FocusBar

• 

  Widget before a study session

• 

Inspiration

Staying focused while studying is difficult, especially with constant distractions like phones and other activities. We wanted to create a tool that gives students immediate, visual feedback on their focus so they can better understand and improve how they study.

What it does

FocusBar is a real-time productivity tracker that uses a webcam and computer vision to classify a student’s focus state as productive, distracted, or neutral. It displays live feedback in a Windows-style widget, tracks time spent in each state, and provides session history and cumulative statistics.

How we built it

FocusBar was built in Python using OpenCV and a custom-trained YOLO model. The vision system detects productive and distracting actions, a logic layer determines the user’s focus state, and a scoring system tracks time-based performance. The UI presents live progress bars, controls, and historical data in a clean desktop widget.

Challenges we ran into

Training a custom object detection model with limited data was challenging, especially when distinguishing similar objects. Translating raw object detections into meaningful focus states also required careful logic to avoid false positives and inconsistent scoring.

Accomplishments that we're proud of

We successfully trained and deployed a custom YOLO model in a real-time application. FocusBar delivers continuous feedback, accurate time-based scoring, and an intuitive interface that turns computer vision into actionable study insights. We are also proud of the threading of the project, something we have not used before and had to learn.

What we learned

We learned how to train and integrate custom computer vision models, work with real-time video input, and design modular systems that connect vision, logic, scoring, and UI components into a cohesive product.

What's next for FocusBar

Next, we plan to increase the number of actions we recognize, allowing more things to be classified as focused v.s. distracted. We also aim to improve the widget and have it give you more detailed feedback about when you study best.

Built With

• opencv
• python
• tkinter