Li, Qingyu und Shi, Yilei und Zhu, Xiao Xiang (2022) Feature and Output Consistency Training for Semi-Supervised Building Footprint Generation. In: International Geoscience and Remote Sensing Symposium (IGARSS), Seiten 171-174. IEEE - Institute of Electrical and Electronics Engineers. IGARSS 2022, 2022-07-17 - 2022-07-22, Kuala Lumpur, Malaysia. doi: 10.1109/IGARSS46834.2022.9883979.
![]() |
PDF
1MB |
Offizielle URL: https://ieeexplore.ieee.org/document/9883979
Kurzfassung
Building footprint maps are important to urban planning and monitoring. However, most existing approaches that fall back on convolutional neural networks (CNNs), require massive annotated samples for network learning. In this research, we propose a novel semi-supervised network, which can help to deal with this issue by leveraging a large amount of unlabeled data. Considering that rich information is also encoded in feature maps, we propose to integrate the consistency of both features and outputs in the end-to-end network training of unlabeled samples on data perturbation, enabling to impose additional constraints. Experiments are conducted on Inria dataset. Our approach is much superior to the state-of-the-art methods in both quantitative and qualitative results.
elib-URL des Eintrags: | https://elib.dlr.de/193313/ | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Dokumentart: | Konferenzbeitrag (Vortrag) | ||||||||||||||||
Titel: | Feature and Output Consistency Training for Semi-Supervised Building Footprint Generation | ||||||||||||||||
Autoren: |
| ||||||||||||||||
Datum: | 2022 | ||||||||||||||||
Erschienen in: | International Geoscience and Remote Sensing Symposium (IGARSS) | ||||||||||||||||
Referierte Publikation: | Ja | ||||||||||||||||
Open Access: | Ja | ||||||||||||||||
Gold Open Access: | Nein | ||||||||||||||||
In SCOPUS: | Ja | ||||||||||||||||
In ISI Web of Science: | Nein | ||||||||||||||||
DOI: | 10.1109/IGARSS46834.2022.9883979 | ||||||||||||||||
Seitenbereich: | Seiten 171-174 | ||||||||||||||||
Verlag: | IEEE - Institute of Electrical and Electronics Engineers | ||||||||||||||||
Status: | veröffentlicht | ||||||||||||||||
Stichwörter: | CNN, urban planning, semi-supervised | ||||||||||||||||
Veranstaltungstitel: | IGARSS 2022 | ||||||||||||||||
Veranstaltungsort: | Kuala Lumpur, Malaysia | ||||||||||||||||
Veranstaltungsart: | internationale Konferenz | ||||||||||||||||
Veranstaltungsbeginn: | 17 Juli 2022 | ||||||||||||||||
Veranstaltungsende: | 22 Juli 2022 | ||||||||||||||||
HGF - Forschungsbereich: | Luftfahrt, Raumfahrt und Verkehr | ||||||||||||||||
HGF - Programm: | Raumfahrt | ||||||||||||||||
HGF - Programmthema: | Erdbeobachtung | ||||||||||||||||
DLR - Schwerpunkt: | Raumfahrt | ||||||||||||||||
DLR - Forschungsgebiet: | R EO - Erdbeobachtung | ||||||||||||||||
DLR - Teilgebiet (Projekt, Vorhaben): | R - Künstliche Intelligenz | ||||||||||||||||
Standort: | Oberpfaffenhofen | ||||||||||||||||
Institute & Einrichtungen: | Institut für Methodik der Fernerkundung > EO Data Science | ||||||||||||||||
Hinterlegt von: | Haschberger, Dr.-Ing. Peter | ||||||||||||||||
Hinterlegt am: | 16 Jan 2023 08:40 | ||||||||||||||||
Letzte Änderung: | 24 Apr 2024 20:54 |
Nur für Mitarbeiter des Archivs: Kontrollseite des Eintrags