Teacher's Pet

Assisting in the advancement of modern education.

Comment

Inspiration

We wanted to enrich our lives and our community as university students. We were inspired to create this because we wanted to ease the lives of both students and professors around the world.

What it does

Teacher's Pet is WebApp / Robot that allows teachers to teach their lessons to students that aren't present without the hassle of setting up a complicated recording system. The camera it is attached to is built with facial recognition to detect the Teacher and start the class. Once the program is ran, students who are viewing the WebApp will then be able to see, hear, read, and follow along the teacher. There is a function that allows the student to take a screen shot of the whiteboard, and there is a designated area that will translate what is on the whiteboard onto the WebApp. Along with that, the WebApp also translates the video audio that's coming in from the live feed, and it also has the feature to follow the teacher as they move around the classroom.

How it was built

Our backend is a Google Cloud compute virtual machine. It performs a variety of tasks:

  • Serving our webapp to users which includes
    • Streaming classroom video
    • Allowing students to take a snapshot of the live feed and transcribe the picture to text (using UIPath)
    • Eventually dictation to text features so student's won't miss a word the professor says.
  • Facial recognition that allows the professor to *Automatically send an email reminder (using UIPath) to students just by walking in front of the camera *Have the camera follow them around the room by tracking their face The frontend is a static html page served by Flask that interacts with our backend.

Challenges I ran into

Streaming video at a reasonable quality and frame rate was a technical challenge for our team. Some of the front-end functions were new as well, but we were able to figure them out and use them to our advantage.

Accomplishments that I'm proud of

None of us have done facial recognition before, so it was very rewarding to see how accurate and versatile we were able to get our algorithm to be. Integrating the front end with our database was also rewarding as it felt like what actual software would do, instead of using some hacky dirty way to get it to work.

What I learned

Some of us learned new languages, new programming environments, but all of us have come to better understand our physical limits with this 18 hour long event.

What's next for Teacher's Pet

First will be to complete the audio transcriptions, then we think it will make a good tool in the business world for video conferences and other online meetings.

Built With

+ 1 more
Share this project:

Updates