Create a virtual environment, activate it and install the PIP requirements.
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
To execute the entire pipeline, do:
python __init__.py
This will download all necessary assets, do the preprocessing and compute the best NN.
To compute the preprocessing of the input images, do:
python preprocessing.py
To work directly on the model:
python model.py
This will download the input folder (if needed), and compute the dataset in output/dataset.csv
.