Using pygame to create a 2d pong game and then using gym and tensorflow to read the pixels on the screen using a CNN then model the actions with a Qlearning RNN to beat the ai opponent
- Gym
- Tensorflow (manually compile for best performance)
- Pygame
- opencv 3 (You need to manually compile)