I want a node to exist for
- every detail that was intentionally created to communicate with a driver (e.g. signs, lights, every dash on every highway)
- every detail that is super-low-hanging fruit for computer vision (e.g. vertical poles)
Let's take the zoom level and turn it up to 23!
I'm working on open source mapping power-tools (not yet released):
- DIY mapping cars (writeups + software + 3D-printed parts for mounting in common cars) (GoPros, Raspberries Pi, ...)
- Software for efficiently extracting detail from that data (computer vision + human confirmation/annotation)
If this overlaps with what you're working on, say hello.