I used posenet to capture coordinates in real-time, and GPT with some few-shot prompting to give feedback. Curious what ideas folks have for different/better approaches here!
Looking for collaborators if anyone wants to hack on this together
Cheers! + super appreicate any thoughts/feedback