The problem SLOE: Supervised Learning in Online Education solves

With the current pandemic going on, education has shifted to online mode, which has a lot of downsides. Amidst all of this, we have a positive service as an add-on to the current medium of online classes. Childhood depression is something that is under looked and has long-term effects, including the mental growth of a person. There are many reasons which lead to a child’s abnormal behavior which cannot be known unless the child tells, like bullying, or getting scolded by a teacher, and even arguments between parents, and especially during these tough times it is really hard to cope up with mental health. Sometimes abnormalities get noticed, but most of the time, child won’t show such traits in front of parents and absorb what she/he is feeling. Mental health is a very serious issue, which needs to be looked upon with utmost care. To create virtual access to a child’s mental health we have come up with our service - SLOE (Supervised Learning in Online Education). As education is online, students sit hours in front of screens, and that’s the only time most of them are interacting with the outer world. During this time, without them noticing we record their behavioral pattern, and notify parents if there is an anomaly in the behavioral pattern of their child, or in simpler terms a sudden change in the behavior of their child during classes.

Challenges we ran into

Integrating the ML model into the app

This was one hard nut to crack. There are technologies like Flask or Streamlit available for easy deployment of a ML model into websites but such is not the case in App Development. We used Firebase to deploy our custom ML model and had to go through a lot of documentation in order to understand the deployment methods and input-output formats for the model which we integrated into the app.