On 13 Oct, P.Satchell -Phillip Satchell wrote:
> This is a bit of an aside but related....
> Does anyone know how they do the real-time Teletext subtitling on the news?
> I've always believed this was done by a very high power speech system (the
> Phillip's one perhaps?) because of the type of mistakes you see. Is there
> some human intervention going on at the same time to fix those errors?
It's not done using speech recognition for the same reasons as you can't do it
for lectures; the technology does not exist.
They use a phonetic keyboard of some sort. I don't know how special it is (ie,
whether it's a chording keyboard), but the essence is that an expert phonetic
typist transcribes the sounds they hear, and a computer finds probably word
sequences that sound like that. I suppose it's similar to the second part of a
speech recogniser but using a human for the first part. I also suppose that the
skill is similar to that required for short-hand.
TTFNr
--
-----------------------------------------------------------------------------
Roger Spooner DPhil Student, Computer Science? York!
Email: [log in to unmask] Web: http://www.york.ac.uk/~riws100/
---------------------------I'll see what I can do----------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|