AI Meets Accessibility: Improving Communication Beyond Traditional Captions

Hello and welcome to This Week in Hearing. My name is Gael Hannan, and I am the co-author of the book Hear & Beyond: Live Skillfully with Hearing Loss. Today we are pleased to welcome back Paul Lee, who is the COO, the Chief Operating Officer of InnoCaption, an accessibility technology company that’s dedicated to helping people impacted by hearing loss break down the communication barriers and their telecommunication barriers in their everyday life. Innocaption is FCC entitled and funded through the Telecommunication Relay Service, and this makes its service free, for eligible people.

Paul, welcome back to “This Week in Hearing.” Thank you very much for having me today, Gael.

You know, those of us who use speech-to-text services on our smartphones, and that would be me with profound hearing loss, would be so excited to hear about some of these new and exciting, I’m gonna call them add-ons, but services for InnoCaption. You know, Paul, I am old enough, to remember being excited about when phones, ringy-dingy phones, got volume control. That was groundbreaking. That was life-changing for me. Mind you, it was a long time ago, and here we are a million years later with all this new stuff. But before we get into the new services, Paul, could you give us the short version of the big picture of what InnoCaption does.

Sure. And I think you gave a really nice intro into this and a segue, but we are, at the end of the day, an accessibility technology company. And what we’re trying to do is make sure that we incorporate the latest innovations in all forms of different technology to create empowering solutions for people with hearing loss. And InnoCaption is a real-time call captioning app that sits on your smartphone. We also offer the app for tablets as well as a web browser-based version as well, and it provides really accurate, fast transcription of your calls.

And we are really unique in that we actually provide the captioning through either fully automated speech recognition technology or also through live stenographers. So these would be trained court reporters, and we empower our users to actually switch between either captioning mode at any time, either before or even during phone calls. And this is really unique and something that sets us apart from any other similar technologies or captioning services specifically for telephones. And as you mentioned, we are FCC certified, which means that anybody in the United States with hearing loss is eligible to use our service free of charge.

That’s amazing. Just briefly, I am not in the United States so I am very envious of that service. And so it’s just the United States. Do you ever have any plans to go beyond the United States?

Yeah, it’s, it’s one where we get requests quite often, especially from our, our neighbors in, in Canada for services like ours and whether we are able to provide it, Given the funding mechanism today is from the FCC, we are constrained to the US hearing loss community for who we can provide service to today. But we are always looking to engage with other government agencies and funds in different countries around the world to the extent that they are interested in providing services like ours, or if we can find other ways to fund similar services too. So it’s definitely something that I would love to have further dialogue with you or anyone up there that has the ability to help us with advocacy or awareness around these types of services and the benefits they can offer.

Oh, that’s great. That’s great. So now I’m very familiar with speech-to-text, but now InnoCaption has text-to-speech services. Can you tell us about them?

Yeah, sure. So this came about because in— at the end of 2024, in December, right around Christmas, the FCC actually gave us a new type of certification So within the telecommunications relay services umbrella that, that you had previously mentioned in terms of how we’re certified and funded, our existing service, the main InnoCaption service, was what is classified as IP Caption Telephone Service, or IPCTS. And this is for people that have hearing loss but are able to speak on their own.

And then there is another category of service called IP Relay. And this was the kind of traditional relay operations where you would call into a call center and then you are connected with an operator in the middle, and then you would type to that operator what you want to say to the other person on the line. They would speak on the person’s behalf and then type back what is being said by the other party.

And with that new certification, it actually allowed us to actually help bring a more modern solution that can basically recreate that type of experience, but in a more efficient and fluid way that empowers our users. And so now what our users are instead of having to call into those relay centers is use our app to make direct calls. And if they are an IP Relay user, they can actually just use text-to-speech for the full conversation if they would like.

And so to make that service even better, we’ve now come up with new features, predictive text-to-speech that actually utilizes the advancements in AI technology to actually make that communication faster and better. And so I wanted to just give a bit of that context for why we started developing all of these new features so quickly over the last year. But basically, the main new feature that we are really excited about that we recently did a press release on is called AI Refine.

And combining that new feature with some of our saved shortcuts and other features that folks that have used text-to-speech services may already be familiar with, we feel like really empowers people to be able to communicate much more quickly and efficiently in calls. And so what that AI Refine feature does is if I’m having a conversation with you on a phone and you ask me a question, Instead of me having to type the full response, the full sentence, character by character, I can actually just type a couple of words to guide the direction of the conversation.

And at the tap of a button, the AI will actually take the context of the conversation and what I am guiding towards and actually help me create a natural-sounding sentence that fits the conversation. I can then edit that conversation if I want to, or in many cases, this is actually very good and pretty contextually aware. So you’re able to then just press a send button and actually send that sentence to be spoken by the AI. And so it’s a really, really helpful tool for people to be able to actually use text-to-speech in a way where it doesn’t create a big lag in conversation because the other party is waiting for you to type so much.