Deep Learning-Based Building Footprint Extraction With Missing Annotations
Ver/ Abrir
Impacto
Scholar |
Otros documentos de la autoría: kang, jian; Fernandez-Beltran, Ruben; Sun, Xian; Ni, Jingen; Plaza, Antonio
Metadatos
Mostrar el registro completo del ítemcomunitat-uji-handle:10234/9
comunitat-uji-handle2:10234/43662
comunitat-uji-handle3:10234/43643
comunitat-uji-handle4:
INVESTIGACIONMetadatos
Título
Deep Learning-Based Building Footprint Extraction With Missing AnnotationsFecha de publicación
2021-04-21Editor
Institute of Electrical and Electronics Engineers; IEEEISSN
1545-598X; 1558-0571Cita bibliográfica
J. Kang, R. Fernandez-Beltran, X. Sun, J. Ni and A. Plaza, "Deep Learning-Based Building Footprint Extraction With Missing Annotations," in IEEE Geoscience and Remote Sensing Letters, doi: 10.1109/LGRS.2021.3072589.Tipo de documento
info:eu-repo/semantics/articleVersión de la editorial
https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=8859Versión
info:eu-repo/semantics/acceptedVersionPalabras clave / Materias
Resumen
Most state-of-the-art deep learning-based methods for extraction of building footprints are aimed at designing proper convolutional neural network (CNN) architectures or loss functions able to effectively predict ... [+]
Most state-of-the-art deep learning-based methods for extraction of building footprints are aimed at designing proper convolutional neural network (CNN) architectures or loss functions able to effectively predict building masks from remote sensing (RS) images. To properly train such CNN models, large-scale and pixel-level building annotations are required. One common approach to obtain scalable benchmark data sets for the segmentation of buildings is to register RS images with auxiliary geospatial information data, such as those available from OpenStreetMaps (OSM). However, due to land-cover changes, urban construction, and delayed geospatial information updating, some building annotations may be missing in the corresponding ground-truth building mask layers. This will likely introduce confusion in the training of CNN models for discriminating between background and building pixels. To solve this important issue, we first formulate the problem as a long-tailed classification one. Then, we introduce a new joint loss function based on three terms: 1) logit adjusted cross entropy (LACE) loss, aimed at discriminating between building and background pixels from a long-tailed label distribution; 2) weighted dice loss, aimed at increasing the F₁ scores of the predicted building masks; and 3) boundary (BD) alignment loss, which is optimized for preserving the fine-grained structure of building boundaries. Our experiments, conducted on two benchmark building segmentation data sets, validate the effectiveness of our newly proposed loss with respect to other state-of-the-art losses commonly used for extracting building footprints. The codes of this letter will be publicly available from https://github.com/jiankang1991/GRSL_BFE_MA. [-]
Entidad financiadora
Ministerio de Ciencia, Innovación y Universidades (Spain) | Generalitat Valenciana | FEDER-Junta de Extremadura | European Union’s Horizon 2020 Research
Código del proyecto o subvención
RTI2018-098651-B-C54 | GV/2020/167 | GR18060 | 734541
Derechos de acceso
© Copyright 2021 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.
http://rightsstatements.org/vocab/InC/1.0/
info:eu-repo/semantics/openAccess
http://rightsstatements.org/vocab/InC/1.0/
info:eu-repo/semantics/openAccess
Aparece en las colecciones
- INIT_Articles [754]