EasyManua.ls Logo

Radio Shack TRS-80 Voxbox - Page 21

Radio Shack TRS-80 Voxbox
36 pages
To Next Page IconTo Next Page
To Next Page IconTo Next Page
To Previous Page IconTo Previous Page
To Previous Page IconTo Previous Page
Loading...
How to
Write Your Own Application
Programs
5.
How
to
Write
Your
Own
Application
Programs
General
Information
The
VOXBOX
allows you to
control the TRS-80
system
or enter
data
into it
with spoken words. Up to 32
different
words
may be
used.
You select what
the words
are
to
do and what
words are to
be used.
What
the words do
(the action they cause or the data
they provide)
is
written into the
program.
The actual
words spoken
are selected
(programmed) by
the user at
the time he runs
the program.
For
convenience
they are
usually the same words,
but they need
not be.
One user may
refer to
the digit
"
1
"
as "ONE" and another
user
as
"UNO"
and to
the word "YES" as
"YES"
or "SI".
As is
true
with human
listeners,
your VOXBOX will
sometimes
make
mistakes
when
listening to you. In general, it should
correctly
recognize
85-95% of
the words you
use, provided
you take reasonable
care
in
speaking
clearly and
distinctly (see Section
3).
In writing
programs, you
should assume that some misunderstanding
will
occur
and provide
for
checks on the data or
verification of
what
was
said at selected
points
in your program, and have the
program
allow
for
correction if
that is important and required
(as
is
done in the
INVENTORY demo).
Technically
speaking,
VOXBOX is a speaker-trained,
isolated-word,
speech
recognition system
with a 32-word vocabulary.
"Speaker-
trained"
means that the
user must enter one or more examples
of each
word or
phrase in
the vocabulary.
It also means that
there are
two
operating
modes for
the device. First, a TRAINING mode in
which
the spoken
examples of each word
are provided,
and then a
RECOG-
NITION
mode
in which
previously trained words are
recognized.
Isolated word
means
that each word or
phrase
must have
a . 1 second
period of
silence
before and after
the word. This is
necessary
to allow
the
computer to
determine
when a word has started
and when it
has
finished.
The
current
vocabulary is stored in a
table in high
memory
(as part
of the
driver
program). Each table
entry consists
of the code for
that
word as you
pronounced it. See figure 3 below.
19

Related product manuals