Inspiration
Our team had two primary inspirations, one was the availability of a large, compatible dataset as almost anyone involved with AI knows, a lack of or bad dataset can end a potentially promising project in a heartbeat, and two was an improvement upon current mental health chatbots as with emotion detection, can give advice previously impossible with asking invasive questions.
What it does
Our project has three main parts, the AI in charge of determining emotion and giving suited advice, the webpage that makes it easy to interact with it, and finally, the flask library and the javascript code that allows the AI to interact with the webpage.
How we built it
We used Google Colaboratory as it gave us free access to GPU's and TPU's that would otherwise have cost thousands of dollars and allowed us to start considering pursuing complex behemoths of AI's such as VGG-16 as well as not have to deal with setting up python virtual environments that need hundreds of packages and libraries to download to each time as Colab does this for us automatically.
Challenges we ran into
One of our most massive challenges was getting the AI that was in Python to interact with the HTML webpage as without a proper server, seemed near impossible till we discovered the Flask framework which allowed for communications albeit with some hurdles along the way.
Accomplishments that we're proud of
The accomplishment that as a team we were most proud of was not just getting a super complicated AI to work incredibly well with our task, but doing so in just a day allowing our attention and efforts to be shifted to other major problems, such as flask and Webhosting.
What we learned
This project taught our team not just what it's like to build a full-stack system with incredibly limited time constraints, but to work with each other and to consider each other's strengths and weaknesses when it came to coding and teamwork, and we already have plans to make our team's synergy significantly better and therefore allow us to pursue projects we could have only dreamed of before
What's next for AI-Powered Emotion Detection and Mental Health Advice
When we eventually revisit this project with much more experience under our belts, we can add true NLP neural networks to our mental health advice, allowing for even more personalized mental health care and pushing the boundaries of what we can do.
Log in or sign up for Devpost to join the conversation.