Adobe Releases Speech-Recognition Indexing

 

 Via several non-official Adobe-centered blogs and news articles, Adobe has released a Soundbooth Beta that showcases a host of new features, including new speech recognition technology that lets users create transcripts of dialogue tracks quickly and search them for words and phrases across their timeline.

 

 Adobe hasn't yet published any official press releases, so this feature may be a tad undocumented and in early beta form to date.

 They also have not disclosed which engine they are using (if it's not propreitary) to parse spoken audio tracks for speech recognition indexing.

 We've blogged about the two most popular such engines, Everyzing's "Ramp" and it's predecessor and core technology, MIT's "Lecture Browser" from CSAIL (Computer Science and Artificial Intelligence Laboratory) and it's very likely this indexing uses these platforms or variations thereof (IOHO), although new "automated indexing of video's spoken audio" technology is emerging from several new entities, including IBM.