Tech/Science

AI System Allows Headphone Users to Listen to Single Person in Crowd

A team of researchers at the University of Washington has developed a groundbreaking artificial intelligence system that allows headphone users to listen to a single person in a crowd by simply looking at them for a few seconds. The system, known as ‘Target Speech Hearing,’ cancels out all other surrounding sounds and plays only the voice of the enrolled speaker in real-time, even as the listener moves around in noisy environments.

The technology addresses the limitations of traditional noise-canceling headphones, such as the lack of control over whom to listen to in a crowded setting. Unlike current models like Apple’s AirPods Pro that adjust sound levels based on wearer’s activities, the UW system enables users to selectively focus on a specific speaker.

Presented at the ACM CHI Conference on Human Factors in Computing Systems, the AI-powered headphones utilize machine learning algorithms to capture and reproduce the vocal patterns of the chosen speaker. By tapping a button while facing the speaker, the system learns and isolates the voice, delivering a clear audio experience even in noisy surroundings.

Lead researcher Shyam Gollakota, a professor at the Paul G. Allen School of Computer Science & Engineering, highlighted the significance of this innovation in modifying auditory perceptions. He emphasized that while AI is commonly associated with chatbots, this project demonstrates the potential to enhance audio experiences for headphone users.

The team’s proof-of-concept device, although not yet commercially available, provides a glimpse into the future of personalized audio technology. By offering the code for others to build upon, the researchers aim to push the boundaries of AI-driven sound manipulation and improve user experiences in various listening environments.

With the ability to enroll specific speakers and filter out unwanted noise, the AI headphones showcase a new level of customization and control for wearers. As advancements in artificial intelligence continue to revolutionize everyday devices, this development opens up possibilities for tailored audio solutions that cater to individual preferences.

LEAVE A RESPONSE

Your email address will not be published. Required fields are marked *