Crowd-powered captioning Like with crowdfunding, the future of live captioning may be powered by a group.
With professional captionists costing as much as $100 per hour, a lower-cost solution is in the works to allow groups of average typists to provide real-time captioning for the deaf and hard of hearing. Legion:Scribe is making that happen.
The system—developed by a team of computer science professors from the University of Rochester, Rochester Institute of Technology and Carnegie Mellon University—allows
multiple users to type pieces of what they hear onto a screen for captioning.
Link to main story: Higher ed takes actions on captions
It then “stitches the partial captions back together automatically,” according to the team.
Three to five people would generally work together to capture all the information being shared, at an accuracy level of nearly 85 percent of a professional captionist’s ability. Each captionist connects via a webpage, where an audio stream offers indicators of when to begin typing. End users simply visit a page to see a live display of captions.
Marcia Layton Turner is a Rochester, New York-based writer.