When considering a pair of premium headphones or earbuds, two critical factors often come to mind: noise isolation and transparency. Active Noise Cancellation (ANC) blocks all external noise, allowing you to hear only the audio from your device. Transparency mode lets some ambient sound through while still playing your music. Apple and Sony have excelled in these technologies, but no current audio gear can focus on a single external sound source.

Imagine wanting to hear only one person’s voice in a crowded room. Researchers at the University of Washington have developed an AI-driven headphone kit that allows you to focus on a specific person’s speech. This technology, called “Target Speech Hearing,” uses a brief visual cue—looking at the person for three to five seconds—to isolate their voice. Remarkably, it works even if you’re moving around and not directly facing the speaker.

Professor Shyam Gollakota from the Paul G. Allen School of Computer Science & Engineering explains, “In this project, we develop AI to modify the auditory perception of anyone wearing headphones, given their preferences. With our devices, you can now hear a single speaker clearly even in a noisy environment with lots of other people talking.”

How It Works

Using commercially available headphones with built-in microphones, the user activates a control while facing the speaker. The microphones capture the targeted speaker’s voice, which reaches both sides of the headphones simultaneously. As the speaker continues talking, providing more data, the AI system becomes better at isolating their voice. The captured audio is transmitted to an integrated computer system, where machine learning software analyzes and locks onto the target voice, delivering it to the listener’s ears, even as they move.

The Technology

Professor Gollakota reveals, “We use off-the-shelf hardware components that are fully available for DIY enthusiasts. We use a Sony WH-1000XM4 noise-cancelling headset, a pair of binaural microphones (Sonic Presence SP15C), and an Orange Pi, all of which are available for folks to tinker with.” This setup demonstrates that the technology can be integrated into existing audio gear. Given recent advances by brands like Apple and Sony with custom audio chips, it’s feasible for them to miniaturize or incorporate this tech into their next-generation devices, enhancing transparency and noise cancellation capabilities.

The AI-powered headphones from the University of Washington represent a significant leap in personalized audio technology, promising a future where you can effortlessly focus on a single voice in any noisy environment.

By Impact Lab