|
Caption Capturing
INTRODUCTION
Consider now how useful it would be to have a transcript of a conversation.
In the previous diagram, we illustrated that the voice recognition
software would transcribe speech into ASCII text. Below we outline
a method for capturing the captions from both sites and integrating them into
a transcript of the interaction.
Diagram 3: Caption Capturing System
|
Click on the yellow component for further information.
|
|
ASSUMPTIONS
This diagram is based upon the following assumptions:
- We will be able to design routines/macro commands capable of compiling the
separate halves of the conversation into an integrated document.
SYSTEM TOUR
We have added a component called a "TextGrabber" to our system. This
component has been placed on the incoming RF path. The TextGrabber decodes the captioning
and outputs ASCII text to a personal computer. The computer, which
is currently performing voice recognition at the local end and therefore already has one
half of the conversation, can now keep a record of both parts. As text from both the
voice recognition software and the TextGrabber are received, the computer will generate
a text file which will contain verbatim text of the entire interaction.
Back to Diagram
COMPONENT DETAILS
TextGrabber
The TextGrabber is a readily available product that accesses caption information
contained within a standard analog television signal. The TextGrabber grabs
the encoded information from the video signal and converts it to ASCII text.
There is both a serial and RS-232 output that can be connected to a standard PC.
Readers interested in additional information can visit the
TextGrabber
site or view this sample
transcript that was produced from a live CNN broadcast using TextGrabber.
Back to Top
"InterCaption: Facilitating Communication Through Captioning"
© K. Acker, T. Lytle, J. Porvin (1998)
Back to VisCom 98
Projects Homepage
|