You can define individual AFP databases in the tasks configuration file (speechserver-tasks.cfg
).
To define a database
In the [Resources]
section, add the name of the database to the list. Prefix the database name with fpdb:
(for a landmark-based database), or fptdb:
(for a template-based database). For example:
[Resources] 0=fpdb:AFP 1=fpdb:ADVERTS 2=fptdb:TEMPLATES
Note: The [Resources]
section also contains all other resource types used by the server, such as language packs used by the speech-to-text operation.
Below the list, type the name of the database inside square brackets to create a configuration section for the database. For example:
[Adverts]
In the new configuration section, add the following parameters.
Pack
|
The name of the database. |
PackDir
|
The path to the directory that contains the database files. If this directory does not already exist, manually create it before you send the action. If the database does not yet exist, IDOL Speech Server creates the database in this directory. |
FxxCacheSize
|
The size (in megabytes) of the cache for the fingerprint indexing file. |
TtxCacheSize
|
The size (in megabytes) of the cache for the FPDB time track indexing file. |
Note: The FxxCacheSize
and TtxCacheSize
parameters are applicable only for the FPDB resource type (that is, for the landmark-based databases).
For example:
[ADVERTS] Pack = Adverts PackDir = C:\resources FxxCacheSize = 2 TtxCacheSize = 200
For more information about these configuration parameters, see the IDOL Speech Server Reference.
When you send the task action, set the AfpDb
action parameter to the name of the database configuration section. For example, AfpDb=ADVERTS
.
Note: You do not need to set the Pack
and PackDir
action parameters, which makes it easier to switch between databases using just the database name.
|