Monday, August 11, 2008

Speech-To-Text Revolution on the Way?

Here's an article by Jan Ozer from on a new feature to be included with Adobe Premiere Pro CS4.

Predicting that Premiere Pro CS4 will be out in early 2009, Ozer writes: "the most significant addition to Premiere Pro is a new speech-to-text function that will revolutionize how video is edited and watched within 6 months of CS4’s ship date."

What is Speech-To-Text? It's an application that can process recorded speech and produce text. With the exception of a few applications which have to be "taught" one's vocal patterns in order to accomplish this, one typically had to manually type or hire a service to produce transcriptions of interviews or speeches. And transcription services aren't cheap.

Ozer predicts: "speech-to-text is such a sufficiently 'disruptive' feature that it will change editing and streaming preferences." He reasons that with the ability to include the transcription as metadata, "A formerly featureless lump of video is transformed into a scanable, searchable object".

Adobe already includes this feature in the beta release of Soundbooth CS4. I haven't had time to work with it, but for the type of work I do, a functional speech-to-text application would save a great deal of time and money.


Rich said...

You can check Adobe's tech in the Soundbooth beta on Adobe Labs, plus Youtube, Blinkx, and others are going to help make this big:


Rich said...

This is part of a general trend too, with Google, Blinkx, and others interested in expanding the power of search and metadata. Here's some notes: