If you enable the vocabulary filtering mode, the postproc
module checks against a list of words that are barred from appearing in the results. The words list might include expletives, names, or any other vocabulary that you want to exclude from the final results. If these words exist in the results, IDOL Speech Server replaces them with a specified term. By default, this term is “<BLEEP>
”; to change the term, set the BarredTerm
parameter in the postproc
module section of the tasks configuration file.
In addition, the text filtering process also supports specific replacement words. Instead of specifying a single replacement word to replace all instances of any word that appears in the barred words list, you can now map different words to different terms by adding a second word to each line in the list. For example:
ANE A&E rubbish dreadful poor
In the first line, ANE
maps to the correct word A&E
. The banned word in the second line maps to the default replacement word (as defined by the BarredTerm
parameter), because there is no specific alternative mapping. In the third line, dreadful
maps to the specified alternative mapping poor
.
You cannot map a single word to multiple replacement words; each line in the list can have a maximum of two entries.
To create a vocabulary filter list
Create a list that contains the words to filter out. Each word or pair of words is a separate list entry.
For more information about IDOL Speech Server's list manager, see Create and Manage Lists.
postproc
module section of the tasks configuration file, set the BarredList
parameter to the list name. For example, BarredList=ListManager/barred
.
|