During COVID-19 pandemic times, online classes felt boring and the online videos seemed like watching pre-recorded videos which did not help much in learning. We missed the in-person class experience, which was more interactive and easy to understand and learn. Many of our friends felt the same plus many of the classes like welding/circuits, it's not possible to conduct online. So we decided to create VizardEdu to help universities/schools to conduct in-person classes during pandemic situations with safety.

What it does

VizardEdu detects if the students have worn masks on their faces or not. VizardEdu also detects if students are sick by thermal imaging, so that if students with high body temperatures can go home and rest. VizardEdu also provides awareness lessons for COVID-19 by offering video lessons on what COVID-19 is and how we can fight COVID-19 and safely conduct in-person classes. The goal of VizardEdu is to help universities provide a safe In-Person Education Environment for students.

How we built it

  1. Used Google Vision API / OpenCV Viola Jons algorithm for face detection.
  2. Used Google Cloud GUI AutoML for creating a model for FaceMask Detection that detects masks on the faces.
  3. Used Flask REST API endpoints hosted on Google App Engine which provided endpoints for detecting masks and thermal imaging results.
  4. Used Google Firebase for getting real-time data of students
  5. Used Google Storage and REACT for frontend UI to render the results of thermal images/face masks/ and face images.

Challenges we ran into

  1. The cloud setup was the biggest operational challenge.
  2. Integrating UI with backend Machine learning models were difficult.
  3. Making the AUTO-ML results real time.
  4. Rendering and processing images on client side was very challenging to achieve real time results.

Accomplishments that we're proud of

  1. Overcoming the above challenges was the biggest challenge.
  2. We used flask api end points to provide realtime face mask detection results of the AUTO-ML model
  3. We used OPENCV.js for rendering and processing images on client side.
  4. We made our entire we app by integrating our development with Google Cloud.
  5. We successfully created a webapp hosted on Google Cloud App Engine which each university can use for detecting masks on the faces of the students.

What we learned

  1. We learned how to use Google Auto ML api and how powerful it is.
  2. We learned about OPENCV.
  3. We learned how to host websites on Google cloud.

What's next for VizardEdu

  1. The future plans for VizardEdu is to check realtime social distancing.
  2. To improve the thermal imaging results.
  3. Extend this product for public places like Malls/ Concerts/ Restaurants / Resorts / Hospitals.

Built With

Share this project: