Even before the COVID-19 pandemic, educational content providers were able to shape minds and transfer skills at the touch of a button. In the past few years, these very same eLearning tools have seen incredible growth around the world. Tools we needed to get through stay-at-home orders became part of the everyday. These same tools evolved and became better than ever by incorporating the latest technologies, including speech-to-text.
Here, we’ll look at how Speechmatics’ best-in-class speech recognition can help your Video Platform solutions aid and assist those teaching the next generation.
It’s no secret that, historically, speech-to-text has performed poorly when it comes to children’s voices. The obvious difference in pitch, the over-enunciation, the fewer common cadences to adults: these factors all contribute to a failing. But the biggest failure comes from under representation. The models previously used for speech recognition were traditionally trained on adult voices due to limited datasets. The solution? Train on a wider range of speakers.
And that’s exactly what Speechmatics has done. By introducing self-supervised learning into our training – expanding to broader datasets of voices, accents, and environments– we’ve seen our accuracy improve hugely. Those under eighteen are finally starting to be understood almost as well as adults.
Having reduced the gap in accuracy between both children and adults, the next obstacle to overcome was to make sure children from all over the world could be heard and understood. We believe everyone should have the right to access the modern technological world regardless of ethnicity, gender, literacy, disability or age. To that end, we’ve made sure our language accuracy is market leading.
We’ve recently added 14 new languages to our language support, taking our total to 48. From Estonian to Ukrainian, from Welsh to Cantonese, we’re incredibly proud of the work we’ve done to make sure many of the world’s children get the same level of support as each other in as many languages as possible.
We believe accuracy is everything. That’s never been more important than when educating children. We’ve known that subtitles can play a huge role in helping develop literacy skills for some time now. The more children get to see words in action, the easier it is for them to understand and repeat them. But the opposite can be true for incorrect captions. How badly does it reflect on your solution if the speech-to-text you’re providing to Video Platforms is inaccurate?
Speechmatics have proven time and again that we’re best-in-class for providing accurate speech-to-text. By choosing us you’ll give your business a great advantage over your competition. Sticking with the status quo and your current vendor risks losing out in the race to provide the best solution you can.
When you’re looking for the right speech-to-text, remember we already power some of the top Video Distribution Platforms in the industry, including those who specialize in eLearning and education. We recently received a glowing review from Udemy, a video-based learning platform, who said, “Having conducted an evaluation of several speech-to-text vendors, we identified one as providing the best quality and soon rolled out automatically generated subtitles for all published English courses on our platform...Two years later, we extended to all Spanish and Portuguese courses. This year, we're adding French, German, Italian and Japanese."
The chosen vendor? Speechmatics, of course.
When it comes to integration, we make setting up as easy as possible with our 6 simple steps. Get in touch today, to find out more about how we can make your Video Distribution Platform the best it can be.
Reggie Mitchell and Rupert Lambert - Video Platform Experts
Book a Meeting today with a Video Platform specialist and we’ll give you the tools you need to differentiate your Video Platform solutions in market and help you deliver on constantly evolving customer expectations.