Coveo Awarded Patent for Automatic Speech Recognition Training Capabilities in Coveo Audio-Video Search Technology
Innovative speech recognition technology converts audio into text,
allowing users to search and find more accurate data from audio and video
files
Coveo Solutions Inc., a global provider of secure enterprise search solutions,
announced today that it won a U.S. patent for technology it developed to
automatically train the language model and modify the dictionary of a speech
recognition engine used with its award-winning Coveo Audio-Video Search (CAVS)
technology.
Coveo’s speech recognition training capabilities ’self-learns’ by
leveraging relevant contextual content related to the audio and video files,
allowing it to better model the enterprise language. Working in unison with
CAVS, the solution converts speech found within audio and video files into
text, so that the text can be indexed to more accurately match user queries
and offer a more productive search experience. With the integration of these
new and innovative speech recognition training capabilities, Coveo offers a
unique search tool that helps businesses and their customers more easily find
content, texts and video simply through textual queries.
“Speech recognition is becoming the standard for audio-visual search, as
recently noted by Google’s executives,” said Laurent Simoneau, Coveo’s chief
executive officer. “But while others in the market are setting the stage, we
are thrilled that Coveo is the first to market with a patented solution that
not only retrieves relevant audio and video, but does so by integrating smart
speech recognition technology capabilities.”
“Text search is a problem that is addressed by many products in the
market, but as more enterprise as well as Web information is stored in audio
or video files, we are losing access to the great swathes of information,”
said Sue Feldman, IDC’s VP for Search and Digital Marketplace Technologies.
“Speech recognition technologies like Coveo’s give us visibility into this
growing collection of non-text information.”
This entry was posted
on Wednesday, September 26th, 2007 at 3:32 pm and is filed under Enterprise Search, Website Search, Audio Search, Video Search.
You can follow any responses to this entry through the RSS 2.0 feed.
Responses are currently closed, but you can trackback from your own site.