Gesture Controlled Rock Paper Scissors

This game proccesses a live video of the users hand gestures, analyzing them for recognized patterns. Landmark detection was applied to a custom dataset of expected gestures, which were then used to train the main model.

The project includes:

• Convolutional neural networks for image feature extraction

• Live video feed

• Instant response from a randomly generated computer 'move'

The system was trained on bursts of image data of my hands.