torchgeo and terratorch
TorchGeo provides the foundational geospatial datasets, preprocessing, and base models that TerraTorch builds upon for fine-tuning specialized geospatial foundation models, making them complementary tools in a machine learning pipeline.
About torchgeo
torchgeo/torchgeo
TorchGeo: datasets, samplers, transforms, and pre-trained models for geospatial data
Handles multi-source geospatial imagery with automatic coordinate reference system (CRS) reprojection and resolution alignment across heterogeneous satellite data like Landsat and Sentinel. Provides geospatial-aware samplers that extract patches using geographic coordinates rather than pixel indices, enabling seamless composition of datasets through intersection and union operators. Integrates with PyTorch Lightning for end-to-end training pipelines on remote sensing tasks like semantic segmentation and change detection.
About terratorch
terrastackai/terratorch
A Python toolkit for fine-tuning Geospatial Foundation Models (GFMs).
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work