Github Repo | Live demo

Finger Rumble is an interactive web application that utilizes the power of deep learning and transfer learning techniques using TensorFlow.js to detect and classify hand gestures captured through the user's webcam. This allows for a hands-free and immersive gameplay experience.

hassan Shoayb Finger Rumble

Usage

  1. Visit the Finger Rumble website here to access the web app.
  2. Allow access to your webcam when prompted to enable gesture detection.
  3. Place your hand in front of the webcam, making sure it's clearly visible.
  4. Then make the gesture for rock while clicking on the 'rock' button to add the rock image samples into memory.
  5. Collect at least 50 rock samples, and then do thesame for Paper, Scissors, Spock and Lizard.
  6. After You've collected your image samples, click on the 'train network' button to train the model of your samples.
  7. Once training is complete, click 'Start Predicting' to see predictions, and 'Stop Predicting' to end.
  8. The app will use its deep learning model to detect and classify your hand gesture as Rock, Paper, Scissors, Spock, or Lizard.
  9. Finally, once you are happy with your model, click 'Download Model' to save the model to your local disk.