Page 166 - Robots Androids and Animatrons : 12 Incredible Projects You Can Build

The operating distance one can speak from the microphone on the
SRC without shouting is about 1 foot (ft). Because of this, when
using the SRC on a mobile robot platform, we incorporate two
small walkie-talkies. The output of one walkie-talkie is connected
to the speech input of the SRC. The other walkie-talkie is used to
speak to the robot via the SRC. This setup eliminates distance
problems and extraneous noise.
Speech recognition is not speech understanding. Just because a
computer can respond to a vocal command does not mean it under-
stands the command spoken. Future voice-recognition systems will
have the ability to distinguish nuances and meaning of words, to
“Do what I mean, not what I say!” However, those systems are still
years away.

Speaker-dependent and speaker-independent speech recognition

Speech recognition is classified into two categories, speaker de-
pendent and speaker independent. Speaker-dependent systems
are trained by the individual who will be using the system. These
systems are capable of achieving a high command count and bet-
ter than 95 percent accuracy for word recognition. The drawback
to this approach is that the system responds accurately only to the 145
individual who trained the system. This is the most common ap-
proach employed in software for personal computers.
A speaker-independent system is trained to respond to a word
regardless of who speaks. Therefore the system must be able to
respond to a large variety of speech patterns, inflections, and
enunciations of the target word. The number of command words is
usually less than for the speaker-dependent system; however, high
accuracy can still be maintained within processing limits. Indus-
trial requirements more often need speaker-independent voice
systems.

Our SRC will be speaker dependent. We can build in a little
speaker independency by allocating more than one word space to
a target word and then programming different word enunciations
in the allocated spaces. Each of these word spaces would trigger
the same command.

Recognition style
Speech-recognition systems have another constraint concerning
the style of speech they can recognize. There are assumed to be
three styles of speech: isolated, connected, and continuous.

Team LRN Speech-controlled mobile robot