🚀 go-pugleaf

In article <4524@xenna.Encore.COM> bzs@Encore.COM (Barry Shein) writes:
>>Good speech recognition hardware can't be more than 5 or 10
>>years away, can it?
>As far as I can tell it's only been 5 or 10 years away for the past
>decade or so, I'd imagine that figure is still correct.

It depends on what you mean by good speech recognition.  It can be argued that
speech recognition is here now.  It also can be argued that we have a long way
to go.

There are a lot of systems in the field that are used for inventory tracking 
(speaker dependent).  Dragon Systems is selling a system on PCs for doctors 
that lets them dictate medical reports (speaker dependent isolated words.)
Also, IBM is beta testing a system for speaker independent isolated words for
medical offices and insurance companies.  Finally Kai Fu Lee and the gang
at CMU (and also now SRI and Lincoln Labs) have demonstrated a system
that does speaker independent, continuous word recognition with a 90% correct
word rate.

You can buy a system today that connects to a Mac and lets you execute 
arbitrary commands based on spoken text.  Thus you can say "save" and have
the Command S key sent to the application.

On the other hand....human like speech recognition has been postulated to
take on the order of a Tera Flop by the IBM people.  The current problems
remaining to be solved (other than processor power) are a bunch of front end 
issues (like noise, multiple speakers speaker adaptation) AND incorporating
natural language so that homonyms and missing works can be filled in.

Drop me a note if you want references to any of this work.

							Malcolm