Marc Rußwurm & Marco Körner, Chair of Remote Sensing Technology Technical University of Munich
This repositotry contains qualitative examples of inference for Satellite Time Series classification for crop type identification.
For the Transformer examples see Transformer.ipynb
For the Recurrence examples see Recurrence.ipynb
Slides are available at presentation.pdf
conda create -n phiweek19 pip jupyter
conda activate phiweek19
pip install -r requirements.txt
This repository shows Results from the testset (orange) of the HOLL region in northern Bavaria. The model parameter were determined based on the training dataset. The hyperparameters were determined based on the performance on a 10% hold-back set from the trainset while training on the 90% of the remaining training dataset.