Wakey Wakey

Wakey Wakey: Stay Alert, Stay Focused – Your Smart Study Companion.

Comment

Inspiration

Many students struggle to stay awake and maintain focus during long study sessions, especially when studying alone. This challenge inspired us to create Wakey Wakey – a productivity app designed to keep students awake, alert, and focused. By combining real-time drowsiness detection with a Pomodoro timer, Wakey Wakey provides an innovative solution for students to combat fatigue and procrastination while studying effectively.

What it does

Wakey Wakey is a smart focus app that ensures students stay productive by:

  1. Drowsiness Detection System:
    • Monitors the user's face through their webcam to detect signs of drowsiness in real-time.
    • Alerts the user with an alarm and visual feedback when they begin to doze off, using a bounding box and status label ("Awake" or "Dozing Off").
  2. Pomodoro Timer:
    • A productivity timer to help students focus using short work intervals and regular breaks.
    • Allows users to customise work and break durations.
    • Plays an alarm to signal the end of each session.

How we built it

  • Frontend:
    • Built using React for the user interface.
    • Styled with Tailwind CSS for a responsive and modern design.
    • Hosted on Vercel for easy deployment and scalability.
  • Backend:
    • Built with Flask to handle drowsiness detection and API endpoints.
    • Utilised YOLOv5 for real-time drowsiness detection using a trained machine learning model.
    • Used OpenCV to process video frames and detect the user's face.
    • Hosted on Heroku to provide a reliable and scalable backend.
  • Machine Learning:
    • Dataset labelled and trained using YOLOv5 for accurate drowsiness detection.
    • The model detects and identifies user states (e.g., "Awake" or "Dozing Off") in real-time.

Challenges we ran into

  • Model Training:
    • Ensuring the YOLOv5 model achieved high accuracy with limited training data.
    • Labelling the dataset manually, which was time-consuming and required precision.
  • Frontend Development:
    • Designing an intuitive interface that integrates real-time drowsiness feedback and the Pomodoro timer seamlessly.
  • Integration:
    • Connecting the React frontend with the Flask backend and ensuring smooth communication.
    • Managing CORS issues during deployment.
  • Deployment:
    • Debugging Python dependencies on Heroku and ensuring the backend ran correctly after deployment.

Accomplishments that we're proud of

  • Successfully creating a fully functional web app that integrates both drowsiness detection and a Pomodoro timer.
  • Achieving high drowsiness detection accuracy using a custom-trained YOLOv5 model.
  • Developing a responsive and user-friendly interface that provides real-time feedback.
  • Overcoming deployment challenges to host the app on both Heroku and Vercel.

What we learned

  • Enhanced our knowledge of React and Tailwind CSS to build modern, responsive web applications.
  • Gained experience in training machine learning models with YOLOv5 and improving their performance through dataset labelling and fine-tuning.
  • Learned how to seamlessly integrate frontend and backend services, managing real-time communication and CORS issues.
  • Improved our understanding of deploying web applications to platforms like Vercel and Heroku.

What's next for Wakey Wakey

  • Integration with Driving Systems:
    • Expand the app's use case to detect drowsiness while driving in vehicles like cars, buses, and trucks.
  • Advanced Features:
    • Add analytics to help users track focus trends and drowsiness over time.
    • Incorporate gamification elements to make productivity more engaging.
  • Multi-Device Support:
    • Extend compatibility to mobile devices and tablets for broader accessibility.
  • Improved Detection:
    • Further refine the YOLOv5 model for even higher accuracy in different environments and lighting conditions.

Built With

Share this project:

Updates