Repository files navigation

rslearn_projects contains the training datasets, model weights, and corresponding codefor machine learning applications built on top ofrslearn at Ai2.

• Model weights and Code: Licensed underApache License 2.0.
• Annotations: Licensed underCC-BY 4.0.

Install rslearn:

git clone https://github.com/allenai/rslearn.gitcd rslearnpip install .[extra]

Install requirements:

cd ..git clone https://github.com/allenai/rslearn_projects.gitcd rslearn_projectspip install -r requirements.txt

rslearn_projects includes tooling that expects model checkpoints and auxiliary files tobe stored in anRSLP_PREFIX directory. Create a file.env to set theRSLP_PREFIXenvironment variable:

mkdir project_dataecho "RSLP_PREFIX=project_data/" > .env

• Sentinel-2 Vessel Detection
• Sentinel-2 Vessel Attribute Prediction
• Landsat Vessel Detection
• Satlas: Solar Farm Segmentation
• Satlas: Wind Turbine Detection
• Satlas: Marine Infrastructure Detection
• Forest Loss Driver Classification
• Maldives Ecosystem Mapping