Code used for the crop classification (CropHarvest) based on multi-view data fusion

• To train a single-view learning model (e.g. Input-level fusion):

python train_singleview.py -s config/singleview_ex.yaml

• To train all the views individually with single-view learning (e.g. for single-view predictions or Ensemble-based fusion):

python train_singleview_pool.py -s config/singleviewpool_ex.yaml

• To train a multi-view learning model (e.g. Feature-level fusion, Decision-level fusion, Gated Fusion, Feature-level fusion with MultiLoss):

python train_multiview.py -s config/multiview_ex.yaml

• To evaluate the model by its predictions (performance):

python evaluate_predictions.py -s config/evaluation_ex.yaml

Please install the required packages with the following command:

pip install -r requirements.txt

for torch

pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113

The data used comes from https://github.com/nasaharvest/cropharvest. However, we also share the structures that we used on Google Drive: https://drive.google.com/drive/folders/1aPlctAL8B5dXSdpM55fr3-RUmAHO3quj

Public repository of our IGARSS 2023 paper.

• 🔒 Final published version
• 🔓 Arxiv
• Presentation

Mena, Francisco, et al. "A comparative assessment of multi-view fusion learning for crop classification." IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2023.

@inproceedings{mena2023comparativeassessmentmultiview,
  title = {A {{Comparative Assessment}} of {{Multi-view Fusion Learning For Crop Classification}}},
  booktitle = {{{IGARSS}} 2023 - 2023 {{IEEE International Geoscience}} and {{Remote Sensing Symposium}}},
  author = {Mena, Francisco and Arenas, Diego and Nuske, Marlon and Dengel, Andreas},
  date = {2023},
  publisher = {{IEEE}},
  doi={10.1109/IGARSS52108.2023.10282138}
}