Future Technologies
INTRODUCTION
In this section we would like to briefly discuss selected examples of
more complex systems that could exist today that would expand upon our
basic model. We also describe how evolving technologies could impact
InterCaption technology.
Multiple User Systems
The InterCaption set up that we discussed earlier features two users.
It is possible, however, to incorporate more users at each site. Delegate
microphone systems allow meeting participants to use individual microphones.
A central control unit, usually used by the meeting facilitator, ensures
that only one microphone is active at any given time.
(Here in an example
of one such system.) A database of voice records could ensure that the
voice recognition software would be able to handle speech from a group of
users. A future possibility would be to have encode user records on identify
cards that could travel with them. The computer routines that integrate the
text from both sites could easily be created to distinguish between
speakers. Advanced captioning features (see below) could help the audience
distinguish between speakers.
Reciprocal Captioning - Read Back Feature
It would be valuable to add speech production technology to our system.
This could be managed by a combination of software and hardware incorporated
into the computer. This "read back" feature would be able to reproduce earlier
interactions and allow users to participate by simply entering text rather than
speaking.
Advanced Captioning
Caption encoders and decoders have more advanced features that exist today
yet are not widely used. For example, captions can be broadcast in
diverse fonts and colors. This functionality would be useful when
captioning multiple speakers or multiple languages.
Improved Speech Recognition
Speech recognition technologies are presently trained on specific users.
It is reasonable to expect that in the future systems may still be trained
on specific users but perhaps that a recognition system would be able to
transcribe the speech of multiple users within one system. (e.g. members
of a company or a family.)
Non-English Speech Recognition
It would be very useful to have automated speech recognition of languages
other than English. It is unclear, however, what forces would drive this
to happen, as this is an issue of marketplace rather than hardware.
Captured Videoconferencing: Synched Voice, Video & Text
We have outlined an approach for capturing and integrating captions on two
sides of a dialogue. Likewise, it would be useful to capture and
digitized the corresponding video and audio. Furthermore, if these six
feeds were linked (voice, video and text on two sides), it would be
possible to use the captured captions to search for keywords in order to
locate a given video segment. Technology exists today that would allow
for a fairly cumbersome, hybridized version of this system - likely
involving a server with capturing capabilities and routines to organize
the information. Greater bandwidth and higher-powered machines would make
this a more realistic goal for an average consumer.
Back to Top
"InterCaption: Facilitating Communication Through Captioning"
© K. Acker, T. Lytle, J. Porvin (1998)
Back to VisCom 98
Projects Homepage
|