Here’s a clever use of a webcam and some facial recognition software — They call it Happy ++ and it will DJ [Pharrell’s] Happy according to how much you’re smiling (or not at all!).
It’s another project to come out of MIT’s Media Lab for a spring event this year by [Rob, Dan & Javier]. The facial tracking software was re-used from an older project, the MIT Mood Meter, which was a clever installation that had several zones on campus tracking the apparent “happiness” of the students walking by.
To create the program they’ve split up the song Happy into its various components. Drums, vocals, band, and the full mix. As the webcam recognizes a smile, it records the intensity, which in turn turns up the vocals and band. If no smiling is present there is only a drum beat.
Perhaps the most interesting part of this project is its interaction with a group — even if people weren’t paying attention, it was still tracking their expressions, and reacting to their “mood”.
If the group erupted in laughter and began smiling, the technology immediately followed suit and the music became more resplendent. It was almost as though the technology was in tune with the effective contours of the conversation and had the ability to react just like everyone else, albeit in a language of its own.
It would be more interesting to have a system that plays dramatic music from an orchestra as it saw people look grim and tight jawed.
Or if they had a drawn gun. Then install it at airports and checkpoints.
Amazing project, but is it just me that thinks the face projected onto the person in the first video looks like the Disclosure face?