YouTube is updating its caption technology to give millions more users access to its content. According to the Google Blog, YouTube and Google Video will now use automatic speech recognition (ASR) in conjunction with YouTube's own captioning system to create automatic captions. The development provides more accessibility for deaf and non-English speaking users.
The captions can be translated into 51 different languages and improve search's ability to locate a direct point in the video. Captions could also solve a potential problem with video search -- namely, that you can't search for the contents of a video without a transcription.
The automatic captions (auto-caps) use the same algorithms used by Google Voice, which are still a bit shaky when it comes to word recognition. But it is expected that the technology will be more accurate in due time. In addition to auto-captions, Google added an auto-timing feature that will help users create their own captions much more quickly.
Using a simple text file, users can type all of the words in the video and submit them to Google's ASR. The ASR then determines when the word is spoken and creates the caption.