# Deep Learning Models for BigEarthNet-S2 with 19 Classes
This repository contains code to use the [BigEarthNet](http://bigearth.net) Sentinel-2 (denoted as BigEarthNet-S2) archive with the nomenclature of 19 classes for deep learning applications. The nomenclature of 19 classes was defined by interpreting and arranging the CORINE Land Cover (CLC) Level-3 nomenclature based on the properties of Sentinel-2 images. This class nomenclature is the product of a collaboration between the [Direção-Geral do Território](http://www.dgterritorio.pt/) in Lisbon, Portugal and the [Remote Sensing Image Analysis (RSiM)](https://www.rsim.tu-berlin.de/) group at TU Berlin, Germany.
A paper describing the creation of the nomenclature of 19 classes is currently under review and will be referenced here in the future. If you are interested in BigEarthNet-S2 with the original CLC Level-3 class nomenclature of 43 classes, please check [here](https://git.tu-berlin.de/rsim/BigEarthNet-S2_43-classes_models).
If you use the BigEarthNet-S2 archive or our pre-trained models, please cite the paper given below:
> G. Sumbul, A. d. Wall, T. Kreuziger, F. Marcelino, H. Costa, P. Benevides, M. Caetano, B. Demir, V. Markl, “[BigEarthNet-MM: A Large Scale Multi-Modal Multi-Label Benchmark Archive for Remote Sensing Image Classification and Retrieval](https://arxiv.org/abs/2105.07921)”, arXiv:2105.07921, 2021.
```
@misc{BigEarthNet-MM,
title={BigEarthNet-MM: A Large Scale Multi-Modal Multi-Label Benchmark Archive for Remote Sensing Image Classification and Retrieval},
author={Gencer Sumbul, Arne de Wall, Tristan Kreuziger, Filipe Marcelino, Hugo Costa, Pedro Benevides, Mário Caetano, Begüm Demir and Volker Markl},
year={2021},
eprint={2105.07921},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
If you are interested in BigEarthNet-S2 with the original CLC Level-3 class nomenclature of 43 classes, please check [here](https://git.tu-berlin.de/rsim/BigEarthNet-S2_43-classes_models).
## Pre-trained Deep Learning Models
We provide code and model weights for the following deep learning models that have been pre-trained on BigEarthNet-S2 with the nomenclature of 19 classes for scene classification:
