose to partition graphs into subsets with certain topological structure and conduct optimization within each subset. The partitioning procedure is guided by the diversity within partitions and randomness over iterations, and we present an interpretation showing why these two factors are essential. The final matching results are calculated over all subsets via an intersection graph. Extensive experimental results on synthetic and real image datasets show that our algorithm notably improves the efficiency without sacrificing the accuracy.",
keywords = "Determinantal point process, Graph clustering, Incremental graph matching, Multi-graph matching",
author = "Tianshu Yu and Junchi Yan and Wei Liu and Baoxin Li",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-01261-8_9",
language = "English (US)",
isbn = "9783030012601",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "142--158",
editor = "Vittorio Ferrari and Cristian Sminchisescu and Yair Weiss and Martial Hebert",
booktitle = "Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings",

}

TY - GEN

T1 - Incremental multi-graph matching via diversity and randomness based graph clustering

AU - Yu, Tianshu

AU - Yan, Junchi

AU - Liu, Wei

AU - Li, Baoxin

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Multi-graph matching refers to finding correspondences across graphs, which are traditionally solved by matching all the graphs in a single batch. However in real-world applications, graphs are often collected incrementally, rather than once for all. In this paper, we present an incremental multi-graph matching approach, which deals with the arriving graph utilizing the previous matching results under the global consistency constraint. When a new graph arrives, rather than re-optimizing over all graphs, we propose to partition graphs into subsets with certain topological structure and conduct optimization within each subset. The partitioning procedure is guided by the diversity within partitions and randomness over iterations, and we present an interpretation showing why these two factors are essential. The final matching results are calculated over all subsets via an intersection graph. Extensive experimental results on synthetic and real image datasets show that our algorithm notably improves the efficiency without sacrificing the accuracy.

AB - Multi-graph matching refers to finding correspondences across graphs, which are traditionally solved by matching all the graphs in a single batch. However in real-world applications, graphs are often collected incrementally, rather than once for all. In this paper, we present an incremental multi-graph matching approach, which deals with the arriving graph utilizing the previous matching results under the global consistency constraint. When a new graph arrives, rather than re-optimizing over all graphs, we propose to partition graphs into subsets with certain topological structure and conduct optimization within each subset. The partitioning procedure is guided by the diversity within partitions and randomness over iterations, and we present an interpretation showing why these two factors are essential. The final matching results are calculated over all subsets via an intersection graph. Extensive experimental results on synthetic and real image datasets show that our algorithm notably improves the efficiency without sacrificing the accuracy.

KW - Determinantal point process

KW - Graph clustering

KW - Incremental graph matching

KW - Multi-graph matching

UR - http://www.scopus.com/inward/record.url?scp=85055499691&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85055499691&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-01261-8_9

DO - 10.1007/978-3-030-01261-8_9

M3 - Conference contribution

SN - 9783030012601

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 142

EP - 158

BT - Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings

A2 - Ferrari, Vittorio

A2 - Sminchisescu, Cristian

A2 - Weiss, Yair

A2 - Hebert, Martial

PB - Springer Verlag

ER -