The phraseprematch module is the main phonetic phrase search module that processes audio files, converts them to phoneme time tracks, and identifies potential hits. It generates a word labels stream (w) containing hit positions, a phoneme data stream (p), and normalized feature frames (nf) associated with the hits.
| Mode | Input | Output | Description |
|---|---|---|---|
NOFILE
|
nf
|
w, p, nf |
Processes the normalized feature frames to perform phrase search. |
READ
|
w, p, nf, fl |
Reads an .fmd file to search for phrases. | |
WRITE
|
nf
|
Writes an .fmd file, but does not perform any phrase search. |
Example:
output <- phraseprematch(_, tr)
| File* | Model* |
| MaxInterWordFrames | Phrase |
| MaxPhraseFrames | PhraseList |
| MaxWordFrames | Threshold |
*You must set the value of these configuration parameters by setting the corresponding action parameters.
|
|