Boyd:

IMO the speech recognition technology being developed for small form
factor telephony, and, their interrelated radio base stations, and,
interconnected voice/data wireline networks will have a longer term
impact than pen-based/slate PCs on the average consumer.

Because of two worker compensation disabilities I have been
(intermittently) using, selling, and installing  speech recognition
software (command and control) applications since the DOS-versions of
Dragon Dictate through some of the latest products on the market -
MS-Windows, Linux, and FreeBSD.

Based on using speech-to-text software control of DOS & MS-Windows
since 1989; I expect to see, within ten years, functional
speech-to-text capabilities for the layman will become the the second
most common interface after the traditional keyboarding and pointing
device approach. I base this on the fact when I first started using
speech-to-text in 1989 there was a vocabulary limit of less than 4,000
recognition words and this was with an experimental MIT project
running on a multi-processor DEC VAX.

FYI: The average American adult uses less than 6,000 words in their
employment. The average American college under-graduate student uses
less than 15,000 words.

Today, off-the-shelf speech recognition have vocabularies of 30,000 to
50,000 words with about 98% accuracy after three weeks of practice
while getting 2 hours of training per day for about 10 days. When you
read about any reviewer of speech-to-text who tries self teaching
prior to using the product for the review. You can figure they will
get at best 25% of the capabilities of a professionally trained user.
I base this on having been involved in both approaches since I became
involved with speech-to-text.

When you have time go searching on the web for wearable computers and
how technicians are using them.

With today's off-the-shelf tools for MS-Windows, Linux, and FreeBSD we
can build proprietary application for speech-to-text input - with
voice output to the user of sequential questions - which is usable by
a repair technician for filling out work orders. The technical
vocabulary will be limited to about 1,200 industry specific words.
However the accuracy rate, after a month with professional training
and practice, will be over 99.99%.

Which means a repair technician could do a complete inspection (with
voice prompts for sequential check list style items) of an aircraft,
tank, missile launch system, building environment (heating/air
conditioning/door control/TV security) system, as well as anything a
public utility technician would do for telephone - gas - electric -
nuclear power generation systems. That 1,200 word vocabulary will
handle nearly every one of their industry specific proprietary words
along with the 6,000 to 8,000 daily used vocabulary of the average
American adult.

Above are my thoughts on the future direction for hardware and
computer control.

IMO the computer's operating system will _NOT_ be nearly as much of an
issue in the future as it has been in the past. I also base this on
looking at the evolution of technical history since 1830's steam
engine powered equipment through today's personal transportation.

Today very few of us can, or would want to attempt to, repair the
anti-locking brake systems available on today's vehicles. Most of us
simply turn the key, turn the steering wheel, and push on the gas
pedal or brake pedal in today's vehicles. The average person doesn't
care about the operating system and protocols used by the computers in
today's cars - Micro$oft would dearly love to to convince all of us it
is important to have all the private vehicles use a form of MS-Windows
and the industries answer is clearly = NOT!!

Below is the opinion of Gordon Moore, co-founder of Intel, regarding
speech recognition.

John Oram

--------

Moore: Speech recognition next hurdle, but a long time away
By David Berlind
July 10, 2002

"... On the next big transforming technology: speech recognition
Good speech recognition will be a transforming capability when it
finally comes into being where you'll be able to talk into your
computer and it will be able to understand what you're saying in
context. It will know if you mean "to" or "two" or "too." Once the
computer understands speech at that level, you'll be able to have an
intelligent conversation with your computer. That can change a lot of
things. First of all, it will make computing available to the people
who are scared off by keyboards and such. Secondly, it will change the
way we use them completely. I don't know if that's 10 years away or 50
years away. I think it's something that certainly will be coming down
the road and it will be really transforming when it does. I suspect
that it's closer to the 50 years than the 10 years to get to the level
that I'm talking about. ..."


http://techupdate.zdnet.com/techupdate/stories/main/0,14179,2873890,00.html



Kenneth Alan Boyd Ramsay wrote:
>
> John:
>
> I was trying to answer the Subject: question above.  The article was
> implying that tablet-based systems would evolve - perhaps as a compromise
> between palm-tops and lap-tops.  I gather a combination of LCD screen
> and light pen or membrane screen/tablet - all in a package about the size
> of a laptop screen.
>
> I, personally, have doubts about voice or handwritten entries compared to
> keyboards.  Although they will eventually work the bugs out, I wonder
> at the cost in "bloatware".  It should be OK as a notepad, given
> sufficient resolution, however.
>
> Thanks for the background.  Knowing where we are coming from gives us a
> better idea where we are headed.
>
> Boyd Ramsay
>
> [EMAIL PROTECTED]
>
> On Sat, 3 Aug 2002, John Oram wrote:
>
> > Boyd:
> >
> > We worked from 1990 to 2001 with Fujitsu Personal Systems Inc. and
> > their proprietary tablet hardware as well as their proprietary
> > versions of DOS & Windows which allowed input with a battery powered
> > pen onto the grid overlay. We did the beta testing for the "AHA!" pen
> > based word processor which Micro$oft bought out in 1993. We also still
> > have contacts with current and retired folks at Fujitsu. The pen
> > software extensions for DOS & Windows were written by Fujit$u which
> > has an OEM relationship with Micro$oft. This technology is over ten
> > years old and works well for specific applications, especially now
> > that the CPU isn't 396 based and the RAM on the motherboards is not
> > limited to 4 MB <BG>
> >
> > So what are your (specific) questions?
> >
> > John Oram
> >
> > Kenneth Alan Boyd Ramsay wrote:
> > >
> > > I just read some hype extolling the virtues? of Windows XP and tablets
> > > or touch screens.  Comments?

To unsubscribe from SURVPC send a message to [EMAIL PROTECTED] with 
unsubscribe SURVPC in the body of the message.
Also, trim this footer from any quoted replies.
More info can be found at;
http://www.softcon.com/archives/SURVPC.html

Reply via email to