“Map with AI” dataset of missing roads in OSM extracted from satellite imagery by non-supervised ML algorithm.
Reference data is available for almost every region in the world.
Disadvantage is the creation date of 2019 -> gets more outdated every day
Define:
Metadata (Name, description, etc.)
Layer definitions
Problems:
lots of data (Africa alone around 100 GB)
rasterize data to a 1x1 km grid (same spatial resolution as GHSPOP) (Mollenweid) -> advantage is that function to deal with such a raster dataset is already in place (raster client)
data is not verified; a lot of incorrect data in desert regions
“Map with AI” dataset of missing roads in OSM extracted from satellite imagery by non-supervised ML algorithm. Reference data is available for almost every region in the world. Disadvantage is the creation date of 2019 -> gets more outdated every day
Define:
Problems: