info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Speech activated scripting without Natlink...
23 views
Skip to first unread message
John Doe
Jan 11, 2014, 8:19:01 PM1/11/14
to
Why are Vocola, Dragonfly, Unimacro, and Advanced Scripting
integrated into speech recognition? So that NaturallySpeaking can
distinguish between what is a command and what is dictation? Speech
activated scripting is light years better than keystroke activated
scripting, but what's the point in integrating the dictation with the
commands? To activate by keystrokes, you press the keystroke
combination and then the script is executed. To activate by voice,
speech recognition must translate it to text anyways. So, just like
using a keystroke combination, you can simply use the output of
speech recognition, as long as you have some way of determining what
is dictation and what is a command. So why not take it from the SR
output?
To be continued...
I'm still rebuilding my computer desk area. Need to finish the mouse
pad/platform. Since I was spending 12 hours a day sitting down, I
bought one of those fitball seating discs. Seems to work reasonably
well when underinflated.
John Doe
Jan 12, 2014, 3:34:52 PM1/12/14
to
As Mark Lillibridge pointed out (the obvious), the major drawback of
sticking a simple macro recorder on the dictation output of speech
recognition is losing the benefit of a much constrained command
vocabulary. In other words... Regular dictation requires a relatively
huge vocabulary. Command is a very small vocabulary. Therefore
command can be much more accurate.
Still, there would be benefits. No need for Natlink. Being able to
use a simple macro recorder. Being able to force dictation
recognition (in US English) by simply holding the control key.
Windows 8 would be usable with DNS.
And maybe I can adjust. By making my commands more English-like.
One thing I really don't understand is why repetition has to be
predefined. If you don't use numbers as command names, and you
shouldn't anyway, you can simply speak a number after a command to
designate how many times (if any) that command should be repeated.
Or, there should be a keyword that designates repetition, like
"times". If you want to repeat a command, you simply say "times" and
a number.
sticking a simple macro recorder on the dictation output of speech
recognition is losing the benefit of a much constrained command
vocabulary. In other words... Regular dictation requires a relatively
huge vocabulary. Command is a very small vocabulary. Therefore
command can be much more accurate.
Still, there would be benefits. No need for Natlink. Being able to
use a simple macro recorder. Being able to force dictation
recognition (in US English) by simply holding the control key.
Windows 8 would be usable with DNS.
And maybe I can adjust. By making my commands more English-like.
One thing I really don't understand is why repetition has to be
predefined. If you don't use numbers as command names, and you
shouldn't anyway, you can simply speak a number after a command to
designate how many times (if any) that command should be repeated.
Or, there should be a keyword that designates repetition, like
"times". If you want to repeat a command, you simply say "times" and
a number.
Mr. Anderson
Jan 12, 2014, 6:31:38 PM1/12/14
to
Macro recorder implies scripting language, which means either Advanced
Scripting or some NatLink variant.
Although I can see some limited value in capturing dictation into your
own separate program, I wouldn't personally go very far down that
path. Reinventing the wheel takes years of one's life.
Mr. Anderson
Scripting or some NatLink variant.
Although I can see some limited value in capturing dictation into your
own separate program, I wouldn't personally go very far down that
path. Reinventing the wheel takes years of one's life.
Mr. Anderson
John Doe
Jan 12, 2014, 9:10:55 PM1/12/14
to
Mr. Anderson <matrixlove yahoo.com> wrote:
> Macro recorder implies scripting language, which means either
> Advanced Scripting or some NatLink variant.
Any macro/scripting language, there are many.
> Macro recorder implies scripting language, which means either
> Advanced Scripting or some NatLink variant.
> Although I can see some limited value in capturing dictation
> into your own separate program, I wouldn't personally go very
> far down that path.
drop it. Apparently the output needs to be taken from DNS at an
earlier stage. I think the problem (besides the otherwise benefit
of a limited command vocabulary) is DNS changing words to fit
into phrases according to English language rules. When I try to
mingle numbers (for repetition) into the phrase, the phrase
radically changes. Looks like it just won't work.
In other words, never mind.
Mr. Anderson
Jan 13, 2014, 7:02:58 PM1/13/14
to
0 new messages