Deep learning with TorchGeo and PyTorch Lightning using custom geodataset скачать в хорошем качестве

Deep learning with TorchGeo and PyTorch Lightning using custom geodataset

A basic yet straightforward tutorial on training (and testing) a geospatial artificial intelligence (GeoAI) model for satellite image semantic segmentation using PyTorch Lightning. TorchGeo is used to handle the georeferenced dataset of images and labels. 2-channels images used are VV and VH polarization images from Sentinel-1 C-band SAR. The labels are from publicly available oil palm maps. Medium article:   / geospatial-deep-learning-using-torchgeo-an...   Oil palm data:    • Open data: Global oil palm plantation map 🌴   TorchGeo paper: https://arxiv.org/abs/2111.08872 Github: https://github.com/microsoft/torchgeo #GeoAI