Singular Hearing
Next-Gen Software Hearing Assistive Technology

The most successful businesses are born from just one seed. For Bruce Sharpe, the founder and CEO of the Canadian firm Singular Hearing, that seed of opportunity was a personal challenge experienced by a family member. Sharpe had seen his father-in-law struggle with hearing loss for years. “Despite having worn hearing aids for a long time, we could see him increasingly withdrawing from family events and social situations. He found it too hard to try to hear with all the noise and commotion that is typical in group settings like parties or at restaurants,” explains Sharpe. This prompted Sharpe to look for better hearing aids but he didn’t have any luck. “Along the way, I learned that hearing in noise was a widespread problem, and despite many advances in hearing assistive technology, had not been adequately addressed.”

Inspired to find a better solution to hearing speech in noise, Sharpe brought himself up to speed on the resurgence of AI-based deep learning and how it could be applied to hearing assistance. He founded Singular Hearing to create software based on these new, fundamental advances. Today, the firm offers an award-winning innovative app, HeardThat, to help those with hearing challenges by delivering clear speech in a noisy environment. “Our app leverages AI techniques and uses deep learning algorithms to intelligently differentiate between speech and noise and separate them out from each other. It eliminates the noise and leaves just the speech, enabling you to hear speech more clearly,” says Sharpe.

Breaking the Mold and Going in a New Direction

What makes HeardThat different from other noise filtering techniques or active noise cancellation devices is it harnesses the power of machine learning to turn a smartphone into a futuristic hearing assistant. Most traditional hearing-assistive products are either amplifiers that make all sounds (including noise) louder or noise-cancellation devices that end up reducing all sounds, including speech. Newer wearable devices are unable to provide the necessary resources to intelligently distinguish between speech and noise. However, they lack the required storage space, battery life, and processing power.

HeardThat doesn’t suppress noise—it separates it from the speech and discards it, leaving just speech with enhanced intelligibility. “Rather than creating new hardware, HeardThat works with the devices people already have, whether that is wired or wireless earbuds or headphones, or Bluetooth-enabled hearing aids or cochlear implants, by improving upon their ability to reduce noise,” remarks John Kenmuir, VP of Marketing, Singular Hearing. The product fills a long-standing gap in the hearing assistance market by serving as affordable, powerful software that takes advantage of a powerful portable computer that people already have—their smartphones.

HeardThat is the result of years of research in neural networks that were trained with thousands of hours of recorded speech and noise to learn what is useful speech and what is just noise. With HeardThat, individuals with hearing loss can engage in conversations they would otherwise have trouble hearing. All a person has to do is download the HeardThat app on their Android phone or iPhone, connect their hearing device to the phone to get started immediately. What’s more? Because HeardThat is an agile and flexible software solution, it can be quickly and continually improved upon. “Based on machine learning, it gets better all the time through use,” adds Kenmuir.

Our app leverages AI techniques and uses deep learning algorithms to intelligently differentiate between speech and noise and separate them out from each other

With its unique capabilities, Singular Hearing continues to generate success stories in the realm of assistive listening. A particular instance that stands out is the story involving a participant at a convention for people with hearing loss held by the Hearing Loss Association of America. “Before the event started, they had background music playing. People around the table were trying to have conversations, and this person next to me said: ‘Why are they playing music? It just completely ruins my ability to hear people talking because of my hearing situation.’ I introduced her to our app and we quickly set it up,” mentions Sharpe. “In two minutes, we had the person going. They kept it on for the whole evening and were able to comfortably take part in the event.” In another case, HeardThat helped a child with learning disabilities focus better on their online classes by eliminating distracting sounds.

Paving the Way for a Better Future

Harnessing its deep expertise in machine learning, audio, and speech processing, Singular Hearing will continue on its quest to create innovative products to help people hear better. “Speech and noise are a problem in various environments and not just for people with hearing loss. We believe our product will be a great success in such situations as well,” asserts Kenmuir. Alongside, the firm is dedicated to continuously updating its app to meet users’ needs. More recently, the firm started a new program, field research, to help advance its application for a variety of real-world situations. In this endeavor, Singular Hearing is seeking Android and iPhone users to assist by using HeardThat in real-life noisy settings and submitting reports on their experience. “Field research will enable our engineers to integrate user experiences and more data into the AI algorithm development. We really believe that offering machine learning, in particular, will make a difference for the future of hearing assistance.