|Instructions - What is LiveDescribe Technology?|
"Dialog Detection" algorithms work hard to detect appropriate description places so you can concentrate on the highly demanding task creating the best descriptions possible for your video content. Manual description marking allows for the user to define spaces that the detection algorithms don’t pick up. Using LiveDescribe it is possible to reduce description creation time from several hours of work for every hour of described, to nearly a one-to-one ratio.
The LiveDescribe System
LiveDescribe is designed to allow near real-time DVI to be added to on-line content by DVI describers (the target user group for LiveDescribe). While there are a number of existing software applications such as Softel ADePT (Softtel-USA, 2001) and Magpie (NCAM, 2003) that support post production DVI, they are not designed to support live description for events such as emergency broadcasts, parades, and other events which are often only viewed once, and relevant during the time in which the event occurs. The real-time description process requires a different approach from post-production description. A real-time describer must record descriptive utterances based on a much shorter duration and constrained view of the material where there is no opportunity to fully preview the material before a description is scripted.
The lead developer on the project is Carmen Branje. Other developers include Jimmy Liu and Kristie Moniz.
View a windows media movie that demonstrates how to use LiveDescribe accompanied by an audio presentation of the different elements of LiveDescribe.
|Last Updated on Friday, 13 April 2007 14:40|