This technology is a hearing aid system that improves hearing attention and spatial localization in particularly crowded and loud environments.
Current hearing aid devices are not able to isolate speech from a specific speaker in a noisy and fluid environment. The inability of hearing aid technologies to localize and amplify specific sounds increases the listening effort of hearing-impaired subjects. To improve hearing attention and speech separation, research efforts have largely relied on static or stationary background noise and talking streams, which are not reflective of most natural conversation environments.
This technology consists of a hearing aid system comprised of a cognitively controlled hearing aid with automatic speech separation and an attention decoder algorithm. This system estimates the localization of overlapping speakers and suppresses background noise in sound space, then uses neural data and a decoding algorithm to predict and amplify the attended speaker. This approach allows a person to better identify, localize, and attend to a conversation, despite noisy environments.
This technology has been validated using a task in which patients are instructed to attend to one of two spatially segregated and moving speakers.
Patent Issued (11,961,533)
Patent Pending (US20240013800)
IR CU23134
Licensing Contact: Beth Kauderer