Details
| Original language | English |
|---|---|
| Title of host publication | KI 2021 |
| Subtitle of host publication | Advances in Artificial Intelligence - 44th German Conference on AI, Proceedings |
| Editors | Stefan Edelkamp, Ralf Möller, Elmar Rueckert |
| Pages | 116-134 |
| Number of pages | 19 |
| ISBN (electronic) | 978-3-030-87626-5 |
| Publication status | Published - 2021 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 12873 LNAI |
| ISSN (Print) | 0302-9743 |
| ISSN (electronic) | 1611-3349 |
Abstract
The semantic segmentation of aerial images enables many useful applications such as tracking city growth, tracking deforestation, or automatically creating and updating maps. However, gathering enough training data to train a proper model for the automated analysis of aerial images is usually too labor-intensive and thus too expensive in most cases. Therefore, domain adaptation techniques are often necessary to be able to adapt existing models or to transfer knowledge from existing datasets to new unlabeled aerial images. Modern adaptation approaches make use of complex architectures involving many model components, losses and loss weights. These approaches are hard to apply in practice since their hyperparameters are hard to optimize for a given adaptation problem. This complexity is the result of trying to separate domain-invariant elements, e.g., structures and shapes, from domain-specific elements, e.g., textures. In this paper, we present a novel model for semantic segmentation, which not only achieves state-of-the-art performance on aerial images, but also inherently learns separate feature representations for shapes and textures. Our goal is to provide a model which can serve as the basis for future domain adaptation approaches which are simpler but still effective. Through end-to-end training our deep learning model learns to map aerial images to feature representations which can be decoded into binary space partitioning trees, a resolution-independent representation of the semantic segmentation, which can then be rendered into a pixelwise semantic segmentation in a differentiable way.
Keywords
- Aerial images, Computer vision, Deep learning, Semantic segmentation, Spatial partitioning
ASJC Scopus subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
- General Computer Science
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
KI 2021: Advances in Artificial Intelligence - 44th German Conference on AI, Proceedings. ed. / Stefan Edelkamp; Ralf Möller; Elmar Rueckert. 2021. p. 116-134 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12873 LNAI).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Semantic Segmentation of Aerial Images Using Binary Space Partitioning
AU - Gritzner, Daniel
AU - Ostermann, Jörn
PY - 2021
Y1 - 2021
N2 - The semantic segmentation of aerial images enables many useful applications such as tracking city growth, tracking deforestation, or automatically creating and updating maps. However, gathering enough training data to train a proper model for the automated analysis of aerial images is usually too labor-intensive and thus too expensive in most cases. Therefore, domain adaptation techniques are often necessary to be able to adapt existing models or to transfer knowledge from existing datasets to new unlabeled aerial images. Modern adaptation approaches make use of complex architectures involving many model components, losses and loss weights. These approaches are hard to apply in practice since their hyperparameters are hard to optimize for a given adaptation problem. This complexity is the result of trying to separate domain-invariant elements, e.g., structures and shapes, from domain-specific elements, e.g., textures. In this paper, we present a novel model for semantic segmentation, which not only achieves state-of-the-art performance on aerial images, but also inherently learns separate feature representations for shapes and textures. Our goal is to provide a model which can serve as the basis for future domain adaptation approaches which are simpler but still effective. Through end-to-end training our deep learning model learns to map aerial images to feature representations which can be decoded into binary space partitioning trees, a resolution-independent representation of the semantic segmentation, which can then be rendered into a pixelwise semantic segmentation in a differentiable way.
AB - The semantic segmentation of aerial images enables many useful applications such as tracking city growth, tracking deforestation, or automatically creating and updating maps. However, gathering enough training data to train a proper model for the automated analysis of aerial images is usually too labor-intensive and thus too expensive in most cases. Therefore, domain adaptation techniques are often necessary to be able to adapt existing models or to transfer knowledge from existing datasets to new unlabeled aerial images. Modern adaptation approaches make use of complex architectures involving many model components, losses and loss weights. These approaches are hard to apply in practice since their hyperparameters are hard to optimize for a given adaptation problem. This complexity is the result of trying to separate domain-invariant elements, e.g., structures and shapes, from domain-specific elements, e.g., textures. In this paper, we present a novel model for semantic segmentation, which not only achieves state-of-the-art performance on aerial images, but also inherently learns separate feature representations for shapes and textures. Our goal is to provide a model which can serve as the basis for future domain adaptation approaches which are simpler but still effective. Through end-to-end training our deep learning model learns to map aerial images to feature representations which can be decoded into binary space partitioning trees, a resolution-independent representation of the semantic segmentation, which can then be rendered into a pixelwise semantic segmentation in a differentiable way.
KW - Aerial images
KW - Computer vision
KW - Deep learning
KW - Semantic segmentation
KW - Spatial partitioning
UR - http://www.scopus.com/inward/record.url?scp=85116898477&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-87626-5_10
DO - 10.1007/978-3-030-87626-5_10
M3 - Conference contribution
SN - 978-3-030-87625-8
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 116
EP - 134
BT - KI 2021
A2 - Edelkamp, Stefan
A2 - Möller, Ralf
A2 - Rueckert, Elmar
ER -