We wanted to create a more efficient way to discover music, that saves time. We were also intrigued by the math behind music discovery algorithms.

What it does

Given a series of images, the user selects images that best match their current mood. Our algorithm will determine the approximate mood of the user and will return a list of songs that they may enjoy accordingly.

How I built it

We learned and utilized web development technologies to make this application possible. For the front end, we used the bootstrap framework to build an intuitive and attractive user interface. For the backend, we used Node.js and Express.js to implement our music discovery algorithm. And we used Microsoft's Azure platform to host our application in the cloud. The Azure services we used were Windows VM, SQL database, and the cloud app platform hosts our entire program.

Challenges I ran into

The challenges we ran into were mostly learning how to use and implement new and cutting edge technologies. Coming in, our team did not have a firm grasp on Node and Express Javascript, and did not know how to use any cloud technologies to host and aid our applications. We had a steep learning curve, but in the end we managed to overcome the odds and we successfully built a cloud hosted application using Node.js.

Accomplishments that I'm proud of

Like we mentioned above, we learned a lot of information and the end result was a cool application.

What I learned

Node.js, Express.js, cool stuff with front-end Javascript, and (last, but not least) how to build and deploy a cloud application using the Azure cloud

What's next for SongInstinct

We plan to refine our algorithm and make our user interface more streamlined so that we can more accurately predict music recommendations based on a users mood-based responses to different images. We also want to plug our app into the Spotify API, so that the user can listen to the music recommendations in our site, using Spotify, without necessarily having to leave SongInstinct

Share this project: