Google has gone upgrading YouTube with new auto-captioning functionality. While support for auto-captioning was introduced quite some time ago, at that point of time, it was only available for a select group of partners.
Using the same speech recognition technology which powers Google Voice search, auto-captioning in YouTube recognizes what is being said in the audio accompanying a video track, and converts into text which is timed and displayed along with the video. While users have been able to manually create captions for their video from quite some time, now the technology for automatically creating captions form their videos is in their hands.
The captions facility is right now only available for video in which people are talking in English, however YouTube does support automatic translation of the generated (or otherwise) caption tracks into multiple languages, making video accessible to all. Over the coming months Google plans to launch support for more languages, which would not only be a boon for those with hearing difficulties, but also for people watching a video in a language they don't understand. All one would need to do is turn on the captions, and then translate the caption into the language of their choice.
However Google allows one to download the generated captions, edit them and upload them back with corrections. One of the amazing technologies which YouTube provides is automatic timing for uploaded captions. You need only to upload the text content of what is being spoken in the video, and YouTube automatically recognizes the timing of the text being spoken by matching the sounds with the words. This way the video uploader is saved the greater amount of effort which goes in adding captions to a video.