Page 9 - index

processor like Word. Certain words were used
for very specific manipulation of the cursor and
the text. Naturally Speaking came on the sce-
ne and became useful sometime around 1999
to 2003, depending on how much you needed
to transcribe documents into the computer.
Early versions had to be “trained” by the user
to recognize their individual voice, and the
speed and accuracy were sometimes accepta-
ble, and sometimes not so much. Things have
really improved since then; now the manufac-
turer, Nuance, claims in its advertising that
“Dragon is 3x faster than typing and it’s 99%
accurate”. So, Voice Recognition software has
really come a long way.
(For those of you, who are not familiar with
Naturally Speaking, it has three primary areas
of functionality: dictation, text-to-speech, and
command input. The user is able to dictate
and have their speech transcribed as written
text, or they can have a document synthesized
as an audio stream, or they can issue com-
mands that are recognized by the program.)
Naturally Speaking is an example of a local
computer application or App. All the computing
needed for it to operate is on the computer
that runs it. Naturally Speaking doesn’t take
advantage of Client-Server technology. If you
attended one of our classes, you will recall
that when an application is implemented with
Client-Server technology, the heavy lifting
(computer processing) is not done locally, but
rather at a Server that is very powerful and
very fast, but remote from the Client. The re-
mote Server is connected to the Client by the
internet, which allows rapid movement of data
between the Client and the Server. So the Cli-
ent App runs on the local computer and is con-
nected to the Server Software, running in the
cloud, via the internet. This combination pro- tion & Control Apps.
vides the total Voice Recognition & Control
System. The client collects input from the user
and sends it to the Server where all the really
complex computing is accomplished. The
Server analyses the input and develops the
responses and sends them to the Client where
the results are presented to the user in audio
and/or display formats.
Naturally Speaking is certainly a useful prod-
uct, but the voice recognition and control that
has really gotten the attention of the public
lately, are the intelligent personal assistants
that are provided by some of the leading com-
puter companies, Apple, Microsoft, Google,
9
and Amazon. Apple was first on the scene
with “Siri”, followed by Google’s “Ok Google”,
then, with Windows 10, came Microsoft’s “Hey
Cortana”, and finally Amazon’s “Alexa”.
All of these are Client-Server implementations.
The Servers are somewhere in the cloud and
the Client resides on your smartphone, in the
case of Siri and “OK Google”, or on your lap-
top (or desktop, or tablet) in the case of “Hey
Cortana”, or on a special device that is placed
centrally located in your home, in the case of
“Alexa”.
All of these assistants use a Natural Language
User Interface to answer questions. You’ll
need a microphone on your device to take ad-
vantage of this capability. The Client app, on
the device, uses the microphone to listen for a
“Wake Phrase”. After this phrase is recog-
nized, the following intercepted speech is then
sent to the Server where it is analyzed via
speech recognition software, and converted to
commands. The Server then uses these com-
mands to gather answers to the original spo-
ken inquiry. All of these assistants can make
recommendations and perform various actions
via their Server capabilities. (For example, a
verbal request for the “weather” might yield
various audible statements about the weather
in your location. Or, a request for “traffic” might
yield audible indications of the traffic in your
location, or possibly maps indicating traffic
problems. Or, a request for the best restaurant
might yield a list of restaurants near your loca-
tion. Or, if you have things set up, the state-
ment “Add eggs to my shopping list” will yield
an updated shopping list including eggs.)
Here are some descriptions (and advertise-
ments) found for each of these Voice Recogni-
· Siri (Speech Interpretation and Recog-
nition Interface) is a computer program
that works as an “intelligent personal
assistant” and “knowledge navigator”,
according to Wikipedia. “The software
adapts to the user’s individual language
usage and individual searches with
continuing use, and returns results that
are individualized”, also from Wikipedia.
“Hey Siri” is the wake phrase, which
can be turned on or off.