MIT Scientists Use Potato Chips To Spy On Conversations, So Step Away From Those Lays
If James Bond snacked, he would snack on potato chips, and then use the potato chip bag to listen to your conversations. At least, that's what MIT researchers are doing — by recording the teeny tiny microvibrations that your potato chip bag makes in response to your voice, or sound in general, scientists can translate these vibrations into their original form, be it speech, music, or other sound. So throw away your bag of munchies before you delve into any deep, dark secrets.
As originally reported by the MIT News Office, this latest breakthrough in spyware came as a joint effort amongst scientists from MIT, Microsoft, and Adobe, who managed to "recover intelligible speech from the vibrations of a potato-chip bag photographed from 15 feet away through soundproof glass." So advanced is their methodology that neither distance nor any soundproofing could stop them — as long as their camera could see the object, they could effectively hear what was being said in its vicinity.
While the experiments worked best with a high-speed camera, as such minute vibrations are best captured by devices that can record up to 6,000 frames per second, researchers also found that a normal digital camera could serve the same purpose. Even at only 60 frames per second, a run-of-the-mill camera, or even a smartphone, could capture the vibrations needed to determine a speaker's gender or the number of speakers present. With enough information, researchers said, the combination of their algorithm with these recordings could even determine a speaker's identity.
It's not just potato chips that can betray your secrets — researchers performed similar tests on plant leaves, aluminum foil, and even the surface of water, and met with comparable results. These phenomenal (and somewhat frightening) results will be presented in a paper at Siggraph, a computer graphics conference, later this year.
The science behind the madness, one of the lead scientists explains, is quite simple. Abe Davis, a grad student at MIT and first author of the paper, told the MIT News Office,
When sound hits an object, it causes the object to vibrate. The motion of this vibration creates a very subtle visual signal that’s usually invisible to the naked eye. People didn’t realize that this information was there.
Seem impossible? I thought so too. But researchers can't decipher what you're saying by looking at a single part of a leaf or a potato chip bag or point on a surface of water. Rather, they must "combin[e] and filter ... all the motion across [an] image" in order to decode the vibrations into sound. In an explanatory video, the scientists played "Mary Had A Little Lamb" to a plant, which was being simultaneously recorded by one of the researchers' cameras.
Unbelievably, by using their algorithm, the team was able to reproduce a somewhat grainy, but unmistakable rendition of the same tune, all by analyzing the plant leaves' motion. The video then shows the team recording the motion of a bag of chips that is separated from the camera by 15 feet and a pane of soundproof glass. This time, the lyrics of "Mary Had A Little Lamb" are spoken, and once again, the team manages to reproduce the sounds. The words aren't crystal clear, but they're there, and considering they were all derived from a bag of potato chips, it's pretty impressive.
This experiment is a continuation of an earlier study conducted at MIT by Michael Rubinstein. In 2012, Rubinstein presented new software at Siggraph that made the invisible visible, all by amplifying the differences in successive video frames that were otherwise imperceivable by the naked eye. The core concept of the experiment was essentially identical to this new study led by Davies — they both examined the microscopic differences between video frames in order to construct a new picture. The main difference is, Rubinstein applied his to visual imaging, and Davies applied his to audio imaging.
As cool as this new technology is, it isn't nearly developed enough to cause any real concern for spying. As Davies told the Washington Post, "Big brother won’t be able to hear anything that anyone ever says all of a sudden," and this technique isn't necessarily a step up from existing sound detection methodologies. However, this practice just might allow us to "discover sound in situations where [we] couldn’t before," Davies says.
Now that researchers have realized the potential hidden in a bag of potato chips, the possibilities are seemingly endless. At the very least, this discovery provides scientists with an entirely novel way of imaging objects. Whereas previously researchers had to "poke and prod" objects in order to determine how they respond to pressure, Davies' team has now discovered that "all you need is to play sound at them."
So while you don't need to run away from your bag of potato chips quite yet, knowing its capabilities might just make you think twice about what you say around it.
Images: Getty Images (2)