Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching

Dongkwon Jin; Jun Tae Lee; Chang Su Kim

doi:10.1007/978-3-030-58565-5_8

Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching

Dongkwon Jin, Jun Tae Lee, Chang Su Kim

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

9 Citations (Scopus)

Abstract

A novel algorithm to detect semantic lines is proposed in this paper. We develop three networks: detection network with mirror attention (D-Net) and comparative ranking and matching networks (R-Net and M-Net). D-Net extracts semantic lines by exploiting rich contextual information. To this end, we design the mirror attention module. Then, through pairwise comparisons of extracted semantic lines, we iteratively select the most semantic line and remove redundant ones overlapping with the selected one. For the pairwise comparisons, we develop R-Net and M-Net in the Siamese architecture. Experiments demonstrate that the proposed algorithm outperforms the conventional semantic line detector significantly. Moreover, we apply the proposed algorithm to detect two important kinds of semantic lines successfully: dominant parallel lines and reflection symmetry axes. Our codes are available at https://github.com/dongkwonjin/Semantic-Line-DRM.

Original language	English
Title of host publication	Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings
Editors	Andrea Vedaldi, Horst Bischof, Thomas Brox, Jan-Michael Frahm
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	119-135
Number of pages	17
ISBN (Print)	9783030585648
DOIs	https://doi.org/10.1007/978-3-030-58565-5_8
Publication status	Published - 2020
Event	16th European Conference on Computer Vision, ECCV 2020 - Glasgow, United Kingdom Duration: 2020 Aug 23 → 2020 Aug 28

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12365 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th European Conference on Computer Vision, ECCV 2020
Country/Territory	United Kingdom
City	Glasgow
Period	20/8/23 → 20/8/28

Bibliographical note

Publisher Copyright:
© 2020, Springer Nature Switzerland AG.

Keywords

Attention
Line detection
Matching
Ranking
Semantic lines

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-58565-5_8

Cite this

Jin, D., Lee, J. T., & Kim, C. S. (2020). Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching. In A. Vedaldi, H. Bischof, T. Brox, & J.-M. Frahm (Eds.), Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings (pp. 119-135). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12365 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-58565-5_8

Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching. / Jin, Dongkwon; Lee, Jun Tae; Kim, Chang Su.
Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings. ed. / Andrea Vedaldi; Horst Bischof; Thomas Brox; Jan-Michael Frahm. Springer Science and Business Media Deutschland GmbH, 2020. p. 119-135 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12365 LNCS).

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

Jin, D, Lee, JT & Kim, CS 2020, Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching. in A Vedaldi, H Bischof, T Brox & J-M Frahm (eds), Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12365 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 119-135, 16th European Conference on Computer Vision, ECCV 2020, Glasgow, United Kingdom, 20/8/23. https://doi.org/10.1007/978-3-030-58565-5_8

Jin D, Lee JT, Kim CS. Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching. In Vedaldi A, Bischof H, Brox T, Frahm JM, editors, Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings. Springer Science and Business Media Deutschland GmbH. 2020. p. 119-135. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-58565-5_8

Jin, Dongkwon ; Lee, Jun Tae ; Kim, Chang Su. / Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching. Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings. editor / Andrea Vedaldi ; Horst Bischof ; Thomas Brox ; Jan-Michael Frahm. Springer Science and Business Media Deutschland GmbH, 2020. pp. 119-135 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{821e21194e3e47f6ab6ccb08ada28ea9,

title = "Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching",

abstract = "A novel algorithm to detect semantic lines is proposed in this paper. We develop three networks: detection network with mirror attention (D-Net) and comparative ranking and matching networks (R-Net and M-Net). D-Net extracts semantic lines by exploiting rich contextual information. To this end, we design the mirror attention module. Then, through pairwise comparisons of extracted semantic lines, we iteratively select the most semantic line and remove redundant ones overlapping with the selected one. For the pairwise comparisons, we develop R-Net and M-Net in the Siamese architecture. Experiments demonstrate that the proposed algorithm outperforms the conventional semantic line detector significantly. Moreover, we apply the proposed algorithm to detect two important kinds of semantic lines successfully: dominant parallel lines and reflection symmetry axes. Our codes are available at https://github.com/dongkwonjin/Semantic-Line-DRM.",

keywords = "Attention, Line detection, Matching, Ranking, Semantic lines",

author = "Dongkwon Jin and Lee, {Jun Tae} and Kim, {Chang Su}",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 16th European Conference on Computer Vision, ECCV 2020 ; Conference date: 23-08-2020 Through 28-08-2020",

year = "2020",

doi = "10.1007/978-3-030-58565-5_8",

language = "English",

isbn = "9783030585648",

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

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "119--135",

editor = "Andrea Vedaldi and Horst Bischof and Thomas Brox and Jan-Michael Frahm",

booktitle = "Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching

AU - Jin, Dongkwon

AU - Lee, Jun Tae

AU - Kim, Chang Su

PY - 2020

Y1 - 2020

N2 - A novel algorithm to detect semantic lines is proposed in this paper. We develop three networks: detection network with mirror attention (D-Net) and comparative ranking and matching networks (R-Net and M-Net). D-Net extracts semantic lines by exploiting rich contextual information. To this end, we design the mirror attention module. Then, through pairwise comparisons of extracted semantic lines, we iteratively select the most semantic line and remove redundant ones overlapping with the selected one. For the pairwise comparisons, we develop R-Net and M-Net in the Siamese architecture. Experiments demonstrate that the proposed algorithm outperforms the conventional semantic line detector significantly. Moreover, we apply the proposed algorithm to detect two important kinds of semantic lines successfully: dominant parallel lines and reflection symmetry axes. Our codes are available at https://github.com/dongkwonjin/Semantic-Line-DRM.

AB - A novel algorithm to detect semantic lines is proposed in this paper. We develop three networks: detection network with mirror attention (D-Net) and comparative ranking and matching networks (R-Net and M-Net). D-Net extracts semantic lines by exploiting rich contextual information. To this end, we design the mirror attention module. Then, through pairwise comparisons of extracted semantic lines, we iteratively select the most semantic line and remove redundant ones overlapping with the selected one. For the pairwise comparisons, we develop R-Net and M-Net in the Siamese architecture. Experiments demonstrate that the proposed algorithm outperforms the conventional semantic line detector significantly. Moreover, we apply the proposed algorithm to detect two important kinds of semantic lines successfully: dominant parallel lines and reflection symmetry axes. Our codes are available at https://github.com/dongkwonjin/Semantic-Line-DRM.

KW - Attention

KW - Line detection

KW - Matching

KW - Ranking

KW - Semantic lines

UR - http://www.scopus.com/inward/record.url?scp=85097371034&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-58565-5_8

DO - 10.1007/978-3-030-58565-5_8

M3 - Conference contribution

AN - SCOPUS:85097371034

SN - 9783030585648

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

SP - 119

EP - 135

BT - Computer Vision – ECCV 2020 - 16th European Conference 2020, Proceedings

A2 - Vedaldi, Andrea

A2 - Bischof, Horst

A2 - Brox, Thomas

A2 - Frahm, Jan-Michael

PB - Springer Science and Business Media Deutschland GmbH

T2 - 16th European Conference on Computer Vision, ECCV 2020

Y2 - 23 August 2020 through 28 August 2020

ER -

Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this