Re: Vista Speech - does doc scanning really improve accuracy?

Mark Conrad <none-> wrote:

> I _did_ check-mark the box that allows user documents to be
> read, but it does not seem to have any effect one way or the
> other.

Except for configuration/settings, computer programs do not work
that way. It is like the claim that a chess program can improve
itself by playing. The idea that recognition will get better might
temporarily encourage the user, but any improvement is not
significant. If you cannot use speech recognition out of the box,
you probably need to improve your voice. A person's diction has the
greatest effect on speech recognition.

By the way. The malicious spammer Marty Martin Markoe has a racket
that goes something like this... Whenever a speech recognition
problem occurs, his first and only effort is to sell a microphone
and/or other hardware to you (more recently, the WSR toolkit, but
that is mainly for getting you to his commercial website). The
spammer with a racket is selling to you a pot of gold at the end of
a rainbow. So when the microphone and/or other hardware you buy does
not make speech recognition usable for you, the spammer attempts to
sell more hardware to you. You see, allegedly everyone else has done
very well after buying hardware from the spammer. But, after selling
hardware to you fails again... the sad sad truth, the dirty low-down
becomes obvious... your voice sucks. So you hang your head and walk
away. If the spammer were interested in helping people, he would not
sell hardware to people (under the illusion that it is going to
improve their speech recognition) without first hearing their voice.

Given current technology, talking to a stupid computer requires good
diction.

Mark Conrad <none-> wrote:

> John Doe <> wrote:

>> Given current technology, talking to a stupid computer
>> requires good diction.
>
> I agree. I am being forced to abandon my hound-dog
> A-mel-i-can English in favor of re-learning my
> native English. (as defined by the Oxford dictionary)
>
> I did not mind so much having to pronounce "neither"
> as "ny-thur" instead of "knee-thur"...
>
> ...however I am hesitant to adopt the way of pronouncing
> "schedule" in the same way our next U.S. president
> (Rush Limbough) pronounces it, as "shed-u-al" ;-)
>
> What dictionary do you favor, when it comes to the
> American version of English?

Merriam-Webster, mostly for being the first to include my
advertising's sense of the word "save".

If you mean for pronunciations, I can live with having to pronounce
stuff the way my stupid computer can understand it. Merriam-Webster
includes audible pronunciations, I have listened to those from time to
time. Like the word "acceptable" with an "ik" "ak" or "ek", in all
cases with a "k" sound. Maybe I should spend more time with
pronunciations, to improve accuracy, but I settle for occasionally
flipping off my computer. On rare occasion the program is clearly
messing up. I can tell that by always having the word(s) pronounced by
DNS when the word is being corrected. Feedback is extremely useful
(and sometimes painful) for improving my speech.

FWIW.

I am glad Microsoft has taken up the chore of including speech
recognition. Should have been done many years ago, probably still
underfunded. Speech recognition is no different to some disabled users
than keyboard input is to those of us who can type. There is no
fundamental difference whatsoever, they are both input methods,
period. The same is true for text-to-speech, as another output method.

Someday... somewhere, over the rainbow... we will be able to
communicate, via speech recognition and translation, with people who
speak a different language without having to learn the foreign
language. Maybe that would be a good project for the United Nations,
or Bill Gates if he wants "high technology to get us out of this
depression" as he says it will.

I have used every type of microphone in the world. Does anybody else know
where I can get one that I have not used before?

"John Doe" <> wrote in message
news:MPfpl.11311$...
> Mark Conrad <none-> wrote:
>
>> John Doe <> wrote:
>
>>> Given current technology, talking to a stupid computer
>>> requires good diction.
>>
>> I agree. I am being forced to abandon my hound-dog
>> A-mel-i-can English in favor of re-learning my
>> native English. (as defined by the Oxford dictionary)
>>
>> I did not mind so much having to pronounce "neither"
>> as "ny-thur" instead of "knee-thur"...
>>
>> ...however I am hesitant to adopt the way of pronouncing
>> "schedule" in the same way our next U.S. president
>> (Rush Limbough) pronounces it, as "shed-u-al" ;-)
>>
>> What dictionary do you favor, when it comes to the
>> American version of English?
>
> Merriam-Webster, mostly for being the first to include my
> advertising's sense of the word "save".
>
> If you mean for pronunciations, I can live with having to pronounce
> stuff the way my stupid computer can understand it. Merriam-Webster
> includes audible pronunciations, I have listened to those from time to
> time. Like the word "acceptable" with an "ik" "ak" or "ek", in all
> cases with a "k" sound. Maybe I should spend more time with
> pronunciations, to improve accuracy, but I settle for occasionally
> flipping off my computer. On rare occasion the program is clearly
> messing up. I can tell that by always having the word(s) pronounced by
> DNS when the word is being corrected. Feedback is extremely useful
> (and sometimes painful) for improving my speech.
>
> FWIW.
>
> I am glad Microsoft has taken up the chore of including speech
> recognition. Should have been done many years ago, probably still
> underfunded. Speech recognition is no different to some disabled users
> than keyboard input is to those of us who can type. There is no
> fundamental difference whatsoever, they are both input methods,
> period. The same is true for text-to-speech, as another output method.
>
> Someday... somewhere, over the rainbow... we will be able to
> communicate, via speech recognition and translation, with people who
> speak a different language without having to learn the foreign
> language. Maybe that would be a good project for the United Nations,
> or Bill Gates if he wants "high technology to get us out of this
> depression" as he says it will.

A forgery.


"John Doe" <jdoe usenetlove.invalid> wrote:

> From: "John Doe" <jdoe usenetlove.invalid>
> References: <250220090155436334%none-of your-business.invalid> <Ofdpl.2541$Lr6.1686 flpi143.ffdc.sbc.com> <250220090901543028%none-of your-business.invalid> <MPfpl.11311$hc1.8926 flpi150.ffdc.sbc.com>
> In-Reply-To: <MPfpl.11311$hc1.8926 flpi150.ffdc.sbc.com>
> Subject: Re: Vista Speech - does doc scanning really improve accuracy?
> Date: Mon, 9 Mar 2009 17:17:20 -0000
> Lines: 1
> MIME-Version: 1.0
> Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original
> Content-Transfer-Encoding: 7bit
> X-Priority: 3
> X-MSMail-Priority: Normal
> Importance: Normal
> X-Newsreader: Microsoft Windows Live Mail 14.0.8064.206
> X-MimeOLE: Produced By Microsoft MimeOLE V14.0.8064.206
> Message-ID: <u#Kl8rNoJHA.5980 TK2MSFTNGP06.phx.gbl>
> Newsgroups: microsoft.public.windows.vista.general
> NNTP-Posting-Host: 79-77-131-228.dynamic.dsl.as9105.com 79.77.131.228
> Path: nlpi059.nbdc.sbc.com!flph199.ffdc.sbc.com!prodigy. com!flph200.ffdc.sbc.com!prodigy.net!newshub.sdsu. edu!feeder.erje.net!news4us.nl!TK2MSFTNGP01.phx.gb l!TK2MSFTNGP06.phx.gbl
> Xref: prodigy.net microsoft.public.windows.vista.general:359982
> X-Received-Date: Mon, 09 Mar 2009 13:22:19 EDT (nlpi059.nbdc.sbc.com)
>
> I have used every type of microphone in the world. Does anybody else know
> where I can get one that I have not used before?
>
> "John Doe" <jdoe usenetlove.invalid> wrote in message
> news:MPfpl.11311$hc1.8926 flpi150.ffdc.sbc.com...
>> Mark Conrad <none-of your-business.invalid> wrote:
>>
>>> John Doe <jdoe usenetlove.invalid> wrote:
>>
>>>> Given current technology, talking to a stupid computer
>>>> requires good diction.
>>>
>>> I agree. I am being forced to abandon my hound-dog
>>> A-mel-i-can English in favor of re-learning my
>>> native English. (as defined by the Oxford dictionary)
>>>
>>> I did not mind so much having to pronounce "neither"
>>> as "ny-thur" instead of "knee-thur"...
>>>
>>> ...however I am hesitant to adopt the way of pronouncing
>>> "schedule" in the same way our next U.S. president
>>> (Rush Limbough) pronounces it, as "shed-u-al" ;-)
>>>
>>> What dictionary do you favor, when it comes to the
>>> American version of English?
>>
>> Merriam-Webster, mostly for being the first to include my
>> advertising's sense of the word "save".
>>
>> If you mean for pronunciations, I can live with having to pronounce
>> stuff the way my stupid computer can understand it. Merriam-Webster
>> includes audible pronunciations, I have listened to those from time to
>> time. Like the word "acceptable" with an "ik" "ak" or "ek", in all
>> cases with a "k" sound. Maybe I should spend more time with
>> pronunciations, to improve accuracy, but I settle for occasionally
>> flipping off my computer. On rare occasion the program is clearly
>> messing up. I can tell that by always having the word(s) pronounced by
>> DNS when the word is being corrected. Feedback is extremely useful
>> (and sometimes painful) for improving my speech.
>>
>> FWIW.
>>
>> I am glad Microsoft has taken up the chore of including speech
>> recognition. Should have been done many years ago, probably still
>> underfunded. Speech recognition is no different to some disabled users
>> than keyboard input is to those of us who can type. There is no
>> fundamental difference whatsoever, they are both input methods,
>> period. The same is true for text-to-speech, as another output method.
>>
>> Someday... somewhere, over the rainbow... we will be able to
>> communicate, via speech recognition and translation, with people who
>> speak a different language without having to learn the foreign
>> language. Maybe that would be a good project for the United Nations,
>> or Bill Gates if he wants "high technology to get us out of this
>> depression" as he says it will.
>
>
>