TorchGeo
A deep learning framework designed specifically for remote sensing image analysis, built on PyTorch. It provides support for remote sensing data processing, geospatial datasets, and pretrained models tailored to Earth observation tasks.
Timeline (last 90 days)
The uGAP team is advancing embedded integration with TorchGeo, with the goal of bringing pretrained weights and geospatial model capabilities directly into the SDK.
TorchGeo is the official extension library for geospatial data in the PyTorch ecosystem, providing Dataset, Sampler, transforms, and pretrained models, with built-in support for reading multi-source satellite imagery such as Sentinel-2, Landsat, and NAIP.
All Facts (2)
The uGAP team is advancing embedded integration with TorchGeo, with the goal of bringing pretrained weights and geospatial model capabilities directly into the SDK.
50%UnverifiedTorchGeo is the official extension library for geospatial data in the PyTorch ecosystem, providing Dataset, Sampler, transforms, and pretrained models, with built-in support for reading multi-source satellite imagery such as Sentinel-2, Landsat, and NAIP.
50%