Last year we released our new live captions and subtitling product. We’d worked closely with AWS’ Transcribe, Translate and Media teams to create a real-time subtitling solution that would harness machine translation and deliver real-time translated subtitles: a completely new way of live subtitling that doesn’t require humans to work.
Live subtitling is not always immediately visible as a feature but its potential significance is huge. User experience is massively variable when it comes to subtitling. Switch on the TV for any live sports event and you’ll get the option for live subtitles in the language being spoken. We’ve got used to that capability which has been around for years, although usually at the familiar low quality. This is an expensive human based capability has been confined to the major broadcasters. Ofcom has reported multiple attempts to improve the quality of these subtitles over the years.
Now with CaptionHub Live, you can integrate a push or pull stream for your live event, and produce 90% accurate subtitles, in 20 languages, with less than 30 seconds latency.
Subtitling live is no longer confined to broadcasters or organisations who are willing to pay for large teams of stenographers. CaptionHub has brought together industry leading automatic speech recognition and machine translation engines that have accelerated localisation in the last two years, in a way that lets any organisation doing live events and video streaming, the capability to stream live subtitles.
CaptionHub Live has become a core focus for CaptionHub and our engineering teams. The product is being developed continually. You can see the current product in action here. If you’d like to hear more, please contact our sales team.