Highlighted projects

Training dataset for PEARL

Development Seed's Data team worked on creating an open training dataset for PEARL, a landcover mapping platform that uses human in the loop machine learning approach. For this, the Data team annotated 7 classes, such as: tree canopy, grass/shrub, bare soil, water, buildings, roads/railroads, and other impervious (service roads, parking lots) in 3 AOI’s of Detroit.

The tool used for this annotation was JOSM, a tool that allowed us to do segmentation labeling of several classes in satellite imagery, in addition, one of the advantages of this tool is it allows you to add and modify boundaries fast and easy.

View of the annotation progress in the JOSM tool.