|
|
| Term |
Definition |
| Broad Vocabulary |
The ability to dictate on any subject matter, as opposed to
only on specific subject matter. |
| Constrained speech recognition program |
Limiting the available vocabulary at any one point in the
dictation process. |
| Continuous-Speech Recognition |
The ability of a program to understand speech spoken without
pauses between the words |
| Discrete Speech Recognition |
The requirement to pause, usually between 1/10th and 2/10ths
second between words |
| Hidden Markov Modeling (HMM) |
A numeric analysis which determines the probability of the
next item in a string of items. Used frequently in continuous-speech recognition systems. |
| Large Vocabulary System |
A vague term which usually denotes the ability of a speech
recognition system to be used for dictation rather than merely for programming or for
command and control. Large vocabulary systems have at least 10,000 words, and usually
between 30,000 and 60,000 in their vocabulary. |
| Phoneme |
The smallest sound utterance. |
| Small Vocabulary System |
A vague term which usually denotes the ability to perform
limited functions. Vocabularies from a few hundred words to a few thousand words fit in
this category. |
| Software Developer's Kit (SDK) |
A program which is utilized by the writers of computer
software to improve or facilitate the writing of their programs. |
| Speaker Adaptive |
The ability of a speech recognition system to increase its
recognition rate by learning from the corrections which are made by the speaker. |
| Speaker Dependent |
A program which requires an enrollment period (usually 15 - 60
minutes) in order to understand a new speaker. |
| Speaker Independent |
A program which does not need an enrollment period to start
recognition. A system which will comprehend any person who speaks into the system as soon
as they start speaking. |
| Speech-to-Text |
The field of computerized voice recognition / speech
recognition. (These terms are usually used interchangeably) |
| Text-to-Speech |
Computerized enunciation of the written word. In a sense, this
is the opposite of speech-to-text. |
| Voice Identification, Voice Verification |
The use of voice recognition technology to identify a person
by their voice, rather than to identify the meaning of the spoken word. |
|
