elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Impressum | Datenschutz | Kontakt | English
Schriftgröße: [-] Text [+]

Uncertainty is not sufficient for identifying noisy labels in training data for binary segmentation of building footprints

Ulman, Hannah und Gütter, Jonas Aaron und Niebling, Julia (2023) Uncertainty is not sufficient for identifying noisy labels in training data for binary segmentation of building footprints. Frontiers in Remote Sensing (3), Seite 1100012. Frontiers Media S.A.. doi: 10.3389/frsen.2022.1100012. ISSN 2673-6187.

[img] PDF - Verlagsversion (veröffentlichte Fassung)
1MB

Offizielle URL: https://www.frontiersin.org/articles/10.3389/frsen.2022.1100012/full

Kurzfassung

Obtaining high quality labels is a major challenge for the application of deep neural networks in the remote sensing domain. A common way of acquiring labels is the usage of crowd sourcing which can provide much needed training data sets but also often contains incorrect labels which can affect the training process of a deep neural network significantly. In this paper, we exploit uncertainty to identify a certain type of label noise for semantic segmentation of buildings in satellite imagery. That type of label noise is known as “omission noise,” i.e., missing labels for whole buildings which still appear in the satellite image. Following the literature, uncertainty during training can help in identifying the “sweet spot” between generalizing well and overfitting to label noise, which is further used to differentiate between noisy and clean labels. The differentiation between clean and noisy labels is based on pixel-wise uncertainty estimation and beta distribution fitting to the uncertainty estimates. For our study, we create a data set for building segmentation with different levels of omission noise to evaluate the impact of the noise level on the performance of the deep neural network during training. In doing so, we show that established uncertainty-based methods to identify noisy labels are in general not sufficient enough for our kind of remote sensing data. On the other hand, for some noise levels, we observe some promising differences between noisy and clean data which opens the possibility to refine the state-of-the-art methods further.

elib-URL des Eintrags:https://elib.dlr.de/193825/
Dokumentart:Zeitschriftenbeitrag
Titel:Uncertainty is not sufficient for identifying noisy labels in training data for binary segmentation of building footprints
Autoren:
AutorenInstitution oder E-Mail-AdresseAutoren-ORCID-iDORCID Put Code
Ulman, Hannahhulman (at) princeton.eduNICHT SPEZIFIZIERTNICHT SPEZIFIZIERT
Gütter, Jonas AaronJonas.Guetter (at) dlr.deNICHT SPEZIFIZIERTNICHT SPEZIFIZIERT
Niebling, JuliaJulia.Niebling (at) dlr.dehttps://orcid.org/0000-0001-5413-2234NICHT SPEZIFIZIERT
Datum:10 Januar 2023
Erschienen in:Frontiers in Remote Sensing
Referierte Publikation:Ja
Open Access:Ja
Gold Open Access:Ja
In SCOPUS:Nein
In ISI Web of Science:Ja
DOI:10.3389/frsen.2022.1100012
Seitenbereich:Seite 1100012
Verlag:Frontiers Media S.A.
Name der Reihe:Image Analysis and Classification
ISSN:2673-6187
Status:veröffentlicht
Stichwörter:deep learning, remote sensing, uncertainty, label noise, segmentation
HGF - Forschungsbereich:Luftfahrt, Raumfahrt und Verkehr
HGF - Programm:Raumfahrt
HGF - Programmthema:Technik für Raumfahrtsysteme
DLR - Schwerpunkt:Raumfahrt
DLR - Forschungsgebiet:R SY - Technik für Raumfahrtsysteme
DLR - Teilgebiet (Projekt, Vorhaben):R - Grundlagenforschung im Bereich Maschinelles Lernen
Standort: Jena
Institute & Einrichtungen:Institut für Datenwissenschaften > Datenanalyse und -intelligenz
Hinterlegt von: Niebling, Julia
Hinterlegt am:14 Feb 2023 14:06
Letzte Änderung:28 Feb 2024 08:24

Nur für Mitarbeiter des Archivs: Kontrollseite des Eintrags

Blättern
Suchen
Hilfe & Kontakt
Informationen
electronic library verwendet EPrints 3.3.12
Gestaltung Webseite und Datenbank: Copyright © Deutsches Zentrum für Luft- und Raumfahrt (DLR). Alle Rechte vorbehalten.