There are surefire ways of confirming your identity: DNA, fingerprints, the shape of your outer ear. Now there’s another method: your style of dancing.
According to new research coming out of the Interdisciplinary Music Research Department at the University of Jyväskylä in Finland:
“Nearly everyone responds to music with movement, whether through subtle toe-tapping or an all-out boogie. A recent discovery shows that our dance style is almost always the same, regardless of the type of music, and a computer can identify the dancer with astounding accuracy.”
Any kind of dancing? Yes. The study involved 73 participants boogieing to blues, country, EDM, jazz, metal, pop, and reggae. The only thing they were told was to listen to the music and do whatever comes naturally.
When the results came in, the algorithm correctly matched the dancer with their identify 94% of the time, suggesting that dance moves are kind of a fingerprint for movement.
One glitch: The algorithm could not distinguish what kind of music people were dancing to. The best it could do was pick the right genre 30% of the time. Metal was particularly hard to pin down.
Does this mean we’re heading for dance-recognition software? We’ll soon see when the robots rise.