Researchers at the University of Bristol built a wearable that translates facial expression into ultrasonic words that can be heard up to 30 meters away.
The current prototype may not be the most subtle of devices, but its creator, Asier Marzo at the University of Bristol, UK, says that later versions of the system could help soldiers communicate without alerting anyone to their presence. Ultrasound travels particularly well underwater so it could also be used to help divers communicate over long distances.
It’s a particularly weird experience for the person on the receiving end of the message, says Marzo. “It’s a very quiet sound, but you don’t know where it’s coming from. It’s like someone is whispering into your ear, but nobody is there.”
To use Marzo’s device, a person first attaches a speaker to their forehead or chest and four electrodes to their lips and jaw. Using a technology known as electromyography, the electrodes capture the electrical signals produced by the facial muscles as a person talks.
Marzo and his colleagues then used a machine learning algorithm trained to recognise which electrical muscle signals were associated with 10 words, including “back”, “stop”, “yes” and “no”. When tested, the system could recognise the correct word more than 80 per cent of the time, even if the person wearing the electrodes was only mouthing them.
The word-recognition system was then paired with wearable speakers that played an ultrasonic recording of the words that a person whispered or mouthed. The speakers emit ultrasound in a narrow beam, at a six-degree angle, so that only people directly in its path would hear the words being spoken.
The ultrasound can be transmitted in the direction the wearer is looking if an eye-tracking camera is attached. For even greater precision, the ultrasound also can be directed using a laser pointer. But pointing a laser at someone is likely to attract attention, says Jennifer Bizley at University College London.
Later versions of the technology could detect a wider range of words and shrink the speaker system so that it could be built into clothing. But for now the device is a fun way to creep out people by whispering in their ears from a distance.