By Eliot Van Buskirk of Evolver.fm.
We are fascinated with new ways to control our tunes, because smartphones get in the way of music apps. We've already been mightily impressed by the way Leap Motion's neat little cameras can let music fans control Spotify by waving their hands around in front of their computers, and also by apps that let you remix or otherwise affect music as it is playing, using gesture controls.
But nothing prepared us for the knowledge that plain old WiFi signals are capable of sensing our gestures around the home or office — no cameras or special motion-sensing equipment required.
Wisee, developed by researchers at the University of Washington, relies on tiny shifts in the frequency of your Wi-Fi signal when you move a hand or a foot. Those shifts don't affect the overall signal much, but these scientists devised an algorithm that can find them.
In the real world, multiple people within the home would be able to employ up to nine gestures (pushing, pulling, etc.) to control music or anything else around the house. It actually takes two gestures; first you have to perform your personal signature gesture to get the system's attention and tell it who you are, like a sort of physical password.
“This is repurposing wireless signals that already exist in new ways," said lead researcher and UW assistant professor Shyam Gollakota. “You can actually use wireless for gesture recognition without needing to deploy more sensors."
According to the team, which also includes UW assistant professor Shwetak Patel, visiting student Qifan Pu, and PhD candidate Sidhant Gupta, you can wave your hands in the air if you'd care to control something from anywhere in the home, because it doesn't need a line of sight. So far, they have achieved 94 percent accuracy in recognizing these gestures.
The video offers additional explanation about this fascinating new reason not to have to pick up a remote control, walk over to our computers, or negotiate a smartphone menu system in order to skip songs, adjust volume, switch playlists, and so on:
We are fascinated with new ways to control our tunes, because smartphones get in the way of music apps. We've already been mightily impressed by the way Leap Motion's neat little cameras can let music fans control Spotify by waving their hands around in front of their computers, and also by apps that let you remix or otherwise affect music as it is playing, using gesture controls.
But nothing prepared us for the knowledge that plain old WiFi signals are capable of sensing our gestures around the home or office — no cameras or special motion-sensing equipment required.
Wisee, developed by researchers at the University of Washington, relies on tiny shifts in the frequency of your Wi-Fi signal when you move a hand or a foot. Those shifts don't affect the overall signal much, but these scientists devised an algorithm that can find them.
In the real world, multiple people within the home would be able to employ up to nine gestures (pushing, pulling, etc.) to control music or anything else around the house. It actually takes two gestures; first you have to perform your personal signature gesture to get the system's attention and tell it who you are, like a sort of physical password.
“This is repurposing wireless signals that already exist in new ways," said lead researcher and UW assistant professor Shyam Gollakota. “You can actually use wireless for gesture recognition without needing to deploy more sensors."
According to the team, which also includes UW assistant professor Shwetak Patel, visiting student Qifan Pu, and PhD candidate Sidhant Gupta, you can wave your hands in the air if you'd care to control something from anywhere in the home, because it doesn't need a line of sight. So far, they have achieved 94 percent accuracy in recognizing these gestures.
The video offers additional explanation about this fascinating new reason not to have to pick up a remote control, walk over to our computers, or negotiate a smartphone menu system in order to skip songs, adjust volume, switch playlists, and so on:
