Google is adding new technology to YouTube that can automatically add text captions to videos using speech recognition.
The website has allowed users to manually add captions to videos for about a year now, but the process requires the author to pen a full transcription of their content and time it properly to the on-screen action. The effort and expense required for this has generally limited the use of captions on YouTube to major broadcasters uploading licensed shows.
Now by utilizing the same voice recognitions used in Google Voice, the website will give viewers an option of inserting machine-generated automatic captions.
The website said it hopes auto captions will not only aid the deaf and hearing impaired, but also allow people across the world to access video content in any of 51 languages. It also noted that a wide rollout of captions may improve YouTube search functionality and even let users jump to the exact part of a video they are looking for.
Warning the results may be spotty at first, YouTube is initially applying the caption technology only to a few channels focusing on education. "We want to make sure we get feedback both from viewers and video owners before we roll them out more broadly," YouTube software engineer Ken Harrenstien said in a blog post.
The company plans to expand the auto-caption feature to other channels and languages later on. Check out a demo below:
In addition to automatic captions, YouTube is also rolling out an "automatic caption timing" across the website for users who include manually created captions. The technology uses speech recognition to figure out when words are being spoken to display the proper text at the right time. ®