To run speech-to-text on specific documents, you can modify the criteria in scripts/IdolSpeech.lua
, or you can use the IdolSpeech
import task and write your own Lua script to identify the documents to process. The IdolSpeech
task only processes documents that have the field AUTN_NEEDS_TRANSCRIPTION
, so your script must add this field to any document that you want to process.
The following example shows how to configure the IdolSpeech
task in the CFS configuration file:
[ImportTasks] Pre0=Lua:Identify_Audio_Files.lua Pre1=IdolSpeech:IdolSpeechSettings [IdolSpeechSettings] IdolSpeechServers=server:15000 IdolSpeechLanguage=ENUK
The Pre0
import task runs a Lua script that determines whether a file is suitable for transcription. You must write this script. The script must add the field AUTN_NEEDS_TRANSCRIPTION
to any documents that you want to process. You can include conditions in the script to filter documents based on the document source, file type, or metadata extracted by KeyView.
The Pre1
import task is the IdolSpeech
task. It specifies the name of a section in the configuration file that contains the settings for the task. In this example the section is named IdolSpeechSettings
.
The IdolSpeechServers
parameter specifies the host name or IP address, and ACI port, of your IDOL Speech Server. To use multiple Speech Servers, see Use Multiple Speech Servers.
The IdolSpeechLanguage
parameter is optional and specifies the language pack to use for transcription. You can set this parameter when all of your audio files are in the same language. If your audio files are in different languages, remove IdolSpeechLanguage
so that Speech Server uses language detection to detect the language for each document. For more information about language identification, see Language Identification.
If you prefer to send files to your IDOL Speech Server by writing them to a shared folder, add the IdolSpeechUseSharedPath
and SharedPath
parameters to the configuration:
[ImportTasks] Pre0=Lua:Identify_Audio_Files.lua Pre1=IdolSpeech:IdolSpeechSettings [IdolSpeechSettings] IdolSpeechServers=server:15000 IdolSpeechLanguage=ENUK IdolSpeechUseSharedPath=true SharedPath=\\server\SharedPath
Setting IdolSpeechUseSharedPath=TRUE
instructs CFS to send files to Speech Server by writing them to the shared folder. specified by the SharedPath
parameter.
For more information about the parameters that you can use to configure this task, refer to the Connector Framework Server Reference.
|