HPE IDOL Speech Server requires supporting files, which are available in packs for different languages and operations. After you complete the installation process, download the relevant packs from the HPE Download Center and save them to the language pack root directory.
Note: The LangPackDir
parameter in the [Paths]
section of the HPE IDOL Speech Server configuration file sets the location of the language pack root directory. If you save the resource packs in a different location to the directory that you specified during installation, you must update this parameter. Alternatively, you can specify the location of individual packs as the value of the PackDir
parameter in each resource configuration section.
You must also configure some of the packs in the speechserver-tasks.cfg
configuration file. For more information, see Configure Language Packs.
HPE IDOL Speech Server resources are classified into types.
fpdb
|
Audio fingerprint databases. These are not downloadable files but are files that are created by several audio fingerprinting tasks. |
lang
|
Basic language packs, which are used for speech-to-text and text normalization. |
langvt
|
Language packs that are configured for processes including phonetic phrase match and audio security. |
sidbase
|
Base speaker identification packs, which are used for speaker identification tasks. |
|