BOP: Benchmark for 6D Object Pose Estimation

Tomas Hodan, Frank Michel, Eric Brachmann, Wadim Kehl, Anders Glent Buch, Dirk Kraft, Bertram Drost, Joel Vidal, Stephan Ihrke, Xenophon Zabulis, Caner Sahin, Fabian Manhardt, Federico Tombari, Tae-Kyun Kim, Jiri Matas, Carsten Rother

Publikation: Bidrag til bog/antologi/rapport/konference-proceeding › Konferencebidrag i proceedings › Forskning › peer review

Resumé

We propose a benchmark for 6D pose estimation of a rigid object from a single RGB-D input image. The training data consists of a texture-mapped 3D object model or images of the object in known 6D poses. The benchmark comprises of: i) eight datasets in a unified format that cover different practical scenarios, including two new datasets focusing on varying lighting conditions, ii) an evaluation methodology with a pose-error function that deals with pose ambiguities, iii) a comprehensive evaluation of 15 diverse recent methods that captures the status quo of the field, and iv) an online evaluation system that is open for continuous submission of new results. The evaluation shows that methods based on point-pair features currently perform best, outperforming template matching methods, learning-based methods and methods based on 3D local features. The project website is available at bop.felk.cvut.cz.
OriginalsprogEngelsk
TitelComputer Vision - ECCV 2018 : 15th European Conference, Munich, Germany, September 8–14, 2018, Proceedings, Part III
RedaktørerVittorio Ferrari, Martial Hebert, Cristian Sminchisescu, Yair Weiss
Vol/bind11207
ForlagSpringer
Publikationsdatosep. 2018
Sider19-35
ISBN (Trykt)978-3-030-01218-2
ISBN (Elektronisk)978-3-030-01219-9
DOI
StatusUdgivet - sep. 2018
Begivenhed15th European Conference on Computer Vision - München, Tyskland
Varighed: 8. sep. 2018 → 14. sep. 2018

Konference

Konference15th European Conference on Computer Vision
LandTyskland
ByMünchen
Periode08/09/2018 → 14/09/2018
NavnLecture Notes in Computer Science
ISSN0302-9743

Fingeraftryk

Template matching
Websites
Textures
Lighting

Citer dette

Hodan, T., Michel, F., Brachmann, E., Kehl, W., Buch, A. G., Kraft, D., ... Rother, C. (2018). BOP: Benchmark for 6D Object Pose Estimation. I V. Ferrari, M. Hebert, C. Sminchisescu, & Y. Weiss (red.), Computer Vision - ECCV 2018: 15th European Conference, Munich, Germany, September 8–14, 2018, Proceedings, Part III (Bind 11207, s. 19-35). Springer. Lecture Notes in Computer Science https://doi.org/10.1007/978-3-030-01249-6_2
Hodan, Tomas ; Michel, Frank ; Brachmann, Eric ; Kehl, Wadim ; Buch, Anders Glent ; Kraft, Dirk ; Drost, Bertram ; Vidal, Joel ; Ihrke, Stephan ; Zabulis, Xenophon ; Sahin, Caner ; Manhardt, Fabian ; Tombari, Federico ; Kim, Tae-Kyun ; Matas, Jiri ; Rother, Carsten. / BOP: Benchmark for 6D Object Pose Estimation. Computer Vision - ECCV 2018: 15th European Conference, Munich, Germany, September 8–14, 2018, Proceedings, Part III. red. / Vittorio Ferrari ; Martial Hebert ; Cristian Sminchisescu ; Yair Weiss. Bind 11207 Springer, 2018. s. 19-35 (Lecture Notes in Computer Science).
@inproceedings{bf15708335ec49b0a16407daced286a9,
title = "BOP: Benchmark for 6D Object Pose Estimation",
abstract = "We propose a benchmark for 6D pose estimation of a rigid object from a single RGB-D input image. The training data consists of a texture-mapped 3D object model or images of the object in known 6D poses. The benchmark comprises of: i) eight datasets in a unified format that cover different practical scenarios, including two new datasets focusing on varying lighting conditions, ii) an evaluation methodology with a pose-error function that deals with pose ambiguities, iii) a comprehensive evaluation of 15 diverse recent methods that captures the status quo of the field, and iv) an online evaluation system that is open for continuous submission of new results. The evaluation shows that methods based on point-pair features currently perform best, outperforming template matching methods, learning-based methods and methods based on 3D local features. The project website is available at bop.felk.cvut.cz.",
author = "Tomas Hodan and Frank Michel and Eric Brachmann and Wadim Kehl and Buch, {Anders Glent} and Dirk Kraft and Bertram Drost and Joel Vidal and Stephan Ihrke and Xenophon Zabulis and Caner Sahin and Fabian Manhardt and Federico Tombari and Tae-Kyun Kim and Jiri Matas and Carsten Rother",
year = "2018",
month = "9",
doi = "10.1007/978-3-030-01249-6_2",
language = "English",
isbn = "978-3-030-01218-2",
volume = "11207",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "19--35",
editor = "Vittorio Ferrari and Martial Hebert and Cristian Sminchisescu and Yair Weiss",
booktitle = "Computer Vision - ECCV 2018",
address = "Germany",

}

Hodan, T, Michel, F, Brachmann, E, Kehl, W, Buch, AG, Kraft, D, Drost, B, Vidal, J, Ihrke, S, Zabulis, X, Sahin, C, Manhardt, F, Tombari, F, Kim, T-K, Matas, J & Rother, C 2018, BOP: Benchmark for 6D Object Pose Estimation. i V Ferrari, M Hebert, C Sminchisescu & Y Weiss (red), Computer Vision - ECCV 2018: 15th European Conference, Munich, Germany, September 8–14, 2018, Proceedings, Part III. bind 11207, Springer, Lecture Notes in Computer Science, s. 19-35, 15th European Conference on Computer Vision, München, Tyskland, 08/09/2018. https://doi.org/10.1007/978-3-030-01249-6_2

BOP: Benchmark for 6D Object Pose Estimation. / Hodan, Tomas; Michel, Frank; Brachmann, Eric; Kehl, Wadim; Buch, Anders Glent; Kraft, Dirk; Drost, Bertram; Vidal, Joel; Ihrke, Stephan; Zabulis, Xenophon; Sahin, Caner; Manhardt, Fabian; Tombari, Federico; Kim, Tae-Kyun; Matas, Jiri; Rother, Carsten.

Computer Vision - ECCV 2018: 15th European Conference, Munich, Germany, September 8–14, 2018, Proceedings, Part III. red. / Vittorio Ferrari; Martial Hebert; Cristian Sminchisescu; Yair Weiss. Bind 11207 Springer, 2018. s. 19-35 (Lecture Notes in Computer Science).

Publikation: Bidrag til bog/antologi/rapport/konference-proceeding › Konferencebidrag i proceedings › Forskning › peer review

TY - GEN

T1 - BOP: Benchmark for 6D Object Pose Estimation

AU - Hodan, Tomas

AU - Michel, Frank

AU - Brachmann, Eric

AU - Kehl, Wadim

AU - Buch, Anders Glent

AU - Kraft, Dirk

AU - Drost, Bertram

AU - Vidal, Joel

AU - Ihrke, Stephan

AU - Zabulis, Xenophon

AU - Sahin, Caner

AU - Manhardt, Fabian

AU - Tombari, Federico

AU - Kim, Tae-Kyun

AU - Matas, Jiri

AU - Rother, Carsten

PY - 2018/9

Y1 - 2018/9

N2 - We propose a benchmark for 6D pose estimation of a rigid object from a single RGB-D input image. The training data consists of a texture-mapped 3D object model or images of the object in known 6D poses. The benchmark comprises of: i) eight datasets in a unified format that cover different practical scenarios, including two new datasets focusing on varying lighting conditions, ii) an evaluation methodology with a pose-error function that deals with pose ambiguities, iii) a comprehensive evaluation of 15 diverse recent methods that captures the status quo of the field, and iv) an online evaluation system that is open for continuous submission of new results. The evaluation shows that methods based on point-pair features currently perform best, outperforming template matching methods, learning-based methods and methods based on 3D local features. The project website is available at bop.felk.cvut.cz.

AB - We propose a benchmark for 6D pose estimation of a rigid object from a single RGB-D input image. The training data consists of a texture-mapped 3D object model or images of the object in known 6D poses. The benchmark comprises of: i) eight datasets in a unified format that cover different practical scenarios, including two new datasets focusing on varying lighting conditions, ii) an evaluation methodology with a pose-error function that deals with pose ambiguities, iii) a comprehensive evaluation of 15 diverse recent methods that captures the status quo of the field, and iv) an online evaluation system that is open for continuous submission of new results. The evaluation shows that methods based on point-pair features currently perform best, outperforming template matching methods, learning-based methods and methods based on 3D local features. The project website is available at bop.felk.cvut.cz.

U2 - 10.1007/978-3-030-01249-6_2

DO - 10.1007/978-3-030-01249-6_2

M3 - Article in proceedings

SN - 978-3-030-01218-2

VL - 11207

T3 - Lecture Notes in Computer Science

SP - 19

EP - 35

BT - Computer Vision - ECCV 2018

A2 - Ferrari, Vittorio

A2 - Hebert, Martial

A2 - Sminchisescu, Cristian

A2 - Weiss, Yair

PB - Springer

ER -

Hodan T, Michel F, Brachmann E, Kehl W, Buch AG, Kraft D et al. BOP: Benchmark for 6D Object Pose Estimation. I Ferrari V, Hebert M, Sminchisescu C, Weiss Y, red., Computer Vision - ECCV 2018: 15th European Conference, Munich, Germany, September 8–14, 2018, Proceedings, Part III. Bind 11207. Springer. 2018. s. 19-35. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-030-01249-6_2

Dokumenter og links