Teachers Pet

Inspiration

With the increasing use of technology, our team was concerned about student's use of AI and loss of focus during classwork, leading to a loss of authenticity. So we wanted to make a fun and annoying program to discuss this.

What it does

This program uses camera detection to be able to detect people. In this case, we were able to specify for teachers. Therefore, when a "teacher" walks by, the user's screen freezes, red light starts flashing on the outside, blares a loud siren with a text-to-speech voice claiming what website the user is utilising (based off of the list of restricted websites that we created). There is also a character, the Hackathon's mascot, Quacker, who is the one talking about what website the user is using in a humorous manner.

How we built it

How we built it: For the detection of the teacher, we used OpenCV together with YOLO (via Ultralytics) to detect people in the camera feed and determine when a teacher enters the frame based on bounding-box size and position. This prevented false positives from brief or distant detections. ## Challenges we ran into:

Accomplishments that we're proud of

We are proud that we were able to finish our project in time, as we thought our idea was quite ambitious and this was our first hackathon. We are also proud of our team delegation as we managed to organise tasks well. We also utilised ElevenLabs API for generating the tts for our mascot, Quacker.

What we learned

We learned how to use many new python modules which we were not previous famillar with such as Yolo and Cv2. Furthermore, this was our first hackathon, so working non-stop for 24 hours was a challenging yet rewarding experience for us.

In the future, we want teachers pet to be enhanced even further by adding a proximity sensor which detects how close or far the teacher is, allowing for more accurate and effective detection. This would minimize false alarms, decreasing the amount of distractions in class, ultimately helping students to focus clearer.

Built With

• cv2
• elevenlabs
• opencv
• pydub
• pyside
• python
• ultralytics
• yolo