Iterative maximum clique clustering based detection filter

Xinyu Zhang; Hao Sheng; Yang Zhang; Jiahui Chen; Yubin Wu; Guangtao Xue; Quanrui Wei

doi:10.1007/978-3-030-04212-7_13

Iterative maximum clique clustering based detection filter

Xinyu Zhang
, Hao Sheng^*
, Yang Zhang
, Jiahui Chen
, Yubin Wu
, Guangtao Xue
, Quanrui Wei

^*Corresponding author for this work

School of Computer Science and Engineering

Beihang University
Shanghai Jiao Tong University
China Electronics Technology Group Corporation

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Object detection is an important research field of computer vision, but getting accurate object detection from a large number of detection candidates has always been a challenge. The most current algorithms use an insufficient Greedy Non-Maximum Suppression (NMS) strategy which heavily relies on the confidence of the detection candidates. This paper proposes the Iterative Detection Filter (IDF) approach, which considers more information of the detection candidates, including overlapping, the confidence generated by the detector, and the ground position perception information of the scene. Through this approach, the detection candidates are mapped to more accurate detections. Our method achieves a significant improvement on the MOT16 and MOT17 datasets, which are widely used in video tracking and detection.

Original language	English
Title of host publication	Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings
Editors	Seiichi Ozawa, Andrew Chi Sing Leung, Long Cheng
Publisher	Springer Verlag
Pages	145-156
Number of pages	12
ISBN (Print)	9783030042110
DOIs	https://doi.org/10.1007/978-3-030-04212-7_13
State	Published - 2018
Event	25th International Conference on Neural Information Processing, ICONIP 2018 - Siem Reap, Cambodia Duration: 13 Dec 2018 → 16 Dec 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11304 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	25th International Conference on Neural Information Processing, ICONIP 2018
Country/Territory	Cambodia
City	Siem Reap
Period	13/12/18 → 16/12/18

Keywords

Detection candidate
Detection filter
Iterative clustering
Maximum clique
Non-Maximum Suppression

Access to Document

10.1007/978-3-030-04212-7_13

Cite this

Zhang, X., Sheng, H., Zhang, Y., Chen, J., Wu, Y., Xue, G., & Wei, Q. (2018). Iterative maximum clique clustering based detection filter. In S. Ozawa, A. C. S. Leung, & L. Cheng (Eds.), Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings (pp. 145-156). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11304 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-04212-7_13

Zhang, Xinyu ; Sheng, Hao ; Zhang, Yang et al. / Iterative maximum clique clustering based detection filter. Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. editor / Seiichi Ozawa ; Andrew Chi Sing Leung ; Long Cheng. Springer Verlag, 2018. pp. 145-156 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3cceaf3058ce42b78185f00d5c1773a5,

title = "Iterative maximum clique clustering based detection filter",

abstract = "Object detection is an important research field of computer vision, but getting accurate object detection from a large number of detection candidates has always been a challenge. The most current algorithms use an insufficient Greedy Non-Maximum Suppression (NMS) strategy which heavily relies on the confidence of the detection candidates. This paper proposes the Iterative Detection Filter (IDF) approach, which considers more information of the detection candidates, including overlapping, the confidence generated by the detector, and the ground position perception information of the scene. Through this approach, the detection candidates are mapped to more accurate detections. Our method achieves a significant improvement on the MOT16 and MOT17 datasets, which are widely used in video tracking and detection.",

keywords = "Detection candidate, Detection filter, Iterative clustering, Maximum clique, Non-Maximum Suppression",

author = "Xinyu Zhang and Hao Sheng and Yang Zhang and Jiahui Chen and Yubin Wu and Guangtao Xue and Quanrui Wei",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2018.; 25th International Conference on Neural Information Processing, ICONIP 2018 ; Conference date: 13-12-2018 Through 16-12-2018",

year = "2018",

doi = "10.1007/978-3-030-04212-7\_13",

language = "英语",

isbn = "9783030042110",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "145--156",

editor = "Seiichi Ozawa and Leung, \{Andrew Chi Sing\} and Long Cheng",

booktitle = "Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings",

address = "德国",

}

Zhang, X, Sheng, H, Zhang, Y, Chen, J, Wu, Y, Xue, G & Wei, Q 2018, Iterative maximum clique clustering based detection filter. in S Ozawa, ACS Leung & L Cheng (eds), Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11304 LNCS, Springer Verlag, pp. 145-156, 25th International Conference on Neural Information Processing, ICONIP 2018, Siem Reap, Cambodia, 13/12/18. https://doi.org/10.1007/978-3-030-04212-7_13

Iterative maximum clique clustering based detection filter. / Zhang, Xinyu; Sheng, Hao; Zhang, Yang et al.
Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. ed. / Seiichi Ozawa; Andrew Chi Sing Leung; Long Cheng. Springer Verlag, 2018. p. 145-156 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11304 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Iterative maximum clique clustering based detection filter

AU - Zhang, Xinyu

AU - Sheng, Hao

AU - Zhang, Yang

AU - Chen, Jiahui

AU - Wu, Yubin

AU - Xue, Guangtao

AU - Wei, Quanrui

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2018.

PY - 2018

Y1 - 2018

N2 - Object detection is an important research field of computer vision, but getting accurate object detection from a large number of detection candidates has always been a challenge. The most current algorithms use an insufficient Greedy Non-Maximum Suppression (NMS) strategy which heavily relies on the confidence of the detection candidates. This paper proposes the Iterative Detection Filter (IDF) approach, which considers more information of the detection candidates, including overlapping, the confidence generated by the detector, and the ground position perception information of the scene. Through this approach, the detection candidates are mapped to more accurate detections. Our method achieves a significant improvement on the MOT16 and MOT17 datasets, which are widely used in video tracking and detection.

AB - Object detection is an important research field of computer vision, but getting accurate object detection from a large number of detection candidates has always been a challenge. The most current algorithms use an insufficient Greedy Non-Maximum Suppression (NMS) strategy which heavily relies on the confidence of the detection candidates. This paper proposes the Iterative Detection Filter (IDF) approach, which considers more information of the detection candidates, including overlapping, the confidence generated by the detector, and the ground position perception information of the scene. Through this approach, the detection candidates are mapped to more accurate detections. Our method achieves a significant improvement on the MOT16 and MOT17 datasets, which are widely used in video tracking and detection.

KW - Detection candidate

KW - Detection filter

KW - Iterative clustering

KW - Maximum clique

KW - Non-Maximum Suppression

UR - https://www.scopus.com/pages/publications/85058989164

U2 - 10.1007/978-3-030-04212-7_13

DO - 10.1007/978-3-030-04212-7_13

M3 - 会议稿件

AN - SCOPUS:85058989164

SN - 9783030042110

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

SP - 145

EP - 156

BT - Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings

A2 - Ozawa, Seiichi

A2 - Leung, Andrew Chi Sing

A2 - Cheng, Long

PB - Springer Verlag

T2 - 25th International Conference on Neural Information Processing, ICONIP 2018

Y2 - 13 December 2018 through 16 December 2018

ER -

Zhang X, Sheng H, Zhang Y, Chen J, Wu Y, Xue G et al. Iterative maximum clique clustering based detection filter. In Ozawa S, Leung ACS, Cheng L, editors, Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. Springer Verlag. 2018. p. 145-156. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-04212-7_13

Iterative maximum clique clustering based detection filter

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this