Music Genie

Inspiration

There are all sorts of music recommendation services out there. Some of them ask you about your favourite artists, while others look at what you listen to. But none of them capture one essential aspect--when you say that you like an artist, you are making a conscious decision to invest yourself in their music. Not so with Music Genie. Music Genie captures the true feelings of the listener, going beyond any existing music service, touching the feelings of the subconscious.

What it does

Music Genie uses the Muse Headband service to receive EEG readings from the wearer. When they listen to a sample song, Muse picks up on how they react--lots of energy means something the user finds enjoyable. No energy? No interest.

How we built it

We built Music Genie as an android app, using the Android and Muse SDK's. We also integrated with the Last.FM API to determine related artists.

Challenges we ran into

The Muse Headband is, as we discovered, inherently unstable. From dropping out mid-session, failing to connect, or running out of battery during testing, we had no end to the issues with handling Muse. Furthermore, as neither of us had programmed in Java, let alone written an android app before, we had our hands full learning the language, as well as struggling with the GUI aspect.

Accomplishments that we're proud of

We are proud of the way that we were able to work for 24 hours straight, without sleep or break, to take an odd idea into a fully-fledged hackathon project. We are proud of the way we overcame each of the challenges listed above, and came out with a better project, and more knowledge, to show for it.

What we learned

We learned something already well known--Occam's razor. After spending an hour troubleshooting why Muse wasn't connecting, only to find out that it ran out of batteries, or after spending 30 minutes trying to get text to display properly, only to discover that our textbox wasn't large enough to support it, we learned that the simplest solution is usually a good first bet.

What's next for Music Genie

Music Genie would benefit enormously from one major feature we didn't have the time to add; machine learning. By integrating our application with Microsoft Azure, as well as generating a large base of test data, we would be able to train the app to more accurately reflect when a song is liked or disliked.

Built With

• android-studio
• java
• last.fm-api
• muse-headband-sdk