Can a machine learn to solve our speech in noise problem?

by Daniel Fink, MD, Chair, The Quiet Coalition

This piece in The Hearing Journal asks, “Can a Machine Learn to Solve our Speech in Noise Problem?”

Maybe yes, maybe no.

The “speech in noise” problem is the difficulty many people with hearing loss–and even people with normal hearing as tested by pure tone audiometry–have  following a conversation if the room (often a restaurant or party) is noisy.

I have that problem, as many adults do, and I also have three problems with this article.

First, talking about a technological solution to the speech in noise problem without discussing how we can interfere with the development of this condition by simply making the world quieter to prevent hearing loss is irresponsible. Imagine public health officials in the 1950s focusing on making better wheelchairs, braces, and crutches for those affected by polio without also working to prevent polio by developing a vaccine. You can’t, because that would have be absurd. To prevent noise-induced hearing loss, we don’t need more research. We don’t need a vaccine. All we need is to make a quieter world, something that has been known for decades.

Second, an even better solution to the speech in noise problem would be to require quieter indoor spaces.

Third, requiring quieter public spaces is exactly what the Americans with Disabilities Act (ADA) requires. People with hearing loss clearly meet the ADA definition of having a disability, and they require “reasonable accommodations” to allow them to fully enjoy (yes, this is the legal standard in ADA) places open to the public. I will be speaking about “Disability Rights of Ambient Noise for People with Auditory Disabilities Under the Americans with Disabilities Act” at the December meeting of the Acoustical Society of America, in New Orleans. I recent learned that my talk will be broadcast live over the internet. Details of how to listen will be posted when they become available.

Dr. Daniel Fink is a leading noise activist based in the Los Angeles area. He serves on the board of the American Tinnitus Association, is the interim chair of Quiet Communities’s Health Advisory Council, and is the founding chair of The Quiet Coalition, an organization of science, health, and legal professionals concerned about the impacts of noise on health, environment, learning, productivity, and quality of life in America.

Comments (3)

  1. Michael Occhipinti

    Well maybe someone should tell someone at the ADA to STOP telling the MBTA in Boston to crank their dings and bells (ostensibly to help the deaf but instead just creating MORE deaf people). I asked WHY the bells in the cars were all over the map, some at 100 dBs and I was told flat out ‘The ADA came in here and insisted’. I was furious, because I find the noises on the Orange line to be extremely stressful, even with my earplugs in. So, as SOON as someone tries to put the kibosh on noise, the noise lovers and apologists will step in and say ‘stop ruining our good time’. Good luck getting any kind of sanity implemented. One would have to take on the leaf blower industry, the jerks that honk their horns unnecessarily, the ridiculously loud and out of code trucks that pass through on a regular basis, the absurdly loud EMT sirens (its not even the Fire engines any more), the jerks with illegal muffler kits on their motorcycles, the absurd trendy restaurants with their minimalist echo chamber dining spaces, and lets not forget the absolute tools who insist on cranking movie trailers to 15 dBs higher than the move. This world is nuts

    Reply
    1. GMB (Post author)

      There are people who are pusing back against the daily assaults of noise. The Quiet Coalition (https://www.thequietcoalition.org) was founded to fight against noise from a health perspective. Change is slow, but it’s coming.

      Reply
  2. Jan L. Mayes

    The problem for machines, including hearing aids, is that the “noise” is actually speech of other talkers in the area. A machine can’t selectively cut some speech and leave other speech unaltered. Another very weak idea contradicting the number one goal of noise prevention that you highlight: noise control aka remove the noise.

    Reply

Leave a Comment

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