This repository contains code to use deep learning models, pre-trained on the BigEarthNet archive with TensorFlow, to train new models, and to evaluate pre-trained models. Note that in addition to original BigEarthNet labels (will be called original in the code), there is a new class nomenclature (will be called BigEarthNet-19 in the code). This repository is compatible with both options.

• For original BigEarthNet labels, it is highly recommended to first check the BigEarthNet Deep Learning Models repository.

• For BigEarthNet-19 labels, it is highly recommended to first check the BigEarthNet-19 Deep Learning Models repository.

• The definitions of ResNet and VGG models are based on their TensorFlow-Slim implementations. The nets folder of TensorFlow-Slim repository with small modifications are already included in the root folder. Please note that license information of this repository does not cover the nets folder, which is licensed under the Apache License 2.0.
• The prep_splits.py script from here for original labels and the prep_splits_19_classes.py script from here for BigEarthNet-19 labels generate TFRecord files for train, validation and test sets from the BigEarthNet archive. To train or evaluate any model, required TFRecord files should be first prepared.
• Config files of each model, which was used to train corresponding model, are given under configs folder. If you want to use those config files, you need to download pre-trained model weights and move to the folders (whose paths are written in the corresponding JSON file).
• TensorFlow package should be installed. All codes are tested with Python 2.7, TensorFlow 1.15.0 and Ubuntu 16.04.

The script train.py expects a JSON configuration file path as a comand line argument. This file contains the following parameters:

• model_name: The name of the Python code containing the corresponding deep learning model. The code must be located under the models directory. The model class will be loaded dynamically based on the model_name parameter: model = importlib.import_module('models.' + args['model_name']).dnn_model(nb_class)
• modality : A flag to indicate which modality will be used during training: S1, S2 or MM. S1 and S2 refer to the BigEarthNet patches from Sentinel-1 and Sentinel-2, respectively while multi-modal (MM) refers to using pairs of S1 and S2 patches together.
• label_type: A flag to indicate which labels will be used during training: original or BigEarthNet-19
• batch_size: Batch size used during training
• nb_epoch: The number of epochs for the training
• learning_rate: The initial learning rate
• out_dir: The path where all log files and checkpoints will be saved.
• save_checkpoint_after_iteration: The iteration after which checkpoint saving should start, i.e., no checkpoints are saved before. Set to zero to always have checkpoints saved.
• save_checkpoint_per_iteration: The number of iterations per which a checkpoint is written, i.e., when iteration_index % save_checkpoint_per_iteration == 0.
• tr_tf_record_files: An array containing TFRecord file(s) for training.
• val_tf_record_files: An array containing TFRecord file(s) for validation (not used for now).
• fine_tune: A flag to indicate if the training of the model will continue from the existing checkpoint whose path will be defined by pretrained_model_path.
• model_file: The base name of a pre-trained model snapshot (i.e., checkpoint).
• shuffle_buffer_size: The number of elements which will be shuffled at the beginning of each epoch. It is not recommended to have large shuffle buffer if you don't have enough space in memory.
• training_size: The size of the training set. If you are using training set suggested in here, it is already set.

The script eval.py expects a JSON configuration file. The needed parameters are as follows:

• model_name: The name of the Python code containing the corresponding deep learning model. The code must be located under the models directory. The model class will be loaded dynamically based on the model_name parameter: model = importlib.import_module('models.' + args['model_name']).dnn_model(nb_class)
• modality : A flag to indicate which modality will be used during evaluation: S1, S2 or MM. S1 and S2 refer to the BigEarthNet patches from Sentinel-1 and Sentinel-2, respectively while multi-modal (MM) refers to using pairs of S1 and S2 patches together.
• batch_size: Batch size used during evaluation
• out_dir: The path where all log files and checkpoints will be saved.
• test_tf_record_files: An array containing TFRecord files for evaluation.
• model_file: The base name of a pre-trained model snapshot (i.e., checkpoint).
• test_size: The size of the test set. If you are using test set suggested in here, it is already set.

Gencer Sümbül http://www.user.tu-berlin.de/gencersumbul/

Tristan Kreuziger https://www.rsim.tu-berlin.de/menue/team/tristan_kreuziger/

The code in this repository except the nets folder is licensed under the MIT License:

MIT License

Copyright (c) 2019 The BigEarthNet Authors

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.