Page 186 -

Speech Recognition 173

with speaker­independent systems. This system is inherently speaker­depen­
dent, meaning that the voice that trained the circuit also uses it. To experiment
with speaker­independent recognition (multiuser), try the following technique.
Set the WD jumper on the main circuit board to the 40­word vocabulary with a
0.96­s word length. Now we will use four word spaces for each command word.
We will arrange the words so that the command words will be recognized by
just decoding the least significant digit (number) on the digital display.
This is accomplished by allocating the word spaces 01, 11, 21, and 31 to the
first target or command word. When the circuit is in recognition mode, we only
decode the least significant digit number, in this case X1 (where X is any num­
ber from 0 to 3) to recognize the target word.
We do this for the remaining word spaces. For instance, the second target
word will use word spaces 02, 12, 22, and 32. We continue in this manner until
all the words are programmed.
If possible, use a different person to speak the word. This will enable the sys­
tem to recognize different voices, inflections, and enunciations of the target
word. The more system resources that are allocated for independent recogni­
tion, the more robust the circuit will become.
There are certain caveats to be aware of. First you are trading off word
vocabulary number for speaker independence. The effective vocabulary drops
from 40 words to 10 words.
The speech interface control circuit shown later may be used in this speaker­
independent experimental capacity.

Voice Security System
This HM2007 wasn’t designed for use in a voice security system. But this
doesn’t prevent you from experimenting with it for that purpose. You may
want to use three or four keywords that must be spoken and recognized in
sequence in order to activate a circuit that opens a lock or allows entry.

Speech Interface Control Circuit
Okay, you have a functioning speech recognition circuit, so now what? You
need a method of allowing those voice commands to activate other electrical
devices or functions. To do this, we need to build a universal speech interface
circuit.
When designing this interface, I weighed options that I thought would make
this interface useful to as many different users as possible. The first parame­
ter I considered was how many outputs the interface should have. I decided
upon 10 outputs. The second consideration was the type of output that the
interface board should provide. Here was a tough choice. I had the option to
make the output an active high signal that the user could use to activate or be
detected. This output could be used on a TTL logic line or CMOS logic line, or
to turn on a transistor switch or power relay in their circuitry.