DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Uncertainty Estimation for Planetary Robotic Terrain Segmentation

Müller, Marcus Gerhard and Durner, Maximilian and Boerdijk, Wout and Blum, Hermann and Gawel, Abel and Stürzl, Wolfgang and Siegwart, Roland and Triebel, Rudolph (2023) Uncertainty Estimation for Planetary Robotic Terrain Segmentation. In: 2023 IEEE Aerospace Conference, AERO 2023, pp. 1-8. IEEE. 2023 IEEE Aerospace Conference, 04-11 Mar 2023, Big Sky, Montana, US. doi: 10.1109/AERO55745.2023.10115611. ISBN 978-166549032-0. ISSN 1095-323X.

[img] PDF

Official URL: https://ieeexplore.ieee.org/document/10115611


Terrain Segmentation information is crucial input for current and future planetary robotic missions. Labeling training data for terrain segmentation is a difficult task and can often cause semantic ambiguity. As a result, large portion of an image usually remains unlabeled. Therefore, it is difficult to evaluate network performance on such regions. Worse is the problem of using such a network for inference, since the quality of predictions cannot be guaranteed if trained with a standard semantic segmentation network. This can be very dangerous for real autonomous robotic missions since the network could predict any of the classes in a particular region, and the robot does not know how much of the prediction to trust. To overcome this issue, we investigate the benefits of uncertainty estimation for terrain segmentation. Knowing how certain the network is about its prediction is an important element for a robust autonomous navigation. In this paper, we present neural networks, which not only give a terrain segmentation prediction, but also an uncertainty estimation. We compare the different methods on the publicly released real world Mars data from the MSL mission.

Item URL in elib:https://elib.dlr.de/195140/
Document Type:Conference or Workshop Item (Speech)
Title:Uncertainty Estimation for Planetary Robotic Terrain Segmentation
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Müller, Marcus GerhardUNSPECIFIEDhttps://orcid.org/0000-0003-4283-6693UNSPECIFIED
Durner, MaximilianUNSPECIFIEDhttps://orcid.org/0000-0001-8885-5334UNSPECIFIED
Boerdijk, WoutUNSPECIFIEDhttps://orcid.org/0000-0003-0789-5970UNSPECIFIED
Blum, HermannAutonomous Systems Lab, ETH Zurichhttps://orcid.org/0000-0002-1713-7877UNSPECIFIED
Gawel, AbelAutonomous Systems Lab, ETH Zurichhttps://orcid.org/0000-0003-2919-4040UNSPECIFIED
Stürzl, WolfgangUNSPECIFIEDhttps://orcid.org/0000-0003-2440-5857UNSPECIFIED
Siegwart, RolandAutonomous Systems Lab, ETH Zurichhttps://orcid.org/0000-0002-2760-7983UNSPECIFIED
Triebel, RudolphUNSPECIFIEDhttps://orcid.org/0000-0002-7975-036XUNSPECIFIED
Date:15 May 2023
Journal or Publication Title:2023 IEEE Aerospace Conference, AERO 2023
Refereed publication:Yes
Open Access:Yes
Gold Open Access:No
In ISI Web of Science:Yes
Page Range:pp. 1-8
Keywords:Planetary exploration rover, Mars Moons eXploration, semantic segmentation, uncertainty estimation, deep learning, machine learning, terrain segmentation
Event Title:2023 IEEE Aerospace Conference
Event Location:Big Sky, Montana, US
Event Type:international Conference
Event Dates:04-11 Mar 2023
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Space
HGF - Program Themes:Robotics
DLR - Research area:Raumfahrt
DLR - Program:R RO - Robotics
DLR - Research theme (Project):R - Multisensory World Modelling (RM) [RO]
Location: Oberpfaffenhofen
Institutes and Institutions:Institute of Robotics and Mechatronics (since 2013) > Perception and Cognition
Deposited By: Müller, Marcus Gerhard
Deposited On:16 May 2023 20:04
Last Modified:14 Nov 2023 09:03

Repository Staff Only: item control page

Help & Contact
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.