In an event that celebrates the innovations of cutting edge technology and entrepreneurial ingenuity, we decided to step off the beaten road... way off. We were all fans of poetry, and we wanted to see how we could combine our passion for CS with it. We made a web app that provides a constant video stream input, and, based off of the user’s mood, generates a logical, continuous poem based on the sentimental value.
As engineers, many of us stay away from poetry- yet there are many poets surprisingly looking in to mathematical and computational poetry- poetry that stems from mathematical constructions and poetry flows inspired from graph theory algorithms. By providing these constrictions, we actually end up in an entirely new field of literature, and as such push the fields of literature, math, CS, and humanity further into unexplored voids.
What it does
Feelings is constantly analyzing the facial expressions of those reading its poetry so it can custom-tailor the sentimentality of the poetry it sends back. It creates poetry using some computational linguistics techniques like n-grams to ensure the words logically fit together.
How I built it
- OpenCV + emotime does sentiment analysis
- Flask handles requests, links video stream with OpenCV
- Vanilla JS/JQuery on the front end with webcam.js to capture video