Mood Matcher
A playlist for every emotion
Two weeks ago, my friend Aidan and I attended PennApps with the hopes of creating a cool hack that we would be proud of and not stress ourselves over and thus Modd Matcher was born.
Inspiration
As two individuals with great music taste, we knew we wanted to create a music hack! We were inspired by how music affects our emotions and emotions affect what kind of music we listen to!
What It Does
Using your webcam, snap a picture of yourself and hit “Generate Playlist”! We will tell you the emotion you’re feeling and what Spotify playlist you should listen to keep the mood going. And of course, we have a feedback form so feel free to send us a message if you have any thoughts, comments, or concerns! We would love to know what we could do to improve!
How We Built It
We used the Google Vision API to analyze facial expressions to recognize which emotion the user is expressing. We then used the Spotify API to generate a playlist that matches the emotion. You can then click the image to open the playlist in Spotify! To connect all of these moving parts, we used Flask so we can combine a Python backend with our HTML/CSS/JS front end.
Challenges We Ran Into
Some challenges we ran into was taking a picture of what is showing on the live webcam view. Another struggle came with how we were going to parse the image uri into the Google Vision API to analyze an image. Having the user upload an image would’ve been easier but would detract from the overall user experience.
Accomplishments We Are Proud Of
We are particularly proud of the webcam and Google Vision API integration. It is our first time using the Google Vision API and using webcam functionalities for a web application.
What We Learned
We learned how to use JSONS, Google Vision API, Spotify API, and JavaScript. We also further developed our skillset in full stack web development!
What’s Next For Mood Matcher
Our next steps for Mood Matcher is to host the website using google app engine or heroku. We also want to create our own emotion analysis algorithm so we can display more than 4 different emotions. We would also like to automate what kind of playlists and/or songs get recommended using the Spotify Audio Analysis features.
Rachelle is a CS + Stat major at Rutgers University. She is interested in entrepreneurship, politics, data science/analytics, and how they all may or may not connect!
