Detection and removal of clouds in satellite imagery is a natural application of machine learning, but freely available training datasets have been in short supply. To address this, Azavea has produced a public dataset consisting of 32 unique Sentinel-2 tiles with cloud labels produced by humans.
Esri’s 2020 Global Land Cover Map and Predictive 2050 Global Land Cover Map forms part of the Living Atlas, and is available under a Creative Commons license. Both products are available in ArcGIS Online as a map service, and for download and viewing.
Radiant Earth Foundation has released “LandCoverNet,” a human-labelled global land cover classification training dataset. This release contains data across Africa, which accounts for ~1/5 of the global dataset.
The International Food Policy Research Institute has published a new version of Spatially-Disaggregated Crop Production Statistics Data for Sub-Saharan Africa for 2017. This new regional update, available to download from the IFPRI Dataverse, marks the fourth generation of the data series.