YouTube Has Speech-to-Text Functionality…and it Works

YouTube Has Speech-to-Text Functionality…and it Works

Editor: The way I read this article is that YouTube has automated speech recognition software that works! If so that’s very cool! The motivation for this technology appears to be the ability to search videos. But whatever the reason, if it leads to captioned videos, that’s great news!

Thanks to NVRC for this story. Please see their notice at the end of this article.


From, Thursday, July 31, 2008

YouTube has introduced rich metadata into select videos, which will significantly impact how online video is discovered and consumed.

Announced in June, we have found the technology is very useful. Last week, we got an update on this from Steve Grove, head of news and politics at YouTube.

YouTube applies a technology in which automated spiders “listen” and transcribe the words in the audio track of a video. This transcription, in the form of metadata, is embedded into the video.

This new speech-to-text technology has been applied to campaign videos from John McCain and Barack Obama as well as non-campaign produced clips. The search box is found on YouTube You Choose page. (It’s on the lower left area of the page.) Visitors can enter words such as “Iraq, “Global Warming,” and “Free Trade.” Also, putting your cursor over the

While this is the first voice-to-text speech application at YouTube and will likely expand, this is an emerging area in video search. Companies including Blinkx, EveryZing and Delve are providing voice-to-text transcriptions in the form of metadata.

Later this year, Adobe, will provide speech-to-text technology into the Flash video production environment, Beet.TV has learned. This is a critically important development in bringing effective search to online video.

I interviewed Steve at Stanford last week where he was a panelist on the Beet.TV online video roundtable.

Here’s story about YouTube and the new voice-to-text features by Anthony Ha at Venture Beat.

To see an illustration and check out links with more information:

YouTube Has Speech-to-Text Functionality…..and it Works


Distributed 2008 by Northern Virginia Resource Center for Deaf and Hard of Hearing Persons (NVRC), 3951 Pender Drive, Suite 130, Fairfax, VA 22030; 703-352-9055 V, 703-352-9056 TTY, 703-352-9058 Fax. You do not need permission to share this information, but please be sure to credit NVRC