Scipediacontent: Scipediacontent moved page Draft Content 294203133 to Audebert et al 2017a

2021-02-02T01:22:21Z

Scipediacontent moved page Draft Content 294203133 to Audebert et al 2017a

Scipediacontent: Created page with " == Abstract == International audience; In this work, we investigate the use of OpenStreetMap data for semantic labeling of Earth Observation images. Deep neural networks hav..."

2021-02-02T01:22:16Z

Created page with " == Abstract == International audience; In this work, we investigate the use of OpenStreetMap data for semantic labeling of Earth Observation images. Deep neural networks hav..."

New page

== Abstract ==

International audience; In this work, we investigate the use of OpenStreetMap data for semantic labeling of Earth Observation images. Deep neural networks have been used in the past for remote sensing data classification from various sensors, including multispectral, hyperspectral, SAR and LiDAR data. While OpenStreetMap has already been used as ground truth data for training such networks, this abundant data source remains rarely exploited as an input information layer. In this paper, we study different use cases and deep network architectures to leverage OpenStreetMap data for semantic labeling of aerial and satellite images. Especially , we look into fusion based architectures and coarse-to-fine segmentation to include the OpenStreetMap layer into multispectral-based deep fully convolutional networks. We illustrate how these methods can be successfully used on two public datasets: ISPRS Potsdam and DFC2017. We show that OpenStreetMap data can efficiently be integrated into the vision-based deep learning models and that it significantly improves both the accuracy performance and the convergence speed of the networks.

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1705.06057 http://arxiv.org/abs/1705.06057]

* [http://arxiv.org/pdf/1705.06057 http://arxiv.org/pdf/1705.06057]

* [http://xplorestaging.ieee.org/ielx7/8014302/8014734/08014933.pdf?arnumber=8014933 http://xplorestaging.ieee.org/ielx7/8014302/8014734/08014933.pdf?arnumber=8014933],
: [http://dx.doi.org/10.1109/cvprw.2017.199 http://dx.doi.org/10.1109/cvprw.2017.199]

* [https://dblp.uni-trier.de/db/journals/corr/corr1705.html#AudebertSL17a https://dblp.uni-trier.de/db/journals/corr/corr1705.html#AudebertSL17a],
: [http://openaccess.thecvf.com/content_cvpr_2017_workshops/w18/papers/Audebert_Joint_Learning_From_CVPR_2017_paper.pdf http://openaccess.thecvf.com/content_cvpr_2017_workshops/w18/papers/Audebert_Joint_Learning_From_CVPR_2017_paper.pdf],
: [https://arxiv.org/abs/1705.06057 https://arxiv.org/abs/1705.06057],
: [https://hal.archives-ouvertes.fr/hal-01523573/document https://hal.archives-ouvertes.fr/hal-01523573/document],
: [http://openaccess.thecvf.com/content_cvpr_2017_workshops/w18/html/Audebert_Joint_Learning_From_CVPR_2017_paper.html http://openaccess.thecvf.com/content_cvpr_2017_workshops/w18/html/Audebert_Joint_Learning_From_CVPR_2017_paper.html],
: [https://ieeexplore.ieee.org/document/8014933 https://ieeexplore.ieee.org/document/8014933],
: [https://ui.adsabs.harvard.edu/abs/2017arXiv170506057A/abstract https://ui.adsabs.harvard.edu/abs/2017arXiv170506057A/abstract],
: [https://hal.archives-ouvertes.fr/hal-01523573 https://hal.archives-ouvertes.fr/hal-01523573],
: [https://academic.microsoft.com/#/detail/2615237590 https://academic.microsoft.com/#/detail/2615237590]

* [https://hal.archives-ouvertes.fr/hal-01523573 https://hal.archives-ouvertes.fr/hal-01523573],
: [https://hal.archives-ouvertes.fr/hal-01523573/document https://hal.archives-ouvertes.fr/hal-01523573/document],
: [https://hal.archives-ouvertes.fr/hal-01523573/file/joint_earthvision_audebert_lesaux_lefevre.pdf https://hal.archives-ouvertes.fr/hal-01523573/file/joint_earthvision_audebert_lesaux_lefevre.pdf]

← Older revision	Revision as of 01:22, 2 February 2021
(No difference)

Audebert et al 2017a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 294203133 to Audebert et al 2017a

Scipediacontent: Created page with " == Abstract == International audience; In this work, we investigate the use of OpenStreetMap data for semantic labeling of Earth Observation images. Deep neural networks hav..."