Simple-dino-game

Teaches agent to play a game like dino-run using python.
Uses next libraries:

• keras (using neural network for learning)
• tkinter (making the game)
• PIL (converting game frames into images)
• numpy (handling images as matrices)

Gameplay of agent after training. Beats the game every single time.

Task list:

• Use distance of obstacle and other relevant params as features.
• Use raw pixel values as features