Welcome to the AI4EO challenge - Enhancing Sentinel-2 Time-Series for Land Management !!
This repository contains a Jupyter notebook, metadata and utility functions to get you started downloading and exploring the data, as well as creating a test submission folder.
Check the challenge webpage for details on how to get started either using the Euro Data Cube (EDC) computational resources, or your own resources.
We provide a Docker Python environment which should be used for the development of your own solution. This environment is already set-up in EDC instances, and can be easily built to run on your own resources.
This environment contains the most popular libraries for geo-spatial processing (rasterio, shapely, gdal, geopandas, eo-learn) and for machine learning (scikit-learn, torch and tensorflow). More info about the libraries included can be found here.
If you are running from the Euro Data Cube workspace, you are ready to go! Open the starter-pack notebook, follow the instructions in the notebook on how to retrieve the credential informations and get started.
If you are using your own resources, consult the challenge webpage to get instructions on how to build the Docker image and get started with the challenge.
For any issue relating to the notebook or the data, open a ticket in the challenge Forum, so that we and the community can adequately support you. For improvements and fixes to the notebook, issues and pull requests can be opened in the AI4EO GitHub repository.