Speech (Voice) recognition and Understanding

Speech (Voice) recognition and
Speech or voice recognition is the process of having the computer
recognizes the normal human voice.
When a speech recognition system is combined with a natural
language processing system, the result is an overall system that not
only recognizes voice input but also understands it.
 Advantages of speech recognition
Ease of access
(Typing skills)
Typing skills – many people may not be able to use computers
effectively
Speaking is easy then typing.
Speed
(speed in speaking
& typing)
Even the most competent typist can speak more quickly than
they can type.
Speed is faster in speaking than typing.
Manual freedom
(Hands free)
There are many situations in which computers might be useful to
people whose hands are otherwise occupied, such as product
assemblers’ pilot of military aircraft, and busy executives.
Remote access We can retrieve information by issuing verbal commands into a

(Verbal commands) telephone.
Accuracy
(Spelling mistakes)
Can reduce the spelling mistakes during typing.
These are minimized with voice input.
 Classifying speech recognizers
There are systems that recognize only individual works and other that
recognize continuous speech.
The systems are further classified as either speaker dependent or
speaker independent.
1) Word recognizers 3) Speaker dependent
2) Continuous speech recognizers 4) Speaker independent
 Word recognizers
A word recognizer is a speech recognition system that identifies
individual words.
Such systems are capable of recognizing only a small vocabulary of
single words or possibly simple phrases.
To give commands or to enter data to a computer using one of these
systems.
 Continuous speck recognizers
These speech recognition units recognize a continuous flow of words.
We can speak to them in complete sentences, and our input will be
recognized and possibly understood.
 Speaker dependent
A speaker dependent system must be customized to the voice of a
particular individual before it can be used.
 Speaker independent
Speaker independent systems mean that anyone can use the system.
Most speaker independent systems are incredibly complete and costly.
Some of the most advanced systems are speech recognition system
and the Sphinx and Janus systems developed at Carnegie Mellon
University.
The Bell Labs system is used, for example, for airline reservations via
human phone input.
IBM’s "Voice Type Simply Speaking" is available for running Windows
95 and applications.