Stay informed with our newsletter.
.webp){kind=spacer}
EN
Revolutionize your focus with AI-driven noise-canceling headphones. Designed to isolate single voices and enhance concentration, these headphones are a game-changer for productivity.
A novel prototype of noise-cancelling headphones introduces a groundbreaking feature: the ability to selectively drown out unwanted background noises while preserving specific sounds, effectively creating an "audio spotlight" on desired voices. Traditional noise-cancelling headphones excel at muffling ambient sounds like engine hum but often inadvertently suppress essential audio frequencies.
Engineered by Shyam Gollakota and his team at the University of Washington, this innovation allows users to customize their auditory experience by prioritizing specific voices or sounds, regardless of frequency. The researchers unveiled their prototype at a recent joint gathering of the Acoustical Society of America and the Canadian Acoustical Association.
Powered by a neural network trained on diverse sound samples, including alarms and birdsong, the device enables users to toggle various sound categories through a companion app. The hardware, comprising off-the-shelf headphones with integrated microphones, captures surrounding noise and processes it through either a small microcontroller or a smartphone running the neural network. The result is an edited audio feed delivered to the headphones, effectively filtering out unwanted sounds in real-time.
Remarkably, this cutting-edge technology operates with remarkable speed, processing audio in just 8 milliseconds, thanks to the streamlined neural network design. Gollakota likens the experience to shining an "audio spotlight" on a particular source, enabling users to focus amidst chaotic surroundings.
By offering unprecedented control over auditory inputs, Gollakota envisions this advancement as the initial stride towards augmenting human acoustic perception. The potential applications are vast, promising enhanced concentration and immersion in any environment.
Source: New Scientist
.png)
