Hugin helps scientists run Machine Learning experiments on geospatial raster data.
Overall Hugin aims to facilitate experimentation with multiple machine learning problems, like:
- Classification
- Segmentation
- Super-Resolution
Currently Hugin builds on top of the Keras machine learning library but it also aims to support, in the future, additional backends like scikit-learn.
Documentation for Hugin is available at https://hugin-eo.readthedocs.io/
This project is supported by the European Space Agency through the ML4EO project and the FUSE4DL research project 432PED from 23/10/2020 (PN-III-P2-2.1-PED-2019-4878)