In the follow-up to Google’s Universal search and Ask.com trolling video content with Blinkx, EveryZing, a podcast search start-up, is using its highly accurate speech-recognition technology to bring content-based searching to videos.
The technology consists of first converting the audio in the media files to text, refining the textual content, and then indexing the text to enable a direct mapping to their occurrence in the video. With an accuracy of 80%, the technology promises great revenue potential with keyword-based searches and targeted ads being in vogue today.
“The big challenge [in online video and audio] … is the opaqueness of media content,” says Tom Wilde, CEO, EveryZing. It’s extremely difficult to know what range of content is inside a video or audio clip. “The problem we want to solve,” he says, “is the discoverability of multimedia within Web search.” EveryZing does this by extracting the content of multimedia files and outputting text so that it can take advantage of the preexisting text-search tools developed by the likes of Google and Yahoo.
Article from Technology Review.
Traditional video search has remained confined to the textual search of the tags, titles, and other data accompanying the video. Recently, social tagging of content has also emerged as a successful cataloging method. EveryZing is not the first firm to use speech recognition to solve the search problem for videos. Blinx, the 2004 start-up that specializes in video search, uses both speech and visual recognition technologies to catalog videos. Also, Blinkx is now powering the video search for Ask.com.
More EveryZinger news is available at:
PodZinger drops the ‘Pod’ (RedHerring)
PodZinger relaunches as EveryZing (BusinessWire)
Not starting out as a firm with the global mission of “cataloging all of the world’s videos,” the company is planning to market to media companies, ramping up its search processes for accuracy and efficiency. So will this be a solo effort, or is EveryZing waiting for someone to come “trolling” along and gobble it up? Join the discussion.