We all know how much work it is to transcribe audio or video to text. Adobe knows as well and is doing something about it. Adobe is set to release a new version of Flash that will feature speech-to-text technology for automatic transcription and it will be incorporated into their video creation applications like Premiere as well.
Earlier this month, we learned that Adobe was working with major video search providers to better understand and be able to index the flash file format.
Adobe VP Jim Guerard has given some insight into the new initiative by Adobe. Flash video automatic transcripts will act as rich metadata and could expand accessibility and video search exponentially.
The inclusion of voice-to-text is not new in the industry but its widespread adoption on Internet could be a huge boost to everything from sales videos to advertisements to commercial on-demand television. Up until now Flash has contained very little text which makes it hard to search and index, but that's all set to change.
Flash Video Interactivity:
While the speech-to-text feature is being implemented at the basic editing level, other new features in Flash will give the videos more interactive layers. This means new ways for viewers to interact with the media built right into the video itself. It could be a new way to engage the viewers instead of the viewing experience being completely passive.
Sure this was all manageable in previous versions of Flash but it required an ActionScripting Guru to pull it off. The new editing tools will give you drag-and-drop access to these new features opening the interactive video market door wide and allowing in the less technically-savvy video creators to make their videos interactive experiences.
Video Indexing Solution:
Another of the new features involves the indexing of video. As we all know this is a difficult area. At present most video is only indexed by what metadata is attached to it and that means more work on the creator's part. Adobe is looking to bring a more standardized approach to video indexing as well as ways to generate more metadata like the speech-to-text transcription so that those videos are easier to find by end users.
To say the future is bright for opening up video to the masses might be an understatement. If it is possible to generate a good amount of data about the video file during the creation and editing process and then carry that data all the way to publication in a text format it means less work on the creator's side because there is only the necessity to generate the data once.
At present while many editing packages allow you to attach notes and the like, much of that data is lost when you move to the publishing side of things and place the videos out on the web. While Adobe is not the only company looking at ways to help monetize and index video, they are a big name and can drive innovation.
Other companies, including Google, Blinkx, Everyzing, Ooyala, and others have for some time been looking at other angles using speech analysis to look inside video as well as image processing to see what is in the video and then index it more accurately.
The Bottom Line:
In the next year or two video should be as easy to find and index on the web as standard HTML pages are today. That means that when you want to get your message out to the world in video format, you'll be able to do it far quicker, cheaper and easier than you can now. This should increase productivity and the bottom line and everyone likes a nice bottom line.