/opendl_gis

Intetion to smoth deep learning implementation with geoespatial data

Primary LanguagePython

opendl_gis

Intetion to smoth deep learning implementation with geoespatial data

The workflow is:

  1. Download tiff images
  2. Create Pytorch Dataset
  3. Transferlearning
  4. Train/Test with Pytorch
  5. Register every iteration in tensorboard

TODOs

#Comment codes properly
#transate variable names to english
#Configure paths to work in anywhere
#Example jupyter notebooks with hyperspectral handling
#rename variables in english
#Example in a jupyter notebook how to work with one 1 band raster
#Example jupyter notebooks with hyperspectral handling and transfer learning #Ensemble models with images from different bands with different resolutions
#Create a free QGIS plugin for semi-supervisionade DL s2
#Example how to work with one 1 band raster
#Code hyperspectral handling
#Code transfer learning with hyperspectral handling

Commits are welcome!