The following schema describes the process for phonetic phrase searching.
[SearchFMD] 0 = w, fl, s, p, nf <- phraseprematch(READ, input) 1 = w2 <- phrasematch(_, w, p, s, nf) 2 = output <- wout(_, w2, fl) DefaultResults = out
0 |
The phraseprematch module reads the preprocessed data file. |
1 |
The phrasematch module searches the audio phrase data from 1 (w , p , s , nf ). |
2 |
The wout module writes the recognized phrase locations resulting from 1 (w2 ) and the file labels from 0 (fl ) to the output file. |
|