CrowdReader | Devpost

Inspiration

At school, we notice that teachers try very hard to adapt their teaching styles to advantage their students, but they unfortunately fail. When thinking about a way to fix this problem, we thought of a way to detect the emotions and engagements of students. We then thought more and realized this idea could be furthered to the political and business world for all sorts of speeches, presentations, and conferences.

What it does

Our web app demos the ability to utilize machine learning to determine the holistic emotion of a crowd based on individual facial expressions.

How we built it

We used Amazon Web Services in python to take in an image and produce confidence levels for various emotions. We took live images from our web app as data and manipulated our output into a graph to be visually appealing

Challenges we ran into

It was difficult to utilize the web camera in our web app and then transferring that data into Flask to be processed by the machine learning technology in our python file.

Accomplishments that we're proud of

Being able to implement machine learning and integrate it into Flask and HTML was truly something we never thought we would be able to do, if not for HackTJ

What we learned

Coming in, we knew very little about Flask and essentially learned all of it in the last 24 hours. We were first introduced to the power of Amazon Web Services in the workshop which inspired us to learn and utilize its machine learning power in our project.