Eye-tracker

This application allows you to scroll a web page with your head movements.

How eye tracker works

1. The user's webcam detects his face and eyes
2. The algorithm detects user blinks
3. After a double blink, the algorithm tracks the user's head movements to get the direction to scroll
4. Then the scroll starts and stops when the user blinks twice

How it works

• Open this DEMO
• Allow access to the camera
• Wait for the algorithm to recognize your face
• Make a double blink to start scroll (the "unlock" icon will appear in the upper right corner)
• Nod up or down
• Make a double blink to stop scroll (the "lock" icon will appear in the upper right corner)

How to use

To clone and run this application, you'll need Git. From your command line:

# Clone this repository
$ git clone https://github.com/Arcady1/Eye-tracker.git

# Go into the repository
$ cd Eye-tracker

# Install dependencies
$ npm install

# Open the index.html file

Credits

This software uses the following open source packages:

• jQuery "^3.5.1"
• Browserify "latest"
• TensorFlow "2.6.0"
• Face landmarks detection model "0.0.1"
• Chart.js "^2.9.4"

Acknowledgments

• TensorFlow Blog
• MediaPipe
• MediaPipe Iris
• MediaPipe Iris MODEL CARD

You may also like...

• Don't touch your face - Face touch detection with BodyPix segmentation
• Doodle Recognition - Web app classsificator based on the Quick, Draw! Dataset.
• Pomodoro Bot - Telegram bot with the pomodoro timer

License

MIT