Efficient 3D Tracking in Urban Environments with Semantic Segmentation

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

In this paper, we present a new 3D tracking approach for self-localization in urban environments. In particular, we build on existing tracking approaches (i.e., visual odometry tracking and SLAM), additionally using the information provided by 2.5D maps of the environment. Since this combination is not straightforward, we adopt ideas from semantic segmentation to find a better alignment between the pose estimated by the tracker and the 2.5D model. Specifically, we show that introducing edges as semantic classes is highly beneficial for our task. In this way, we can reduce tracker inaccuracies and prevent drifting, thus increasing the tracker’s stability. We evaluate our approach for two different challenging scenarios, also showing that it is generally applicable in different application domains and that we are not limited to a specific tracking method.
Original languageEnglish
Title of host publicationProceedings of the British Machine Vision Conference (BMVC)
Publication statusPublished - 2017
EventBritish Machine Vision Conference 2017 - London, United Kingdom
Duration: 4 Sep 20177 Sep 2017

Conference

ConferenceBritish Machine Vision Conference 2017
Abbreviated titleBMVC 2017
CountryUnited Kingdom
CityLondon
Period4/09/177/09/17

Fingerprint

Semantics

Cite this

Hirzer, M., Arth, C., Roth, P. M., & Lepetit, V. (2017). Efficient 3D Tracking in Urban Environments with Semantic Segmentation. In Proceedings of the British Machine Vision Conference (BMVC)

Efficient 3D Tracking in Urban Environments with Semantic Segmentation. / Hirzer, Martin; Arth, Clemens; Roth, Peter M.; Lepetit, Vincent.

Proceedings of the British Machine Vision Conference (BMVC). 2017.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Hirzer, M, Arth, C, Roth, PM & Lepetit, V 2017, Efficient 3D Tracking in Urban Environments with Semantic Segmentation. in Proceedings of the British Machine Vision Conference (BMVC). British Machine Vision Conference 2017, London, United Kingdom, 4/09/17.
Hirzer M, Arth C, Roth PM, Lepetit V. Efficient 3D Tracking in Urban Environments with Semantic Segmentation. In Proceedings of the British Machine Vision Conference (BMVC). 2017
Hirzer, Martin ; Arth, Clemens ; Roth, Peter M. ; Lepetit, Vincent. / Efficient 3D Tracking in Urban Environments with Semantic Segmentation. Proceedings of the British Machine Vision Conference (BMVC). 2017.
@inproceedings{55c4d6808bef4bc1b76112f37310fa6e,
title = "Efficient 3D Tracking in Urban Environments with Semantic Segmentation",
abstract = "In this paper, we present a new 3D tracking approach for self-localization in urban environments. In particular, we build on existing tracking approaches (i.e., visual odometry tracking and SLAM), additionally using the information provided by 2.5D maps of the environment. Since this combination is not straightforward, we adopt ideas from semantic segmentation to find a better alignment between the pose estimated by the tracker and the 2.5D model. Specifically, we show that introducing edges as semantic classes is highly beneficial for our task. In this way, we can reduce tracker inaccuracies and prevent drifting, thus increasing the tracker’s stability. We evaluate our approach for two different challenging scenarios, also showing that it is generally applicable in different application domains and that we are not limited to a specific tracking method.",
author = "Martin Hirzer and Clemens Arth and Roth, {Peter M.} and Vincent Lepetit",
year = "2017",
language = "English",
booktitle = "Proceedings of the British Machine Vision Conference (BMVC)",

}

TY - GEN

T1 - Efficient 3D Tracking in Urban Environments with Semantic Segmentation

AU - Hirzer, Martin

AU - Arth, Clemens

AU - Roth, Peter M.

AU - Lepetit, Vincent

PY - 2017

Y1 - 2017

N2 - In this paper, we present a new 3D tracking approach for self-localization in urban environments. In particular, we build on existing tracking approaches (i.e., visual odometry tracking and SLAM), additionally using the information provided by 2.5D maps of the environment. Since this combination is not straightforward, we adopt ideas from semantic segmentation to find a better alignment between the pose estimated by the tracker and the 2.5D model. Specifically, we show that introducing edges as semantic classes is highly beneficial for our task. In this way, we can reduce tracker inaccuracies and prevent drifting, thus increasing the tracker’s stability. We evaluate our approach for two different challenging scenarios, also showing that it is generally applicable in different application domains and that we are not limited to a specific tracking method.

AB - In this paper, we present a new 3D tracking approach for self-localization in urban environments. In particular, we build on existing tracking approaches (i.e., visual odometry tracking and SLAM), additionally using the information provided by 2.5D maps of the environment. Since this combination is not straightforward, we adopt ideas from semantic segmentation to find a better alignment between the pose estimated by the tracker and the 2.5D model. Specifically, we show that introducing edges as semantic classes is highly beneficial for our task. In this way, we can reduce tracker inaccuracies and prevent drifting, thus increasing the tracker’s stability. We evaluate our approach for two different challenging scenarios, also showing that it is generally applicable in different application domains and that we are not limited to a specific tracking method.

M3 - Conference contribution

BT - Proceedings of the British Machine Vision Conference (BMVC)

ER -