The Smart Class application was a protoype project my supervisor tasked me with building as a possible solution to helping teachers during the pandemic have an alternative method to tracking student engagement during live synchronous online classes.
A Flutter app that will send images to an API, the API will classify the level of engagement of the student and save the result to a small Database. Later on, the API will provide the app with an endpoint to have a metric report of the engagement throughout the class. The API codes that the app communicates with can be found in this other repository which was used to store the ML model code that generated the model that classified sent images. @KimRaicho developed and trained the ML model that was used for classification of images. The repo also houses the FLASK API and server code within the server folder which I contributed to by writing the Flask API code and designing the simple database we were going to use.
Moving back to the FE, before building the application I made some designs using Figma.
If you have Android Studio, it is simply a matter of cloning the project and clicking play. In order to fully be able to test the uploading of images and creation of a report, you'll need to run the server whose code can be found in the aforementioned repository. In the readME of the repo, you'll also see that there is a link to the final model used.
The demo file was 3mb too large so here is a link to it instead. Smart Class Demo