Skip to content

README.md

Latest commit

 

History

History
51 lines (33 loc) · 1.37 KB

README.md

File metadata and controls

51 lines (33 loc) · 1.37 KB

DCAI: GeoCroissant Recipes

What is GeoCroissant?

GeoCroissant extends the MLCommons Croissant metadata standard with geospatial concepts for GeoAI workflows. It introduces support for spatial and temporal coverage, coordinate reference systems, spatial resolution, band configuration, and time-series metadata.

Requirements

Install the required dependency:

pip install mlcroissant

Usage

Load a GeoCroissant dataset using the mlcroissant Python library:

import mlcroissant as mlc

dataset = mlc.Dataset("geocroissant.json")

Conformance

GeoCroissant datasets must declare conformance to both the Croissant and GeoCroissant specifications:

"dct:conformsTo": [
  "http://mlcommons.org/croissant/1.1",
  "http://mlcommons.org/croissant/geo/1.0"
]

Authors

Rajat Shinde, Manil Maskey, AG Stephens, Harsh Shinde, Joseph Edgerton, Dr. Tejasri Nampally.,  
Douglas Fils, Edenna Chen, Claus Weiland, Pedram Ghamisi, Gerald Fenoy,  
Yuhan Douglas Rao, Omar Benjelloun, and Elena Simperl

GeoCroissant Working Group · croissant-geo@mlcommons.org

Acknowledgements

  • MLCommons GeoCroissant Working Group
  • MLCommons Croissant Working Group
  • Open Geospatial Consortium (OGC) GeoAI Domain Working Group