After looking at the challenges presented by the streamers, we wanted to do an idea that would be fun for the viewers and the streamer. This idea allowed for more interaction between streamer and viewer and allowing viewers to be more involved with the stream.
What it does
Our AI detects emotion from the streamer and the viewers get to vote what twitch emote will rain upon in the stream.
How we built it
We used sentiment analysis to register and detect the emotions of the streamer. We made the animations and did the physics using matter.js and built a voting system
Challenges we ran into
Two main challenges were actually connecting openCV to a live stream and implementing everything into extensions.
Accomplishments that we're proud of
We were able to get the emotion detection running on a livestream.
What we learned
We learned how to utilize runtime data.
What's next for Streamote
A better model with a better confidence rate and better scalability