The ultimate rock-paper-scissor detection using Machine Learning. I used Google Mediapipe API to detect all the 21 hand landmarks and using those landmark's (x, y) coordinate I calculated the distance. √ (x2 – x1) + (y2 – y1) that's the formula in case you didn't go to school. 21 points so there are total 210 possible combination and all of those distance is what the Machine Learning model was seeing to predict rock, paper or scissor. Which is a little overkill but at the end it's worth it.
Code link:
https://github.com/deep-roy-python/Th...