Sleep Detector

A Flask web application that checks for closed eyes via the browser camera. The image data is processed with machine learning models in the Python server backend. The result is then sent back to the client and updated live on the webpage with colors and text.

Contains settings to adjust both the face detection threshold (e.g. if the app isn't detecting your face, you can turn it up) or sensitivity for closed eyes (i.e. how sensitive the app is to shut eyes, should be calibrated depending on the environment and lighting).

Run the main.py file to start the flask server locally.

Credits

Thanks to Sarcovora, amritapasu, caasib, and elainejiangg, for collaborating on the machine learning component of this project at MIT BeaverWorks. The original project is here.

Server Side Python Dependencies

• numpy
• OpenCV
• Flask
• SocketIO
• Torchvision
• facenet_models

Client Side

• Uses jQuery and SocketIO

Examples