How will you approach to use Kinect to detect the different set of poses?
I have developed near the app, recognised the face emotions. So in that app I counted the percentage of consiliense the gamer's face control points with preset control points. For every emotions there was a treshold to be counted as working (for example 80% for smile preset).I will use the same math
Proposal:
Hello! I have an experience with Kinect 1 and 2. A week ago I have finished the project for Hasbro, where 6 robots have repeated the motions of the players, it was an adv project for a Hamley's shop in Moscow.
So, I can do it all - the video playing when idle, player detection, press buttons with a hand (the fist guesture works not very good especially for children, i made it like hand is just over the button for a second - so it is pressed). About the dance poses I will write as an answer on your question.
Some questions for you:
1. Daily code sending. As a freelancer, it is not very good idea. I'm ready of course to send you builds everyday with new options and features. But the code will be sended after the whole project done. The code could be sended everyday if we had a signed paper, but not as freelancer. I hope you will understand, and of course it will not late the project.
2. About animated gif - why animated? The player should take a static positions from stop-frames, so what should we animate?