Granularity / Spread of Input Dataset

[tobiaskauer]

Questions:

• How many locations are in the two biggest datasets (from my knowledgeNY / Boston)
• Where are they?
• Can we generate beautified scenarios (six scenarios, each with a realistic image, labels and calulated metrics) for each of these locations?

[tobiaskauer]

@sagarjoglekar Can you also give me the geolocations for the input and the output images?

[sagarjoglekar]

Boston Results : https://www.dropbox.com/s/tg4tvqrk8c5wcjp/BostonMatches.tar?dl=0

[tobiaskauer]

closed. for boston we have:

• ~1,3k locations with a trueskill score, of which
• ~800 have labels (this will maybe grow to the full 1.3k), of which
• 130 have a more beautiful alternative