In this lab, you will create an interactive Rock-Paper-Scissors game, in which the user can play with the computer using the camera.
A keras model created using Teachable-Machine is located within this repository at source/keras_model
. This has been trained to recognise the three hand gestures associated wth rock paper scissors, as well as a fourth null gesture.
To replace this model with one that has been trained on your gestures, follow the link above and create the Rock, Paper, Scissor, and Nothing classes, using your own images.
A testing script was provided to ensure this model works, located at source/RPS-Template.py
to execute it, an appropriate conda environment with opencv-python, keras, and tensorflow is required.
located at source/rock_paper_scissors.py
is a simple game that will play Rock-Paper-Scissors. This can be ran without any dependancies.
located at source/CV_RPS.py
is the rock-paper-scissor game combined with the keras neural network. This requires the same environment as the keras model above does.