A Mood-Aware Playlist for Music Hack Day
If you’re into music and technology, and never attended Music Hack Day, you’re missing out. Not just on a full weekend of intense music hacking, but also meeting all sorts of great people. At Music Hack Day Paris we’ve come across an interesting mix of artists, designers, back-end, front-end and app developers. But most of all, we’ve returned – be it totally exhausted – with better knowledge of the Deezer & Gracenote APIs, and totally proud of our little mood-aware automated DJ hack, Moodji.
Now what has music to do with mood? Everything! When we’re feeling down, we want a tune to comfort us. When we’re all excited, there’s need for music just shouting ‘#gogogo!”. Music has the power to change our mood for the better. Music just rocks our mood. That’s why our hack..
Moodji, the little mood-aware automated DJ hack
Based on the Argus Labs Mobile SDK for ambient sensing, the Gracenote Rhythm API, and the Deezer API, we’ve made Moodji. Our Moodji hack senses in real-time how you’re feeling, and creates an adaptive playlist filled with music you like that ‘s ideal to listen to in your current mood.
For the Moodji live demo, we decided to showcase the music suggestions based on Ben’s mood. Opposed to emotion, mood is something that evolves granularly over time, influenced by a variety of variables; from how much you sleep, over who you interact with to what’s the weather like. So for demo purposes, we had to move things up a bit. Rather than go with running a few preset stories through the system (safe, but boring), we’d had to find something that would change Ben’s detected mood almost instantly. The solution was quite simple – an massive SMS insult bomb, with the help of the hackday audience.
The demo effect meant that – of course – the text messages arrived delayed. But the all came through eventually! We’ve got to see Ben’s mood – and the music accordingly – evolve from neutral, to active, to angry. The follow up positive messages got him back to his usual positive self. These mood changes got us the following tracks through the Gracenote API:
- Neutral: U Don’t Have To Call by Usher
- Alert: Live for the Night by Krewella
- Angry: Through The Fire And Flames by Dragonforce
- Positive: Can’t We Be Sweathearts from The Cleftones
If you’re curious to learn more about our approach of combining automatically detected mood states with music, or about what else regarding human behaviour and mindset we can deduct through from smartphones and other consumer sensor devices, drop an email to firstname.lastname@example.org, or poke @arguslabs on Twitter.
Of course, we weren’t the only ones that had some great stuff to show off by demo time. For example, have a look at Eugenio Tacchini’s HER/Music for example, a hack inspired by the film Her.
The conversational music curator fluently and friendly responded to questions such as “play me something more recent.”
There’s an overview of the 25 projects submitted on Hacker League –do check it out for some more music magic.
Overall, we’ve had an amazing time, learned a whole bunch of new things, managed a few firsts (for me, that was first ever GitHub commit) and got to know some great people & APIs more closely.
So thank you @Deezer for organising this wonderful experience, and @Spotify for the light-filtering swag. Those sunglasses came in handy to let the eyes recover from 36 hours of hectic hacking.
It was awesome!
*Demo Effect, 2 nouns. From ‘demo’, a public demonstration of the capabilities of something, typically computer software, and ‘effect’, a change which is a result or consequence of an action or other cause. Combined, these two nouns refer to the fact of device working perfectly always except when demonstrated for an audience. Something unexpected that goes wrong only when someone else than just oneself is seeing. Somewhat connected to Murphy’s law.