allophone-based recognition
| Recognition technique in which the models of words (or expressions) to be recognized are made up by concatenating the models of the language's phonemes. The words may be transcribed into phonemes from a glossary or generated from an automatic text-based phonetization module.
The most common alternative to this type of modeling is whole-word based modeling.
| |
| automatic speech recognition (ASR)
| Technology used to identify (and sometimes also interpret) automatically a command, whether it is a keyword or a complete phrase pronounced by a user. The speech recognition may be in speaker-dependent or
speaker-independent modes.
| |
|
