STAR Analytical Services awarded NIH grant to develop system for hearing-impaired listeners

STAR Analytical Services announces the award of a 3-year $1,500,000 grant from the National Institutes of Health to develop an assistive conversation system for hearing-impaired listeners, including many aging baby-boomers, who have difficulty understanding and participating in conversations in noisy environments. The proposed project will extend STAR's previous work on novel techniques to suppress distracting sounds and enhance conversational speech. The resulting appliance will enable listeners to create an immersive virtual acoustic reality. In 2010, STAR received a small Business Innovation Research (SBIR) award from NIDCD, an institute of NIH, to improve the lives of people with communication disorders. This new grant extends that project into phase two.

This innovative research, conducted by Senior Research Fellow Dr. Richard Goldhor, Chief Scientist Dr. Joel MacAuslan, and UMass-Dartmouth Professor Karen Payton and their team, will improve on algorithms for audio analysis and develop the device. For a demonstration of this processing, with a stirring speech and a "fascinating" discussion of the mathematical techniques, review the samples here: http://staranalyticalservices.com/audio/JFK_Sample_Output_1.wav

"The goal of the design," says MacAuslan, "is to automate much of the complexity of acoustic source separation, to make it simple for a user to control the system through an app and to deliver an improved conversation capability. The app will be simple to set up, simple to integrate with other personal devices, and most importantly, simple to suppress irrelevant background sounds." Other examples of the utility of this approach include making it simple to attend to sources of interest, such as conversation partners; simple to accommodate novel sources: e.g., a waiter coming to a table in a restaurant; simple to hear important information, even from unexpected sources; and simple to amplify only certain sources, such as nearby ones.