14 DAY TRIAL // It's been quite a few years since Google introduce its automatic captioning service to YouTube, and in the meantime, it seems that over 1 billion videos benefited from it.
Back in 2009, Google announced YouTube was going to have an automatic captioning service. In the beginning, the results were less than fortunate, but now things have improved greatly as Google's coding advanced. There are still mistakes in there, don't get us wrong, but they're far less annoying than they were in the beginning. The purpose of it all was to make these videos accessible to those who have problems hearing or cannot hear at all.
According to YouTube's announcement, there are now one billion videos with captions, while people watch them 15 million times a day.
"A major goal for the team has been improving the accuracy of automatic captions -- something that is not easy to do for a platform of YouTube’s size and diversity of content. Key to the success of this endeavor was improving our speech recognition, machine learning algorithms, and expanding our training data. All together, those technological efforts have resulted in a 50 percent leap in accuracy for automatic captions in English, which is getting us closer and closer to human transcription error rates," writes Liat Kaver, YouTube product manager.
More work, more languages
The company says they will continue working on improving the accuracy of captions since this remains an important goal. YouTube also wants to extend the work to all of the ten supported languages so more videos can be captioned. As always, they don't plan to do it on their own, but hope they'll have the help of the community of creators and viewers everywhere.
"Ideally, every video would have an automatic caption track generated by our system and then reviewed and edited by the creator. With the improvements we've made to the automated speech recognition, this is now easier than ever," Kaver writes.
Here's photo evidence of how the caption system has evolved. The first picture shows an older version of the feature, while the second shows the current caption system.
