The problem Abhisargah solves

Abhisargah is an Android-AI based mental health assessment tool, which securely collects and analyses user data to keep a record of their mental health while upgrading its own performance. Unlike any other mental health assessment tool, our app has lots of different features described below:

AI-based Mental Health Assessment tool

Cloud-based storage for security and accessibility

Chat-bot for better UI and better connection with user.

Tableau Data Visualization tool

AR based data collection

Unique ID for every user, which will keep the results secure.

Mental Health-related articles and yoga sessions linked to the app

Test results based on 24 Key indicators

Test results of user to be accessible by their parents and children

Due to connection with cloud it also ensure security of personal data of users.

Challenges we ran into

Building this app was a roller coster ride for every member of our team . We encountered many issues while developing this app ,Major issues that we encountered are:

Integrating our App with Machine Learning model: There exists number of ways for integrating Applications made by some older languages like Java and Kotlin, but with Flutter there aren't many reliable libraries to do this task.So to do this we had to go through bunch libraries and other solutions. Finally after brainstorming and asking bunch of people we found out this library that accepts model in tensorflow which allows us to locally model and predict the dataset which we gathered and turning it into desirable result.

This led to our second issue : Converting our ML model into tensorflow model, since
our team mates had previous learning strictly to machine learning but not with RNN , they had to spend rigourous amount of time and effort in order to convert the model which they build before into tensorflow model.

Data Collection: This is probably the most difficult hurdle that we came accross.In this we had to collect datas of different user groups , searching through various kaggle datasets to get meaningful data. Our team in this process consulted many therapist and some teachers in order to get some more features that they can use and we constructed all the datas and . This however is not completed and it will be trained and will get better in the future and hopefully we would be able to diagnose many other mental health problems. For this future data collection we are planning to use AR(Augmented Reality)to gather data to expand this application boundaries.