Opti-Tracker: Never Loose Attention

Abstract

The Opti-Tracker project is an intelligent eye and face behaviour monitoring application designed to enhance user safety and productivity. Built using Python, PyQt5, and OpenCV, this application processes real-time video streams to detect signs of sleepiness and non-attentiveness based on facial landmark analysis. Leveraging advanced computer vision techniques, the system tracks key metrics such as eyelid closure and eye position, alerting users when thresholds for fatigue or distraction are exceeded.

The user interface is intuitive, offering features like camera selection, start/stop controls, and a dynamically resized video display. By integrating responsive visual feedback and auditory alerts, Opti-Tracker provides a seamless monitoring experience. This application has significant potential for applications in driver safety, workplace monitoring, and productivity enhancement, making it a valuable tool in environments where sustained attention is critical.

Technologies

The main technologies used here are -

• PyQT → User interface of the desktop application.
• OpenCV → Capturing video from the machine, and image manipulation.
• Mediapipe → Pre-trained models and landmark detection on each frame of the video.
• PyDub → Play Audio files through system speakers.
  

A short high level system design diagram is given below

User Flow Diagram

User flow diagram..

User Interface Overview

Here is an interface overview of the application (NOTE: It’s a desktop application so look and feel might be different based on the operating system theme, and desktop environment, the below picture is from a MacOS operating system based on UNIX).

[MORE DETAILED BREAKDOWN... WIP]

Conclusion

The Opti-Tracker project utilizes computer vision to monitor real-time facial behavior, detecting fatigue and distraction through facial landmark analysis. It enhances safety and productivity by providing timely alerts for inattentiveness or sleepiness. With a user-friendly interface, real-time performance, and scalable design, the system is adaptable for diverse applications where focus and attention are critical.

[VIDEO SHOWCASE]
Also there is a full video walkthrough of the whole process uploaded in youtube.