8000
Skip to content

Latest commit

 

History

5 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Swath for Earth Observation ML

Rationale:

  • We need a quick way to load ONNX models and make inference on edge devices. These devices often require low power. Solafune's precipitation competition penalizes the expense of using big models, for example.

  • Managing virtual environments, geopandas and python etc is a nightmare on edge devices. Idea is to have a self contained executable that will leverage GDAL and load the rasters, make the transformations, load the ONNX model and run the inference.

Roadmap:

  • Ideally this project would be written in Jai as a self contained program from compilation to configuration. Generate bindings for GDAL, ONNX and make an end-to-end one unit of EO-ML.
  • Add NVIDIA Jetson CUDA support and/or bindings for fast inference on edge devices.

About

Georeferenced raster inference in C over GDAL, PROJ, and ONNX Runtime.

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages

0