Annotated datasets have become one of the most crucial preconditions for the development and evaluation of machine learning-based methods designed for the automated interpretation of remote sensing data. In this paper, we review the historic development of such datasets, discuss their features based on a few selected examples, and address open issues for future developments.
@article{arxiv.2105.11726,
title = {There is no data like more data -- current status of machine learning datasets in remote sensing},
author = {Michael Schmitt and Seyed Ali Ahmadi and Ronny Hänsch},
journal= {arXiv preprint arXiv:2105.11726},
year = {2021}
}
Comments
accepted for the Proceedings of IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2021