This project uses hand recognition with the help of TensorFlow to simulate keyboard inputs that can be used to drive cars in a generic car driving game.
How does it work?
- A frame is taken as in input from the Webcam video stream.
- Neural networks is used to detect the position of the hands (for details check https://github.com/victordibia/handtracking)
- The controls work in two modes:
- When two hands are detected, the locatio of both the hannds are used to drive and steer.
- When 1 hand is detected, the user can cchange gears to reverese.
- The inputs are done with the help of pynput in python.
- Video preview: https://youtu.be/V222C7h71tQ