You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Labellers use both JOSM and MapEdit to perform labeling and the same models that help prioritize areas could also help prioritize where labelers should focus their attention. To enable this, where possible the models will be enhanced to also output segmentation masks that can be used as an overlay layer for labelers. The segmentation masks will not be of high enough fidelity to output building footprints, but could aid labeling. As a first step, Azavea will take existing output from one of the model runs and demonstrate its usage as a layer in MapEdit. If this looks promising additional steps will be taken to make it easier to easily import prediction layers into the editing user interfaces. This will include improving Franklin for visualizing model results and improving model running to place its results in a STAC API.
Includes
Create JOSM or MapEdit plugin for Granary
The text was updated successfully, but these errors were encountered:
Labellers use both JOSM and MapEdit to perform labeling and the same models that help prioritize areas could also help prioritize where labelers should focus their attention. To enable this, where possible the models will be enhanced to also output segmentation masks that can be used as an overlay layer for labelers. The segmentation masks will not be of high enough fidelity to output building footprints, but could aid labeling. As a first step, Azavea will take existing output from one of the model runs and demonstrate its usage as a layer in MapEdit. If this looks promising additional steps will be taken to make it easier to easily import prediction layers into the editing user interfaces. This will include improving Franklin for visualizing model results and improving model running to place its results in a STAC API.
Includes
The text was updated successfully, but these errors were encountered: